diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,92862 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 5.0, + "eval_steps": 500, + "global_step": 13260, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0003770739064856712, + "grad_norm": 5.614122643897961, + "learning_rate": 6.033182503770739e-08, + "loss": 0.9623, + "step": 1 + }, + { + "epoch": 0.0007541478129713424, + "grad_norm": 5.5346283902561995, + "learning_rate": 1.2066365007541479e-07, + "loss": 0.954, + "step": 2 + }, + { + "epoch": 0.0011312217194570137, + "grad_norm": 5.508075565063379, + "learning_rate": 1.809954751131222e-07, + "loss": 0.9405, + "step": 3 + }, + { + "epoch": 0.0015082956259426848, + "grad_norm": 5.600471741887271, + "learning_rate": 2.4132730015082957e-07, + "loss": 0.9801, + "step": 4 + }, + { + "epoch": 0.001885369532428356, + "grad_norm": 5.553490994460785, + "learning_rate": 3.01659125188537e-07, + "loss": 0.9495, + "step": 5 + }, + { + "epoch": 0.0022624434389140274, + "grad_norm": 5.599503340008059, + "learning_rate": 3.619909502262444e-07, + "loss": 0.9647, + "step": 6 + }, + { + "epoch": 0.0026395173453996985, + "grad_norm": 5.541186020994753, + "learning_rate": 4.223227752639518e-07, + "loss": 0.9601, + "step": 7 + }, + { + "epoch": 0.0030165912518853697, + "grad_norm": 5.460325978239511, + "learning_rate": 4.826546003016591e-07, + "loss": 0.9458, + "step": 8 + }, + { + "epoch": 0.003393665158371041, + "grad_norm": 5.4008680686639545, + "learning_rate": 5.429864253393665e-07, + "loss": 0.9359, + "step": 9 + }, + { + "epoch": 0.003770739064856712, + "grad_norm": 5.1913431067540605, + "learning_rate": 6.03318250377074e-07, + "loss": 0.9612, + "step": 10 + }, + { + "epoch": 0.004147812971342383, + "grad_norm": 5.266977769178586, + "learning_rate": 6.636500754147814e-07, + "loss": 0.962, + "step": 11 + }, + { + "epoch": 0.004524886877828055, + "grad_norm": 5.124048622285298, + "learning_rate": 7.239819004524888e-07, + "loss": 0.9406, + "step": 12 + }, + { + "epoch": 0.004901960784313725, + "grad_norm": 4.167697136950075, + "learning_rate": 7.843137254901962e-07, + "loss": 0.9072, + "step": 13 + }, + { + "epoch": 0.005279034690799397, + "grad_norm": 4.236799199814104, + "learning_rate": 8.446455505279036e-07, + "loss": 0.9314, + "step": 14 + }, + { + "epoch": 0.005656108597285068, + "grad_norm": 4.09891132930575, + "learning_rate": 9.049773755656109e-07, + "loss": 0.8961, + "step": 15 + }, + { + "epoch": 0.006033182503770739, + "grad_norm": 3.9290217553491704, + "learning_rate": 9.653092006033183e-07, + "loss": 0.9089, + "step": 16 + }, + { + "epoch": 0.00641025641025641, + "grad_norm": 3.8244311286807893, + "learning_rate": 1.0256410256410257e-06, + "loss": 0.8971, + "step": 17 + }, + { + "epoch": 0.006787330316742082, + "grad_norm": 2.279272835872739, + "learning_rate": 1.085972850678733e-06, + "loss": 0.8706, + "step": 18 + }, + { + "epoch": 0.007164404223227752, + "grad_norm": 2.229093346253853, + "learning_rate": 1.1463046757164404e-06, + "loss": 0.8509, + "step": 19 + }, + { + "epoch": 0.007541478129713424, + "grad_norm": 2.183414587266228, + "learning_rate": 1.206636500754148e-06, + "loss": 0.8458, + "step": 20 + }, + { + "epoch": 0.007918552036199095, + "grad_norm": 2.085358234005872, + "learning_rate": 1.2669683257918552e-06, + "loss": 0.858, + "step": 21 + }, + { + "epoch": 0.008295625942684766, + "grad_norm": 1.9946922904004925, + "learning_rate": 1.3273001508295628e-06, + "loss": 0.8613, + "step": 22 + }, + { + "epoch": 0.008672699849170438, + "grad_norm": 1.8834999808308333, + "learning_rate": 1.3876319758672702e-06, + "loss": 0.8431, + "step": 23 + }, + { + "epoch": 0.00904977375565611, + "grad_norm": 1.640249255867922, + "learning_rate": 1.4479638009049775e-06, + "loss": 0.8426, + "step": 24 + }, + { + "epoch": 0.00942684766214178, + "grad_norm": 2.2845490535571686, + "learning_rate": 1.5082956259426847e-06, + "loss": 0.8205, + "step": 25 + }, + { + "epoch": 0.00980392156862745, + "grad_norm": 2.5495662824242133, + "learning_rate": 1.5686274509803923e-06, + "loss": 0.8126, + "step": 26 + }, + { + "epoch": 0.010180995475113122, + "grad_norm": 2.580062567187357, + "learning_rate": 1.6289592760180997e-06, + "loss": 0.8253, + "step": 27 + }, + { + "epoch": 0.010558069381598794, + "grad_norm": 2.5284849063286705, + "learning_rate": 1.6892911010558073e-06, + "loss": 0.8075, + "step": 28 + }, + { + "epoch": 0.010935143288084464, + "grad_norm": 2.2931201342069722, + "learning_rate": 1.7496229260935144e-06, + "loss": 0.7977, + "step": 29 + }, + { + "epoch": 0.011312217194570135, + "grad_norm": 2.0609959413373993, + "learning_rate": 1.8099547511312218e-06, + "loss": 0.8017, + "step": 30 + }, + { + "epoch": 0.011689291101055807, + "grad_norm": 1.7442022295260522, + "learning_rate": 1.8702865761689292e-06, + "loss": 0.7958, + "step": 31 + }, + { + "epoch": 0.012066365007541479, + "grad_norm": 1.1796524771385322, + "learning_rate": 1.9306184012066366e-06, + "loss": 0.7852, + "step": 32 + }, + { + "epoch": 0.012443438914027148, + "grad_norm": 1.042685722819972, + "learning_rate": 1.9909502262443437e-06, + "loss": 0.7699, + "step": 33 + }, + { + "epoch": 0.01282051282051282, + "grad_norm": 1.1873067463784812, + "learning_rate": 2.0512820512820513e-06, + "loss": 0.7694, + "step": 34 + }, + { + "epoch": 0.013197586726998492, + "grad_norm": 1.1365935260801048, + "learning_rate": 2.111613876319759e-06, + "loss": 0.7525, + "step": 35 + }, + { + "epoch": 0.013574660633484163, + "grad_norm": 0.9687384361707908, + "learning_rate": 2.171945701357466e-06, + "loss": 0.7519, + "step": 36 + }, + { + "epoch": 0.013951734539969835, + "grad_norm": 0.7943829443655575, + "learning_rate": 2.2322775263951737e-06, + "loss": 0.7414, + "step": 37 + }, + { + "epoch": 0.014328808446455505, + "grad_norm": 0.7351845727973827, + "learning_rate": 2.292609351432881e-06, + "loss": 0.7319, + "step": 38 + }, + { + "epoch": 0.014705882352941176, + "grad_norm": 0.7639537379200323, + "learning_rate": 2.3529411764705885e-06, + "loss": 0.7361, + "step": 39 + }, + { + "epoch": 0.015082956259426848, + "grad_norm": 0.6876784614629111, + "learning_rate": 2.413273001508296e-06, + "loss": 0.7269, + "step": 40 + }, + { + "epoch": 0.01546003016591252, + "grad_norm": 0.6382762211049149, + "learning_rate": 2.4736048265460032e-06, + "loss": 0.7198, + "step": 41 + }, + { + "epoch": 0.01583710407239819, + "grad_norm": 0.5552712670269359, + "learning_rate": 2.5339366515837104e-06, + "loss": 0.7255, + "step": 42 + }, + { + "epoch": 0.016214177978883863, + "grad_norm": 0.5442406641707824, + "learning_rate": 2.5942684766214184e-06, + "loss": 0.7006, + "step": 43 + }, + { + "epoch": 0.016591251885369532, + "grad_norm": 0.6374174831261075, + "learning_rate": 2.6546003016591256e-06, + "loss": 0.7068, + "step": 44 + }, + { + "epoch": 0.016968325791855202, + "grad_norm": 0.6293528656815067, + "learning_rate": 2.7149321266968327e-06, + "loss": 0.6988, + "step": 45 + }, + { + "epoch": 0.017345399698340876, + "grad_norm": 0.5491083083435888, + "learning_rate": 2.7752639517345403e-06, + "loss": 0.6986, + "step": 46 + }, + { + "epoch": 0.017722473604826545, + "grad_norm": 0.4250094985686511, + "learning_rate": 2.8355957767722475e-06, + "loss": 0.6901, + "step": 47 + }, + { + "epoch": 0.01809954751131222, + "grad_norm": 0.4052592505905265, + "learning_rate": 2.895927601809955e-06, + "loss": 0.6768, + "step": 48 + }, + { + "epoch": 0.01847662141779789, + "grad_norm": 0.4784439261884374, + "learning_rate": 2.9562594268476623e-06, + "loss": 0.6784, + "step": 49 + }, + { + "epoch": 0.01885369532428356, + "grad_norm": 0.4784907566632004, + "learning_rate": 3.0165912518853694e-06, + "loss": 0.6902, + "step": 50 + }, + { + "epoch": 0.019230769230769232, + "grad_norm": 0.439900784694477, + "learning_rate": 3.0769230769230774e-06, + "loss": 0.6719, + "step": 51 + }, + { + "epoch": 0.0196078431372549, + "grad_norm": 0.38597781397085723, + "learning_rate": 3.1372549019607846e-06, + "loss": 0.685, + "step": 52 + }, + { + "epoch": 0.01998491704374057, + "grad_norm": 0.31076279922380046, + "learning_rate": 3.1975867269984918e-06, + "loss": 0.6714, + "step": 53 + }, + { + "epoch": 0.020361990950226245, + "grad_norm": 0.3223502388007259, + "learning_rate": 3.2579185520361994e-06, + "loss": 0.6842, + "step": 54 + }, + { + "epoch": 0.020739064856711915, + "grad_norm": 0.35242945376764434, + "learning_rate": 3.3182503770739065e-06, + "loss": 0.6603, + "step": 55 + }, + { + "epoch": 0.021116138763197588, + "grad_norm": 0.35399634051517714, + "learning_rate": 3.3785822021116145e-06, + "loss": 0.6803, + "step": 56 + }, + { + "epoch": 0.021493212669683258, + "grad_norm": 0.28571656869153067, + "learning_rate": 3.4389140271493217e-06, + "loss": 0.6667, + "step": 57 + }, + { + "epoch": 0.021870286576168928, + "grad_norm": 0.27572399351213894, + "learning_rate": 3.499245852187029e-06, + "loss": 0.6441, + "step": 58 + }, + { + "epoch": 0.0222473604826546, + "grad_norm": 0.32154542645259976, + "learning_rate": 3.5595776772247365e-06, + "loss": 0.6512, + "step": 59 + }, + { + "epoch": 0.02262443438914027, + "grad_norm": 0.2721247758554184, + "learning_rate": 3.6199095022624436e-06, + "loss": 0.6647, + "step": 60 + }, + { + "epoch": 0.023001508295625944, + "grad_norm": 0.25886793775367134, + "learning_rate": 3.6802413273001512e-06, + "loss": 0.6469, + "step": 61 + }, + { + "epoch": 0.023378582202111614, + "grad_norm": 0.24853346568900633, + "learning_rate": 3.7405731523378584e-06, + "loss": 0.6389, + "step": 62 + }, + { + "epoch": 0.023755656108597284, + "grad_norm": 0.24846858297191, + "learning_rate": 3.8009049773755656e-06, + "loss": 0.6412, + "step": 63 + }, + { + "epoch": 0.024132730015082957, + "grad_norm": 0.255753132589998, + "learning_rate": 3.861236802413273e-06, + "loss": 0.6637, + "step": 64 + }, + { + "epoch": 0.024509803921568627, + "grad_norm": 0.25420684848634, + "learning_rate": 3.92156862745098e-06, + "loss": 0.6616, + "step": 65 + }, + { + "epoch": 0.024886877828054297, + "grad_norm": 0.22823452678387043, + "learning_rate": 3.9819004524886875e-06, + "loss": 0.6465, + "step": 66 + }, + { + "epoch": 0.02526395173453997, + "grad_norm": 0.2338106747342013, + "learning_rate": 4.0422322775263955e-06, + "loss": 0.6389, + "step": 67 + }, + { + "epoch": 0.02564102564102564, + "grad_norm": 0.24794035281294352, + "learning_rate": 4.102564102564103e-06, + "loss": 0.6505, + "step": 68 + }, + { + "epoch": 0.026018099547511313, + "grad_norm": 0.21408360191731488, + "learning_rate": 4.162895927601811e-06, + "loss": 0.659, + "step": 69 + }, + { + "epoch": 0.026395173453996983, + "grad_norm": 0.1941973352243594, + "learning_rate": 4.223227752639518e-06, + "loss": 0.6402, + "step": 70 + }, + { + "epoch": 0.026772247360482653, + "grad_norm": 0.2230078259517713, + "learning_rate": 4.283559577677225e-06, + "loss": 0.6319, + "step": 71 + }, + { + "epoch": 0.027149321266968326, + "grad_norm": 0.19275978055853862, + "learning_rate": 4.343891402714932e-06, + "loss": 0.6279, + "step": 72 + }, + { + "epoch": 0.027526395173453996, + "grad_norm": 0.19522438993603042, + "learning_rate": 4.404223227752639e-06, + "loss": 0.6403, + "step": 73 + }, + { + "epoch": 0.02790346907993967, + "grad_norm": 0.1805007931813921, + "learning_rate": 4.464555052790347e-06, + "loss": 0.6325, + "step": 74 + }, + { + "epoch": 0.02828054298642534, + "grad_norm": 0.2119624747373943, + "learning_rate": 4.5248868778280546e-06, + "loss": 0.6362, + "step": 75 + }, + { + "epoch": 0.02865761689291101, + "grad_norm": 0.17740531418594224, + "learning_rate": 4.585218702865762e-06, + "loss": 0.6281, + "step": 76 + }, + { + "epoch": 0.029034690799396683, + "grad_norm": 0.18799506157024637, + "learning_rate": 4.64555052790347e-06, + "loss": 0.6335, + "step": 77 + }, + { + "epoch": 0.029411764705882353, + "grad_norm": 0.17604173866776288, + "learning_rate": 4.705882352941177e-06, + "loss": 0.6219, + "step": 78 + }, + { + "epoch": 0.029788838612368026, + "grad_norm": 0.1837234999807563, + "learning_rate": 4.766214177978885e-06, + "loss": 0.6428, + "step": 79 + }, + { + "epoch": 0.030165912518853696, + "grad_norm": 0.1991927463042432, + "learning_rate": 4.826546003016592e-06, + "loss": 0.6413, + "step": 80 + }, + { + "epoch": 0.030542986425339366, + "grad_norm": 0.17211050055685104, + "learning_rate": 4.886877828054299e-06, + "loss": 0.6081, + "step": 81 + }, + { + "epoch": 0.03092006033182504, + "grad_norm": 0.18741546596262823, + "learning_rate": 4.9472096530920064e-06, + "loss": 0.6242, + "step": 82 + }, + { + "epoch": 0.03129713423831071, + "grad_norm": 0.17408288533207283, + "learning_rate": 5.007541478129714e-06, + "loss": 0.6184, + "step": 83 + }, + { + "epoch": 0.03167420814479638, + "grad_norm": 0.19837044376910315, + "learning_rate": 5.067873303167421e-06, + "loss": 0.6245, + "step": 84 + }, + { + "epoch": 0.03205128205128205, + "grad_norm": 0.15810864165392968, + "learning_rate": 5.128205128205128e-06, + "loss": 0.6242, + "step": 85 + }, + { + "epoch": 0.032428355957767725, + "grad_norm": 0.19258413941694802, + "learning_rate": 5.188536953242837e-06, + "loss": 0.6164, + "step": 86 + }, + { + "epoch": 0.032805429864253395, + "grad_norm": 0.18503927985302768, + "learning_rate": 5.248868778280544e-06, + "loss": 0.6114, + "step": 87 + }, + { + "epoch": 0.033182503770739065, + "grad_norm": 0.155472694900789, + "learning_rate": 5.309200603318251e-06, + "loss": 0.6164, + "step": 88 + }, + { + "epoch": 0.033559577677224735, + "grad_norm": 0.15996122513327246, + "learning_rate": 5.369532428355958e-06, + "loss": 0.6165, + "step": 89 + }, + { + "epoch": 0.033936651583710405, + "grad_norm": 0.1831202624037049, + "learning_rate": 5.4298642533936655e-06, + "loss": 0.6039, + "step": 90 + }, + { + "epoch": 0.03431372549019608, + "grad_norm": 0.1571718554495097, + "learning_rate": 5.4901960784313735e-06, + "loss": 0.6132, + "step": 91 + }, + { + "epoch": 0.03469079939668175, + "grad_norm": 0.1504841336032594, + "learning_rate": 5.550527903469081e-06, + "loss": 0.6018, + "step": 92 + }, + { + "epoch": 0.03506787330316742, + "grad_norm": 0.16381284503174218, + "learning_rate": 5.610859728506788e-06, + "loss": 0.6135, + "step": 93 + }, + { + "epoch": 0.03544494720965309, + "grad_norm": 0.16419009740195079, + "learning_rate": 5.671191553544495e-06, + "loss": 0.6061, + "step": 94 + }, + { + "epoch": 0.03582202111613876, + "grad_norm": 0.2227824197253181, + "learning_rate": 5.731523378582202e-06, + "loss": 0.6214, + "step": 95 + }, + { + "epoch": 0.03619909502262444, + "grad_norm": 0.2751462663010355, + "learning_rate": 5.79185520361991e-06, + "loss": 0.611, + "step": 96 + }, + { + "epoch": 0.03657616892911011, + "grad_norm": 0.29464277866111926, + "learning_rate": 5.852187028657617e-06, + "loss": 0.6125, + "step": 97 + }, + { + "epoch": 0.03695324283559578, + "grad_norm": 0.24882133142844326, + "learning_rate": 5.9125188536953245e-06, + "loss": 0.6151, + "step": 98 + }, + { + "epoch": 0.03733031674208145, + "grad_norm": 0.20354588861000328, + "learning_rate": 5.972850678733032e-06, + "loss": 0.6049, + "step": 99 + }, + { + "epoch": 0.03770739064856712, + "grad_norm": 0.28718784204865017, + "learning_rate": 6.033182503770739e-06, + "loss": 0.6157, + "step": 100 + }, + { + "epoch": 0.03808446455505279, + "grad_norm": 0.21540346698135138, + "learning_rate": 6.093514328808446e-06, + "loss": 0.6141, + "step": 101 + }, + { + "epoch": 0.038461538461538464, + "grad_norm": 0.1832933204280512, + "learning_rate": 6.153846153846155e-06, + "loss": 0.608, + "step": 102 + }, + { + "epoch": 0.038838612368024134, + "grad_norm": 0.2311438726328487, + "learning_rate": 6.214177978883862e-06, + "loss": 0.6111, + "step": 103 + }, + { + "epoch": 0.0392156862745098, + "grad_norm": 0.1734443946709334, + "learning_rate": 6.274509803921569e-06, + "loss": 0.6035, + "step": 104 + }, + { + "epoch": 0.03959276018099547, + "grad_norm": 0.19954213432759152, + "learning_rate": 6.334841628959276e-06, + "loss": 0.6051, + "step": 105 + }, + { + "epoch": 0.03996983408748114, + "grad_norm": 0.21674946634731077, + "learning_rate": 6.3951734539969835e-06, + "loss": 0.6131, + "step": 106 + }, + { + "epoch": 0.04034690799396682, + "grad_norm": 0.16250961138552467, + "learning_rate": 6.4555052790346916e-06, + "loss": 0.5957, + "step": 107 + }, + { + "epoch": 0.04072398190045249, + "grad_norm": 0.2244591947401309, + "learning_rate": 6.515837104072399e-06, + "loss": 0.6042, + "step": 108 + }, + { + "epoch": 0.04110105580693816, + "grad_norm": 0.18857380282430852, + "learning_rate": 6.576168929110106e-06, + "loss": 0.6065, + "step": 109 + }, + { + "epoch": 0.04147812971342383, + "grad_norm": 0.1817667903211212, + "learning_rate": 6.636500754147813e-06, + "loss": 0.6143, + "step": 110 + }, + { + "epoch": 0.0418552036199095, + "grad_norm": 0.1954469933114291, + "learning_rate": 6.69683257918552e-06, + "loss": 0.5997, + "step": 111 + }, + { + "epoch": 0.042232277526395176, + "grad_norm": 0.1509168410253301, + "learning_rate": 6.757164404223229e-06, + "loss": 0.6194, + "step": 112 + }, + { + "epoch": 0.042609351432880846, + "grad_norm": 0.17225583659137259, + "learning_rate": 6.817496229260936e-06, + "loss": 0.5936, + "step": 113 + }, + { + "epoch": 0.042986425339366516, + "grad_norm": 0.14189749137256388, + "learning_rate": 6.8778280542986434e-06, + "loss": 0.5966, + "step": 114 + }, + { + "epoch": 0.043363499245852186, + "grad_norm": 0.16275045556293566, + "learning_rate": 6.938159879336351e-06, + "loss": 0.5848, + "step": 115 + }, + { + "epoch": 0.043740573152337855, + "grad_norm": 0.16228748665251988, + "learning_rate": 6.998491704374058e-06, + "loss": 0.5943, + "step": 116 + }, + { + "epoch": 0.04411764705882353, + "grad_norm": 0.18552087570968728, + "learning_rate": 7.058823529411766e-06, + "loss": 0.6065, + "step": 117 + }, + { + "epoch": 0.0444947209653092, + "grad_norm": 0.19685467922717909, + "learning_rate": 7.119155354449473e-06, + "loss": 0.5893, + "step": 118 + }, + { + "epoch": 0.04487179487179487, + "grad_norm": 0.17613903673740228, + "learning_rate": 7.17948717948718e-06, + "loss": 0.5939, + "step": 119 + }, + { + "epoch": 0.04524886877828054, + "grad_norm": 0.21535791887834285, + "learning_rate": 7.239819004524887e-06, + "loss": 0.5919, + "step": 120 + }, + { + "epoch": 0.04562594268476621, + "grad_norm": 0.19565876137086893, + "learning_rate": 7.3001508295625945e-06, + "loss": 0.5994, + "step": 121 + }, + { + "epoch": 0.04600301659125189, + "grad_norm": 0.19079234956529986, + "learning_rate": 7.3604826546003025e-06, + "loss": 0.6019, + "step": 122 + }, + { + "epoch": 0.04638009049773756, + "grad_norm": 0.19820461001358874, + "learning_rate": 7.42081447963801e-06, + "loss": 0.5865, + "step": 123 + }, + { + "epoch": 0.04675716440422323, + "grad_norm": 0.23688620627205897, + "learning_rate": 7.481146304675717e-06, + "loss": 0.5846, + "step": 124 + }, + { + "epoch": 0.0471342383107089, + "grad_norm": 0.22767006683152782, + "learning_rate": 7.541478129713424e-06, + "loss": 0.5808, + "step": 125 + }, + { + "epoch": 0.04751131221719457, + "grad_norm": 0.19085527823390494, + "learning_rate": 7.601809954751131e-06, + "loss": 0.5851, + "step": 126 + }, + { + "epoch": 0.047888386123680245, + "grad_norm": 0.21595662520062966, + "learning_rate": 7.66214177978884e-06, + "loss": 0.5927, + "step": 127 + }, + { + "epoch": 0.048265460030165915, + "grad_norm": 0.28022040295421335, + "learning_rate": 7.722473604826546e-06, + "loss": 0.5716, + "step": 128 + }, + { + "epoch": 0.048642533936651584, + "grad_norm": 0.29146140710564616, + "learning_rate": 7.782805429864253e-06, + "loss": 0.5919, + "step": 129 + }, + { + "epoch": 0.049019607843137254, + "grad_norm": 0.24369109196643904, + "learning_rate": 7.84313725490196e-06, + "loss": 0.5799, + "step": 130 + }, + { + "epoch": 0.049396681749622924, + "grad_norm": 0.21596467882510184, + "learning_rate": 7.903469079939668e-06, + "loss": 0.5892, + "step": 131 + }, + { + "epoch": 0.049773755656108594, + "grad_norm": 0.27378002294563936, + "learning_rate": 7.963800904977375e-06, + "loss": 0.6049, + "step": 132 + }, + { + "epoch": 0.05015082956259427, + "grad_norm": 0.29725109237954905, + "learning_rate": 8.024132730015084e-06, + "loss": 0.5971, + "step": 133 + }, + { + "epoch": 0.05052790346907994, + "grad_norm": 0.2620075176024356, + "learning_rate": 8.084464555052791e-06, + "loss": 0.5798, + "step": 134 + }, + { + "epoch": 0.05090497737556561, + "grad_norm": 0.19847795372935012, + "learning_rate": 8.144796380090498e-06, + "loss": 0.5868, + "step": 135 + }, + { + "epoch": 0.05128205128205128, + "grad_norm": 0.30968142258441483, + "learning_rate": 8.205128205128205e-06, + "loss": 0.5912, + "step": 136 + }, + { + "epoch": 0.05165912518853695, + "grad_norm": 0.2621024516582372, + "learning_rate": 8.265460030165913e-06, + "loss": 0.5926, + "step": 137 + }, + { + "epoch": 0.05203619909502263, + "grad_norm": 0.2593496955690314, + "learning_rate": 8.325791855203621e-06, + "loss": 0.5833, + "step": 138 + }, + { + "epoch": 0.0524132730015083, + "grad_norm": 0.26601164763252416, + "learning_rate": 8.386123680241329e-06, + "loss": 0.5772, + "step": 139 + }, + { + "epoch": 0.05279034690799397, + "grad_norm": 0.2195300398550268, + "learning_rate": 8.446455505279036e-06, + "loss": 0.5888, + "step": 140 + }, + { + "epoch": 0.053167420814479636, + "grad_norm": 0.21217739026909513, + "learning_rate": 8.506787330316743e-06, + "loss": 0.5892, + "step": 141 + }, + { + "epoch": 0.053544494720965306, + "grad_norm": 0.32399942974519036, + "learning_rate": 8.56711915535445e-06, + "loss": 0.5868, + "step": 142 + }, + { + "epoch": 0.05392156862745098, + "grad_norm": 0.4027899138371131, + "learning_rate": 8.627450980392157e-06, + "loss": 0.5869, + "step": 143 + }, + { + "epoch": 0.05429864253393665, + "grad_norm": 0.4420536499649134, + "learning_rate": 8.687782805429864e-06, + "loss": 0.5877, + "step": 144 + }, + { + "epoch": 0.05467571644042232, + "grad_norm": 0.5290295589010047, + "learning_rate": 8.748114630467572e-06, + "loss": 0.587, + "step": 145 + }, + { + "epoch": 0.05505279034690799, + "grad_norm": 0.5469502796969072, + "learning_rate": 8.808446455505279e-06, + "loss": 0.5802, + "step": 146 + }, + { + "epoch": 0.05542986425339366, + "grad_norm": 0.27115893192511425, + "learning_rate": 8.868778280542986e-06, + "loss": 0.5827, + "step": 147 + }, + { + "epoch": 0.05580693815987934, + "grad_norm": 0.3504057926409703, + "learning_rate": 8.929110105580695e-06, + "loss": 0.5704, + "step": 148 + }, + { + "epoch": 0.05618401206636501, + "grad_norm": 0.5985673394806671, + "learning_rate": 8.989441930618402e-06, + "loss": 0.5842, + "step": 149 + }, + { + "epoch": 0.05656108597285068, + "grad_norm": 0.46779618154417363, + "learning_rate": 9.049773755656109e-06, + "loss": 0.5859, + "step": 150 + }, + { + "epoch": 0.05693815987933635, + "grad_norm": 0.2638333083882535, + "learning_rate": 9.110105580693816e-06, + "loss": 0.578, + "step": 151 + }, + { + "epoch": 0.05731523378582202, + "grad_norm": 0.40700034532414137, + "learning_rate": 9.170437405731523e-06, + "loss": 0.5638, + "step": 152 + }, + { + "epoch": 0.057692307692307696, + "grad_norm": 0.4786651958186014, + "learning_rate": 9.230769230769232e-06, + "loss": 0.5801, + "step": 153 + }, + { + "epoch": 0.058069381598793365, + "grad_norm": 0.2955240307335612, + "learning_rate": 9.29110105580694e-06, + "loss": 0.5581, + "step": 154 + }, + { + "epoch": 0.058446455505279035, + "grad_norm": 0.406058210482642, + "learning_rate": 9.351432880844647e-06, + "loss": 0.5813, + "step": 155 + }, + { + "epoch": 0.058823529411764705, + "grad_norm": 0.4185186662822101, + "learning_rate": 9.411764705882354e-06, + "loss": 0.573, + "step": 156 + }, + { + "epoch": 0.059200603318250375, + "grad_norm": 0.3033670131597742, + "learning_rate": 9.472096530920061e-06, + "loss": 0.5572, + "step": 157 + }, + { + "epoch": 0.05957767722473605, + "grad_norm": 0.42114660402711496, + "learning_rate": 9.53242835595777e-06, + "loss": 0.5803, + "step": 158 + }, + { + "epoch": 0.05995475113122172, + "grad_norm": 0.40388693103234896, + "learning_rate": 9.592760180995477e-06, + "loss": 0.5858, + "step": 159 + }, + { + "epoch": 0.06033182503770739, + "grad_norm": 0.32520641504383074, + "learning_rate": 9.653092006033184e-06, + "loss": 0.5758, + "step": 160 + }, + { + "epoch": 0.06070889894419306, + "grad_norm": 0.4198271390910551, + "learning_rate": 9.713423831070891e-06, + "loss": 0.5743, + "step": 161 + }, + { + "epoch": 0.06108597285067873, + "grad_norm": 0.38782734632284377, + "learning_rate": 9.773755656108599e-06, + "loss": 0.5802, + "step": 162 + }, + { + "epoch": 0.0614630467571644, + "grad_norm": 0.29943901816979823, + "learning_rate": 9.834087481146306e-06, + "loss": 0.578, + "step": 163 + }, + { + "epoch": 0.06184012066365008, + "grad_norm": 0.26556647606659195, + "learning_rate": 9.894419306184013e-06, + "loss": 0.57, + "step": 164 + }, + { + "epoch": 0.06221719457013575, + "grad_norm": 0.3085968140716229, + "learning_rate": 9.95475113122172e-06, + "loss": 0.5752, + "step": 165 + }, + { + "epoch": 0.06259426847662142, + "grad_norm": 0.31294690671392056, + "learning_rate": 1.0015082956259427e-05, + "loss": 0.5845, + "step": 166 + }, + { + "epoch": 0.06297134238310709, + "grad_norm": 0.18608242094169755, + "learning_rate": 1.0075414781297134e-05, + "loss": 0.5873, + "step": 167 + }, + { + "epoch": 0.06334841628959276, + "grad_norm": 0.26372876777896287, + "learning_rate": 1.0135746606334842e-05, + "loss": 0.5858, + "step": 168 + }, + { + "epoch": 0.06372549019607843, + "grad_norm": 0.24389882358266, + "learning_rate": 1.0196078431372549e-05, + "loss": 0.5783, + "step": 169 + }, + { + "epoch": 0.0641025641025641, + "grad_norm": 0.24706649191682362, + "learning_rate": 1.0256410256410256e-05, + "loss": 0.5802, + "step": 170 + }, + { + "epoch": 0.06447963800904978, + "grad_norm": 0.2671796995099201, + "learning_rate": 1.0316742081447966e-05, + "loss": 0.5751, + "step": 171 + }, + { + "epoch": 0.06485671191553545, + "grad_norm": 0.29151265994188347, + "learning_rate": 1.0377073906485674e-05, + "loss": 0.5622, + "step": 172 + }, + { + "epoch": 0.06523378582202112, + "grad_norm": 0.3180884139291169, + "learning_rate": 1.043740573152338e-05, + "loss": 0.5767, + "step": 173 + }, + { + "epoch": 0.06561085972850679, + "grad_norm": 0.34221060934556474, + "learning_rate": 1.0497737556561088e-05, + "loss": 0.5704, + "step": 174 + }, + { + "epoch": 0.06598793363499246, + "grad_norm": 0.3546356721847451, + "learning_rate": 1.0558069381598795e-05, + "loss": 0.571, + "step": 175 + }, + { + "epoch": 0.06636500754147813, + "grad_norm": 0.39066322274448156, + "learning_rate": 1.0618401206636502e-05, + "loss": 0.5687, + "step": 176 + }, + { + "epoch": 0.0667420814479638, + "grad_norm": 0.44285685944405395, + "learning_rate": 1.067873303167421e-05, + "loss": 0.568, + "step": 177 + }, + { + "epoch": 0.06711915535444947, + "grad_norm": 0.4160508832483384, + "learning_rate": 1.0739064856711917e-05, + "loss": 0.5817, + "step": 178 + }, + { + "epoch": 0.06749622926093514, + "grad_norm": 0.40014395541929143, + "learning_rate": 1.0799396681749624e-05, + "loss": 0.5839, + "step": 179 + }, + { + "epoch": 0.06787330316742081, + "grad_norm": 0.3271476214755996, + "learning_rate": 1.0859728506787331e-05, + "loss": 0.5886, + "step": 180 + }, + { + "epoch": 0.06825037707390648, + "grad_norm": 0.29399709855452855, + "learning_rate": 1.0920060331825038e-05, + "loss": 0.5806, + "step": 181 + }, + { + "epoch": 0.06862745098039216, + "grad_norm": 0.3595193913672239, + "learning_rate": 1.0980392156862747e-05, + "loss": 0.5751, + "step": 182 + }, + { + "epoch": 0.06900452488687783, + "grad_norm": 0.31038422812557637, + "learning_rate": 1.1040723981900454e-05, + "loss": 0.5752, + "step": 183 + }, + { + "epoch": 0.0693815987933635, + "grad_norm": 0.34942262985182587, + "learning_rate": 1.1101055806938161e-05, + "loss": 0.5691, + "step": 184 + }, + { + "epoch": 0.06975867269984917, + "grad_norm": 0.3374488248725956, + "learning_rate": 1.1161387631975868e-05, + "loss": 0.5736, + "step": 185 + }, + { + "epoch": 0.07013574660633484, + "grad_norm": 0.18465473173703645, + "learning_rate": 1.1221719457013576e-05, + "loss": 0.5636, + "step": 186 + }, + { + "epoch": 0.07051282051282051, + "grad_norm": 0.3048571669265055, + "learning_rate": 1.1282051282051283e-05, + "loss": 0.5514, + "step": 187 + }, + { + "epoch": 0.07088989441930618, + "grad_norm": 0.3846062369692668, + "learning_rate": 1.134238310708899e-05, + "loss": 0.5669, + "step": 188 + }, + { + "epoch": 0.07126696832579185, + "grad_norm": 0.4415738087763715, + "learning_rate": 1.1402714932126697e-05, + "loss": 0.567, + "step": 189 + }, + { + "epoch": 0.07164404223227752, + "grad_norm": 0.444678460491635, + "learning_rate": 1.1463046757164404e-05, + "loss": 0.5708, + "step": 190 + }, + { + "epoch": 0.07202111613876319, + "grad_norm": 0.45055161361980306, + "learning_rate": 1.1523378582202111e-05, + "loss": 0.5729, + "step": 191 + }, + { + "epoch": 0.07239819004524888, + "grad_norm": 0.44808343268983847, + "learning_rate": 1.158371040723982e-05, + "loss": 0.5682, + "step": 192 + }, + { + "epoch": 0.07277526395173455, + "grad_norm": 0.4392101255680459, + "learning_rate": 1.1644042232277528e-05, + "loss": 0.5798, + "step": 193 + }, + { + "epoch": 0.07315233785822021, + "grad_norm": 0.43368421849819744, + "learning_rate": 1.1704374057315235e-05, + "loss": 0.5607, + "step": 194 + }, + { + "epoch": 0.07352941176470588, + "grad_norm": 0.30626668511025895, + "learning_rate": 1.1764705882352942e-05, + "loss": 0.5715, + "step": 195 + }, + { + "epoch": 0.07390648567119155, + "grad_norm": 0.4269584248897345, + "learning_rate": 1.1825037707390649e-05, + "loss": 0.5681, + "step": 196 + }, + { + "epoch": 0.07428355957767722, + "grad_norm": 0.621448148033233, + "learning_rate": 1.1885369532428356e-05, + "loss": 0.5688, + "step": 197 + }, + { + "epoch": 0.0746606334841629, + "grad_norm": 0.5649324355887121, + "learning_rate": 1.1945701357466063e-05, + "loss": 0.5759, + "step": 198 + }, + { + "epoch": 0.07503770739064856, + "grad_norm": 0.43599326935906013, + "learning_rate": 1.200603318250377e-05, + "loss": 0.5612, + "step": 199 + }, + { + "epoch": 0.07541478129713423, + "grad_norm": 0.3498718698211955, + "learning_rate": 1.2066365007541478e-05, + "loss": 0.5716, + "step": 200 + }, + { + "epoch": 0.0757918552036199, + "grad_norm": 0.4229305577191288, + "learning_rate": 1.2126696832579185e-05, + "loss": 0.5656, + "step": 201 + }, + { + "epoch": 0.07616892911010557, + "grad_norm": 0.41171472930236636, + "learning_rate": 1.2187028657616892e-05, + "loss": 0.5629, + "step": 202 + }, + { + "epoch": 0.07654600301659126, + "grad_norm": 0.39822444984040384, + "learning_rate": 1.2247360482654603e-05, + "loss": 0.5795, + "step": 203 + }, + { + "epoch": 0.07692307692307693, + "grad_norm": 0.32789577831384015, + "learning_rate": 1.230769230769231e-05, + "loss": 0.5584, + "step": 204 + }, + { + "epoch": 0.0773001508295626, + "grad_norm": 0.3589832090256609, + "learning_rate": 1.2368024132730017e-05, + "loss": 0.5648, + "step": 205 + }, + { + "epoch": 0.07767722473604827, + "grad_norm": 0.35939064862017384, + "learning_rate": 1.2428355957767724e-05, + "loss": 0.5727, + "step": 206 + }, + { + "epoch": 0.07805429864253394, + "grad_norm": 0.372842367339838, + "learning_rate": 1.2488687782805431e-05, + "loss": 0.5706, + "step": 207 + }, + { + "epoch": 0.0784313725490196, + "grad_norm": 0.32624881848449117, + "learning_rate": 1.2549019607843138e-05, + "loss": 0.5711, + "step": 208 + }, + { + "epoch": 0.07880844645550528, + "grad_norm": 0.44624452159290307, + "learning_rate": 1.2609351432880846e-05, + "loss": 0.5587, + "step": 209 + }, + { + "epoch": 0.07918552036199095, + "grad_norm": 0.43916813577451913, + "learning_rate": 1.2669683257918553e-05, + "loss": 0.5748, + "step": 210 + }, + { + "epoch": 0.07956259426847662, + "grad_norm": 0.47393860960844275, + "learning_rate": 1.273001508295626e-05, + "loss": 0.5562, + "step": 211 + }, + { + "epoch": 0.07993966817496229, + "grad_norm": 0.5105753983712781, + "learning_rate": 1.2790346907993967e-05, + "loss": 0.5642, + "step": 212 + }, + { + "epoch": 0.08031674208144797, + "grad_norm": 0.6880143977512962, + "learning_rate": 1.2850678733031676e-05, + "loss": 0.5641, + "step": 213 + }, + { + "epoch": 0.08069381598793364, + "grad_norm": 0.785028571707085, + "learning_rate": 1.2911010558069383e-05, + "loss": 0.561, + "step": 214 + }, + { + "epoch": 0.08107088989441931, + "grad_norm": 0.7185904816404678, + "learning_rate": 1.297134238310709e-05, + "loss": 0.5697, + "step": 215 + }, + { + "epoch": 0.08144796380090498, + "grad_norm": 0.5635299974866919, + "learning_rate": 1.3031674208144797e-05, + "loss": 0.5506, + "step": 216 + }, + { + "epoch": 0.08182503770739065, + "grad_norm": 0.5580202782093856, + "learning_rate": 1.3092006033182505e-05, + "loss": 0.5606, + "step": 217 + }, + { + "epoch": 0.08220211161387632, + "grad_norm": 0.6397078280977471, + "learning_rate": 1.3152337858220212e-05, + "loss": 0.5743, + "step": 218 + }, + { + "epoch": 0.08257918552036199, + "grad_norm": 0.7070603387974327, + "learning_rate": 1.3212669683257919e-05, + "loss": 0.5691, + "step": 219 + }, + { + "epoch": 0.08295625942684766, + "grad_norm": 0.6333869442495119, + "learning_rate": 1.3273001508295626e-05, + "loss": 0.5517, + "step": 220 + }, + { + "epoch": 0.08333333333333333, + "grad_norm": 0.4434809957114788, + "learning_rate": 1.3333333333333333e-05, + "loss": 0.5701, + "step": 221 + }, + { + "epoch": 0.083710407239819, + "grad_norm": 0.3370137007781036, + "learning_rate": 1.339366515837104e-05, + "loss": 0.5491, + "step": 222 + }, + { + "epoch": 0.08408748114630468, + "grad_norm": 0.5397596618019787, + "learning_rate": 1.3453996983408751e-05, + "loss": 0.5635, + "step": 223 + }, + { + "epoch": 0.08446455505279035, + "grad_norm": 0.6503756316648931, + "learning_rate": 1.3514328808446458e-05, + "loss": 0.561, + "step": 224 + }, + { + "epoch": 0.08484162895927602, + "grad_norm": 0.701297714470785, + "learning_rate": 1.3574660633484165e-05, + "loss": 0.5655, + "step": 225 + }, + { + "epoch": 0.08521870286576169, + "grad_norm": 0.5492106909335944, + "learning_rate": 1.3634992458521873e-05, + "loss": 0.556, + "step": 226 + }, + { + "epoch": 0.08559577677224736, + "grad_norm": 0.5236107236410761, + "learning_rate": 1.369532428355958e-05, + "loss": 0.5693, + "step": 227 + }, + { + "epoch": 0.08597285067873303, + "grad_norm": 0.6134886051888156, + "learning_rate": 1.3755656108597287e-05, + "loss": 0.5534, + "step": 228 + }, + { + "epoch": 0.0863499245852187, + "grad_norm": 0.6267060874852842, + "learning_rate": 1.3815987933634994e-05, + "loss": 0.5775, + "step": 229 + }, + { + "epoch": 0.08672699849170437, + "grad_norm": 0.5406523417799868, + "learning_rate": 1.3876319758672701e-05, + "loss": 0.5688, + "step": 230 + }, + { + "epoch": 0.08710407239819004, + "grad_norm": 0.4726689134401687, + "learning_rate": 1.3936651583710408e-05, + "loss": 0.5678, + "step": 231 + }, + { + "epoch": 0.08748114630467571, + "grad_norm": 0.43790147402826163, + "learning_rate": 1.3996983408748116e-05, + "loss": 0.5533, + "step": 232 + }, + { + "epoch": 0.08785822021116138, + "grad_norm": 0.437950706243473, + "learning_rate": 1.4057315233785823e-05, + "loss": 0.5564, + "step": 233 + }, + { + "epoch": 0.08823529411764706, + "grad_norm": 0.39896315832761775, + "learning_rate": 1.4117647058823532e-05, + "loss": 0.5666, + "step": 234 + }, + { + "epoch": 0.08861236802413273, + "grad_norm": 0.4625945126912622, + "learning_rate": 1.4177978883861239e-05, + "loss": 0.5613, + "step": 235 + }, + { + "epoch": 0.0889894419306184, + "grad_norm": 0.6085351068924789, + "learning_rate": 1.4238310708898946e-05, + "loss": 0.5481, + "step": 236 + }, + { + "epoch": 0.08936651583710407, + "grad_norm": 0.7291761157834109, + "learning_rate": 1.4298642533936653e-05, + "loss": 0.552, + "step": 237 + }, + { + "epoch": 0.08974358974358974, + "grad_norm": 0.7858350926569604, + "learning_rate": 1.435897435897436e-05, + "loss": 0.5569, + "step": 238 + }, + { + "epoch": 0.09012066365007541, + "grad_norm": 0.7649270027387413, + "learning_rate": 1.4419306184012067e-05, + "loss": 0.5523, + "step": 239 + }, + { + "epoch": 0.09049773755656108, + "grad_norm": 0.6935280540415244, + "learning_rate": 1.4479638009049775e-05, + "loss": 0.5454, + "step": 240 + }, + { + "epoch": 0.09087481146304675, + "grad_norm": 0.6303577522947715, + "learning_rate": 1.4539969834087482e-05, + "loss": 0.5593, + "step": 241 + }, + { + "epoch": 0.09125188536953242, + "grad_norm": 0.4653068676291331, + "learning_rate": 1.4600301659125189e-05, + "loss": 0.5584, + "step": 242 + }, + { + "epoch": 0.0916289592760181, + "grad_norm": 0.4670798345873098, + "learning_rate": 1.4660633484162896e-05, + "loss": 0.549, + "step": 243 + }, + { + "epoch": 0.09200603318250378, + "grad_norm": 0.5889279265986069, + "learning_rate": 1.4720965309200605e-05, + "loss": 0.5437, + "step": 244 + }, + { + "epoch": 0.09238310708898945, + "grad_norm": 0.614735615293968, + "learning_rate": 1.4781297134238312e-05, + "loss": 0.5444, + "step": 245 + }, + { + "epoch": 0.09276018099547512, + "grad_norm": 0.5949576121013005, + "learning_rate": 1.484162895927602e-05, + "loss": 0.5616, + "step": 246 + }, + { + "epoch": 0.09313725490196079, + "grad_norm": 0.4691174695420404, + "learning_rate": 1.4901960784313726e-05, + "loss": 0.5571, + "step": 247 + }, + { + "epoch": 0.09351432880844646, + "grad_norm": 0.4187342659466297, + "learning_rate": 1.4962292609351434e-05, + "loss": 0.5512, + "step": 248 + }, + { + "epoch": 0.09389140271493213, + "grad_norm": 0.6608525142798336, + "learning_rate": 1.502262443438914e-05, + "loss": 0.551, + "step": 249 + }, + { + "epoch": 0.0942684766214178, + "grad_norm": 0.9851305321458627, + "learning_rate": 1.5082956259426848e-05, + "loss": 0.5544, + "step": 250 + }, + { + "epoch": 0.09464555052790347, + "grad_norm": 0.9303752045361074, + "learning_rate": 1.5143288084464555e-05, + "loss": 0.5523, + "step": 251 + }, + { + "epoch": 0.09502262443438914, + "grad_norm": 0.5549393253061631, + "learning_rate": 1.5203619909502262e-05, + "loss": 0.5353, + "step": 252 + }, + { + "epoch": 0.0953996983408748, + "grad_norm": 0.3855055034197569, + "learning_rate": 1.526395173453997e-05, + "loss": 0.5525, + "step": 253 + }, + { + "epoch": 0.09577677224736049, + "grad_norm": 0.7251010119852346, + "learning_rate": 1.532428355957768e-05, + "loss": 0.5663, + "step": 254 + }, + { + "epoch": 0.09615384615384616, + "grad_norm": 0.8579192248635441, + "learning_rate": 1.5384615384615387e-05, + "loss": 0.5564, + "step": 255 + }, + { + "epoch": 0.09653092006033183, + "grad_norm": 0.7816825007185926, + "learning_rate": 1.5444947209653093e-05, + "loss": 0.5496, + "step": 256 + }, + { + "epoch": 0.0969079939668175, + "grad_norm": 0.49758181728998424, + "learning_rate": 1.55052790346908e-05, + "loss": 0.5569, + "step": 257 + }, + { + "epoch": 0.09728506787330317, + "grad_norm": 0.36159807078106626, + "learning_rate": 1.5565610859728507e-05, + "loss": 0.5488, + "step": 258 + }, + { + "epoch": 0.09766214177978884, + "grad_norm": 0.5628460796558962, + "learning_rate": 1.5625942684766216e-05, + "loss": 0.5532, + "step": 259 + }, + { + "epoch": 0.09803921568627451, + "grad_norm": 0.7130877140420384, + "learning_rate": 1.568627450980392e-05, + "loss": 0.5362, + "step": 260 + }, + { + "epoch": 0.09841628959276018, + "grad_norm": 0.6887200228505592, + "learning_rate": 1.574660633484163e-05, + "loss": 0.5489, + "step": 261 + }, + { + "epoch": 0.09879336349924585, + "grad_norm": 0.5636231666069549, + "learning_rate": 1.5806938159879336e-05, + "loss": 0.5483, + "step": 262 + }, + { + "epoch": 0.09917043740573152, + "grad_norm": 0.43624247835374313, + "learning_rate": 1.5867269984917045e-05, + "loss": 0.5487, + "step": 263 + }, + { + "epoch": 0.09954751131221719, + "grad_norm": 0.5555269305117205, + "learning_rate": 1.592760180995475e-05, + "loss": 0.5553, + "step": 264 + }, + { + "epoch": 0.09992458521870287, + "grad_norm": 0.6417512003556118, + "learning_rate": 1.5987933634992462e-05, + "loss": 0.5521, + "step": 265 + }, + { + "epoch": 0.10030165912518854, + "grad_norm": 0.5543065296275551, + "learning_rate": 1.6048265460030168e-05, + "loss": 0.5527, + "step": 266 + }, + { + "epoch": 0.10067873303167421, + "grad_norm": 0.49133207178470845, + "learning_rate": 1.6108597285067877e-05, + "loss": 0.559, + "step": 267 + }, + { + "epoch": 0.10105580693815988, + "grad_norm": 0.572089203314267, + "learning_rate": 1.6168929110105582e-05, + "loss": 0.5552, + "step": 268 + }, + { + "epoch": 0.10143288084464555, + "grad_norm": 0.6081048846528053, + "learning_rate": 1.622926093514329e-05, + "loss": 0.5386, + "step": 269 + }, + { + "epoch": 0.10180995475113122, + "grad_norm": 0.6224972556821279, + "learning_rate": 1.6289592760180996e-05, + "loss": 0.5515, + "step": 270 + }, + { + "epoch": 0.10218702865761689, + "grad_norm": 0.5967854547873926, + "learning_rate": 1.6349924585218705e-05, + "loss": 0.5214, + "step": 271 + }, + { + "epoch": 0.10256410256410256, + "grad_norm": 0.5715876345441009, + "learning_rate": 1.641025641025641e-05, + "loss": 0.5501, + "step": 272 + }, + { + "epoch": 0.10294117647058823, + "grad_norm": 0.5160783222942807, + "learning_rate": 1.647058823529412e-05, + "loss": 0.5491, + "step": 273 + }, + { + "epoch": 0.1033182503770739, + "grad_norm": 0.4681397098606609, + "learning_rate": 1.6530920060331825e-05, + "loss": 0.5381, + "step": 274 + }, + { + "epoch": 0.10369532428355958, + "grad_norm": 0.44008206972025743, + "learning_rate": 1.6591251885369534e-05, + "loss": 0.5483, + "step": 275 + }, + { + "epoch": 0.10407239819004525, + "grad_norm": 0.4843099784368105, + "learning_rate": 1.6651583710407243e-05, + "loss": 0.5407, + "step": 276 + }, + { + "epoch": 0.10444947209653092, + "grad_norm": 0.5914401660473909, + "learning_rate": 1.6711915535444948e-05, + "loss": 0.5348, + "step": 277 + }, + { + "epoch": 0.1048265460030166, + "grad_norm": 0.4839590804114454, + "learning_rate": 1.6772247360482657e-05, + "loss": 0.5572, + "step": 278 + }, + { + "epoch": 0.10520361990950226, + "grad_norm": 0.3640377253900058, + "learning_rate": 1.6832579185520363e-05, + "loss": 0.5632, + "step": 279 + }, + { + "epoch": 0.10558069381598793, + "grad_norm": 0.46124411365403584, + "learning_rate": 1.689291101055807e-05, + "loss": 0.5677, + "step": 280 + }, + { + "epoch": 0.1059577677224736, + "grad_norm": 0.6801701301654982, + "learning_rate": 1.6953242835595777e-05, + "loss": 0.5504, + "step": 281 + }, + { + "epoch": 0.10633484162895927, + "grad_norm": 0.7217643457628259, + "learning_rate": 1.7013574660633486e-05, + "loss": 0.5514, + "step": 282 + }, + { + "epoch": 0.10671191553544494, + "grad_norm": 0.746951481921502, + "learning_rate": 1.707390648567119e-05, + "loss": 0.5541, + "step": 283 + }, + { + "epoch": 0.10708898944193061, + "grad_norm": 0.9107864376019389, + "learning_rate": 1.71342383107089e-05, + "loss": 0.5411, + "step": 284 + }, + { + "epoch": 0.1074660633484163, + "grad_norm": 1.1461534499288535, + "learning_rate": 1.719457013574661e-05, + "loss": 0.5524, + "step": 285 + }, + { + "epoch": 0.10784313725490197, + "grad_norm": 0.8051347150112012, + "learning_rate": 1.7254901960784314e-05, + "loss": 0.5494, + "step": 286 + }, + { + "epoch": 0.10822021116138764, + "grad_norm": 0.4276811886641824, + "learning_rate": 1.7315233785822023e-05, + "loss": 0.5389, + "step": 287 + }, + { + "epoch": 0.1085972850678733, + "grad_norm": 0.7956211927040981, + "learning_rate": 1.737556561085973e-05, + "loss": 0.548, + "step": 288 + }, + { + "epoch": 0.10897435897435898, + "grad_norm": 1.0353999285370188, + "learning_rate": 1.7435897435897438e-05, + "loss": 0.5456, + "step": 289 + }, + { + "epoch": 0.10935143288084465, + "grad_norm": 0.5264903802818677, + "learning_rate": 1.7496229260935143e-05, + "loss": 0.5533, + "step": 290 + }, + { + "epoch": 0.10972850678733032, + "grad_norm": 0.6036395532902895, + "learning_rate": 1.7556561085972852e-05, + "loss": 0.5397, + "step": 291 + }, + { + "epoch": 0.11010558069381599, + "grad_norm": 0.9720168636027958, + "learning_rate": 1.7616892911010557e-05, + "loss": 0.5444, + "step": 292 + }, + { + "epoch": 0.11048265460030166, + "grad_norm": 0.7816921302833592, + "learning_rate": 1.7677224736048266e-05, + "loss": 0.5343, + "step": 293 + }, + { + "epoch": 0.11085972850678733, + "grad_norm": 0.5709679352512201, + "learning_rate": 1.7737556561085972e-05, + "loss": 0.5576, + "step": 294 + }, + { + "epoch": 0.111236802413273, + "grad_norm": 0.5899055506976723, + "learning_rate": 1.779788838612368e-05, + "loss": 0.5389, + "step": 295 + }, + { + "epoch": 0.11161387631975868, + "grad_norm": 0.634246050597436, + "learning_rate": 1.785822021116139e-05, + "loss": 0.5541, + "step": 296 + }, + { + "epoch": 0.11199095022624435, + "grad_norm": 0.5797012285150372, + "learning_rate": 1.79185520361991e-05, + "loss": 0.5456, + "step": 297 + }, + { + "epoch": 0.11236802413273002, + "grad_norm": 0.6133048348510126, + "learning_rate": 1.7978883861236804e-05, + "loss": 0.5431, + "step": 298 + }, + { + "epoch": 0.11274509803921569, + "grad_norm": 0.6415877919768428, + "learning_rate": 1.8039215686274513e-05, + "loss": 0.5382, + "step": 299 + }, + { + "epoch": 0.11312217194570136, + "grad_norm": 0.6496393180815279, + "learning_rate": 1.8099547511312218e-05, + "loss": 0.5262, + "step": 300 + }, + { + "epoch": 0.11349924585218703, + "grad_norm": 0.6204977758220354, + "learning_rate": 1.8159879336349927e-05, + "loss": 0.5532, + "step": 301 + }, + { + "epoch": 0.1138763197586727, + "grad_norm": 0.6819889461069473, + "learning_rate": 1.8220211161387633e-05, + "loss": 0.5352, + "step": 302 + }, + { + "epoch": 0.11425339366515837, + "grad_norm": 0.5962848700754978, + "learning_rate": 1.828054298642534e-05, + "loss": 0.5355, + "step": 303 + }, + { + "epoch": 0.11463046757164404, + "grad_norm": 0.46307910794790963, + "learning_rate": 1.8340874811463047e-05, + "loss": 0.5342, + "step": 304 + }, + { + "epoch": 0.11500754147812971, + "grad_norm": 0.5613349797030054, + "learning_rate": 1.8401206636500756e-05, + "loss": 0.5685, + "step": 305 + }, + { + "epoch": 0.11538461538461539, + "grad_norm": 0.3583330585881259, + "learning_rate": 1.8461538461538465e-05, + "loss": 0.5397, + "step": 306 + }, + { + "epoch": 0.11576168929110106, + "grad_norm": 0.45704373188681013, + "learning_rate": 1.852187028657617e-05, + "loss": 0.5398, + "step": 307 + }, + { + "epoch": 0.11613876319758673, + "grad_norm": 0.5195709533051902, + "learning_rate": 1.858220211161388e-05, + "loss": 0.5456, + "step": 308 + }, + { + "epoch": 0.1165158371040724, + "grad_norm": 0.5315383570448426, + "learning_rate": 1.8642533936651584e-05, + "loss": 0.5472, + "step": 309 + }, + { + "epoch": 0.11689291101055807, + "grad_norm": 0.5564299992698981, + "learning_rate": 1.8702865761689293e-05, + "loss": 0.5316, + "step": 310 + }, + { + "epoch": 0.11726998491704374, + "grad_norm": 0.6787266308342124, + "learning_rate": 1.8763197586727e-05, + "loss": 0.5363, + "step": 311 + }, + { + "epoch": 0.11764705882352941, + "grad_norm": 0.8456520950445957, + "learning_rate": 1.8823529411764708e-05, + "loss": 0.5445, + "step": 312 + }, + { + "epoch": 0.11802413273001508, + "grad_norm": 1.0117321256368699, + "learning_rate": 1.8883861236802413e-05, + "loss": 0.545, + "step": 313 + }, + { + "epoch": 0.11840120663650075, + "grad_norm": 1.005679604371225, + "learning_rate": 1.8944193061840122e-05, + "loss": 0.5324, + "step": 314 + }, + { + "epoch": 0.11877828054298642, + "grad_norm": 0.9038515221562614, + "learning_rate": 1.9004524886877827e-05, + "loss": 0.5286, + "step": 315 + }, + { + "epoch": 0.1191553544494721, + "grad_norm": 0.7649699299562862, + "learning_rate": 1.906485671191554e-05, + "loss": 0.5402, + "step": 316 + }, + { + "epoch": 0.11953242835595777, + "grad_norm": 0.8050537577708033, + "learning_rate": 1.9125188536953245e-05, + "loss": 0.5424, + "step": 317 + }, + { + "epoch": 0.11990950226244344, + "grad_norm": 0.8886728564640513, + "learning_rate": 1.9185520361990954e-05, + "loss": 0.5482, + "step": 318 + }, + { + "epoch": 0.12028657616892911, + "grad_norm": 0.7698429879460099, + "learning_rate": 1.924585218702866e-05, + "loss": 0.5376, + "step": 319 + }, + { + "epoch": 0.12066365007541478, + "grad_norm": 0.7170751082500985, + "learning_rate": 1.930618401206637e-05, + "loss": 0.5545, + "step": 320 + }, + { + "epoch": 0.12104072398190045, + "grad_norm": 0.750095251989184, + "learning_rate": 1.9366515837104074e-05, + "loss": 0.5255, + "step": 321 + }, + { + "epoch": 0.12141779788838612, + "grad_norm": 0.7577448972569838, + "learning_rate": 1.9426847662141783e-05, + "loss": 0.554, + "step": 322 + }, + { + "epoch": 0.12179487179487179, + "grad_norm": 0.7619086943968296, + "learning_rate": 1.9487179487179488e-05, + "loss": 0.5373, + "step": 323 + }, + { + "epoch": 0.12217194570135746, + "grad_norm": 0.6132401628722957, + "learning_rate": 1.9547511312217197e-05, + "loss": 0.5446, + "step": 324 + }, + { + "epoch": 0.12254901960784313, + "grad_norm": 0.4426163380144261, + "learning_rate": 1.9607843137254903e-05, + "loss": 0.5548, + "step": 325 + }, + { + "epoch": 0.1229260935143288, + "grad_norm": 0.6320215997827915, + "learning_rate": 1.966817496229261e-05, + "loss": 0.546, + "step": 326 + }, + { + "epoch": 0.12330316742081449, + "grad_norm": 0.7470755171203476, + "learning_rate": 1.972850678733032e-05, + "loss": 0.5358, + "step": 327 + }, + { + "epoch": 0.12368024132730016, + "grad_norm": 0.5285389115422098, + "learning_rate": 1.9788838612368026e-05, + "loss": 0.5279, + "step": 328 + }, + { + "epoch": 0.12405731523378583, + "grad_norm": 0.41787202630341314, + "learning_rate": 1.9849170437405735e-05, + "loss": 0.5338, + "step": 329 + }, + { + "epoch": 0.1244343891402715, + "grad_norm": 0.6769381806287961, + "learning_rate": 1.990950226244344e-05, + "loss": 0.5506, + "step": 330 + }, + { + "epoch": 0.12481146304675717, + "grad_norm": 0.7180202320433325, + "learning_rate": 1.996983408748115e-05, + "loss": 0.534, + "step": 331 + }, + { + "epoch": 0.12518853695324283, + "grad_norm": 0.44870935215470026, + "learning_rate": 2.0030165912518854e-05, + "loss": 0.5419, + "step": 332 + }, + { + "epoch": 0.1255656108597285, + "grad_norm": 0.5018705951626008, + "learning_rate": 2.0090497737556563e-05, + "loss": 0.5417, + "step": 333 + }, + { + "epoch": 0.12594268476621417, + "grad_norm": 0.6723688607282032, + "learning_rate": 2.015082956259427e-05, + "loss": 0.54, + "step": 334 + }, + { + "epoch": 0.12631975867269984, + "grad_norm": 0.802623811796234, + "learning_rate": 2.0211161387631978e-05, + "loss": 0.5257, + "step": 335 + }, + { + "epoch": 0.12669683257918551, + "grad_norm": 0.8662045504957842, + "learning_rate": 2.0271493212669683e-05, + "loss": 0.5476, + "step": 336 + }, + { + "epoch": 0.12707390648567118, + "grad_norm": 0.8208306688184609, + "learning_rate": 2.0331825037707392e-05, + "loss": 0.5394, + "step": 337 + }, + { + "epoch": 0.12745098039215685, + "grad_norm": 0.8709835621999373, + "learning_rate": 2.0392156862745097e-05, + "loss": 0.5332, + "step": 338 + }, + { + "epoch": 0.12782805429864252, + "grad_norm": 0.9470431689366736, + "learning_rate": 2.0452488687782806e-05, + "loss": 0.547, + "step": 339 + }, + { + "epoch": 0.1282051282051282, + "grad_norm": 0.807800935607607, + "learning_rate": 2.0512820512820512e-05, + "loss": 0.5371, + "step": 340 + }, + { + "epoch": 0.12858220211161386, + "grad_norm": 0.4436641314235256, + "learning_rate": 2.057315233785822e-05, + "loss": 0.5267, + "step": 341 + }, + { + "epoch": 0.12895927601809956, + "grad_norm": 0.6251160796351121, + "learning_rate": 2.0633484162895933e-05, + "loss": 0.5347, + "step": 342 + }, + { + "epoch": 0.12933634992458523, + "grad_norm": 1.0106939105660857, + "learning_rate": 2.0693815987933638e-05, + "loss": 0.5364, + "step": 343 + }, + { + "epoch": 0.1297134238310709, + "grad_norm": 0.8708869547695157, + "learning_rate": 2.0754147812971347e-05, + "loss": 0.5399, + "step": 344 + }, + { + "epoch": 0.13009049773755657, + "grad_norm": 0.47402081418162684, + "learning_rate": 2.0814479638009053e-05, + "loss": 0.5265, + "step": 345 + }, + { + "epoch": 0.13046757164404224, + "grad_norm": 0.7034385833149872, + "learning_rate": 2.087481146304676e-05, + "loss": 0.5382, + "step": 346 + }, + { + "epoch": 0.1308446455505279, + "grad_norm": 0.775714433402047, + "learning_rate": 2.0935143288084467e-05, + "loss": 0.5428, + "step": 347 + }, + { + "epoch": 0.13122171945701358, + "grad_norm": 0.5431064349801128, + "learning_rate": 2.0995475113122176e-05, + "loss": 0.5457, + "step": 348 + }, + { + "epoch": 0.13159879336349925, + "grad_norm": 0.36033538595332765, + "learning_rate": 2.105580693815988e-05, + "loss": 0.5275, + "step": 349 + }, + { + "epoch": 0.13197586726998492, + "grad_norm": 0.543202786558742, + "learning_rate": 2.111613876319759e-05, + "loss": 0.5095, + "step": 350 + }, + { + "epoch": 0.1323529411764706, + "grad_norm": 0.6057365442454049, + "learning_rate": 2.1176470588235296e-05, + "loss": 0.5345, + "step": 351 + }, + { + "epoch": 0.13273001508295626, + "grad_norm": 0.6374467611937762, + "learning_rate": 2.1236802413273005e-05, + "loss": 0.5274, + "step": 352 + }, + { + "epoch": 0.13310708898944193, + "grad_norm": 0.7825493100655928, + "learning_rate": 2.129713423831071e-05, + "loss": 0.528, + "step": 353 + }, + { + "epoch": 0.1334841628959276, + "grad_norm": 1.02740170128605, + "learning_rate": 2.135746606334842e-05, + "loss": 0.5295, + "step": 354 + }, + { + "epoch": 0.13386123680241327, + "grad_norm": 1.0525674712574902, + "learning_rate": 2.1417797888386124e-05, + "loss": 0.5482, + "step": 355 + }, + { + "epoch": 0.13423831070889894, + "grad_norm": 0.836084570628571, + "learning_rate": 2.1478129713423833e-05, + "loss": 0.5428, + "step": 356 + }, + { + "epoch": 0.1346153846153846, + "grad_norm": 0.7070455303122072, + "learning_rate": 2.153846153846154e-05, + "loss": 0.5399, + "step": 357 + }, + { + "epoch": 0.13499245852187028, + "grad_norm": 0.5307654442508366, + "learning_rate": 2.1598793363499248e-05, + "loss": 0.5301, + "step": 358 + }, + { + "epoch": 0.13536953242835595, + "grad_norm": 0.5029676523750929, + "learning_rate": 2.1659125188536953e-05, + "loss": 0.5175, + "step": 359 + }, + { + "epoch": 0.13574660633484162, + "grad_norm": 0.5978859492989534, + "learning_rate": 2.1719457013574662e-05, + "loss": 0.5327, + "step": 360 + }, + { + "epoch": 0.1361236802413273, + "grad_norm": 0.670046416739267, + "learning_rate": 2.1779788838612367e-05, + "loss": 0.5342, + "step": 361 + }, + { + "epoch": 0.13650075414781296, + "grad_norm": 0.6198753215586429, + "learning_rate": 2.1840120663650076e-05, + "loss": 0.5401, + "step": 362 + }, + { + "epoch": 0.13687782805429866, + "grad_norm": 0.5752377165599012, + "learning_rate": 2.1900452488687785e-05, + "loss": 0.5344, + "step": 363 + }, + { + "epoch": 0.13725490196078433, + "grad_norm": 0.8770182094077874, + "learning_rate": 2.1960784313725494e-05, + "loss": 0.5312, + "step": 364 + }, + { + "epoch": 0.13763197586727, + "grad_norm": 1.3004708487533077, + "learning_rate": 2.2021116138763203e-05, + "loss": 0.5358, + "step": 365 + }, + { + "epoch": 0.13800904977375567, + "grad_norm": 0.5388232251304353, + "learning_rate": 2.2081447963800908e-05, + "loss": 0.54, + "step": 366 + }, + { + "epoch": 0.13838612368024134, + "grad_norm": 0.9249001948901423, + "learning_rate": 2.2141779788838617e-05, + "loss": 0.5283, + "step": 367 + }, + { + "epoch": 0.138763197586727, + "grad_norm": 1.239226312817128, + "learning_rate": 2.2202111613876323e-05, + "loss": 0.5232, + "step": 368 + }, + { + "epoch": 0.13914027149321267, + "grad_norm": 0.39558715763925306, + "learning_rate": 2.226244343891403e-05, + "loss": 0.5194, + "step": 369 + }, + { + "epoch": 0.13951734539969834, + "grad_norm": 1.0422789691871563, + "learning_rate": 2.2322775263951737e-05, + "loss": 0.5166, + "step": 370 + }, + { + "epoch": 0.13989441930618401, + "grad_norm": 1.3253864003016393, + "learning_rate": 2.2383107088989446e-05, + "loss": 0.5419, + "step": 371 + }, + { + "epoch": 0.14027149321266968, + "grad_norm": 0.4883191454885404, + "learning_rate": 2.244343891402715e-05, + "loss": 0.5193, + "step": 372 + }, + { + "epoch": 0.14064856711915535, + "grad_norm": 1.3426307746707304, + "learning_rate": 2.250377073906486e-05, + "loss": 0.5384, + "step": 373 + }, + { + "epoch": 0.14102564102564102, + "grad_norm": 0.7162125283949953, + "learning_rate": 2.2564102564102566e-05, + "loss": 0.5347, + "step": 374 + }, + { + "epoch": 0.1414027149321267, + "grad_norm": 0.9517515760166454, + "learning_rate": 2.2624434389140274e-05, + "loss": 0.5409, + "step": 375 + }, + { + "epoch": 0.14177978883861236, + "grad_norm": 0.9901668415592955, + "learning_rate": 2.268476621417798e-05, + "loss": 0.5393, + "step": 376 + }, + { + "epoch": 0.14215686274509803, + "grad_norm": 0.550514344675212, + "learning_rate": 2.274509803921569e-05, + "loss": 0.5279, + "step": 377 + }, + { + "epoch": 0.1425339366515837, + "grad_norm": 0.6496745197769902, + "learning_rate": 2.2805429864253394e-05, + "loss": 0.5435, + "step": 378 + }, + { + "epoch": 0.14291101055806937, + "grad_norm": 0.5680136583383624, + "learning_rate": 2.2865761689291103e-05, + "loss": 0.5324, + "step": 379 + }, + { + "epoch": 0.14328808446455504, + "grad_norm": 0.5709940060282162, + "learning_rate": 2.292609351432881e-05, + "loss": 0.5403, + "step": 380 + }, + { + "epoch": 0.1436651583710407, + "grad_norm": 0.7207116679334922, + "learning_rate": 2.2986425339366517e-05, + "loss": 0.526, + "step": 381 + }, + { + "epoch": 0.14404223227752638, + "grad_norm": 0.5692791268646312, + "learning_rate": 2.3046757164404223e-05, + "loss": 0.5339, + "step": 382 + }, + { + "epoch": 0.14441930618401205, + "grad_norm": 0.5407966892798871, + "learning_rate": 2.3107088989441932e-05, + "loss": 0.5319, + "step": 383 + }, + { + "epoch": 0.14479638009049775, + "grad_norm": 0.6866690327914815, + "learning_rate": 2.316742081447964e-05, + "loss": 0.514, + "step": 384 + }, + { + "epoch": 0.14517345399698342, + "grad_norm": 0.5898879076291751, + "learning_rate": 2.322775263951735e-05, + "loss": 0.5187, + "step": 385 + }, + { + "epoch": 0.1455505279034691, + "grad_norm": 0.5470005305427551, + "learning_rate": 2.3288084464555055e-05, + "loss": 0.5279, + "step": 386 + }, + { + "epoch": 0.14592760180995476, + "grad_norm": 0.7806615408602505, + "learning_rate": 2.3348416289592764e-05, + "loss": 0.5505, + "step": 387 + }, + { + "epoch": 0.14630467571644043, + "grad_norm": 0.823946674488269, + "learning_rate": 2.340874811463047e-05, + "loss": 0.5338, + "step": 388 + }, + { + "epoch": 0.1466817496229261, + "grad_norm": 0.84443057867474, + "learning_rate": 2.3469079939668178e-05, + "loss": 0.5382, + "step": 389 + }, + { + "epoch": 0.14705882352941177, + "grad_norm": 1.0683695773036288, + "learning_rate": 2.3529411764705884e-05, + "loss": 0.5341, + "step": 390 + }, + { + "epoch": 0.14743589743589744, + "grad_norm": 0.9098533085143852, + "learning_rate": 2.3589743589743593e-05, + "loss": 0.5337, + "step": 391 + }, + { + "epoch": 0.1478129713423831, + "grad_norm": 0.6673862275395585, + "learning_rate": 2.3650075414781298e-05, + "loss": 0.5226, + "step": 392 + }, + { + "epoch": 0.14819004524886878, + "grad_norm": 0.7684260815656159, + "learning_rate": 2.3710407239819007e-05, + "loss": 0.5262, + "step": 393 + }, + { + "epoch": 0.14856711915535445, + "grad_norm": 0.8656114539656322, + "learning_rate": 2.3770739064856712e-05, + "loss": 0.5381, + "step": 394 + }, + { + "epoch": 0.14894419306184012, + "grad_norm": 0.8649804756772275, + "learning_rate": 2.383107088989442e-05, + "loss": 0.5226, + "step": 395 + }, + { + "epoch": 0.1493212669683258, + "grad_norm": 0.7973108597770733, + "learning_rate": 2.3891402714932127e-05, + "loss": 0.5241, + "step": 396 + }, + { + "epoch": 0.14969834087481146, + "grad_norm": 0.8846530557706691, + "learning_rate": 2.3951734539969836e-05, + "loss": 0.5262, + "step": 397 + }, + { + "epoch": 0.15007541478129713, + "grad_norm": 1.0192529895115987, + "learning_rate": 2.401206636500754e-05, + "loss": 0.5456, + "step": 398 + }, + { + "epoch": 0.1504524886877828, + "grad_norm": 0.8867947504616269, + "learning_rate": 2.407239819004525e-05, + "loss": 0.5433, + "step": 399 + }, + { + "epoch": 0.15082956259426847, + "grad_norm": 1.0602009214207582, + "learning_rate": 2.4132730015082955e-05, + "loss": 0.525, + "step": 400 + }, + { + "epoch": 0.15120663650075414, + "grad_norm": 1.0129861741239785, + "learning_rate": 2.4193061840120664e-05, + "loss": 0.5449, + "step": 401 + }, + { + "epoch": 0.1515837104072398, + "grad_norm": 0.5647524596389712, + "learning_rate": 2.425339366515837e-05, + "loss": 0.5353, + "step": 402 + }, + { + "epoch": 0.15196078431372548, + "grad_norm": 0.6892167763940965, + "learning_rate": 2.431372549019608e-05, + "loss": 0.5295, + "step": 403 + }, + { + "epoch": 0.15233785822021115, + "grad_norm": 1.048419618304232, + "learning_rate": 2.4374057315233784e-05, + "loss": 0.5286, + "step": 404 + }, + { + "epoch": 0.15271493212669685, + "grad_norm": 0.677660589372063, + "learning_rate": 2.4434389140271496e-05, + "loss": 0.5286, + "step": 405 + }, + { + "epoch": 0.15309200603318251, + "grad_norm": 0.917817570872978, + "learning_rate": 2.4494720965309205e-05, + "loss": 0.536, + "step": 406 + }, + { + "epoch": 0.15346907993966818, + "grad_norm": 1.1726584715037884, + "learning_rate": 2.455505279034691e-05, + "loss": 0.5392, + "step": 407 + }, + { + "epoch": 0.15384615384615385, + "grad_norm": 0.7114599212145515, + "learning_rate": 2.461538461538462e-05, + "loss": 0.5264, + "step": 408 + }, + { + "epoch": 0.15422322775263952, + "grad_norm": 1.0747882913666442, + "learning_rate": 2.4675716440422325e-05, + "loss": 0.5428, + "step": 409 + }, + { + "epoch": 0.1546003016591252, + "grad_norm": 0.8422373022788147, + "learning_rate": 2.4736048265460034e-05, + "loss": 0.5349, + "step": 410 + }, + { + "epoch": 0.15497737556561086, + "grad_norm": 0.9253711275625769, + "learning_rate": 2.479638009049774e-05, + "loss": 0.5198, + "step": 411 + }, + { + "epoch": 0.15535444947209653, + "grad_norm": 1.1584499737533465, + "learning_rate": 2.4856711915535448e-05, + "loss": 0.5397, + "step": 412 + }, + { + "epoch": 0.1557315233785822, + "grad_norm": 0.694465207780418, + "learning_rate": 2.4917043740573154e-05, + "loss": 0.5187, + "step": 413 + }, + { + "epoch": 0.15610859728506787, + "grad_norm": 0.797073087487603, + "learning_rate": 2.4977375565610862e-05, + "loss": 0.5319, + "step": 414 + }, + { + "epoch": 0.15648567119155354, + "grad_norm": 0.8993142361085866, + "learning_rate": 2.5037707390648568e-05, + "loss": 0.5347, + "step": 415 + }, + { + "epoch": 0.1568627450980392, + "grad_norm": 1.0786922837232342, + "learning_rate": 2.5098039215686277e-05, + "loss": 0.537, + "step": 416 + }, + { + "epoch": 0.15723981900452488, + "grad_norm": 0.919507926795887, + "learning_rate": 2.5158371040723982e-05, + "loss": 0.5339, + "step": 417 + }, + { + "epoch": 0.15761689291101055, + "grad_norm": 0.9271914587969777, + "learning_rate": 2.521870286576169e-05, + "loss": 0.5427, + "step": 418 + }, + { + "epoch": 0.15799396681749622, + "grad_norm": 0.8341900584617097, + "learning_rate": 2.5279034690799397e-05, + "loss": 0.5271, + "step": 419 + }, + { + "epoch": 0.1583710407239819, + "grad_norm": 0.8637393334378645, + "learning_rate": 2.5339366515837106e-05, + "loss": 0.5234, + "step": 420 + }, + { + "epoch": 0.15874811463046756, + "grad_norm": 0.933936865525467, + "learning_rate": 2.539969834087481e-05, + "loss": 0.5152, + "step": 421 + }, + { + "epoch": 0.15912518853695323, + "grad_norm": 0.7224495262038839, + "learning_rate": 2.546003016591252e-05, + "loss": 0.5286, + "step": 422 + }, + { + "epoch": 0.1595022624434389, + "grad_norm": 0.6194749610000853, + "learning_rate": 2.5520361990950225e-05, + "loss": 0.531, + "step": 423 + }, + { + "epoch": 0.15987933634992457, + "grad_norm": 0.5466277766554448, + "learning_rate": 2.5580693815987934e-05, + "loss": 0.5264, + "step": 424 + }, + { + "epoch": 0.16025641025641027, + "grad_norm": 0.7026828044457961, + "learning_rate": 2.5641025641025646e-05, + "loss": 0.5278, + "step": 425 + }, + { + "epoch": 0.16063348416289594, + "grad_norm": 0.9226299938073272, + "learning_rate": 2.5701357466063352e-05, + "loss": 0.5124, + "step": 426 + }, + { + "epoch": 0.1610105580693816, + "grad_norm": 0.8904264167502456, + "learning_rate": 2.576168929110106e-05, + "loss": 0.5254, + "step": 427 + }, + { + "epoch": 0.16138763197586728, + "grad_norm": 0.9507441353352831, + "learning_rate": 2.5822021116138766e-05, + "loss": 0.5347, + "step": 428 + }, + { + "epoch": 0.16176470588235295, + "grad_norm": 0.780171060794259, + "learning_rate": 2.5882352941176475e-05, + "loss": 0.5321, + "step": 429 + }, + { + "epoch": 0.16214177978883862, + "grad_norm": 0.6054494873361168, + "learning_rate": 2.594268476621418e-05, + "loss": 0.5265, + "step": 430 + }, + { + "epoch": 0.1625188536953243, + "grad_norm": 0.59250337934879, + "learning_rate": 2.600301659125189e-05, + "loss": 0.5264, + "step": 431 + }, + { + "epoch": 0.16289592760180996, + "grad_norm": 0.5922721341870341, + "learning_rate": 2.6063348416289595e-05, + "loss": 0.5178, + "step": 432 + }, + { + "epoch": 0.16327300150829563, + "grad_norm": 0.576515090903087, + "learning_rate": 2.6123680241327304e-05, + "loss": 0.5209, + "step": 433 + }, + { + "epoch": 0.1636500754147813, + "grad_norm": 0.6587882471840134, + "learning_rate": 2.618401206636501e-05, + "loss": 0.5281, + "step": 434 + }, + { + "epoch": 0.16402714932126697, + "grad_norm": 0.7852170827639879, + "learning_rate": 2.6244343891402718e-05, + "loss": 0.5341, + "step": 435 + }, + { + "epoch": 0.16440422322775264, + "grad_norm": 0.9659693801454377, + "learning_rate": 2.6304675716440424e-05, + "loss": 0.5404, + "step": 436 + }, + { + "epoch": 0.1647812971342383, + "grad_norm": 1.0887805920840894, + "learning_rate": 2.6365007541478132e-05, + "loss": 0.521, + "step": 437 + }, + { + "epoch": 0.16515837104072398, + "grad_norm": 0.8857050765238099, + "learning_rate": 2.6425339366515838e-05, + "loss": 0.5306, + "step": 438 + }, + { + "epoch": 0.16553544494720965, + "grad_norm": 0.5792178546784454, + "learning_rate": 2.6485671191553547e-05, + "loss": 0.5138, + "step": 439 + }, + { + "epoch": 0.16591251885369532, + "grad_norm": 0.5794842817424628, + "learning_rate": 2.6546003016591252e-05, + "loss": 0.5248, + "step": 440 + }, + { + "epoch": 0.166289592760181, + "grad_norm": 0.792254971903189, + "learning_rate": 2.660633484162896e-05, + "loss": 0.5342, + "step": 441 + }, + { + "epoch": 0.16666666666666666, + "grad_norm": 0.9833580503812501, + "learning_rate": 2.6666666666666667e-05, + "loss": 0.5378, + "step": 442 + }, + { + "epoch": 0.16704374057315233, + "grad_norm": 1.0045103227779268, + "learning_rate": 2.6726998491704375e-05, + "loss": 0.5318, + "step": 443 + }, + { + "epoch": 0.167420814479638, + "grad_norm": 0.8917681534393571, + "learning_rate": 2.678733031674208e-05, + "loss": 0.5288, + "step": 444 + }, + { + "epoch": 0.16779788838612367, + "grad_norm": 0.9229863911122994, + "learning_rate": 2.684766214177979e-05, + "loss": 0.526, + "step": 445 + }, + { + "epoch": 0.16817496229260936, + "grad_norm": 1.1546422549555346, + "learning_rate": 2.6907993966817502e-05, + "loss": 0.5268, + "step": 446 + }, + { + "epoch": 0.16855203619909503, + "grad_norm": 0.8718673335269225, + "learning_rate": 2.6968325791855208e-05, + "loss": 0.5327, + "step": 447 + }, + { + "epoch": 0.1689291101055807, + "grad_norm": 0.6468984327466095, + "learning_rate": 2.7028657616892916e-05, + "loss": 0.53, + "step": 448 + }, + { + "epoch": 0.16930618401206637, + "grad_norm": 0.4902283942870501, + "learning_rate": 2.7088989441930622e-05, + "loss": 0.507, + "step": 449 + }, + { + "epoch": 0.16968325791855204, + "grad_norm": 0.6754300752188921, + "learning_rate": 2.714932126696833e-05, + "loss": 0.5198, + "step": 450 + }, + { + "epoch": 0.1700603318250377, + "grad_norm": 0.8506983600674197, + "learning_rate": 2.7209653092006036e-05, + "loss": 0.5303, + "step": 451 + }, + { + "epoch": 0.17043740573152338, + "grad_norm": 0.9670528048421904, + "learning_rate": 2.7269984917043745e-05, + "loss": 0.5185, + "step": 452 + }, + { + "epoch": 0.17081447963800905, + "grad_norm": 1.0511946577537075, + "learning_rate": 2.733031674208145e-05, + "loss": 0.5239, + "step": 453 + }, + { + "epoch": 0.17119155354449472, + "grad_norm": 0.7621250170270317, + "learning_rate": 2.739064856711916e-05, + "loss": 0.5254, + "step": 454 + }, + { + "epoch": 0.1715686274509804, + "grad_norm": 0.6201361134998161, + "learning_rate": 2.7450980392156865e-05, + "loss": 0.5366, + "step": 455 + }, + { + "epoch": 0.17194570135746606, + "grad_norm": 0.6291227560746181, + "learning_rate": 2.7511312217194574e-05, + "loss": 0.5245, + "step": 456 + }, + { + "epoch": 0.17232277526395173, + "grad_norm": 0.6640097445720589, + "learning_rate": 2.757164404223228e-05, + "loss": 0.5105, + "step": 457 + }, + { + "epoch": 0.1726998491704374, + "grad_norm": 0.5603387992283033, + "learning_rate": 2.7631975867269988e-05, + "loss": 0.5156, + "step": 458 + }, + { + "epoch": 0.17307692307692307, + "grad_norm": 0.49588519938113995, + "learning_rate": 2.7692307692307694e-05, + "loss": 0.5061, + "step": 459 + }, + { + "epoch": 0.17345399698340874, + "grad_norm": 0.6023705572111043, + "learning_rate": 2.7752639517345402e-05, + "loss": 0.5038, + "step": 460 + }, + { + "epoch": 0.1738310708898944, + "grad_norm": 0.6472586995197627, + "learning_rate": 2.7812971342383108e-05, + "loss": 0.5215, + "step": 461 + }, + { + "epoch": 0.17420814479638008, + "grad_norm": 0.5606125755538321, + "learning_rate": 2.7873303167420817e-05, + "loss": 0.5168, + "step": 462 + }, + { + "epoch": 0.17458521870286575, + "grad_norm": 0.5472731127240856, + "learning_rate": 2.7933634992458522e-05, + "loss": 0.5316, + "step": 463 + }, + { + "epoch": 0.17496229260935142, + "grad_norm": 0.62198527373118, + "learning_rate": 2.799396681749623e-05, + "loss": 0.5172, + "step": 464 + }, + { + "epoch": 0.1753393665158371, + "grad_norm": 0.6802580358454702, + "learning_rate": 2.8054298642533937e-05, + "loss": 0.5209, + "step": 465 + }, + { + "epoch": 0.17571644042232276, + "grad_norm": 0.6939585069870416, + "learning_rate": 2.8114630467571645e-05, + "loss": 0.5329, + "step": 466 + }, + { + "epoch": 0.17609351432880846, + "grad_norm": 0.7471738158454391, + "learning_rate": 2.8174962292609354e-05, + "loss": 0.5303, + "step": 467 + }, + { + "epoch": 0.17647058823529413, + "grad_norm": 0.9653849255559597, + "learning_rate": 2.8235294117647063e-05, + "loss": 0.5128, + "step": 468 + }, + { + "epoch": 0.1768476621417798, + "grad_norm": 1.1362882422231515, + "learning_rate": 2.829562594268477e-05, + "loss": 0.5229, + "step": 469 + }, + { + "epoch": 0.17722473604826547, + "grad_norm": 0.7638829290554418, + "learning_rate": 2.8355957767722477e-05, + "loss": 0.52, + "step": 470 + }, + { + "epoch": 0.17760180995475114, + "grad_norm": 0.6686360871348043, + "learning_rate": 2.8416289592760183e-05, + "loss": 0.5262, + "step": 471 + }, + { + "epoch": 0.1779788838612368, + "grad_norm": 0.9193075132499012, + "learning_rate": 2.8476621417797892e-05, + "loss": 0.528, + "step": 472 + }, + { + "epoch": 0.17835595776772248, + "grad_norm": 1.0786977290107285, + "learning_rate": 2.8536953242835597e-05, + "loss": 0.5197, + "step": 473 + }, + { + "epoch": 0.17873303167420815, + "grad_norm": 0.7048132672675862, + "learning_rate": 2.8597285067873306e-05, + "loss": 0.5433, + "step": 474 + }, + { + "epoch": 0.17911010558069382, + "grad_norm": 0.6450547219296359, + "learning_rate": 2.865761689291101e-05, + "loss": 0.5233, + "step": 475 + }, + { + "epoch": 0.1794871794871795, + "grad_norm": 0.932634015837647, + "learning_rate": 2.871794871794872e-05, + "loss": 0.5367, + "step": 476 + }, + { + "epoch": 0.17986425339366516, + "grad_norm": 0.8604940283430506, + "learning_rate": 2.8778280542986426e-05, + "loss": 0.522, + "step": 477 + }, + { + "epoch": 0.18024132730015083, + "grad_norm": 0.5465211401521544, + "learning_rate": 2.8838612368024135e-05, + "loss": 0.5217, + "step": 478 + }, + { + "epoch": 0.1806184012066365, + "grad_norm": 0.8077145812615553, + "learning_rate": 2.889894419306184e-05, + "loss": 0.5228, + "step": 479 + }, + { + "epoch": 0.18099547511312217, + "grad_norm": 0.9615095444133367, + "learning_rate": 2.895927601809955e-05, + "loss": 0.5202, + "step": 480 + }, + { + "epoch": 0.18137254901960784, + "grad_norm": 0.5199327818369935, + "learning_rate": 2.9019607843137258e-05, + "loss": 0.5218, + "step": 481 + }, + { + "epoch": 0.1817496229260935, + "grad_norm": 0.6752719938064006, + "learning_rate": 2.9079939668174963e-05, + "loss": 0.5268, + "step": 482 + }, + { + "epoch": 0.18212669683257918, + "grad_norm": 0.7495052502471382, + "learning_rate": 2.9140271493212672e-05, + "loss": 0.5235, + "step": 483 + }, + { + "epoch": 0.18250377073906485, + "grad_norm": 0.48674736852326556, + "learning_rate": 2.9200603318250378e-05, + "loss": 0.5105, + "step": 484 + }, + { + "epoch": 0.18288084464555052, + "grad_norm": 0.6228661473054954, + "learning_rate": 2.9260935143288087e-05, + "loss": 0.5253, + "step": 485 + }, + { + "epoch": 0.1832579185520362, + "grad_norm": 0.6998002895773153, + "learning_rate": 2.9321266968325792e-05, + "loss": 0.5332, + "step": 486 + }, + { + "epoch": 0.18363499245852188, + "grad_norm": 0.6461298285815458, + "learning_rate": 2.9381598793363504e-05, + "loss": 0.5177, + "step": 487 + }, + { + "epoch": 0.18401206636500755, + "grad_norm": 0.80766550830779, + "learning_rate": 2.944193061840121e-05, + "loss": 0.5276, + "step": 488 + }, + { + "epoch": 0.18438914027149322, + "grad_norm": 0.6507576255041448, + "learning_rate": 2.950226244343892e-05, + "loss": 0.5121, + "step": 489 + }, + { + "epoch": 0.1847662141779789, + "grad_norm": 0.5691037725546034, + "learning_rate": 2.9562594268476624e-05, + "loss": 0.5215, + "step": 490 + }, + { + "epoch": 0.18514328808446456, + "grad_norm": 0.6840937944594588, + "learning_rate": 2.9622926093514333e-05, + "loss": 0.5297, + "step": 491 + }, + { + "epoch": 0.18552036199095023, + "grad_norm": 0.8166849436177234, + "learning_rate": 2.968325791855204e-05, + "loss": 0.5252, + "step": 492 + }, + { + "epoch": 0.1858974358974359, + "grad_norm": 1.1587405657738152, + "learning_rate": 2.9743589743589747e-05, + "loss": 0.5225, + "step": 493 + }, + { + "epoch": 0.18627450980392157, + "grad_norm": 0.8510108808660409, + "learning_rate": 2.9803921568627453e-05, + "loss": 0.5198, + "step": 494 + }, + { + "epoch": 0.18665158371040724, + "grad_norm": 0.446645743465664, + "learning_rate": 2.9864253393665162e-05, + "loss": 0.5156, + "step": 495 + }, + { + "epoch": 0.1870286576168929, + "grad_norm": 0.4707899523269208, + "learning_rate": 2.9924585218702867e-05, + "loss": 0.519, + "step": 496 + }, + { + "epoch": 0.18740573152337858, + "grad_norm": 0.6144830893724954, + "learning_rate": 2.9984917043740576e-05, + "loss": 0.5129, + "step": 497 + }, + { + "epoch": 0.18778280542986425, + "grad_norm": 0.7073375178508576, + "learning_rate": 3.004524886877828e-05, + "loss": 0.5246, + "step": 498 + }, + { + "epoch": 0.18815987933634992, + "grad_norm": 0.9790377265588477, + "learning_rate": 3.010558069381599e-05, + "loss": 0.5154, + "step": 499 + }, + { + "epoch": 0.1885369532428356, + "grad_norm": 1.28219651172438, + "learning_rate": 3.0165912518853696e-05, + "loss": 0.5253, + "step": 500 + }, + { + "epoch": 0.18891402714932126, + "grad_norm": 0.7571923260315351, + "learning_rate": 3.0226244343891405e-05, + "loss": 0.5136, + "step": 501 + }, + { + "epoch": 0.18929110105580693, + "grad_norm": 0.7350357520642378, + "learning_rate": 3.028657616892911e-05, + "loss": 0.5187, + "step": 502 + }, + { + "epoch": 0.1896681749622926, + "grad_norm": 0.9529699748440582, + "learning_rate": 3.034690799396682e-05, + "loss": 0.5209, + "step": 503 + }, + { + "epoch": 0.19004524886877827, + "grad_norm": 1.1503461113193032, + "learning_rate": 3.0407239819004525e-05, + "loss": 0.5369, + "step": 504 + }, + { + "epoch": 0.19042232277526394, + "grad_norm": 0.5805951751468477, + "learning_rate": 3.0467571644042233e-05, + "loss": 0.5112, + "step": 505 + }, + { + "epoch": 0.1907993966817496, + "grad_norm": 0.8177776972564045, + "learning_rate": 3.052790346907994e-05, + "loss": 0.5252, + "step": 506 + }, + { + "epoch": 0.19117647058823528, + "grad_norm": 1.1917306890055646, + "learning_rate": 3.0588235294117644e-05, + "loss": 0.5309, + "step": 507 + }, + { + "epoch": 0.19155354449472098, + "grad_norm": 0.5468019419744119, + "learning_rate": 3.064856711915536e-05, + "loss": 0.5176, + "step": 508 + }, + { + "epoch": 0.19193061840120665, + "grad_norm": 0.8629098734418542, + "learning_rate": 3.070889894419307e-05, + "loss": 0.5253, + "step": 509 + }, + { + "epoch": 0.19230769230769232, + "grad_norm": 1.0441138129479457, + "learning_rate": 3.0769230769230774e-05, + "loss": 0.518, + "step": 510 + }, + { + "epoch": 0.192684766214178, + "grad_norm": 0.6658280509981802, + "learning_rate": 3.082956259426848e-05, + "loss": 0.5091, + "step": 511 + }, + { + "epoch": 0.19306184012066366, + "grad_norm": 0.6015584486355624, + "learning_rate": 3.0889894419306185e-05, + "loss": 0.5165, + "step": 512 + }, + { + "epoch": 0.19343891402714933, + "grad_norm": 0.8336129047876748, + "learning_rate": 3.09502262443439e-05, + "loss": 0.5179, + "step": 513 + }, + { + "epoch": 0.193815987933635, + "grad_norm": 0.756617927829012, + "learning_rate": 3.10105580693816e-05, + "loss": 0.5248, + "step": 514 + }, + { + "epoch": 0.19419306184012067, + "grad_norm": 0.6135360270279898, + "learning_rate": 3.107088989441931e-05, + "loss": 0.5252, + "step": 515 + }, + { + "epoch": 0.19457013574660634, + "grad_norm": 0.6696847713421032, + "learning_rate": 3.1131221719457014e-05, + "loss": 0.5104, + "step": 516 + }, + { + "epoch": 0.194947209653092, + "grad_norm": 0.7910852107873164, + "learning_rate": 3.1191553544494726e-05, + "loss": 0.5268, + "step": 517 + }, + { + "epoch": 0.19532428355957768, + "grad_norm": 0.967010501082144, + "learning_rate": 3.125188536953243e-05, + "loss": 0.5138, + "step": 518 + }, + { + "epoch": 0.19570135746606335, + "grad_norm": 1.288621723987154, + "learning_rate": 3.131221719457014e-05, + "loss": 0.5255, + "step": 519 + }, + { + "epoch": 0.19607843137254902, + "grad_norm": 0.9068680505348491, + "learning_rate": 3.137254901960784e-05, + "loss": 0.5112, + "step": 520 + }, + { + "epoch": 0.1964555052790347, + "grad_norm": 0.769907366497192, + "learning_rate": 3.1432880844645555e-05, + "loss": 0.5207, + "step": 521 + }, + { + "epoch": 0.19683257918552036, + "grad_norm": 0.5749941110891146, + "learning_rate": 3.149321266968326e-05, + "loss": 0.5087, + "step": 522 + }, + { + "epoch": 0.19720965309200603, + "grad_norm": 0.5413488730144038, + "learning_rate": 3.1553544494720966e-05, + "loss": 0.5237, + "step": 523 + }, + { + "epoch": 0.1975867269984917, + "grad_norm": 0.6957104182432634, + "learning_rate": 3.161387631975867e-05, + "loss": 0.5215, + "step": 524 + }, + { + "epoch": 0.19796380090497737, + "grad_norm": 0.6436813462589682, + "learning_rate": 3.1674208144796384e-05, + "loss": 0.5143, + "step": 525 + }, + { + "epoch": 0.19834087481146304, + "grad_norm": 0.5767860847385149, + "learning_rate": 3.173453996983409e-05, + "loss": 0.5099, + "step": 526 + }, + { + "epoch": 0.1987179487179487, + "grad_norm": 0.6184096079649865, + "learning_rate": 3.1794871794871795e-05, + "loss": 0.5174, + "step": 527 + }, + { + "epoch": 0.19909502262443438, + "grad_norm": 0.6581241290876918, + "learning_rate": 3.18552036199095e-05, + "loss": 0.5127, + "step": 528 + }, + { + "epoch": 0.19947209653092007, + "grad_norm": 0.8044914540283841, + "learning_rate": 3.191553544494721e-05, + "loss": 0.514, + "step": 529 + }, + { + "epoch": 0.19984917043740574, + "grad_norm": 1.0859562333946777, + "learning_rate": 3.1975867269984924e-05, + "loss": 0.5241, + "step": 530 + }, + { + "epoch": 0.2002262443438914, + "grad_norm": 1.2316443764120404, + "learning_rate": 3.203619909502263e-05, + "loss": 0.5129, + "step": 531 + }, + { + "epoch": 0.20060331825037708, + "grad_norm": 0.7854798655741155, + "learning_rate": 3.2096530920060335e-05, + "loss": 0.5189, + "step": 532 + }, + { + "epoch": 0.20098039215686275, + "grad_norm": 0.6847908346774234, + "learning_rate": 3.215686274509804e-05, + "loss": 0.5131, + "step": 533 + }, + { + "epoch": 0.20135746606334842, + "grad_norm": 0.48745984150393923, + "learning_rate": 3.221719457013575e-05, + "loss": 0.5236, + "step": 534 + }, + { + "epoch": 0.2017345399698341, + "grad_norm": 0.6631424818064214, + "learning_rate": 3.227752639517346e-05, + "loss": 0.517, + "step": 535 + }, + { + "epoch": 0.20211161387631976, + "grad_norm": 1.022208978207881, + "learning_rate": 3.2337858220211164e-05, + "loss": 0.5196, + "step": 536 + }, + { + "epoch": 0.20248868778280543, + "grad_norm": 1.3672788006637113, + "learning_rate": 3.239819004524887e-05, + "loss": 0.5145, + "step": 537 + }, + { + "epoch": 0.2028657616892911, + "grad_norm": 0.5661726878052495, + "learning_rate": 3.245852187028658e-05, + "loss": 0.5165, + "step": 538 + }, + { + "epoch": 0.20324283559577677, + "grad_norm": 0.682761128765322, + "learning_rate": 3.251885369532429e-05, + "loss": 0.5213, + "step": 539 + }, + { + "epoch": 0.20361990950226244, + "grad_norm": 1.1992017670689215, + "learning_rate": 3.257918552036199e-05, + "loss": 0.5139, + "step": 540 + }, + { + "epoch": 0.2039969834087481, + "grad_norm": 1.0063617293554592, + "learning_rate": 3.26395173453997e-05, + "loss": 0.5103, + "step": 541 + }, + { + "epoch": 0.20437405731523378, + "grad_norm": 0.9310736390499247, + "learning_rate": 3.269984917043741e-05, + "loss": 0.5239, + "step": 542 + }, + { + "epoch": 0.20475113122171945, + "grad_norm": 0.8692518202287649, + "learning_rate": 3.2760180995475116e-05, + "loss": 0.5001, + "step": 543 + }, + { + "epoch": 0.20512820512820512, + "grad_norm": 0.9057992392758819, + "learning_rate": 3.282051282051282e-05, + "loss": 0.5183, + "step": 544 + }, + { + "epoch": 0.2055052790346908, + "grad_norm": 0.9626569235215452, + "learning_rate": 3.288084464555053e-05, + "loss": 0.515, + "step": 545 + }, + { + "epoch": 0.20588235294117646, + "grad_norm": 1.1720599721647318, + "learning_rate": 3.294117647058824e-05, + "loss": 0.5148, + "step": 546 + }, + { + "epoch": 0.20625942684766213, + "grad_norm": 1.4323293561283905, + "learning_rate": 3.3001508295625945e-05, + "loss": 0.5206, + "step": 547 + }, + { + "epoch": 0.2066365007541478, + "grad_norm": 0.7661607012582133, + "learning_rate": 3.306184012066365e-05, + "loss": 0.5157, + "step": 548 + }, + { + "epoch": 0.20701357466063347, + "grad_norm": 1.3466004984175923, + "learning_rate": 3.3122171945701356e-05, + "loss": 0.526, + "step": 549 + }, + { + "epoch": 0.20739064856711917, + "grad_norm": 1.035846544379246, + "learning_rate": 3.318250377073907e-05, + "loss": 0.5043, + "step": 550 + }, + { + "epoch": 0.20776772247360484, + "grad_norm": 1.2384100137833287, + "learning_rate": 3.324283559577678e-05, + "loss": 0.5203, + "step": 551 + }, + { + "epoch": 0.2081447963800905, + "grad_norm": 0.9657341563356918, + "learning_rate": 3.3303167420814486e-05, + "loss": 0.5165, + "step": 552 + }, + { + "epoch": 0.20852187028657618, + "grad_norm": 1.110823592824664, + "learning_rate": 3.336349924585219e-05, + "loss": 0.5115, + "step": 553 + }, + { + "epoch": 0.20889894419306185, + "grad_norm": 0.8611658615751949, + "learning_rate": 3.3423831070889897e-05, + "loss": 0.4999, + "step": 554 + }, + { + "epoch": 0.20927601809954752, + "grad_norm": 0.9865198973427227, + "learning_rate": 3.348416289592761e-05, + "loss": 0.5129, + "step": 555 + }, + { + "epoch": 0.2096530920060332, + "grad_norm": 0.935345445145998, + "learning_rate": 3.3544494720965314e-05, + "loss": 0.519, + "step": 556 + }, + { + "epoch": 0.21003016591251886, + "grad_norm": 0.7078093963380401, + "learning_rate": 3.360482654600302e-05, + "loss": 0.5199, + "step": 557 + }, + { + "epoch": 0.21040723981900453, + "grad_norm": 0.7659428704322047, + "learning_rate": 3.3665158371040725e-05, + "loss": 0.5253, + "step": 558 + }, + { + "epoch": 0.2107843137254902, + "grad_norm": 0.9296572362641379, + "learning_rate": 3.372549019607844e-05, + "loss": 0.5192, + "step": 559 + }, + { + "epoch": 0.21116138763197587, + "grad_norm": 1.0779247664289053, + "learning_rate": 3.378582202111614e-05, + "loss": 0.5069, + "step": 560 + }, + { + "epoch": 0.21153846153846154, + "grad_norm": 1.1205802565087224, + "learning_rate": 3.384615384615385e-05, + "loss": 0.5096, + "step": 561 + }, + { + "epoch": 0.2119155354449472, + "grad_norm": 0.75595101156539, + "learning_rate": 3.3906485671191554e-05, + "loss": 0.5019, + "step": 562 + }, + { + "epoch": 0.21229260935143288, + "grad_norm": 0.5427993847335281, + "learning_rate": 3.3966817496229266e-05, + "loss": 0.5092, + "step": 563 + }, + { + "epoch": 0.21266968325791855, + "grad_norm": 0.6733862693290518, + "learning_rate": 3.402714932126697e-05, + "loss": 0.5051, + "step": 564 + }, + { + "epoch": 0.21304675716440422, + "grad_norm": 0.8198923524127851, + "learning_rate": 3.408748114630468e-05, + "loss": 0.5309, + "step": 565 + }, + { + "epoch": 0.21342383107088989, + "grad_norm": 0.8556769744408598, + "learning_rate": 3.414781297134238e-05, + "loss": 0.5067, + "step": 566 + }, + { + "epoch": 0.21380090497737556, + "grad_norm": 0.759872072666077, + "learning_rate": 3.4208144796380095e-05, + "loss": 0.5257, + "step": 567 + }, + { + "epoch": 0.21417797888386123, + "grad_norm": 0.8699914860769392, + "learning_rate": 3.42684766214178e-05, + "loss": 0.5224, + "step": 568 + }, + { + "epoch": 0.2145550527903469, + "grad_norm": 1.0116242069862573, + "learning_rate": 3.4328808446455506e-05, + "loss": 0.5403, + "step": 569 + }, + { + "epoch": 0.2149321266968326, + "grad_norm": 1.033400258088885, + "learning_rate": 3.438914027149322e-05, + "loss": 0.5193, + "step": 570 + }, + { + "epoch": 0.21530920060331826, + "grad_norm": 0.8298722094658479, + "learning_rate": 3.4449472096530923e-05, + "loss": 0.5077, + "step": 571 + }, + { + "epoch": 0.21568627450980393, + "grad_norm": 0.7497029195333125, + "learning_rate": 3.450980392156863e-05, + "loss": 0.5215, + "step": 572 + }, + { + "epoch": 0.2160633484162896, + "grad_norm": 0.74372896202511, + "learning_rate": 3.457013574660634e-05, + "loss": 0.5328, + "step": 573 + }, + { + "epoch": 0.21644042232277527, + "grad_norm": 1.0542551481871862, + "learning_rate": 3.463046757164405e-05, + "loss": 0.5039, + "step": 574 + }, + { + "epoch": 0.21681749622926094, + "grad_norm": 1.1207026856323106, + "learning_rate": 3.469079939668175e-05, + "loss": 0.5222, + "step": 575 + }, + { + "epoch": 0.2171945701357466, + "grad_norm": 0.5894993095886089, + "learning_rate": 3.475113122171946e-05, + "loss": 0.5229, + "step": 576 + }, + { + "epoch": 0.21757164404223228, + "grad_norm": 0.5735203344920038, + "learning_rate": 3.481146304675717e-05, + "loss": 0.5167, + "step": 577 + }, + { + "epoch": 0.21794871794871795, + "grad_norm": 0.8794245928013755, + "learning_rate": 3.4871794871794875e-05, + "loss": 0.5132, + "step": 578 + }, + { + "epoch": 0.21832579185520362, + "grad_norm": 1.156608605952243, + "learning_rate": 3.493212669683258e-05, + "loss": 0.5136, + "step": 579 + }, + { + "epoch": 0.2187028657616893, + "grad_norm": 0.7369753733557245, + "learning_rate": 3.4992458521870286e-05, + "loss": 0.5209, + "step": 580 + }, + { + "epoch": 0.21907993966817496, + "grad_norm": 0.4463563806073032, + "learning_rate": 3.5052790346908e-05, + "loss": 0.508, + "step": 581 + }, + { + "epoch": 0.21945701357466063, + "grad_norm": 0.7294978801012003, + "learning_rate": 3.5113122171945704e-05, + "loss": 0.5069, + "step": 582 + }, + { + "epoch": 0.2198340874811463, + "grad_norm": 0.9938013800316022, + "learning_rate": 3.517345399698341e-05, + "loss": 0.5131, + "step": 583 + }, + { + "epoch": 0.22021116138763197, + "grad_norm": 1.2030340756423037, + "learning_rate": 3.5233785822021115e-05, + "loss": 0.5105, + "step": 584 + }, + { + "epoch": 0.22058823529411764, + "grad_norm": 1.0202341066866019, + "learning_rate": 3.529411764705883e-05, + "loss": 0.517, + "step": 585 + }, + { + "epoch": 0.2209653092006033, + "grad_norm": 1.2048124370549171, + "learning_rate": 3.535444947209653e-05, + "loss": 0.5152, + "step": 586 + }, + { + "epoch": 0.22134238310708898, + "grad_norm": 0.7216395924178158, + "learning_rate": 3.541478129713424e-05, + "loss": 0.5192, + "step": 587 + }, + { + "epoch": 0.22171945701357465, + "grad_norm": 0.6814743246616025, + "learning_rate": 3.5475113122171944e-05, + "loss": 0.5137, + "step": 588 + }, + { + "epoch": 0.22209653092006032, + "grad_norm": 1.018495649800309, + "learning_rate": 3.5535444947209656e-05, + "loss": 0.5009, + "step": 589 + }, + { + "epoch": 0.222473604826546, + "grad_norm": 0.939008876102782, + "learning_rate": 3.559577677224736e-05, + "loss": 0.5072, + "step": 590 + }, + { + "epoch": 0.2228506787330317, + "grad_norm": 0.639533618752926, + "learning_rate": 3.5656108597285074e-05, + "loss": 0.5309, + "step": 591 + }, + { + "epoch": 0.22322775263951736, + "grad_norm": 0.6612203390580229, + "learning_rate": 3.571644042232278e-05, + "loss": 0.5048, + "step": 592 + }, + { + "epoch": 0.22360482654600303, + "grad_norm": 0.8182377916401323, + "learning_rate": 3.5776772247360485e-05, + "loss": 0.5186, + "step": 593 + }, + { + "epoch": 0.2239819004524887, + "grad_norm": 1.0133304881848944, + "learning_rate": 3.58371040723982e-05, + "loss": 0.5155, + "step": 594 + }, + { + "epoch": 0.22435897435897437, + "grad_norm": 1.1919455597746862, + "learning_rate": 3.58974358974359e-05, + "loss": 0.5145, + "step": 595 + }, + { + "epoch": 0.22473604826546004, + "grad_norm": 0.7591385539884121, + "learning_rate": 3.595776772247361e-05, + "loss": 0.5076, + "step": 596 + }, + { + "epoch": 0.2251131221719457, + "grad_norm": 0.7797019145470101, + "learning_rate": 3.601809954751131e-05, + "loss": 0.5212, + "step": 597 + }, + { + "epoch": 0.22549019607843138, + "grad_norm": 0.8722313956078716, + "learning_rate": 3.6078431372549025e-05, + "loss": 0.5299, + "step": 598 + }, + { + "epoch": 0.22586726998491705, + "grad_norm": 0.8312521913002245, + "learning_rate": 3.613876319758673e-05, + "loss": 0.5056, + "step": 599 + }, + { + "epoch": 0.22624434389140272, + "grad_norm": 0.7910874797922932, + "learning_rate": 3.6199095022624436e-05, + "loss": 0.503, + "step": 600 + }, + { + "epoch": 0.22662141779788839, + "grad_norm": 0.7461989379980156, + "learning_rate": 3.625942684766214e-05, + "loss": 0.5122, + "step": 601 + }, + { + "epoch": 0.22699849170437406, + "grad_norm": 0.6563228445619591, + "learning_rate": 3.6319758672699854e-05, + "loss": 0.5112, + "step": 602 + }, + { + "epoch": 0.22737556561085973, + "grad_norm": 0.9289170981167056, + "learning_rate": 3.638009049773756e-05, + "loss": 0.5107, + "step": 603 + }, + { + "epoch": 0.2277526395173454, + "grad_norm": 1.1075867397185888, + "learning_rate": 3.6440422322775265e-05, + "loss": 0.5349, + "step": 604 + }, + { + "epoch": 0.22812971342383107, + "grad_norm": 0.8027140670533149, + "learning_rate": 3.650075414781297e-05, + "loss": 0.509, + "step": 605 + }, + { + "epoch": 0.22850678733031674, + "grad_norm": 0.7866998352791615, + "learning_rate": 3.656108597285068e-05, + "loss": 0.5279, + "step": 606 + }, + { + "epoch": 0.2288838612368024, + "grad_norm": 0.7813894440419381, + "learning_rate": 3.662141779788839e-05, + "loss": 0.5152, + "step": 607 + }, + { + "epoch": 0.22926093514328807, + "grad_norm": 0.5465510970337469, + "learning_rate": 3.6681749622926094e-05, + "loss": 0.5103, + "step": 608 + }, + { + "epoch": 0.22963800904977374, + "grad_norm": 0.6402803381404601, + "learning_rate": 3.67420814479638e-05, + "loss": 0.5123, + "step": 609 + }, + { + "epoch": 0.23001508295625941, + "grad_norm": 0.6137437110740943, + "learning_rate": 3.680241327300151e-05, + "loss": 0.5079, + "step": 610 + }, + { + "epoch": 0.23039215686274508, + "grad_norm": 0.6477304345090505, + "learning_rate": 3.686274509803922e-05, + "loss": 0.5182, + "step": 611 + }, + { + "epoch": 0.23076923076923078, + "grad_norm": 0.7610355668392307, + "learning_rate": 3.692307692307693e-05, + "loss": 0.5228, + "step": 612 + }, + { + "epoch": 0.23114630467571645, + "grad_norm": 0.7882626836832346, + "learning_rate": 3.6983408748114635e-05, + "loss": 0.5206, + "step": 613 + }, + { + "epoch": 0.23152337858220212, + "grad_norm": 0.7885658980891393, + "learning_rate": 3.704374057315234e-05, + "loss": 0.5095, + "step": 614 + }, + { + "epoch": 0.2319004524886878, + "grad_norm": 0.8057559844547482, + "learning_rate": 3.710407239819005e-05, + "loss": 0.5071, + "step": 615 + }, + { + "epoch": 0.23227752639517346, + "grad_norm": 0.9161856931063713, + "learning_rate": 3.716440422322776e-05, + "loss": 0.5118, + "step": 616 + }, + { + "epoch": 0.23265460030165913, + "grad_norm": 0.9947410364721513, + "learning_rate": 3.722473604826546e-05, + "loss": 0.5172, + "step": 617 + }, + { + "epoch": 0.2330316742081448, + "grad_norm": 1.1664500729792104, + "learning_rate": 3.728506787330317e-05, + "loss": 0.5136, + "step": 618 + }, + { + "epoch": 0.23340874811463047, + "grad_norm": 0.9264660092636934, + "learning_rate": 3.734539969834088e-05, + "loss": 0.5135, + "step": 619 + }, + { + "epoch": 0.23378582202111614, + "grad_norm": 1.0008160680488354, + "learning_rate": 3.7405731523378587e-05, + "loss": 0.5241, + "step": 620 + }, + { + "epoch": 0.2341628959276018, + "grad_norm": 1.339177916474467, + "learning_rate": 3.746606334841629e-05, + "loss": 0.5197, + "step": 621 + }, + { + "epoch": 0.23453996983408748, + "grad_norm": 0.6453207008283555, + "learning_rate": 3.7526395173454e-05, + "loss": 0.5125, + "step": 622 + }, + { + "epoch": 0.23491704374057315, + "grad_norm": 0.6544949395333255, + "learning_rate": 3.758672699849171e-05, + "loss": 0.5184, + "step": 623 + }, + { + "epoch": 0.23529411764705882, + "grad_norm": 1.000409645880309, + "learning_rate": 3.7647058823529415e-05, + "loss": 0.5313, + "step": 624 + }, + { + "epoch": 0.2356711915535445, + "grad_norm": 1.4665758721265967, + "learning_rate": 3.770739064856712e-05, + "loss": 0.5206, + "step": 625 + }, + { + "epoch": 0.23604826546003016, + "grad_norm": 0.6145509071988942, + "learning_rate": 3.7767722473604826e-05, + "loss": 0.5188, + "step": 626 + }, + { + "epoch": 0.23642533936651583, + "grad_norm": 1.3107059335263007, + "learning_rate": 3.782805429864254e-05, + "loss": 0.5121, + "step": 627 + }, + { + "epoch": 0.2368024132730015, + "grad_norm": 1.0661048766300443, + "learning_rate": 3.7888386123680244e-05, + "loss": 0.5172, + "step": 628 + }, + { + "epoch": 0.23717948717948717, + "grad_norm": 1.3212088359285705, + "learning_rate": 3.794871794871795e-05, + "loss": 0.5191, + "step": 629 + }, + { + "epoch": 0.23755656108597284, + "grad_norm": 0.7864098367482385, + "learning_rate": 3.8009049773755655e-05, + "loss": 0.5144, + "step": 630 + }, + { + "epoch": 0.2379336349924585, + "grad_norm": 1.395058327939899, + "learning_rate": 3.806938159879337e-05, + "loss": 0.5108, + "step": 631 + }, + { + "epoch": 0.2383107088989442, + "grad_norm": 0.6616736433710896, + "learning_rate": 3.812971342383108e-05, + "loss": 0.5, + "step": 632 + }, + { + "epoch": 0.23868778280542988, + "grad_norm": 0.8088872400504385, + "learning_rate": 3.8190045248868785e-05, + "loss": 0.5249, + "step": 633 + }, + { + "epoch": 0.23906485671191555, + "grad_norm": 1.015687839743295, + "learning_rate": 3.825037707390649e-05, + "loss": 0.53, + "step": 634 + }, + { + "epoch": 0.23944193061840122, + "grad_norm": 0.9807873641061713, + "learning_rate": 3.8310708898944196e-05, + "loss": 0.5091, + "step": 635 + }, + { + "epoch": 0.2398190045248869, + "grad_norm": 1.0414708080434543, + "learning_rate": 3.837104072398191e-05, + "loss": 0.513, + "step": 636 + }, + { + "epoch": 0.24019607843137256, + "grad_norm": 0.9812861717770868, + "learning_rate": 3.8431372549019614e-05, + "loss": 0.5181, + "step": 637 + }, + { + "epoch": 0.24057315233785823, + "grad_norm": 1.0290012120520826, + "learning_rate": 3.849170437405732e-05, + "loss": 0.5096, + "step": 638 + }, + { + "epoch": 0.2409502262443439, + "grad_norm": 0.7063309571098696, + "learning_rate": 3.8552036199095024e-05, + "loss": 0.5153, + "step": 639 + }, + { + "epoch": 0.24132730015082957, + "grad_norm": 0.5788776778308621, + "learning_rate": 3.861236802413274e-05, + "loss": 0.4991, + "step": 640 + }, + { + "epoch": 0.24170437405731524, + "grad_norm": 0.7258115842255983, + "learning_rate": 3.867269984917044e-05, + "loss": 0.5018, + "step": 641 + }, + { + "epoch": 0.2420814479638009, + "grad_norm": 0.820198716376149, + "learning_rate": 3.873303167420815e-05, + "loss": 0.5104, + "step": 642 + }, + { + "epoch": 0.24245852187028658, + "grad_norm": 0.8390694564145562, + "learning_rate": 3.879336349924585e-05, + "loss": 0.5133, + "step": 643 + }, + { + "epoch": 0.24283559577677225, + "grad_norm": 1.0944326103164097, + "learning_rate": 3.8853695324283565e-05, + "loss": 0.5173, + "step": 644 + }, + { + "epoch": 0.24321266968325791, + "grad_norm": 1.0067524864681492, + "learning_rate": 3.891402714932127e-05, + "loss": 0.4874, + "step": 645 + }, + { + "epoch": 0.24358974358974358, + "grad_norm": 1.0334509363528008, + "learning_rate": 3.8974358974358976e-05, + "loss": 0.5185, + "step": 646 + }, + { + "epoch": 0.24396681749622925, + "grad_norm": 0.9676756748725418, + "learning_rate": 3.903469079939668e-05, + "loss": 0.5016, + "step": 647 + }, + { + "epoch": 0.24434389140271492, + "grad_norm": 0.9355186346200685, + "learning_rate": 3.9095022624434394e-05, + "loss": 0.5102, + "step": 648 + }, + { + "epoch": 0.2447209653092006, + "grad_norm": 0.9470011451595359, + "learning_rate": 3.91553544494721e-05, + "loss": 0.5116, + "step": 649 + }, + { + "epoch": 0.24509803921568626, + "grad_norm": 0.9784960535098289, + "learning_rate": 3.9215686274509805e-05, + "loss": 0.5069, + "step": 650 + }, + { + "epoch": 0.24547511312217193, + "grad_norm": 1.1660640148666102, + "learning_rate": 3.927601809954751e-05, + "loss": 0.5147, + "step": 651 + }, + { + "epoch": 0.2458521870286576, + "grad_norm": 1.0052854814276226, + "learning_rate": 3.933634992458522e-05, + "loss": 0.5149, + "step": 652 + }, + { + "epoch": 0.2462292609351433, + "grad_norm": 0.892909918329878, + "learning_rate": 3.939668174962293e-05, + "loss": 0.508, + "step": 653 + }, + { + "epoch": 0.24660633484162897, + "grad_norm": 0.7188233462998999, + "learning_rate": 3.945701357466064e-05, + "loss": 0.5092, + "step": 654 + }, + { + "epoch": 0.24698340874811464, + "grad_norm": 0.9041208395460688, + "learning_rate": 3.9517345399698346e-05, + "loss": 0.5047, + "step": 655 + }, + { + "epoch": 0.2473604826546003, + "grad_norm": 0.8821597989546563, + "learning_rate": 3.957767722473605e-05, + "loss": 0.5093, + "step": 656 + }, + { + "epoch": 0.24773755656108598, + "grad_norm": 0.9170822335572869, + "learning_rate": 3.963800904977376e-05, + "loss": 0.5196, + "step": 657 + }, + { + "epoch": 0.24811463046757165, + "grad_norm": 1.0495892627029446, + "learning_rate": 3.969834087481147e-05, + "loss": 0.4999, + "step": 658 + }, + { + "epoch": 0.24849170437405732, + "grad_norm": 0.9166569658719925, + "learning_rate": 3.9758672699849175e-05, + "loss": 0.5151, + "step": 659 + }, + { + "epoch": 0.248868778280543, + "grad_norm": 0.669305670979591, + "learning_rate": 3.981900452488688e-05, + "loss": 0.5072, + "step": 660 + }, + { + "epoch": 0.24924585218702866, + "grad_norm": 0.6023393749383104, + "learning_rate": 3.9879336349924586e-05, + "loss": 0.5029, + "step": 661 + }, + { + "epoch": 0.24962292609351433, + "grad_norm": 0.830060970167815, + "learning_rate": 3.99396681749623e-05, + "loss": 0.5065, + "step": 662 + }, + { + "epoch": 0.25, + "grad_norm": 1.0499621767586056, + "learning_rate": 4e-05, + "loss": 0.521, + "step": 663 + }, + { + "epoch": 0.25037707390648567, + "grad_norm": 1.0649925904364537, + "learning_rate": 4.006033182503771e-05, + "loss": 0.5173, + "step": 664 + }, + { + "epoch": 0.25075414781297134, + "grad_norm": 0.9319869705649906, + "learning_rate": 4.012066365007542e-05, + "loss": 0.5073, + "step": 665 + }, + { + "epoch": 0.251131221719457, + "grad_norm": 0.7891292288237327, + "learning_rate": 4.0180995475113126e-05, + "loss": 0.5091, + "step": 666 + }, + { + "epoch": 0.2515082956259427, + "grad_norm": 0.7964715846135405, + "learning_rate": 4.024132730015083e-05, + "loss": 0.5147, + "step": 667 + }, + { + "epoch": 0.25188536953242835, + "grad_norm": 0.8490333038735938, + "learning_rate": 4.030165912518854e-05, + "loss": 0.5123, + "step": 668 + }, + { + "epoch": 0.252262443438914, + "grad_norm": 0.8330496556337817, + "learning_rate": 4.036199095022625e-05, + "loss": 0.5055, + "step": 669 + }, + { + "epoch": 0.2526395173453997, + "grad_norm": 0.8811190792429804, + "learning_rate": 4.0422322775263955e-05, + "loss": 0.5041, + "step": 670 + }, + { + "epoch": 0.25301659125188536, + "grad_norm": 0.9525998734353431, + "learning_rate": 4.048265460030166e-05, + "loss": 0.5121, + "step": 671 + }, + { + "epoch": 0.25339366515837103, + "grad_norm": 0.918762919938664, + "learning_rate": 4.0542986425339366e-05, + "loss": 0.5171, + "step": 672 + }, + { + "epoch": 0.2537707390648567, + "grad_norm": 0.7734350770911819, + "learning_rate": 4.060331825037708e-05, + "loss": 0.4967, + "step": 673 + }, + { + "epoch": 0.25414781297134237, + "grad_norm": 0.7310806082570724, + "learning_rate": 4.0663650075414784e-05, + "loss": 0.5005, + "step": 674 + }, + { + "epoch": 0.25452488687782804, + "grad_norm": 0.9129336147049775, + "learning_rate": 4.072398190045249e-05, + "loss": 0.4984, + "step": 675 + }, + { + "epoch": 0.2549019607843137, + "grad_norm": 0.9700874891554266, + "learning_rate": 4.0784313725490195e-05, + "loss": 0.5102, + "step": 676 + }, + { + "epoch": 0.2552790346907994, + "grad_norm": 0.8806938608475243, + "learning_rate": 4.084464555052791e-05, + "loss": 0.5276, + "step": 677 + }, + { + "epoch": 0.25565610859728505, + "grad_norm": 0.7131877307027, + "learning_rate": 4.090497737556561e-05, + "loss": 0.5118, + "step": 678 + }, + { + "epoch": 0.2560331825037707, + "grad_norm": 0.8472215289293024, + "learning_rate": 4.096530920060332e-05, + "loss": 0.4988, + "step": 679 + }, + { + "epoch": 0.2564102564102564, + "grad_norm": 1.1697367522152806, + "learning_rate": 4.1025641025641023e-05, + "loss": 0.5177, + "step": 680 + }, + { + "epoch": 0.25678733031674206, + "grad_norm": 0.898081137822097, + "learning_rate": 4.1085972850678736e-05, + "loss": 0.5053, + "step": 681 + }, + { + "epoch": 0.2571644042232277, + "grad_norm": 0.7490317586166658, + "learning_rate": 4.114630467571644e-05, + "loss": 0.5121, + "step": 682 + }, + { + "epoch": 0.2575414781297134, + "grad_norm": 0.6227155964413209, + "learning_rate": 4.120663650075415e-05, + "loss": 0.5043, + "step": 683 + }, + { + "epoch": 0.2579185520361991, + "grad_norm": 0.7815960843811846, + "learning_rate": 4.1266968325791866e-05, + "loss": 0.4953, + "step": 684 + }, + { + "epoch": 0.2582956259426848, + "grad_norm": 0.9597919088475204, + "learning_rate": 4.132730015082957e-05, + "loss": 0.5019, + "step": 685 + }, + { + "epoch": 0.25867269984917046, + "grad_norm": 0.845668558983369, + "learning_rate": 4.1387631975867277e-05, + "loss": 0.5164, + "step": 686 + }, + { + "epoch": 0.25904977375565613, + "grad_norm": 0.7189029833710002, + "learning_rate": 4.144796380090498e-05, + "loss": 0.5104, + "step": 687 + }, + { + "epoch": 0.2594268476621418, + "grad_norm": 0.9997791510772692, + "learning_rate": 4.1508295625942694e-05, + "loss": 0.5042, + "step": 688 + }, + { + "epoch": 0.25980392156862747, + "grad_norm": 1.162420297283108, + "learning_rate": 4.15686274509804e-05, + "loss": 0.5035, + "step": 689 + }, + { + "epoch": 0.26018099547511314, + "grad_norm": 0.7549278805370252, + "learning_rate": 4.1628959276018105e-05, + "loss": 0.5146, + "step": 690 + }, + { + "epoch": 0.2605580693815988, + "grad_norm": 0.6977547674384297, + "learning_rate": 4.168929110105581e-05, + "loss": 0.5133, + "step": 691 + }, + { + "epoch": 0.2609351432880845, + "grad_norm": 0.7592408218289051, + "learning_rate": 4.174962292609352e-05, + "loss": 0.4905, + "step": 692 + }, + { + "epoch": 0.26131221719457015, + "grad_norm": 0.8169480036617114, + "learning_rate": 4.180995475113123e-05, + "loss": 0.5182, + "step": 693 + }, + { + "epoch": 0.2616892911010558, + "grad_norm": 0.9542934137261023, + "learning_rate": 4.1870286576168934e-05, + "loss": 0.5235, + "step": 694 + }, + { + "epoch": 0.2620663650075415, + "grad_norm": 1.025184352730565, + "learning_rate": 4.193061840120664e-05, + "loss": 0.5093, + "step": 695 + }, + { + "epoch": 0.26244343891402716, + "grad_norm": 0.9899504634496423, + "learning_rate": 4.199095022624435e-05, + "loss": 0.5172, + "step": 696 + }, + { + "epoch": 0.26282051282051283, + "grad_norm": 1.0987345034204952, + "learning_rate": 4.205128205128206e-05, + "loss": 0.511, + "step": 697 + }, + { + "epoch": 0.2631975867269985, + "grad_norm": 0.826320225833394, + "learning_rate": 4.211161387631976e-05, + "loss": 0.5055, + "step": 698 + }, + { + "epoch": 0.26357466063348417, + "grad_norm": 0.6880307004120179, + "learning_rate": 4.217194570135747e-05, + "loss": 0.505, + "step": 699 + }, + { + "epoch": 0.26395173453996984, + "grad_norm": 0.7139619040835475, + "learning_rate": 4.223227752639518e-05, + "loss": 0.5074, + "step": 700 + }, + { + "epoch": 0.2643288084464555, + "grad_norm": 0.7934264565006127, + "learning_rate": 4.2292609351432886e-05, + "loss": 0.5206, + "step": 701 + }, + { + "epoch": 0.2647058823529412, + "grad_norm": 0.7397580121981251, + "learning_rate": 4.235294117647059e-05, + "loss": 0.5176, + "step": 702 + }, + { + "epoch": 0.26508295625942685, + "grad_norm": 0.6887484778854754, + "learning_rate": 4.24132730015083e-05, + "loss": 0.5053, + "step": 703 + }, + { + "epoch": 0.2654600301659125, + "grad_norm": 0.8403907674194017, + "learning_rate": 4.247360482654601e-05, + "loss": 0.5117, + "step": 704 + }, + { + "epoch": 0.2658371040723982, + "grad_norm": 1.2955520396265343, + "learning_rate": 4.2533936651583714e-05, + "loss": 0.5156, + "step": 705 + }, + { + "epoch": 0.26621417797888386, + "grad_norm": 0.9377244960456969, + "learning_rate": 4.259426847662142e-05, + "loss": 0.5116, + "step": 706 + }, + { + "epoch": 0.26659125188536953, + "grad_norm": 0.7834673654788632, + "learning_rate": 4.2654600301659125e-05, + "loss": 0.5134, + "step": 707 + }, + { + "epoch": 0.2669683257918552, + "grad_norm": 0.9220047343602175, + "learning_rate": 4.271493212669684e-05, + "loss": 0.5009, + "step": 708 + }, + { + "epoch": 0.26734539969834087, + "grad_norm": 1.2109261600347183, + "learning_rate": 4.277526395173454e-05, + "loss": 0.5042, + "step": 709 + }, + { + "epoch": 0.26772247360482654, + "grad_norm": 0.7458456133979725, + "learning_rate": 4.283559577677225e-05, + "loss": 0.5023, + "step": 710 + }, + { + "epoch": 0.2680995475113122, + "grad_norm": 0.9619204028766324, + "learning_rate": 4.2895927601809954e-05, + "loss": 0.4968, + "step": 711 + }, + { + "epoch": 0.2684766214177979, + "grad_norm": 1.0212720684190686, + "learning_rate": 4.2956259426847666e-05, + "loss": 0.5127, + "step": 712 + }, + { + "epoch": 0.26885369532428355, + "grad_norm": 0.8489796142268702, + "learning_rate": 4.301659125188537e-05, + "loss": 0.5078, + "step": 713 + }, + { + "epoch": 0.2692307692307692, + "grad_norm": 1.0985882902088802, + "learning_rate": 4.307692307692308e-05, + "loss": 0.5159, + "step": 714 + }, + { + "epoch": 0.2696078431372549, + "grad_norm": 0.7485776449226195, + "learning_rate": 4.313725490196078e-05, + "loss": 0.5125, + "step": 715 + }, + { + "epoch": 0.26998491704374056, + "grad_norm": 0.8784324920275286, + "learning_rate": 4.3197586726998495e-05, + "loss": 0.5099, + "step": 716 + }, + { + "epoch": 0.2703619909502262, + "grad_norm": 0.8923852047810921, + "learning_rate": 4.32579185520362e-05, + "loss": 0.5008, + "step": 717 + }, + { + "epoch": 0.2707390648567119, + "grad_norm": 0.6508813350281071, + "learning_rate": 4.3318250377073906e-05, + "loss": 0.4942, + "step": 718 + }, + { + "epoch": 0.27111613876319757, + "grad_norm": 0.8016592018816051, + "learning_rate": 4.337858220211161e-05, + "loss": 0.5049, + "step": 719 + }, + { + "epoch": 0.27149321266968324, + "grad_norm": 0.7567654400035373, + "learning_rate": 4.3438914027149324e-05, + "loss": 0.4958, + "step": 720 + }, + { + "epoch": 0.2718702865761689, + "grad_norm": 0.6565782260827147, + "learning_rate": 4.349924585218703e-05, + "loss": 0.4932, + "step": 721 + }, + { + "epoch": 0.2722473604826546, + "grad_norm": 0.967627920153312, + "learning_rate": 4.3559577677224735e-05, + "loss": 0.4954, + "step": 722 + }, + { + "epoch": 0.27262443438914025, + "grad_norm": 1.1378830971261298, + "learning_rate": 4.361990950226244e-05, + "loss": 0.5139, + "step": 723 + }, + { + "epoch": 0.2730015082956259, + "grad_norm": 0.5696475577720911, + "learning_rate": 4.368024132730015e-05, + "loss": 0.4912, + "step": 724 + }, + { + "epoch": 0.2733785822021116, + "grad_norm": 0.7704495082746453, + "learning_rate": 4.374057315233786e-05, + "loss": 0.4999, + "step": 725 + }, + { + "epoch": 0.2737556561085973, + "grad_norm": 0.8684150534424665, + "learning_rate": 4.380090497737557e-05, + "loss": 0.4942, + "step": 726 + }, + { + "epoch": 0.274132730015083, + "grad_norm": 0.8161583148315334, + "learning_rate": 4.386123680241328e-05, + "loss": 0.5142, + "step": 727 + }, + { + "epoch": 0.27450980392156865, + "grad_norm": 0.8996608043131235, + "learning_rate": 4.392156862745099e-05, + "loss": 0.4994, + "step": 728 + }, + { + "epoch": 0.2748868778280543, + "grad_norm": 1.083050618624497, + "learning_rate": 4.398190045248869e-05, + "loss": 0.5014, + "step": 729 + }, + { + "epoch": 0.27526395173454, + "grad_norm": 1.088057688391252, + "learning_rate": 4.4042232277526406e-05, + "loss": 0.5044, + "step": 730 + }, + { + "epoch": 0.27564102564102566, + "grad_norm": 0.7481916335705241, + "learning_rate": 4.410256410256411e-05, + "loss": 0.4966, + "step": 731 + }, + { + "epoch": 0.27601809954751133, + "grad_norm": 0.6278567987216656, + "learning_rate": 4.4162895927601817e-05, + "loss": 0.5222, + "step": 732 + }, + { + "epoch": 0.276395173453997, + "grad_norm": 0.6695470080216677, + "learning_rate": 4.422322775263952e-05, + "loss": 0.5128, + "step": 733 + }, + { + "epoch": 0.27677224736048267, + "grad_norm": 0.6966494100149412, + "learning_rate": 4.4283559577677234e-05, + "loss": 0.4978, + "step": 734 + }, + { + "epoch": 0.27714932126696834, + "grad_norm": 0.8482151654655828, + "learning_rate": 4.434389140271494e-05, + "loss": 0.502, + "step": 735 + }, + { + "epoch": 0.277526395173454, + "grad_norm": 1.1292406037017975, + "learning_rate": 4.4404223227752645e-05, + "loss": 0.5173, + "step": 736 + }, + { + "epoch": 0.2779034690799397, + "grad_norm": 1.0849270236048327, + "learning_rate": 4.446455505279035e-05, + "loss": 0.5099, + "step": 737 + }, + { + "epoch": 0.27828054298642535, + "grad_norm": 0.9479803710145223, + "learning_rate": 4.452488687782806e-05, + "loss": 0.5058, + "step": 738 + }, + { + "epoch": 0.278657616892911, + "grad_norm": 0.8806309967491087, + "learning_rate": 4.458521870286577e-05, + "loss": 0.5004, + "step": 739 + }, + { + "epoch": 0.2790346907993967, + "grad_norm": 0.8185562009697873, + "learning_rate": 4.4645550527903474e-05, + "loss": 0.4981, + "step": 740 + }, + { + "epoch": 0.27941176470588236, + "grad_norm": 0.7865046948454764, + "learning_rate": 4.470588235294118e-05, + "loss": 0.5179, + "step": 741 + }, + { + "epoch": 0.27978883861236803, + "grad_norm": 0.8201512656616999, + "learning_rate": 4.476621417797889e-05, + "loss": 0.4974, + "step": 742 + }, + { + "epoch": 0.2801659125188537, + "grad_norm": 0.7361759157519244, + "learning_rate": 4.48265460030166e-05, + "loss": 0.4916, + "step": 743 + }, + { + "epoch": 0.28054298642533937, + "grad_norm": 0.6808789768411301, + "learning_rate": 4.48868778280543e-05, + "loss": 0.5025, + "step": 744 + }, + { + "epoch": 0.28092006033182504, + "grad_norm": 0.8291778659849435, + "learning_rate": 4.494720965309201e-05, + "loss": 0.5057, + "step": 745 + }, + { + "epoch": 0.2812971342383107, + "grad_norm": 0.7917140574409488, + "learning_rate": 4.500754147812972e-05, + "loss": 0.4981, + "step": 746 + }, + { + "epoch": 0.2816742081447964, + "grad_norm": 0.6719520599860472, + "learning_rate": 4.5067873303167426e-05, + "loss": 0.501, + "step": 747 + }, + { + "epoch": 0.28205128205128205, + "grad_norm": 0.8665217055130805, + "learning_rate": 4.512820512820513e-05, + "loss": 0.502, + "step": 748 + }, + { + "epoch": 0.2824283559577677, + "grad_norm": 1.103153433490193, + "learning_rate": 4.518853695324284e-05, + "loss": 0.5076, + "step": 749 + }, + { + "epoch": 0.2828054298642534, + "grad_norm": 0.9445346146267433, + "learning_rate": 4.524886877828055e-05, + "loss": 0.5028, + "step": 750 + }, + { + "epoch": 0.28318250377073906, + "grad_norm": 1.0075466741451804, + "learning_rate": 4.5309200603318254e-05, + "loss": 0.5139, + "step": 751 + }, + { + "epoch": 0.28355957767722473, + "grad_norm": 1.1429788459896475, + "learning_rate": 4.536953242835596e-05, + "loss": 0.5098, + "step": 752 + }, + { + "epoch": 0.2839366515837104, + "grad_norm": 0.664301767513823, + "learning_rate": 4.5429864253393665e-05, + "loss": 0.5053, + "step": 753 + }, + { + "epoch": 0.28431372549019607, + "grad_norm": 0.5272298384908539, + "learning_rate": 4.549019607843138e-05, + "loss": 0.4985, + "step": 754 + }, + { + "epoch": 0.28469079939668174, + "grad_norm": 0.5321654370946767, + "learning_rate": 4.555052790346908e-05, + "loss": 0.5027, + "step": 755 + }, + { + "epoch": 0.2850678733031674, + "grad_norm": 0.8277760032728366, + "learning_rate": 4.561085972850679e-05, + "loss": 0.5086, + "step": 756 + }, + { + "epoch": 0.2854449472096531, + "grad_norm": 1.1540558800815042, + "learning_rate": 4.5671191553544494e-05, + "loss": 0.5051, + "step": 757 + }, + { + "epoch": 0.28582202111613875, + "grad_norm": 0.7224387806969684, + "learning_rate": 4.5731523378582206e-05, + "loss": 0.4945, + "step": 758 + }, + { + "epoch": 0.2861990950226244, + "grad_norm": 0.5242235292761531, + "learning_rate": 4.579185520361991e-05, + "loss": 0.5046, + "step": 759 + }, + { + "epoch": 0.2865761689291101, + "grad_norm": 0.5944039417043698, + "learning_rate": 4.585218702865762e-05, + "loss": 0.5017, + "step": 760 + }, + { + "epoch": 0.28695324283559576, + "grad_norm": 0.692104267128947, + "learning_rate": 4.591251885369532e-05, + "loss": 0.5038, + "step": 761 + }, + { + "epoch": 0.2873303167420814, + "grad_norm": 0.6866320955744407, + "learning_rate": 4.5972850678733035e-05, + "loss": 0.5076, + "step": 762 + }, + { + "epoch": 0.2877073906485671, + "grad_norm": 0.664351877751294, + "learning_rate": 4.603318250377074e-05, + "loss": 0.4993, + "step": 763 + }, + { + "epoch": 0.28808446455505277, + "grad_norm": 0.8700614864856983, + "learning_rate": 4.6093514328808446e-05, + "loss": 0.5058, + "step": 764 + }, + { + "epoch": 0.28846153846153844, + "grad_norm": 1.0788162797722372, + "learning_rate": 4.615384615384615e-05, + "loss": 0.5068, + "step": 765 + }, + { + "epoch": 0.2888386123680241, + "grad_norm": 1.0215783003627323, + "learning_rate": 4.6214177978883864e-05, + "loss": 0.4868, + "step": 766 + }, + { + "epoch": 0.28921568627450983, + "grad_norm": 0.9631340297641102, + "learning_rate": 4.6274509803921576e-05, + "loss": 0.4918, + "step": 767 + }, + { + "epoch": 0.2895927601809955, + "grad_norm": 1.0592539290420113, + "learning_rate": 4.633484162895928e-05, + "loss": 0.5136, + "step": 768 + }, + { + "epoch": 0.28996983408748117, + "grad_norm": 0.9087718047411013, + "learning_rate": 4.6395173453996994e-05, + "loss": 0.503, + "step": 769 + }, + { + "epoch": 0.29034690799396684, + "grad_norm": 0.8129538257161784, + "learning_rate": 4.64555052790347e-05, + "loss": 0.5104, + "step": 770 + }, + { + "epoch": 0.2907239819004525, + "grad_norm": 0.7802666317002079, + "learning_rate": 4.6515837104072405e-05, + "loss": 0.5005, + "step": 771 + }, + { + "epoch": 0.2911010558069382, + "grad_norm": 0.7991128419280614, + "learning_rate": 4.657616892911011e-05, + "loss": 0.5155, + "step": 772 + }, + { + "epoch": 0.29147812971342385, + "grad_norm": 0.6776854749060812, + "learning_rate": 4.663650075414782e-05, + "loss": 0.4959, + "step": 773 + }, + { + "epoch": 0.2918552036199095, + "grad_norm": 0.5781894818638714, + "learning_rate": 4.669683257918553e-05, + "loss": 0.5037, + "step": 774 + }, + { + "epoch": 0.2922322775263952, + "grad_norm": 0.6110180082804871, + "learning_rate": 4.675716440422323e-05, + "loss": 0.5008, + "step": 775 + }, + { + "epoch": 0.29260935143288086, + "grad_norm": 0.6737347160033889, + "learning_rate": 4.681749622926094e-05, + "loss": 0.522, + "step": 776 + }, + { + "epoch": 0.29298642533936653, + "grad_norm": 0.8144606035599683, + "learning_rate": 4.687782805429865e-05, + "loss": 0.5135, + "step": 777 + }, + { + "epoch": 0.2933634992458522, + "grad_norm": 1.0106129975873708, + "learning_rate": 4.6938159879336356e-05, + "loss": 0.5067, + "step": 778 + }, + { + "epoch": 0.29374057315233787, + "grad_norm": 1.1085773405816832, + "learning_rate": 4.699849170437406e-05, + "loss": 0.5043, + "step": 779 + }, + { + "epoch": 0.29411764705882354, + "grad_norm": 0.788519174113078, + "learning_rate": 4.705882352941177e-05, + "loss": 0.5059, + "step": 780 + }, + { + "epoch": 0.2944947209653092, + "grad_norm": 0.6917141728088446, + "learning_rate": 4.711915535444948e-05, + "loss": 0.4885, + "step": 781 + }, + { + "epoch": 0.2948717948717949, + "grad_norm": 0.8188265657940159, + "learning_rate": 4.7179487179487185e-05, + "loss": 0.5135, + "step": 782 + }, + { + "epoch": 0.29524886877828055, + "grad_norm": 0.7501458307558968, + "learning_rate": 4.723981900452489e-05, + "loss": 0.5031, + "step": 783 + }, + { + "epoch": 0.2956259426847662, + "grad_norm": 0.651161235443772, + "learning_rate": 4.7300150829562596e-05, + "loss": 0.5084, + "step": 784 + }, + { + "epoch": 0.2960030165912519, + "grad_norm": 0.6713978947030245, + "learning_rate": 4.736048265460031e-05, + "loss": 0.515, + "step": 785 + }, + { + "epoch": 0.29638009049773756, + "grad_norm": 0.8512740699800453, + "learning_rate": 4.7420814479638014e-05, + "loss": 0.5017, + "step": 786 + }, + { + "epoch": 0.29675716440422323, + "grad_norm": 0.9730225738256755, + "learning_rate": 4.748114630467572e-05, + "loss": 0.4989, + "step": 787 + }, + { + "epoch": 0.2971342383107089, + "grad_norm": 1.0139501542539584, + "learning_rate": 4.7541478129713425e-05, + "loss": 0.4946, + "step": 788 + }, + { + "epoch": 0.29751131221719457, + "grad_norm": 1.0016670308825621, + "learning_rate": 4.760180995475114e-05, + "loss": 0.5031, + "step": 789 + }, + { + "epoch": 0.29788838612368024, + "grad_norm": 1.080994068457492, + "learning_rate": 4.766214177978884e-05, + "loss": 0.5055, + "step": 790 + }, + { + "epoch": 0.2982654600301659, + "grad_norm": 1.0107853599688381, + "learning_rate": 4.772247360482655e-05, + "loss": 0.5062, + "step": 791 + }, + { + "epoch": 0.2986425339366516, + "grad_norm": 1.0731872483490088, + "learning_rate": 4.778280542986425e-05, + "loss": 0.4973, + "step": 792 + }, + { + "epoch": 0.29901960784313725, + "grad_norm": 0.6953657508919528, + "learning_rate": 4.7843137254901966e-05, + "loss": 0.5109, + "step": 793 + }, + { + "epoch": 0.2993966817496229, + "grad_norm": 0.4636406121483226, + "learning_rate": 4.790346907993967e-05, + "loss": 0.4945, + "step": 794 + }, + { + "epoch": 0.2997737556561086, + "grad_norm": 0.6072787022083715, + "learning_rate": 4.7963800904977377e-05, + "loss": 0.5006, + "step": 795 + }, + { + "epoch": 0.30015082956259426, + "grad_norm": 0.7804171395120932, + "learning_rate": 4.802413273001508e-05, + "loss": 0.4994, + "step": 796 + }, + { + "epoch": 0.3005279034690799, + "grad_norm": 0.8187755002266776, + "learning_rate": 4.8084464555052794e-05, + "loss": 0.503, + "step": 797 + }, + { + "epoch": 0.3009049773755656, + "grad_norm": 0.7615871620836511, + "learning_rate": 4.81447963800905e-05, + "loss": 0.4915, + "step": 798 + }, + { + "epoch": 0.30128205128205127, + "grad_norm": 0.6511365979662428, + "learning_rate": 4.8205128205128205e-05, + "loss": 0.5047, + "step": 799 + }, + { + "epoch": 0.30165912518853694, + "grad_norm": 0.7269569706851688, + "learning_rate": 4.826546003016591e-05, + "loss": 0.4962, + "step": 800 + }, + { + "epoch": 0.3020361990950226, + "grad_norm": 0.7279530195656335, + "learning_rate": 4.832579185520362e-05, + "loss": 0.5023, + "step": 801 + }, + { + "epoch": 0.3024132730015083, + "grad_norm": 0.5180178907406704, + "learning_rate": 4.838612368024133e-05, + "loss": 0.5017, + "step": 802 + }, + { + "epoch": 0.30279034690799395, + "grad_norm": 0.4425023478626619, + "learning_rate": 4.8446455505279034e-05, + "loss": 0.4917, + "step": 803 + }, + { + "epoch": 0.3031674208144796, + "grad_norm": 0.5770775247471239, + "learning_rate": 4.850678733031674e-05, + "loss": 0.5004, + "step": 804 + }, + { + "epoch": 0.3035444947209653, + "grad_norm": 0.73219197777679, + "learning_rate": 4.856711915535445e-05, + "loss": 0.4912, + "step": 805 + }, + { + "epoch": 0.30392156862745096, + "grad_norm": 0.9690080415288054, + "learning_rate": 4.862745098039216e-05, + "loss": 0.4964, + "step": 806 + }, + { + "epoch": 0.3042986425339366, + "grad_norm": 1.30318553999985, + "learning_rate": 4.868778280542986e-05, + "loss": 0.5012, + "step": 807 + }, + { + "epoch": 0.3046757164404223, + "grad_norm": 0.7249608049459065, + "learning_rate": 4.874811463046757e-05, + "loss": 0.4932, + "step": 808 + }, + { + "epoch": 0.305052790346908, + "grad_norm": 0.9900348658523896, + "learning_rate": 4.880844645550529e-05, + "loss": 0.512, + "step": 809 + }, + { + "epoch": 0.3054298642533937, + "grad_norm": 1.117717845977899, + "learning_rate": 4.886877828054299e-05, + "loss": 0.5083, + "step": 810 + }, + { + "epoch": 0.30580693815987936, + "grad_norm": 0.6261497278671593, + "learning_rate": 4.8929110105580705e-05, + "loss": 0.4835, + "step": 811 + }, + { + "epoch": 0.30618401206636503, + "grad_norm": 1.0049562990173726, + "learning_rate": 4.898944193061841e-05, + "loss": 0.5022, + "step": 812 + }, + { + "epoch": 0.3065610859728507, + "grad_norm": 0.9748538921544919, + "learning_rate": 4.9049773755656116e-05, + "loss": 0.502, + "step": 813 + }, + { + "epoch": 0.30693815987933637, + "grad_norm": 1.006006308175813, + "learning_rate": 4.911010558069382e-05, + "loss": 0.4969, + "step": 814 + }, + { + "epoch": 0.30731523378582204, + "grad_norm": 0.8382849082677398, + "learning_rate": 4.9170437405731533e-05, + "loss": 0.5035, + "step": 815 + }, + { + "epoch": 0.3076923076923077, + "grad_norm": 0.8311153543059534, + "learning_rate": 4.923076923076924e-05, + "loss": 0.4997, + "step": 816 + }, + { + "epoch": 0.3080693815987934, + "grad_norm": 1.1221486897249844, + "learning_rate": 4.9291101055806944e-05, + "loss": 0.5008, + "step": 817 + }, + { + "epoch": 0.30844645550527905, + "grad_norm": 0.6614011511632739, + "learning_rate": 4.935143288084465e-05, + "loss": 0.5028, + "step": 818 + }, + { + "epoch": 0.3088235294117647, + "grad_norm": 0.6747589439193913, + "learning_rate": 4.941176470588236e-05, + "loss": 0.5019, + "step": 819 + }, + { + "epoch": 0.3092006033182504, + "grad_norm": 0.7108670381557826, + "learning_rate": 4.947209653092007e-05, + "loss": 0.4955, + "step": 820 + }, + { + "epoch": 0.30957767722473606, + "grad_norm": 0.6343365147713492, + "learning_rate": 4.953242835595777e-05, + "loss": 0.51, + "step": 821 + }, + { + "epoch": 0.30995475113122173, + "grad_norm": 0.6719681661542236, + "learning_rate": 4.959276018099548e-05, + "loss": 0.5052, + "step": 822 + }, + { + "epoch": 0.3103318250377074, + "grad_norm": 0.6418897115480495, + "learning_rate": 4.965309200603319e-05, + "loss": 0.5067, + "step": 823 + }, + { + "epoch": 0.31070889894419307, + "grad_norm": 0.739359676843324, + "learning_rate": 4.9713423831070896e-05, + "loss": 0.5082, + "step": 824 + }, + { + "epoch": 0.31108597285067874, + "grad_norm": 0.817465133116278, + "learning_rate": 4.97737556561086e-05, + "loss": 0.4951, + "step": 825 + }, + { + "epoch": 0.3114630467571644, + "grad_norm": 0.9548617195484783, + "learning_rate": 4.983408748114631e-05, + "loss": 0.5005, + "step": 826 + }, + { + "epoch": 0.3118401206636501, + "grad_norm": 0.8405101113023482, + "learning_rate": 4.989441930618402e-05, + "loss": 0.5055, + "step": 827 + }, + { + "epoch": 0.31221719457013575, + "grad_norm": 0.6944640940260212, + "learning_rate": 4.9954751131221725e-05, + "loss": 0.4883, + "step": 828 + }, + { + "epoch": 0.3125942684766214, + "grad_norm": 0.7690653160216466, + "learning_rate": 5.001508295625943e-05, + "loss": 0.5029, + "step": 829 + }, + { + "epoch": 0.3129713423831071, + "grad_norm": 0.862190671145783, + "learning_rate": 5.0075414781297136e-05, + "loss": 0.5081, + "step": 830 + }, + { + "epoch": 0.31334841628959276, + "grad_norm": 1.0369138500530604, + "learning_rate": 5.013574660633485e-05, + "loss": 0.4927, + "step": 831 + }, + { + "epoch": 0.3137254901960784, + "grad_norm": 1.0535074837941494, + "learning_rate": 5.0196078431372554e-05, + "loss": 0.4962, + "step": 832 + }, + { + "epoch": 0.3141025641025641, + "grad_norm": 0.9831498299500862, + "learning_rate": 5.025641025641026e-05, + "loss": 0.5066, + "step": 833 + }, + { + "epoch": 0.31447963800904977, + "grad_norm": 1.0754997564314628, + "learning_rate": 5.0316742081447965e-05, + "loss": 0.4859, + "step": 834 + }, + { + "epoch": 0.31485671191553544, + "grad_norm": 0.9527038838239744, + "learning_rate": 5.037707390648568e-05, + "loss": 0.499, + "step": 835 + }, + { + "epoch": 0.3152337858220211, + "grad_norm": 0.9186943482908543, + "learning_rate": 5.043740573152338e-05, + "loss": 0.4904, + "step": 836 + }, + { + "epoch": 0.3156108597285068, + "grad_norm": 0.8857785418849724, + "learning_rate": 5.049773755656109e-05, + "loss": 0.5055, + "step": 837 + }, + { + "epoch": 0.31598793363499245, + "grad_norm": 0.7119882414880951, + "learning_rate": 5.055806938159879e-05, + "loss": 0.493, + "step": 838 + }, + { + "epoch": 0.3163650075414781, + "grad_norm": 0.4450509690563639, + "learning_rate": 5.0618401206636506e-05, + "loss": 0.4962, + "step": 839 + }, + { + "epoch": 0.3167420814479638, + "grad_norm": 0.46111654413510395, + "learning_rate": 5.067873303167421e-05, + "loss": 0.4805, + "step": 840 + }, + { + "epoch": 0.31711915535444946, + "grad_norm": 0.5268799451952569, + "learning_rate": 5.0739064856711916e-05, + "loss": 0.4993, + "step": 841 + }, + { + "epoch": 0.3174962292609351, + "grad_norm": 0.4768142708988169, + "learning_rate": 5.079939668174962e-05, + "loss": 0.497, + "step": 842 + }, + { + "epoch": 0.3178733031674208, + "grad_norm": 0.4042347623920423, + "learning_rate": 5.0859728506787334e-05, + "loss": 0.5028, + "step": 843 + }, + { + "epoch": 0.31825037707390647, + "grad_norm": 0.5228882313646609, + "learning_rate": 5.092006033182504e-05, + "loss": 0.4952, + "step": 844 + }, + { + "epoch": 0.31862745098039214, + "grad_norm": 0.5508316460022625, + "learning_rate": 5.0980392156862745e-05, + "loss": 0.5128, + "step": 845 + }, + { + "epoch": 0.3190045248868778, + "grad_norm": 0.589864309348432, + "learning_rate": 5.104072398190045e-05, + "loss": 0.4957, + "step": 846 + }, + { + "epoch": 0.3193815987933635, + "grad_norm": 0.7536929260323892, + "learning_rate": 5.110105580693816e-05, + "loss": 0.491, + "step": 847 + }, + { + "epoch": 0.31975867269984914, + "grad_norm": 0.993034965501514, + "learning_rate": 5.116138763197587e-05, + "loss": 0.5056, + "step": 848 + }, + { + "epoch": 0.3201357466063348, + "grad_norm": 1.3339127172899559, + "learning_rate": 5.1221719457013574e-05, + "loss": 0.5075, + "step": 849 + }, + { + "epoch": 0.32051282051282054, + "grad_norm": 0.6386097815198015, + "learning_rate": 5.128205128205129e-05, + "loss": 0.4906, + "step": 850 + }, + { + "epoch": 0.3208898944193062, + "grad_norm": 0.49227060377953963, + "learning_rate": 5.1342383107089e-05, + "loss": 0.4905, + "step": 851 + }, + { + "epoch": 0.3212669683257919, + "grad_norm": 0.9269459062328911, + "learning_rate": 5.1402714932126704e-05, + "loss": 0.49, + "step": 852 + }, + { + "epoch": 0.32164404223227755, + "grad_norm": 1.425884334525326, + "learning_rate": 5.146304675716441e-05, + "loss": 0.4958, + "step": 853 + }, + { + "epoch": 0.3220211161387632, + "grad_norm": 0.5367433680998255, + "learning_rate": 5.152337858220212e-05, + "loss": 0.5085, + "step": 854 + }, + { + "epoch": 0.3223981900452489, + "grad_norm": 0.8959858356622888, + "learning_rate": 5.158371040723983e-05, + "loss": 0.5082, + "step": 855 + }, + { + "epoch": 0.32277526395173456, + "grad_norm": 1.3186701106777399, + "learning_rate": 5.164404223227753e-05, + "loss": 0.4885, + "step": 856 + }, + { + "epoch": 0.32315233785822023, + "grad_norm": 0.5609652299298706, + "learning_rate": 5.170437405731524e-05, + "loss": 0.4849, + "step": 857 + }, + { + "epoch": 0.3235294117647059, + "grad_norm": 1.080700412551734, + "learning_rate": 5.176470588235295e-05, + "loss": 0.4974, + "step": 858 + }, + { + "epoch": 0.32390648567119157, + "grad_norm": 1.0265797171926545, + "learning_rate": 5.1825037707390656e-05, + "loss": 0.4985, + "step": 859 + }, + { + "epoch": 0.32428355957767724, + "grad_norm": 0.8020170069374954, + "learning_rate": 5.188536953242836e-05, + "loss": 0.4991, + "step": 860 + }, + { + "epoch": 0.3246606334841629, + "grad_norm": 0.9536184376751431, + "learning_rate": 5.194570135746607e-05, + "loss": 0.5071, + "step": 861 + }, + { + "epoch": 0.3250377073906486, + "grad_norm": 0.8398705674914443, + "learning_rate": 5.200603318250378e-05, + "loss": 0.502, + "step": 862 + }, + { + "epoch": 0.32541478129713425, + "grad_norm": 0.9061007582653983, + "learning_rate": 5.2066365007541484e-05, + "loss": 0.4917, + "step": 863 + }, + { + "epoch": 0.3257918552036199, + "grad_norm": 1.1551416683433964, + "learning_rate": 5.212669683257919e-05, + "loss": 0.5017, + "step": 864 + }, + { + "epoch": 0.3261689291101056, + "grad_norm": 0.7944940637004783, + "learning_rate": 5.2187028657616895e-05, + "loss": 0.5146, + "step": 865 + }, + { + "epoch": 0.32654600301659126, + "grad_norm": 0.8333954591699354, + "learning_rate": 5.224736048265461e-05, + "loss": 0.4908, + "step": 866 + }, + { + "epoch": 0.3269230769230769, + "grad_norm": 0.6765423383413012, + "learning_rate": 5.230769230769231e-05, + "loss": 0.5124, + "step": 867 + }, + { + "epoch": 0.3273001508295626, + "grad_norm": 0.8140042804640544, + "learning_rate": 5.236802413273002e-05, + "loss": 0.4954, + "step": 868 + }, + { + "epoch": 0.32767722473604827, + "grad_norm": 0.9929961328073594, + "learning_rate": 5.2428355957767724e-05, + "loss": 0.5089, + "step": 869 + }, + { + "epoch": 0.32805429864253394, + "grad_norm": 1.1360134354373164, + "learning_rate": 5.2488687782805436e-05, + "loss": 0.4821, + "step": 870 + }, + { + "epoch": 0.3284313725490196, + "grad_norm": 0.8646162324366451, + "learning_rate": 5.254901960784314e-05, + "loss": 0.4974, + "step": 871 + }, + { + "epoch": 0.3288084464555053, + "grad_norm": 0.7446280843334614, + "learning_rate": 5.260935143288085e-05, + "loss": 0.5, + "step": 872 + }, + { + "epoch": 0.32918552036199095, + "grad_norm": 0.7724199574722778, + "learning_rate": 5.266968325791855e-05, + "loss": 0.4892, + "step": 873 + }, + { + "epoch": 0.3295625942684766, + "grad_norm": 0.7437128431001179, + "learning_rate": 5.2730015082956265e-05, + "loss": 0.5098, + "step": 874 + }, + { + "epoch": 0.3299396681749623, + "grad_norm": 0.6430149605654962, + "learning_rate": 5.279034690799397e-05, + "loss": 0.501, + "step": 875 + }, + { + "epoch": 0.33031674208144796, + "grad_norm": 0.8277758166599012, + "learning_rate": 5.2850678733031676e-05, + "loss": 0.4915, + "step": 876 + }, + { + "epoch": 0.3306938159879336, + "grad_norm": 0.9671019450678979, + "learning_rate": 5.291101055806938e-05, + "loss": 0.5069, + "step": 877 + }, + { + "epoch": 0.3310708898944193, + "grad_norm": 1.0747871668247297, + "learning_rate": 5.2971342383107094e-05, + "loss": 0.4954, + "step": 878 + }, + { + "epoch": 0.33144796380090497, + "grad_norm": 0.9541666389979623, + "learning_rate": 5.30316742081448e-05, + "loss": 0.4978, + "step": 879 + }, + { + "epoch": 0.33182503770739064, + "grad_norm": 1.0037444214935802, + "learning_rate": 5.3092006033182505e-05, + "loss": 0.5027, + "step": 880 + }, + { + "epoch": 0.3322021116138763, + "grad_norm": 1.1012496635757731, + "learning_rate": 5.315233785822021e-05, + "loss": 0.4944, + "step": 881 + }, + { + "epoch": 0.332579185520362, + "grad_norm": 0.8912418580429053, + "learning_rate": 5.321266968325792e-05, + "loss": 0.4967, + "step": 882 + }, + { + "epoch": 0.33295625942684764, + "grad_norm": 0.7802947537801932, + "learning_rate": 5.327300150829563e-05, + "loss": 0.499, + "step": 883 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 0.7603398841901109, + "learning_rate": 5.333333333333333e-05, + "loss": 0.4969, + "step": 884 + }, + { + "epoch": 0.333710407239819, + "grad_norm": 0.7385291314564274, + "learning_rate": 5.339366515837104e-05, + "loss": 0.4876, + "step": 885 + }, + { + "epoch": 0.33408748114630465, + "grad_norm": 0.6991040995034289, + "learning_rate": 5.345399698340875e-05, + "loss": 0.4995, + "step": 886 + }, + { + "epoch": 0.3344645550527903, + "grad_norm": 0.6955880738297595, + "learning_rate": 5.3514328808446456e-05, + "loss": 0.4939, + "step": 887 + }, + { + "epoch": 0.334841628959276, + "grad_norm": 0.7495774437494238, + "learning_rate": 5.357466063348416e-05, + "loss": 0.5017, + "step": 888 + }, + { + "epoch": 0.33521870286576166, + "grad_norm": 0.8236825583766182, + "learning_rate": 5.363499245852187e-05, + "loss": 0.5153, + "step": 889 + }, + { + "epoch": 0.33559577677224733, + "grad_norm": 0.8525703834067683, + "learning_rate": 5.369532428355958e-05, + "loss": 0.4948, + "step": 890 + }, + { + "epoch": 0.33597285067873306, + "grad_norm": 0.7589007424487039, + "learning_rate": 5.375565610859729e-05, + "loss": 0.4909, + "step": 891 + }, + { + "epoch": 0.33634992458521873, + "grad_norm": 0.7943021064755984, + "learning_rate": 5.3815987933635004e-05, + "loss": 0.4982, + "step": 892 + }, + { + "epoch": 0.3367269984917044, + "grad_norm": 0.8358503893204744, + "learning_rate": 5.387631975867271e-05, + "loss": 0.509, + "step": 893 + }, + { + "epoch": 0.33710407239819007, + "grad_norm": 0.6377138947343135, + "learning_rate": 5.3936651583710415e-05, + "loss": 0.5031, + "step": 894 + }, + { + "epoch": 0.33748114630467574, + "grad_norm": 0.5418159688611615, + "learning_rate": 5.399698340874812e-05, + "loss": 0.4949, + "step": 895 + }, + { + "epoch": 0.3378582202111614, + "grad_norm": 0.6354714874187987, + "learning_rate": 5.405731523378583e-05, + "loss": 0.4894, + "step": 896 + }, + { + "epoch": 0.3382352941176471, + "grad_norm": 0.6399758636536206, + "learning_rate": 5.411764705882354e-05, + "loss": 0.5003, + "step": 897 + }, + { + "epoch": 0.33861236802413275, + "grad_norm": 0.659092290595574, + "learning_rate": 5.4177978883861244e-05, + "loss": 0.4911, + "step": 898 + }, + { + "epoch": 0.3389894419306184, + "grad_norm": 0.7287721365140515, + "learning_rate": 5.423831070889895e-05, + "loss": 0.4855, + "step": 899 + }, + { + "epoch": 0.3393665158371041, + "grad_norm": 0.8024780672906126, + "learning_rate": 5.429864253393666e-05, + "loss": 0.5093, + "step": 900 + }, + { + "epoch": 0.33974358974358976, + "grad_norm": 1.0010513104044962, + "learning_rate": 5.435897435897437e-05, + "loss": 0.5011, + "step": 901 + }, + { + "epoch": 0.3401206636500754, + "grad_norm": 1.2071006473032466, + "learning_rate": 5.441930618401207e-05, + "loss": 0.4909, + "step": 902 + }, + { + "epoch": 0.3404977375565611, + "grad_norm": 0.7573351091660278, + "learning_rate": 5.447963800904978e-05, + "loss": 0.5176, + "step": 903 + }, + { + "epoch": 0.34087481146304677, + "grad_norm": 0.6994047607545065, + "learning_rate": 5.453996983408749e-05, + "loss": 0.494, + "step": 904 + }, + { + "epoch": 0.34125188536953244, + "grad_norm": 0.8811080471819026, + "learning_rate": 5.4600301659125196e-05, + "loss": 0.4977, + "step": 905 + }, + { + "epoch": 0.3416289592760181, + "grad_norm": 0.8943840476534125, + "learning_rate": 5.46606334841629e-05, + "loss": 0.4919, + "step": 906 + }, + { + "epoch": 0.3420060331825038, + "grad_norm": 0.9612995631910768, + "learning_rate": 5.4720965309200607e-05, + "loss": 0.4829, + "step": 907 + }, + { + "epoch": 0.34238310708898945, + "grad_norm": 0.9599385895619331, + "learning_rate": 5.478129713423832e-05, + "loss": 0.4876, + "step": 908 + }, + { + "epoch": 0.3427601809954751, + "grad_norm": 0.796605772861369, + "learning_rate": 5.4841628959276024e-05, + "loss": 0.5007, + "step": 909 + }, + { + "epoch": 0.3431372549019608, + "grad_norm": 0.7206176086879026, + "learning_rate": 5.490196078431373e-05, + "loss": 0.5045, + "step": 910 + }, + { + "epoch": 0.34351432880844646, + "grad_norm": 0.8620240118274003, + "learning_rate": 5.4962292609351435e-05, + "loss": 0.5061, + "step": 911 + }, + { + "epoch": 0.3438914027149321, + "grad_norm": 1.1343997407781403, + "learning_rate": 5.502262443438915e-05, + "loss": 0.4906, + "step": 912 + }, + { + "epoch": 0.3442684766214178, + "grad_norm": 0.9403070974557919, + "learning_rate": 5.508295625942685e-05, + "loss": 0.491, + "step": 913 + }, + { + "epoch": 0.34464555052790347, + "grad_norm": 0.9431003736075959, + "learning_rate": 5.514328808446456e-05, + "loss": 0.4991, + "step": 914 + }, + { + "epoch": 0.34502262443438914, + "grad_norm": 0.999751727646435, + "learning_rate": 5.5203619909502264e-05, + "loss": 0.4997, + "step": 915 + }, + { + "epoch": 0.3453996983408748, + "grad_norm": 1.000195274910837, + "learning_rate": 5.5263951734539976e-05, + "loss": 0.4888, + "step": 916 + }, + { + "epoch": 0.3457767722473605, + "grad_norm": 0.9623382732034219, + "learning_rate": 5.532428355957768e-05, + "loss": 0.4796, + "step": 917 + }, + { + "epoch": 0.34615384615384615, + "grad_norm": 0.9490476784758983, + "learning_rate": 5.538461538461539e-05, + "loss": 0.4868, + "step": 918 + }, + { + "epoch": 0.3465309200603318, + "grad_norm": 0.9891852790189476, + "learning_rate": 5.544494720965309e-05, + "loss": 0.498, + "step": 919 + }, + { + "epoch": 0.3469079939668175, + "grad_norm": 1.061276077661698, + "learning_rate": 5.5505279034690805e-05, + "loss": 0.5154, + "step": 920 + }, + { + "epoch": 0.34728506787330315, + "grad_norm": 0.941377180671855, + "learning_rate": 5.556561085972851e-05, + "loss": 0.4978, + "step": 921 + }, + { + "epoch": 0.3476621417797888, + "grad_norm": 0.8020054334787672, + "learning_rate": 5.5625942684766216e-05, + "loss": 0.4854, + "step": 922 + }, + { + "epoch": 0.3480392156862745, + "grad_norm": 0.711265817078838, + "learning_rate": 5.568627450980392e-05, + "loss": 0.5046, + "step": 923 + }, + { + "epoch": 0.34841628959276016, + "grad_norm": 0.6591460584984773, + "learning_rate": 5.5746606334841633e-05, + "loss": 0.4908, + "step": 924 + }, + { + "epoch": 0.34879336349924583, + "grad_norm": 0.6623426296341324, + "learning_rate": 5.580693815987934e-05, + "loss": 0.4924, + "step": 925 + }, + { + "epoch": 0.3491704374057315, + "grad_norm": 0.7493751599068856, + "learning_rate": 5.5867269984917044e-05, + "loss": 0.4895, + "step": 926 + }, + { + "epoch": 0.3495475113122172, + "grad_norm": 0.8412826915228752, + "learning_rate": 5.592760180995475e-05, + "loss": 0.5001, + "step": 927 + }, + { + "epoch": 0.34992458521870284, + "grad_norm": 0.9198354537169514, + "learning_rate": 5.598793363499246e-05, + "loss": 0.4896, + "step": 928 + }, + { + "epoch": 0.3503016591251885, + "grad_norm": 0.8368228530487816, + "learning_rate": 5.604826546003017e-05, + "loss": 0.4789, + "step": 929 + }, + { + "epoch": 0.3506787330316742, + "grad_norm": 0.7930912105899031, + "learning_rate": 5.610859728506787e-05, + "loss": 0.4834, + "step": 930 + }, + { + "epoch": 0.35105580693815985, + "grad_norm": 0.8349643141766847, + "learning_rate": 5.616892911010558e-05, + "loss": 0.496, + "step": 931 + }, + { + "epoch": 0.3514328808446455, + "grad_norm": 0.8565155318956049, + "learning_rate": 5.622926093514329e-05, + "loss": 0.4886, + "step": 932 + }, + { + "epoch": 0.35180995475113125, + "grad_norm": 0.8731333215961097, + "learning_rate": 5.6289592760181e-05, + "loss": 0.4868, + "step": 933 + }, + { + "epoch": 0.3521870286576169, + "grad_norm": 0.9786711668430589, + "learning_rate": 5.634992458521871e-05, + "loss": 0.4828, + "step": 934 + }, + { + "epoch": 0.3525641025641026, + "grad_norm": 1.1322494235991496, + "learning_rate": 5.641025641025642e-05, + "loss": 0.5122, + "step": 935 + }, + { + "epoch": 0.35294117647058826, + "grad_norm": 0.7027046393012663, + "learning_rate": 5.6470588235294126e-05, + "loss": 0.4997, + "step": 936 + }, + { + "epoch": 0.35331825037707393, + "grad_norm": 0.6760824649782298, + "learning_rate": 5.653092006033183e-05, + "loss": 0.4943, + "step": 937 + }, + { + "epoch": 0.3536953242835596, + "grad_norm": 0.9655487888058569, + "learning_rate": 5.659125188536954e-05, + "loss": 0.4934, + "step": 938 + }, + { + "epoch": 0.35407239819004527, + "grad_norm": 1.038089318109261, + "learning_rate": 5.665158371040725e-05, + "loss": 0.4795, + "step": 939 + }, + { + "epoch": 0.35444947209653094, + "grad_norm": 0.6971847835010317, + "learning_rate": 5.6711915535444955e-05, + "loss": 0.4999, + "step": 940 + }, + { + "epoch": 0.3548265460030166, + "grad_norm": 0.6453457708461844, + "learning_rate": 5.677224736048266e-05, + "loss": 0.4982, + "step": 941 + }, + { + "epoch": 0.3552036199095023, + "grad_norm": 0.7427417309011406, + "learning_rate": 5.6832579185520366e-05, + "loss": 0.4954, + "step": 942 + }, + { + "epoch": 0.35558069381598795, + "grad_norm": 0.6468907261103926, + "learning_rate": 5.689291101055808e-05, + "loss": 0.4903, + "step": 943 + }, + { + "epoch": 0.3559577677224736, + "grad_norm": 0.5886873990770067, + "learning_rate": 5.6953242835595784e-05, + "loss": 0.4807, + "step": 944 + }, + { + "epoch": 0.3563348416289593, + "grad_norm": 0.6243010586564788, + "learning_rate": 5.701357466063349e-05, + "loss": 0.4994, + "step": 945 + }, + { + "epoch": 0.35671191553544496, + "grad_norm": 0.7016062975171797, + "learning_rate": 5.7073906485671195e-05, + "loss": 0.5028, + "step": 946 + }, + { + "epoch": 0.3570889894419306, + "grad_norm": 0.54924341678057, + "learning_rate": 5.713423831070891e-05, + "loss": 0.4974, + "step": 947 + }, + { + "epoch": 0.3574660633484163, + "grad_norm": 0.4747262649825036, + "learning_rate": 5.719457013574661e-05, + "loss": 0.4901, + "step": 948 + }, + { + "epoch": 0.35784313725490197, + "grad_norm": 0.7905522886486742, + "learning_rate": 5.725490196078432e-05, + "loss": 0.4836, + "step": 949 + }, + { + "epoch": 0.35822021116138764, + "grad_norm": 0.8467048207265504, + "learning_rate": 5.731523378582202e-05, + "loss": 0.4787, + "step": 950 + }, + { + "epoch": 0.3585972850678733, + "grad_norm": 0.774234413660124, + "learning_rate": 5.7375565610859735e-05, + "loss": 0.4967, + "step": 951 + }, + { + "epoch": 0.358974358974359, + "grad_norm": 0.9330162763371682, + "learning_rate": 5.743589743589744e-05, + "loss": 0.4975, + "step": 952 + }, + { + "epoch": 0.35935143288084465, + "grad_norm": 1.0723723182012748, + "learning_rate": 5.7496229260935146e-05, + "loss": 0.4949, + "step": 953 + }, + { + "epoch": 0.3597285067873303, + "grad_norm": 0.9213014433864551, + "learning_rate": 5.755656108597285e-05, + "loss": 0.4901, + "step": 954 + }, + { + "epoch": 0.360105580693816, + "grad_norm": 1.00569177846033, + "learning_rate": 5.7616892911010564e-05, + "loss": 0.4968, + "step": 955 + }, + { + "epoch": 0.36048265460030166, + "grad_norm": 1.0024194083418525, + "learning_rate": 5.767722473604827e-05, + "loss": 0.4934, + "step": 956 + }, + { + "epoch": 0.3608597285067873, + "grad_norm": 0.8977814405717678, + "learning_rate": 5.7737556561085975e-05, + "loss": 0.4933, + "step": 957 + }, + { + "epoch": 0.361236802413273, + "grad_norm": 0.9163870958715572, + "learning_rate": 5.779788838612368e-05, + "loss": 0.5008, + "step": 958 + }, + { + "epoch": 0.36161387631975866, + "grad_norm": 1.0096381827145562, + "learning_rate": 5.785822021116139e-05, + "loss": 0.5019, + "step": 959 + }, + { + "epoch": 0.36199095022624433, + "grad_norm": 0.9038071746162362, + "learning_rate": 5.79185520361991e-05, + "loss": 0.4915, + "step": 960 + }, + { + "epoch": 0.36236802413273, + "grad_norm": 0.6314509657883325, + "learning_rate": 5.7978883861236804e-05, + "loss": 0.4895, + "step": 961 + }, + { + "epoch": 0.3627450980392157, + "grad_norm": 0.4892036653211975, + "learning_rate": 5.8039215686274516e-05, + "loss": 0.4926, + "step": 962 + }, + { + "epoch": 0.36312217194570134, + "grad_norm": 0.6307948357557274, + "learning_rate": 5.809954751131222e-05, + "loss": 0.4842, + "step": 963 + }, + { + "epoch": 0.363499245852187, + "grad_norm": 0.7609931725053473, + "learning_rate": 5.815987933634993e-05, + "loss": 0.4848, + "step": 964 + }, + { + "epoch": 0.3638763197586727, + "grad_norm": 0.7878340845948, + "learning_rate": 5.822021116138763e-05, + "loss": 0.4848, + "step": 965 + }, + { + "epoch": 0.36425339366515835, + "grad_norm": 0.8135770961577151, + "learning_rate": 5.8280542986425345e-05, + "loss": 0.4879, + "step": 966 + }, + { + "epoch": 0.364630467571644, + "grad_norm": 0.8554548465205372, + "learning_rate": 5.834087481146305e-05, + "loss": 0.4893, + "step": 967 + }, + { + "epoch": 0.3650075414781297, + "grad_norm": 0.9265978542100496, + "learning_rate": 5.8401206636500756e-05, + "loss": 0.4962, + "step": 968 + }, + { + "epoch": 0.36538461538461536, + "grad_norm": 1.0537618017594306, + "learning_rate": 5.846153846153846e-05, + "loss": 0.4865, + "step": 969 + }, + { + "epoch": 0.36576168929110103, + "grad_norm": 0.9654525068210636, + "learning_rate": 5.852187028657617e-05, + "loss": 0.4975, + "step": 970 + }, + { + "epoch": 0.3661387631975867, + "grad_norm": 0.8638635810842284, + "learning_rate": 5.858220211161388e-05, + "loss": 0.498, + "step": 971 + }, + { + "epoch": 0.3665158371040724, + "grad_norm": 0.8203594500826578, + "learning_rate": 5.8642533936651584e-05, + "loss": 0.4984, + "step": 972 + }, + { + "epoch": 0.36689291101055804, + "grad_norm": 0.7653209985342293, + "learning_rate": 5.870286576168929e-05, + "loss": 0.5086, + "step": 973 + }, + { + "epoch": 0.36726998491704377, + "grad_norm": 0.9706349505677377, + "learning_rate": 5.876319758672701e-05, + "loss": 0.4923, + "step": 974 + }, + { + "epoch": 0.36764705882352944, + "grad_norm": 1.161528672385691, + "learning_rate": 5.8823529411764714e-05, + "loss": 0.5096, + "step": 975 + }, + { + "epoch": 0.3680241327300151, + "grad_norm": 0.6423574327561705, + "learning_rate": 5.888386123680242e-05, + "loss": 0.4903, + "step": 976 + }, + { + "epoch": 0.3684012066365008, + "grad_norm": 0.7034046834482739, + "learning_rate": 5.894419306184013e-05, + "loss": 0.4939, + "step": 977 + }, + { + "epoch": 0.36877828054298645, + "grad_norm": 0.9317506149400315, + "learning_rate": 5.900452488687784e-05, + "loss": 0.4998, + "step": 978 + }, + { + "epoch": 0.3691553544494721, + "grad_norm": 1.0959249399317395, + "learning_rate": 5.906485671191554e-05, + "loss": 0.4963, + "step": 979 + }, + { + "epoch": 0.3695324283559578, + "grad_norm": 0.9173257631980015, + "learning_rate": 5.912518853695325e-05, + "loss": 0.481, + "step": 980 + }, + { + "epoch": 0.36990950226244346, + "grad_norm": 0.710993531668539, + "learning_rate": 5.918552036199096e-05, + "loss": 0.4923, + "step": 981 + }, + { + "epoch": 0.3702865761689291, + "grad_norm": 0.6775436145838276, + "learning_rate": 5.9245852187028666e-05, + "loss": 0.49, + "step": 982 + }, + { + "epoch": 0.3706636500754148, + "grad_norm": 0.8193407239387279, + "learning_rate": 5.930618401206637e-05, + "loss": 0.4855, + "step": 983 + }, + { + "epoch": 0.37104072398190047, + "grad_norm": 0.7292365104347951, + "learning_rate": 5.936651583710408e-05, + "loss": 0.4878, + "step": 984 + }, + { + "epoch": 0.37141779788838614, + "grad_norm": 0.6085072502236696, + "learning_rate": 5.942684766214179e-05, + "loss": 0.4809, + "step": 985 + }, + { + "epoch": 0.3717948717948718, + "grad_norm": 0.5837530065779712, + "learning_rate": 5.9487179487179495e-05, + "loss": 0.4903, + "step": 986 + }, + { + "epoch": 0.3721719457013575, + "grad_norm": 0.7137217201726928, + "learning_rate": 5.95475113122172e-05, + "loss": 0.4788, + "step": 987 + }, + { + "epoch": 0.37254901960784315, + "grad_norm": 0.7420668048288089, + "learning_rate": 5.9607843137254906e-05, + "loss": 0.4918, + "step": 988 + }, + { + "epoch": 0.3729260935143288, + "grad_norm": 0.7598110090313601, + "learning_rate": 5.966817496229262e-05, + "loss": 0.4979, + "step": 989 + }, + { + "epoch": 0.3733031674208145, + "grad_norm": 0.9814933221333819, + "learning_rate": 5.9728506787330323e-05, + "loss": 0.4858, + "step": 990 + }, + { + "epoch": 0.37368024132730016, + "grad_norm": 1.27456411326936, + "learning_rate": 5.978883861236803e-05, + "loss": 0.4903, + "step": 991 + }, + { + "epoch": 0.3740573152337858, + "grad_norm": 0.708501606195646, + "learning_rate": 5.9849170437405734e-05, + "loss": 0.4873, + "step": 992 + }, + { + "epoch": 0.3744343891402715, + "grad_norm": 0.574951592977524, + "learning_rate": 5.990950226244345e-05, + "loss": 0.4839, + "step": 993 + }, + { + "epoch": 0.37481146304675717, + "grad_norm": 0.5638795005946449, + "learning_rate": 5.996983408748115e-05, + "loss": 0.5033, + "step": 994 + }, + { + "epoch": 0.37518853695324283, + "grad_norm": 0.7878125237627697, + "learning_rate": 6.003016591251886e-05, + "loss": 0.4953, + "step": 995 + }, + { + "epoch": 0.3755656108597285, + "grad_norm": 0.9869915533893524, + "learning_rate": 6.009049773755656e-05, + "loss": 0.4909, + "step": 996 + }, + { + "epoch": 0.3759426847662142, + "grad_norm": 1.0554994911906126, + "learning_rate": 6.0150829562594275e-05, + "loss": 0.4865, + "step": 997 + }, + { + "epoch": 0.37631975867269984, + "grad_norm": 0.9290269326331366, + "learning_rate": 6.021116138763198e-05, + "loss": 0.4828, + "step": 998 + }, + { + "epoch": 0.3766968325791855, + "grad_norm": 0.727479031865973, + "learning_rate": 6.0271493212669686e-05, + "loss": 0.4935, + "step": 999 + }, + { + "epoch": 0.3770739064856712, + "grad_norm": 0.6262664252716501, + "learning_rate": 6.033182503770739e-05, + "loss": 0.5055, + "step": 1000 + }, + { + "epoch": 0.37745098039215685, + "grad_norm": 0.694043635220054, + "learning_rate": 6.0392156862745104e-05, + "loss": 0.5079, + "step": 1001 + }, + { + "epoch": 0.3778280542986425, + "grad_norm": 0.7691653795774377, + "learning_rate": 6.045248868778281e-05, + "loss": 0.4864, + "step": 1002 + }, + { + "epoch": 0.3782051282051282, + "grad_norm": 0.8657671618391574, + "learning_rate": 6.0512820512820515e-05, + "loss": 0.5101, + "step": 1003 + }, + { + "epoch": 0.37858220211161386, + "grad_norm": 0.9341795642203037, + "learning_rate": 6.057315233785822e-05, + "loss": 0.509, + "step": 1004 + }, + { + "epoch": 0.37895927601809953, + "grad_norm": 1.0030075556460847, + "learning_rate": 6.063348416289593e-05, + "loss": 0.4894, + "step": 1005 + }, + { + "epoch": 0.3793363499245852, + "grad_norm": 1.0774790717424447, + "learning_rate": 6.069381598793364e-05, + "loss": 0.4907, + "step": 1006 + }, + { + "epoch": 0.3797134238310709, + "grad_norm": 1.1300655360207987, + "learning_rate": 6.0754147812971344e-05, + "loss": 0.4994, + "step": 1007 + }, + { + "epoch": 0.38009049773755654, + "grad_norm": 0.8373001591004682, + "learning_rate": 6.081447963800905e-05, + "loss": 0.4929, + "step": 1008 + }, + { + "epoch": 0.3804675716440422, + "grad_norm": 0.7126166744374712, + "learning_rate": 6.087481146304676e-05, + "loss": 0.4893, + "step": 1009 + }, + { + "epoch": 0.3808446455505279, + "grad_norm": 0.8220206635237941, + "learning_rate": 6.093514328808447e-05, + "loss": 0.494, + "step": 1010 + }, + { + "epoch": 0.38122171945701355, + "grad_norm": 1.0480935109833591, + "learning_rate": 6.099547511312217e-05, + "loss": 0.4976, + "step": 1011 + }, + { + "epoch": 0.3815987933634992, + "grad_norm": 1.0678645613837394, + "learning_rate": 6.105580693815988e-05, + "loss": 0.4908, + "step": 1012 + }, + { + "epoch": 0.3819758672699849, + "grad_norm": 0.7475360475682536, + "learning_rate": 6.111613876319759e-05, + "loss": 0.4731, + "step": 1013 + }, + { + "epoch": 0.38235294117647056, + "grad_norm": 0.7281258847633267, + "learning_rate": 6.117647058823529e-05, + "loss": 0.4937, + "step": 1014 + }, + { + "epoch": 0.38273001508295623, + "grad_norm": 0.926968226345761, + "learning_rate": 6.1236802413273e-05, + "loss": 0.5058, + "step": 1015 + }, + { + "epoch": 0.38310708898944196, + "grad_norm": 1.0137399059658696, + "learning_rate": 6.129713423831071e-05, + "loss": 0.4975, + "step": 1016 + }, + { + "epoch": 0.3834841628959276, + "grad_norm": 0.8384164995158832, + "learning_rate": 6.135746606334843e-05, + "loss": 0.4866, + "step": 1017 + }, + { + "epoch": 0.3838612368024133, + "grad_norm": 0.6924998355673924, + "learning_rate": 6.141779788838614e-05, + "loss": 0.4935, + "step": 1018 + }, + { + "epoch": 0.38423831070889897, + "grad_norm": 0.7658021011617201, + "learning_rate": 6.147812971342384e-05, + "loss": 0.4987, + "step": 1019 + }, + { + "epoch": 0.38461538461538464, + "grad_norm": 0.729485632449907, + "learning_rate": 6.153846153846155e-05, + "loss": 0.485, + "step": 1020 + }, + { + "epoch": 0.3849924585218703, + "grad_norm": 0.61873018241436, + "learning_rate": 6.159879336349925e-05, + "loss": 0.49, + "step": 1021 + }, + { + "epoch": 0.385369532428356, + "grad_norm": 0.5844911400113839, + "learning_rate": 6.165912518853696e-05, + "loss": 0.4857, + "step": 1022 + }, + { + "epoch": 0.38574660633484165, + "grad_norm": 0.47977904658416226, + "learning_rate": 6.171945701357467e-05, + "loss": 0.4915, + "step": 1023 + }, + { + "epoch": 0.3861236802413273, + "grad_norm": 0.4067026091215392, + "learning_rate": 6.177978883861237e-05, + "loss": 0.4994, + "step": 1024 + }, + { + "epoch": 0.386500754147813, + "grad_norm": 0.47474288419921223, + "learning_rate": 6.184012066365008e-05, + "loss": 0.4806, + "step": 1025 + }, + { + "epoch": 0.38687782805429866, + "grad_norm": 0.4450302087844997, + "learning_rate": 6.19004524886878e-05, + "loss": 0.4909, + "step": 1026 + }, + { + "epoch": 0.3872549019607843, + "grad_norm": 0.33983843485135357, + "learning_rate": 6.19607843137255e-05, + "loss": 0.4755, + "step": 1027 + }, + { + "epoch": 0.38763197586727, + "grad_norm": 0.36983780636389696, + "learning_rate": 6.20211161387632e-05, + "loss": 0.4872, + "step": 1028 + }, + { + "epoch": 0.38800904977375567, + "grad_norm": 0.3820546680699495, + "learning_rate": 6.208144796380092e-05, + "loss": 0.4978, + "step": 1029 + }, + { + "epoch": 0.38838612368024134, + "grad_norm": 0.5199502085180359, + "learning_rate": 6.214177978883862e-05, + "loss": 0.4983, + "step": 1030 + }, + { + "epoch": 0.388763197586727, + "grad_norm": 0.6961975925872178, + "learning_rate": 6.220211161387633e-05, + "loss": 0.4892, + "step": 1031 + }, + { + "epoch": 0.3891402714932127, + "grad_norm": 0.6990075018210086, + "learning_rate": 6.226244343891403e-05, + "loss": 0.4777, + "step": 1032 + }, + { + "epoch": 0.38951734539969834, + "grad_norm": 0.662359092095744, + "learning_rate": 6.232277526395174e-05, + "loss": 0.4916, + "step": 1033 + }, + { + "epoch": 0.389894419306184, + "grad_norm": 0.8474305654788516, + "learning_rate": 6.238310708898945e-05, + "loss": 0.4998, + "step": 1034 + }, + { + "epoch": 0.3902714932126697, + "grad_norm": 1.0030744930772226, + "learning_rate": 6.244343891402715e-05, + "loss": 0.4795, + "step": 1035 + }, + { + "epoch": 0.39064856711915535, + "grad_norm": 1.4121506110001654, + "learning_rate": 6.250377073906486e-05, + "loss": 0.4914, + "step": 1036 + }, + { + "epoch": 0.391025641025641, + "grad_norm": 0.498882878306055, + "learning_rate": 6.256410256410258e-05, + "loss": 0.4723, + "step": 1037 + }, + { + "epoch": 0.3914027149321267, + "grad_norm": 0.9666017797542729, + "learning_rate": 6.262443438914027e-05, + "loss": 0.4793, + "step": 1038 + }, + { + "epoch": 0.39177978883861236, + "grad_norm": 1.6439753152480379, + "learning_rate": 6.268476621417799e-05, + "loss": 0.4994, + "step": 1039 + }, + { + "epoch": 0.39215686274509803, + "grad_norm": 0.6064051835639982, + "learning_rate": 6.274509803921569e-05, + "loss": 0.481, + "step": 1040 + }, + { + "epoch": 0.3925339366515837, + "grad_norm": 1.774265828803087, + "learning_rate": 6.28054298642534e-05, + "loss": 0.5005, + "step": 1041 + }, + { + "epoch": 0.3929110105580694, + "grad_norm": 0.7332032557414802, + "learning_rate": 6.286576168929111e-05, + "loss": 0.4864, + "step": 1042 + }, + { + "epoch": 0.39328808446455504, + "grad_norm": 1.9198933693088678, + "learning_rate": 6.292609351432881e-05, + "loss": 0.5023, + "step": 1043 + }, + { + "epoch": 0.3936651583710407, + "grad_norm": 1.2938275357384, + "learning_rate": 6.298642533936652e-05, + "loss": 0.5118, + "step": 1044 + }, + { + "epoch": 0.3940422322775264, + "grad_norm": 1.5933996005314923, + "learning_rate": 6.304675716440423e-05, + "loss": 0.5014, + "step": 1045 + }, + { + "epoch": 0.39441930618401205, + "grad_norm": 1.466841981347575, + "learning_rate": 6.310708898944193e-05, + "loss": 0.5195, + "step": 1046 + }, + { + "epoch": 0.3947963800904977, + "grad_norm": 1.013630802051211, + "learning_rate": 6.316742081447964e-05, + "loss": 0.4911, + "step": 1047 + }, + { + "epoch": 0.3951734539969834, + "grad_norm": 1.106370114785758, + "learning_rate": 6.322775263951734e-05, + "loss": 0.4941, + "step": 1048 + }, + { + "epoch": 0.39555052790346906, + "grad_norm": 0.8110326719633039, + "learning_rate": 6.328808446455505e-05, + "loss": 0.5017, + "step": 1049 + }, + { + "epoch": 0.39592760180995473, + "grad_norm": 0.8164183509825681, + "learning_rate": 6.334841628959277e-05, + "loss": 0.4921, + "step": 1050 + }, + { + "epoch": 0.3963046757164404, + "grad_norm": 0.7120348529312289, + "learning_rate": 6.340874811463047e-05, + "loss": 0.4855, + "step": 1051 + }, + { + "epoch": 0.39668174962292607, + "grad_norm": 0.6872969786223055, + "learning_rate": 6.346907993966818e-05, + "loss": 0.475, + "step": 1052 + }, + { + "epoch": 0.39705882352941174, + "grad_norm": 0.6949175251767835, + "learning_rate": 6.352941176470589e-05, + "loss": 0.4922, + "step": 1053 + }, + { + "epoch": 0.3974358974358974, + "grad_norm": 0.5808677143795777, + "learning_rate": 6.358974358974359e-05, + "loss": 0.4841, + "step": 1054 + }, + { + "epoch": 0.3978129713423831, + "grad_norm": 0.6802849786016265, + "learning_rate": 6.36500754147813e-05, + "loss": 0.4762, + "step": 1055 + }, + { + "epoch": 0.39819004524886875, + "grad_norm": 0.5106896521413651, + "learning_rate": 6.3710407239819e-05, + "loss": 0.4913, + "step": 1056 + }, + { + "epoch": 0.3985671191553545, + "grad_norm": 0.8006085890269992, + "learning_rate": 6.377073906485673e-05, + "loss": 0.494, + "step": 1057 + }, + { + "epoch": 0.39894419306184015, + "grad_norm": 0.5365342141443274, + "learning_rate": 6.383107088989442e-05, + "loss": 0.4908, + "step": 1058 + }, + { + "epoch": 0.3993212669683258, + "grad_norm": 0.663864929117036, + "learning_rate": 6.389140271493214e-05, + "loss": 0.5026, + "step": 1059 + }, + { + "epoch": 0.3996983408748115, + "grad_norm": 0.6885491861775999, + "learning_rate": 6.395173453996985e-05, + "loss": 0.4979, + "step": 1060 + }, + { + "epoch": 0.40007541478129716, + "grad_norm": 0.6907501512820813, + "learning_rate": 6.401206636500755e-05, + "loss": 0.4791, + "step": 1061 + }, + { + "epoch": 0.4004524886877828, + "grad_norm": 0.7907462886986237, + "learning_rate": 6.407239819004526e-05, + "loss": 0.4906, + "step": 1062 + }, + { + "epoch": 0.4008295625942685, + "grad_norm": 1.0443655253594095, + "learning_rate": 6.413273001508296e-05, + "loss": 0.4863, + "step": 1063 + }, + { + "epoch": 0.40120663650075417, + "grad_norm": 1.0560675502899666, + "learning_rate": 6.419306184012067e-05, + "loss": 0.4873, + "step": 1064 + }, + { + "epoch": 0.40158371040723984, + "grad_norm": 1.0022823825948426, + "learning_rate": 6.425339366515838e-05, + "loss": 0.4887, + "step": 1065 + }, + { + "epoch": 0.4019607843137255, + "grad_norm": 1.0059322831388684, + "learning_rate": 6.431372549019608e-05, + "loss": 0.4751, + "step": 1066 + }, + { + "epoch": 0.4023378582202112, + "grad_norm": 0.9937623740142493, + "learning_rate": 6.43740573152338e-05, + "loss": 0.4963, + "step": 1067 + }, + { + "epoch": 0.40271493212669685, + "grad_norm": 0.953353042206792, + "learning_rate": 6.44343891402715e-05, + "loss": 0.4929, + "step": 1068 + }, + { + "epoch": 0.4030920060331825, + "grad_norm": 0.8770206447687253, + "learning_rate": 6.44947209653092e-05, + "loss": 0.4912, + "step": 1069 + }, + { + "epoch": 0.4034690799396682, + "grad_norm": 0.7826688295518346, + "learning_rate": 6.455505279034692e-05, + "loss": 0.4844, + "step": 1070 + }, + { + "epoch": 0.40384615384615385, + "grad_norm": 0.7383235409295061, + "learning_rate": 6.461538461538462e-05, + "loss": 0.4973, + "step": 1071 + }, + { + "epoch": 0.4042232277526395, + "grad_norm": 0.9778077553522683, + "learning_rate": 6.467571644042233e-05, + "loss": 0.5005, + "step": 1072 + }, + { + "epoch": 0.4046003016591252, + "grad_norm": 1.0903145745311016, + "learning_rate": 6.473604826546004e-05, + "loss": 0.4956, + "step": 1073 + }, + { + "epoch": 0.40497737556561086, + "grad_norm": 0.7357848489707534, + "learning_rate": 6.479638009049774e-05, + "loss": 0.4769, + "step": 1074 + }, + { + "epoch": 0.40535444947209653, + "grad_norm": 0.5783577323571955, + "learning_rate": 6.485671191553545e-05, + "loss": 0.4982, + "step": 1075 + }, + { + "epoch": 0.4057315233785822, + "grad_norm": 0.6244386397880324, + "learning_rate": 6.491704374057316e-05, + "loss": 0.488, + "step": 1076 + }, + { + "epoch": 0.4061085972850679, + "grad_norm": 0.6960731062892747, + "learning_rate": 6.497737556561086e-05, + "loss": 0.4857, + "step": 1077 + }, + { + "epoch": 0.40648567119155354, + "grad_norm": 0.722212976368741, + "learning_rate": 6.503770739064857e-05, + "loss": 0.4841, + "step": 1078 + }, + { + "epoch": 0.4068627450980392, + "grad_norm": 0.7522602825807809, + "learning_rate": 6.509803921568627e-05, + "loss": 0.4875, + "step": 1079 + }, + { + "epoch": 0.4072398190045249, + "grad_norm": 0.6894783990419834, + "learning_rate": 6.515837104072399e-05, + "loss": 0.4933, + "step": 1080 + }, + { + "epoch": 0.40761689291101055, + "grad_norm": 0.6241892880116224, + "learning_rate": 6.52187028657617e-05, + "loss": 0.5035, + "step": 1081 + }, + { + "epoch": 0.4079939668174962, + "grad_norm": 0.5551767186636406, + "learning_rate": 6.52790346907994e-05, + "loss": 0.4973, + "step": 1082 + }, + { + "epoch": 0.4083710407239819, + "grad_norm": 0.5727523077084846, + "learning_rate": 6.533936651583711e-05, + "loss": 0.492, + "step": 1083 + }, + { + "epoch": 0.40874811463046756, + "grad_norm": 0.5551622633439878, + "learning_rate": 6.539969834087482e-05, + "loss": 0.4853, + "step": 1084 + }, + { + "epoch": 0.40912518853695323, + "grad_norm": 0.5513047373063492, + "learning_rate": 6.546003016591252e-05, + "loss": 0.4886, + "step": 1085 + }, + { + "epoch": 0.4095022624434389, + "grad_norm": 0.6708228107620624, + "learning_rate": 6.552036199095023e-05, + "loss": 0.4907, + "step": 1086 + }, + { + "epoch": 0.4098793363499246, + "grad_norm": 0.725054312351277, + "learning_rate": 6.558069381598793e-05, + "loss": 0.4885, + "step": 1087 + }, + { + "epoch": 0.41025641025641024, + "grad_norm": 0.8490822759946116, + "learning_rate": 6.564102564102564e-05, + "loss": 0.485, + "step": 1088 + }, + { + "epoch": 0.4106334841628959, + "grad_norm": 1.1095643263099026, + "learning_rate": 6.570135746606336e-05, + "loss": 0.4951, + "step": 1089 + }, + { + "epoch": 0.4110105580693816, + "grad_norm": 0.8855925326859895, + "learning_rate": 6.576168929110105e-05, + "loss": 0.4757, + "step": 1090 + }, + { + "epoch": 0.41138763197586725, + "grad_norm": 0.6703812637294, + "learning_rate": 6.582202111613877e-05, + "loss": 0.481, + "step": 1091 + }, + { + "epoch": 0.4117647058823529, + "grad_norm": 0.5445896041385985, + "learning_rate": 6.588235294117648e-05, + "loss": 0.4966, + "step": 1092 + }, + { + "epoch": 0.4121417797888386, + "grad_norm": 0.47616031346951965, + "learning_rate": 6.594268476621418e-05, + "loss": 0.4863, + "step": 1093 + }, + { + "epoch": 0.41251885369532426, + "grad_norm": 0.5846420146036593, + "learning_rate": 6.600301659125189e-05, + "loss": 0.4917, + "step": 1094 + }, + { + "epoch": 0.41289592760180993, + "grad_norm": 0.679226646082533, + "learning_rate": 6.606334841628959e-05, + "loss": 0.5044, + "step": 1095 + }, + { + "epoch": 0.4132730015082956, + "grad_norm": 0.7559717778844453, + "learning_rate": 6.61236802413273e-05, + "loss": 0.494, + "step": 1096 + }, + { + "epoch": 0.41365007541478127, + "grad_norm": 0.941243996539562, + "learning_rate": 6.618401206636501e-05, + "loss": 0.4869, + "step": 1097 + }, + { + "epoch": 0.41402714932126694, + "grad_norm": 1.1708758849242278, + "learning_rate": 6.624434389140271e-05, + "loss": 0.5024, + "step": 1098 + }, + { + "epoch": 0.41440422322775267, + "grad_norm": 0.6229962687413534, + "learning_rate": 6.630467571644044e-05, + "loss": 0.4904, + "step": 1099 + }, + { + "epoch": 0.41478129713423834, + "grad_norm": 0.4803584534171727, + "learning_rate": 6.636500754147814e-05, + "loss": 0.486, + "step": 1100 + }, + { + "epoch": 0.415158371040724, + "grad_norm": 0.8523196424860839, + "learning_rate": 6.642533936651585e-05, + "loss": 0.4828, + "step": 1101 + }, + { + "epoch": 0.4155354449472097, + "grad_norm": 0.9247226156060528, + "learning_rate": 6.648567119155356e-05, + "loss": 0.4983, + "step": 1102 + }, + { + "epoch": 0.41591251885369535, + "grad_norm": 0.7575523601477657, + "learning_rate": 6.654600301659126e-05, + "loss": 0.4861, + "step": 1103 + }, + { + "epoch": 0.416289592760181, + "grad_norm": 0.6298396024096967, + "learning_rate": 6.660633484162897e-05, + "loss": 0.4871, + "step": 1104 + }, + { + "epoch": 0.4166666666666667, + "grad_norm": 0.6386705087588872, + "learning_rate": 6.666666666666667e-05, + "loss": 0.4954, + "step": 1105 + }, + { + "epoch": 0.41704374057315236, + "grad_norm": 0.6924239743839539, + "learning_rate": 6.672699849170438e-05, + "loss": 0.4843, + "step": 1106 + }, + { + "epoch": 0.417420814479638, + "grad_norm": 0.7958342233333011, + "learning_rate": 6.67873303167421e-05, + "loss": 0.4944, + "step": 1107 + }, + { + "epoch": 0.4177978883861237, + "grad_norm": 0.9868858250506829, + "learning_rate": 6.684766214177979e-05, + "loss": 0.4911, + "step": 1108 + }, + { + "epoch": 0.41817496229260936, + "grad_norm": 1.2392016746425, + "learning_rate": 6.69079939668175e-05, + "loss": 0.491, + "step": 1109 + }, + { + "epoch": 0.41855203619909503, + "grad_norm": 0.631684902826481, + "learning_rate": 6.696832579185522e-05, + "loss": 0.4851, + "step": 1110 + }, + { + "epoch": 0.4189291101055807, + "grad_norm": 0.6284390937797663, + "learning_rate": 6.702865761689292e-05, + "loss": 0.4766, + "step": 1111 + }, + { + "epoch": 0.4193061840120664, + "grad_norm": 0.7430954549339348, + "learning_rate": 6.708898944193063e-05, + "loss": 0.4851, + "step": 1112 + }, + { + "epoch": 0.41968325791855204, + "grad_norm": 0.8378012165223365, + "learning_rate": 6.714932126696833e-05, + "loss": 0.5005, + "step": 1113 + }, + { + "epoch": 0.4200603318250377, + "grad_norm": 0.88819745502195, + "learning_rate": 6.720965309200604e-05, + "loss": 0.4823, + "step": 1114 + }, + { + "epoch": 0.4204374057315234, + "grad_norm": 0.7929204799900993, + "learning_rate": 6.726998491704375e-05, + "loss": 0.4895, + "step": 1115 + }, + { + "epoch": 0.42081447963800905, + "grad_norm": 0.6479954875273558, + "learning_rate": 6.733031674208145e-05, + "loss": 0.4849, + "step": 1116 + }, + { + "epoch": 0.4211915535444947, + "grad_norm": 0.6606067572636036, + "learning_rate": 6.739064856711916e-05, + "loss": 0.4998, + "step": 1117 + }, + { + "epoch": 0.4215686274509804, + "grad_norm": 0.6910717300245863, + "learning_rate": 6.745098039215687e-05, + "loss": 0.4957, + "step": 1118 + }, + { + "epoch": 0.42194570135746606, + "grad_norm": 0.8172185293172616, + "learning_rate": 6.751131221719457e-05, + "loss": 0.493, + "step": 1119 + }, + { + "epoch": 0.42232277526395173, + "grad_norm": 0.9837895511330845, + "learning_rate": 6.757164404223229e-05, + "loss": 0.489, + "step": 1120 + }, + { + "epoch": 0.4226998491704374, + "grad_norm": 0.9889668043775006, + "learning_rate": 6.763197586726998e-05, + "loss": 0.4719, + "step": 1121 + }, + { + "epoch": 0.4230769230769231, + "grad_norm": 1.0709791921215526, + "learning_rate": 6.76923076923077e-05, + "loss": 0.4793, + "step": 1122 + }, + { + "epoch": 0.42345399698340874, + "grad_norm": 1.0260759268959565, + "learning_rate": 6.775263951734541e-05, + "loss": 0.4796, + "step": 1123 + }, + { + "epoch": 0.4238310708898944, + "grad_norm": 1.0554806689528877, + "learning_rate": 6.781297134238311e-05, + "loss": 0.4836, + "step": 1124 + }, + { + "epoch": 0.4242081447963801, + "grad_norm": 0.9470263366776898, + "learning_rate": 6.787330316742082e-05, + "loss": 0.5035, + "step": 1125 + }, + { + "epoch": 0.42458521870286575, + "grad_norm": 0.7738734249007297, + "learning_rate": 6.793363499245853e-05, + "loss": 0.4919, + "step": 1126 + }, + { + "epoch": 0.4249622926093514, + "grad_norm": 0.700783602286893, + "learning_rate": 6.799396681749623e-05, + "loss": 0.4968, + "step": 1127 + }, + { + "epoch": 0.4253393665158371, + "grad_norm": 0.7139931114301348, + "learning_rate": 6.805429864253394e-05, + "loss": 0.4811, + "step": 1128 + }, + { + "epoch": 0.42571644042232276, + "grad_norm": 0.6505684066630788, + "learning_rate": 6.811463046757164e-05, + "loss": 0.4794, + "step": 1129 + }, + { + "epoch": 0.42609351432880843, + "grad_norm": 0.49823953397642834, + "learning_rate": 6.817496229260935e-05, + "loss": 0.4886, + "step": 1130 + }, + { + "epoch": 0.4264705882352941, + "grad_norm": 0.4712016823127072, + "learning_rate": 6.823529411764707e-05, + "loss": 0.4863, + "step": 1131 + }, + { + "epoch": 0.42684766214177977, + "grad_norm": 0.6031860422041968, + "learning_rate": 6.829562594268477e-05, + "loss": 0.4891, + "step": 1132 + }, + { + "epoch": 0.42722473604826544, + "grad_norm": 0.7071446686351914, + "learning_rate": 6.835595776772248e-05, + "loss": 0.4894, + "step": 1133 + }, + { + "epoch": 0.4276018099547511, + "grad_norm": 0.6506242001579449, + "learning_rate": 6.841628959276019e-05, + "loss": 0.4826, + "step": 1134 + }, + { + "epoch": 0.4279788838612368, + "grad_norm": 0.5629466240552495, + "learning_rate": 6.847662141779789e-05, + "loss": 0.4732, + "step": 1135 + }, + { + "epoch": 0.42835595776772245, + "grad_norm": 0.5681010652977355, + "learning_rate": 6.85369532428356e-05, + "loss": 0.4893, + "step": 1136 + }, + { + "epoch": 0.4287330316742081, + "grad_norm": 0.89601594893839, + "learning_rate": 6.85972850678733e-05, + "loss": 0.4967, + "step": 1137 + }, + { + "epoch": 0.4291101055806938, + "grad_norm": 1.1615049453373074, + "learning_rate": 6.865761689291101e-05, + "loss": 0.4889, + "step": 1138 + }, + { + "epoch": 0.42948717948717946, + "grad_norm": 0.8435664154089875, + "learning_rate": 6.871794871794872e-05, + "loss": 0.4907, + "step": 1139 + }, + { + "epoch": 0.4298642533936652, + "grad_norm": 0.765813299061489, + "learning_rate": 6.877828054298644e-05, + "loss": 0.4823, + "step": 1140 + }, + { + "epoch": 0.43024132730015086, + "grad_norm": 0.676799192784314, + "learning_rate": 6.883861236802415e-05, + "loss": 0.4841, + "step": 1141 + }, + { + "epoch": 0.4306184012066365, + "grad_norm": 0.6273847110103469, + "learning_rate": 6.889894419306185e-05, + "loss": 0.4858, + "step": 1142 + }, + { + "epoch": 0.4309954751131222, + "grad_norm": 0.600127959863881, + "learning_rate": 6.895927601809956e-05, + "loss": 0.4751, + "step": 1143 + }, + { + "epoch": 0.43137254901960786, + "grad_norm": 0.7154529237734639, + "learning_rate": 6.901960784313726e-05, + "loss": 0.4887, + "step": 1144 + }, + { + "epoch": 0.43174962292609353, + "grad_norm": 0.9543354472442214, + "learning_rate": 6.907993966817497e-05, + "loss": 0.4867, + "step": 1145 + }, + { + "epoch": 0.4321266968325792, + "grad_norm": 1.1463486304333983, + "learning_rate": 6.914027149321268e-05, + "loss": 0.5004, + "step": 1146 + }, + { + "epoch": 0.4325037707390649, + "grad_norm": 0.8966146568679646, + "learning_rate": 6.920060331825038e-05, + "loss": 0.4994, + "step": 1147 + }, + { + "epoch": 0.43288084464555054, + "grad_norm": 0.675679774866756, + "learning_rate": 6.92609351432881e-05, + "loss": 0.4952, + "step": 1148 + }, + { + "epoch": 0.4332579185520362, + "grad_norm": 0.5021963246223926, + "learning_rate": 6.93212669683258e-05, + "loss": 0.4781, + "step": 1149 + }, + { + "epoch": 0.4336349924585219, + "grad_norm": 0.6216235805810283, + "learning_rate": 6.93815987933635e-05, + "loss": 0.4868, + "step": 1150 + }, + { + "epoch": 0.43401206636500755, + "grad_norm": 0.791234955204009, + "learning_rate": 6.944193061840122e-05, + "loss": 0.4827, + "step": 1151 + }, + { + "epoch": 0.4343891402714932, + "grad_norm": 0.8071008822147748, + "learning_rate": 6.950226244343892e-05, + "loss": 0.4686, + "step": 1152 + }, + { + "epoch": 0.4347662141779789, + "grad_norm": 0.7373061404111442, + "learning_rate": 6.956259426847663e-05, + "loss": 0.4789, + "step": 1153 + }, + { + "epoch": 0.43514328808446456, + "grad_norm": 0.8200370910543675, + "learning_rate": 6.962292609351434e-05, + "loss": 0.4865, + "step": 1154 + }, + { + "epoch": 0.43552036199095023, + "grad_norm": 0.9415821034813441, + "learning_rate": 6.968325791855204e-05, + "loss": 0.4787, + "step": 1155 + }, + { + "epoch": 0.4358974358974359, + "grad_norm": 1.1068640972910204, + "learning_rate": 6.974358974358975e-05, + "loss": 0.4847, + "step": 1156 + }, + { + "epoch": 0.4362745098039216, + "grad_norm": 0.952074245119849, + "learning_rate": 6.980392156862746e-05, + "loss": 0.4934, + "step": 1157 + }, + { + "epoch": 0.43665158371040724, + "grad_norm": 0.8843653158261156, + "learning_rate": 6.986425339366516e-05, + "loss": 0.4843, + "step": 1158 + }, + { + "epoch": 0.4370286576168929, + "grad_norm": 0.8891188848465101, + "learning_rate": 6.992458521870287e-05, + "loss": 0.4896, + "step": 1159 + }, + { + "epoch": 0.4374057315233786, + "grad_norm": 1.0454779080798366, + "learning_rate": 6.998491704374057e-05, + "loss": 0.4825, + "step": 1160 + }, + { + "epoch": 0.43778280542986425, + "grad_norm": 1.0739199063842861, + "learning_rate": 7.004524886877828e-05, + "loss": 0.4874, + "step": 1161 + }, + { + "epoch": 0.4381598793363499, + "grad_norm": 0.7482688050991799, + "learning_rate": 7.0105580693816e-05, + "loss": 0.4874, + "step": 1162 + }, + { + "epoch": 0.4385369532428356, + "grad_norm": 0.5707467720911787, + "learning_rate": 7.01659125188537e-05, + "loss": 0.4993, + "step": 1163 + }, + { + "epoch": 0.43891402714932126, + "grad_norm": 0.6592701628727324, + "learning_rate": 7.022624434389141e-05, + "loss": 0.486, + "step": 1164 + }, + { + "epoch": 0.43929110105580693, + "grad_norm": 0.8189649816675325, + "learning_rate": 7.028657616892912e-05, + "loss": 0.4841, + "step": 1165 + }, + { + "epoch": 0.4396681749622926, + "grad_norm": 0.8546473970126792, + "learning_rate": 7.034690799396682e-05, + "loss": 0.4918, + "step": 1166 + }, + { + "epoch": 0.44004524886877827, + "grad_norm": 0.7393735279915095, + "learning_rate": 7.040723981900453e-05, + "loss": 0.4915, + "step": 1167 + }, + { + "epoch": 0.44042232277526394, + "grad_norm": 0.6848437538859511, + "learning_rate": 7.046757164404223e-05, + "loss": 0.4859, + "step": 1168 + }, + { + "epoch": 0.4407993966817496, + "grad_norm": 0.8058301985468115, + "learning_rate": 7.052790346907994e-05, + "loss": 0.4866, + "step": 1169 + }, + { + "epoch": 0.4411764705882353, + "grad_norm": 0.866668750771634, + "learning_rate": 7.058823529411765e-05, + "loss": 0.4756, + "step": 1170 + }, + { + "epoch": 0.44155354449472095, + "grad_norm": 0.7764328271916255, + "learning_rate": 7.064856711915535e-05, + "loss": 0.4782, + "step": 1171 + }, + { + "epoch": 0.4419306184012066, + "grad_norm": 0.7518266187438104, + "learning_rate": 7.070889894419307e-05, + "loss": 0.4824, + "step": 1172 + }, + { + "epoch": 0.4423076923076923, + "grad_norm": 0.7769383925182823, + "learning_rate": 7.076923076923078e-05, + "loss": 0.4805, + "step": 1173 + }, + { + "epoch": 0.44268476621417796, + "grad_norm": 0.6864436906989314, + "learning_rate": 7.082956259426848e-05, + "loss": 0.4742, + "step": 1174 + }, + { + "epoch": 0.44306184012066363, + "grad_norm": 0.6629949015321746, + "learning_rate": 7.088989441930619e-05, + "loss": 0.4657, + "step": 1175 + }, + { + "epoch": 0.4434389140271493, + "grad_norm": 0.6745886186691782, + "learning_rate": 7.095022624434389e-05, + "loss": 0.4898, + "step": 1176 + }, + { + "epoch": 0.44381598793363497, + "grad_norm": 0.5742694767335156, + "learning_rate": 7.10105580693816e-05, + "loss": 0.4859, + "step": 1177 + }, + { + "epoch": 0.44419306184012064, + "grad_norm": 0.6239679038613902, + "learning_rate": 7.107088989441931e-05, + "loss": 0.5119, + "step": 1178 + }, + { + "epoch": 0.4445701357466063, + "grad_norm": 0.6168037133650212, + "learning_rate": 7.113122171945701e-05, + "loss": 0.4726, + "step": 1179 + }, + { + "epoch": 0.444947209653092, + "grad_norm": 0.6313116226050427, + "learning_rate": 7.119155354449472e-05, + "loss": 0.477, + "step": 1180 + }, + { + "epoch": 0.4453242835595777, + "grad_norm": 0.6934643178057698, + "learning_rate": 7.125188536953243e-05, + "loss": 0.4765, + "step": 1181 + }, + { + "epoch": 0.4457013574660634, + "grad_norm": 0.6124508066966218, + "learning_rate": 7.131221719457015e-05, + "loss": 0.4979, + "step": 1182 + }, + { + "epoch": 0.44607843137254904, + "grad_norm": 0.7769735520551019, + "learning_rate": 7.137254901960786e-05, + "loss": 0.4912, + "step": 1183 + }, + { + "epoch": 0.4464555052790347, + "grad_norm": 1.1511369530415365, + "learning_rate": 7.143288084464556e-05, + "loss": 0.4959, + "step": 1184 + }, + { + "epoch": 0.4468325791855204, + "grad_norm": 0.8119750514436437, + "learning_rate": 7.149321266968327e-05, + "loss": 0.4735, + "step": 1185 + }, + { + "epoch": 0.44720965309200605, + "grad_norm": 0.7575573599056287, + "learning_rate": 7.155354449472097e-05, + "loss": 0.5046, + "step": 1186 + }, + { + "epoch": 0.4475867269984917, + "grad_norm": 0.8912369466267854, + "learning_rate": 7.161387631975868e-05, + "loss": 0.4789, + "step": 1187 + }, + { + "epoch": 0.4479638009049774, + "grad_norm": 0.8863768275468289, + "learning_rate": 7.16742081447964e-05, + "loss": 0.4976, + "step": 1188 + }, + { + "epoch": 0.44834087481146306, + "grad_norm": 1.1098358078338002, + "learning_rate": 7.173453996983409e-05, + "loss": 0.489, + "step": 1189 + }, + { + "epoch": 0.44871794871794873, + "grad_norm": 0.9027598688588758, + "learning_rate": 7.17948717948718e-05, + "loss": 0.4718, + "step": 1190 + }, + { + "epoch": 0.4490950226244344, + "grad_norm": 0.7689825683917816, + "learning_rate": 7.185520361990952e-05, + "loss": 0.4974, + "step": 1191 + }, + { + "epoch": 0.4494720965309201, + "grad_norm": 0.5409371946252679, + "learning_rate": 7.191553544494722e-05, + "loss": 0.4846, + "step": 1192 + }, + { + "epoch": 0.44984917043740574, + "grad_norm": 0.42264515926933527, + "learning_rate": 7.197586726998493e-05, + "loss": 0.4838, + "step": 1193 + }, + { + "epoch": 0.4502262443438914, + "grad_norm": 0.4142257467916799, + "learning_rate": 7.203619909502263e-05, + "loss": 0.4791, + "step": 1194 + }, + { + "epoch": 0.4506033182503771, + "grad_norm": 0.48007640146992064, + "learning_rate": 7.209653092006034e-05, + "loss": 0.4747, + "step": 1195 + }, + { + "epoch": 0.45098039215686275, + "grad_norm": 0.5803042538093169, + "learning_rate": 7.215686274509805e-05, + "loss": 0.4886, + "step": 1196 + }, + { + "epoch": 0.4513574660633484, + "grad_norm": 0.6073814750634307, + "learning_rate": 7.221719457013575e-05, + "loss": 0.4722, + "step": 1197 + }, + { + "epoch": 0.4517345399698341, + "grad_norm": 0.6670850526851113, + "learning_rate": 7.227752639517346e-05, + "loss": 0.478, + "step": 1198 + }, + { + "epoch": 0.45211161387631976, + "grad_norm": 0.8631150441750365, + "learning_rate": 7.233785822021117e-05, + "loss": 0.4983, + "step": 1199 + }, + { + "epoch": 0.45248868778280543, + "grad_norm": 0.9327382155107675, + "learning_rate": 7.239819004524887e-05, + "loss": 0.4724, + "step": 1200 + }, + { + "epoch": 0.4528657616892911, + "grad_norm": 1.0519856389139752, + "learning_rate": 7.245852187028659e-05, + "loss": 0.4829, + "step": 1201 + }, + { + "epoch": 0.45324283559577677, + "grad_norm": 1.1781375473271134, + "learning_rate": 7.251885369532428e-05, + "loss": 0.4764, + "step": 1202 + }, + { + "epoch": 0.45361990950226244, + "grad_norm": 0.6403605341371914, + "learning_rate": 7.2579185520362e-05, + "loss": 0.4858, + "step": 1203 + }, + { + "epoch": 0.4539969834087481, + "grad_norm": 0.4742035204489584, + "learning_rate": 7.263951734539971e-05, + "loss": 0.4757, + "step": 1204 + }, + { + "epoch": 0.4543740573152338, + "grad_norm": 0.7300015811288513, + "learning_rate": 7.269984917043741e-05, + "loss": 0.4691, + "step": 1205 + }, + { + "epoch": 0.45475113122171945, + "grad_norm": 0.9453650376462809, + "learning_rate": 7.276018099547512e-05, + "loss": 0.4863, + "step": 1206 + }, + { + "epoch": 0.4551282051282051, + "grad_norm": 1.005152472898676, + "learning_rate": 7.282051282051283e-05, + "loss": 0.4855, + "step": 1207 + }, + { + "epoch": 0.4555052790346908, + "grad_norm": 0.8942749457578317, + "learning_rate": 7.288084464555053e-05, + "loss": 0.4927, + "step": 1208 + }, + { + "epoch": 0.45588235294117646, + "grad_norm": 0.824215574338513, + "learning_rate": 7.294117647058824e-05, + "loss": 0.4801, + "step": 1209 + }, + { + "epoch": 0.45625942684766213, + "grad_norm": 0.7670489342807044, + "learning_rate": 7.300150829562594e-05, + "loss": 0.491, + "step": 1210 + }, + { + "epoch": 0.4566365007541478, + "grad_norm": 0.7437694701926957, + "learning_rate": 7.306184012066365e-05, + "loss": 0.4922, + "step": 1211 + }, + { + "epoch": 0.45701357466063347, + "grad_norm": 0.6660368911604929, + "learning_rate": 7.312217194570137e-05, + "loss": 0.4818, + "step": 1212 + }, + { + "epoch": 0.45739064856711914, + "grad_norm": 0.736284763558501, + "learning_rate": 7.318250377073906e-05, + "loss": 0.4906, + "step": 1213 + }, + { + "epoch": 0.4577677224736048, + "grad_norm": 0.7726218251471543, + "learning_rate": 7.324283559577678e-05, + "loss": 0.499, + "step": 1214 + }, + { + "epoch": 0.4581447963800905, + "grad_norm": 0.6212735813217507, + "learning_rate": 7.330316742081449e-05, + "loss": 0.4773, + "step": 1215 + }, + { + "epoch": 0.45852187028657615, + "grad_norm": 0.563471726814476, + "learning_rate": 7.336349924585219e-05, + "loss": 0.4898, + "step": 1216 + }, + { + "epoch": 0.4588989441930618, + "grad_norm": 0.6532985293888912, + "learning_rate": 7.34238310708899e-05, + "loss": 0.484, + "step": 1217 + }, + { + "epoch": 0.4592760180995475, + "grad_norm": 0.7453114230903738, + "learning_rate": 7.34841628959276e-05, + "loss": 0.4879, + "step": 1218 + }, + { + "epoch": 0.45965309200603316, + "grad_norm": 0.7040297137541054, + "learning_rate": 7.354449472096531e-05, + "loss": 0.4711, + "step": 1219 + }, + { + "epoch": 0.46003016591251883, + "grad_norm": 0.7937261813203652, + "learning_rate": 7.360482654600302e-05, + "loss": 0.4887, + "step": 1220 + }, + { + "epoch": 0.4604072398190045, + "grad_norm": 0.89599632602296, + "learning_rate": 7.366515837104072e-05, + "loss": 0.4838, + "step": 1221 + }, + { + "epoch": 0.46078431372549017, + "grad_norm": 0.8772450216041185, + "learning_rate": 7.372549019607843e-05, + "loss": 0.4991, + "step": 1222 + }, + { + "epoch": 0.4611613876319759, + "grad_norm": 0.8721537037385185, + "learning_rate": 7.378582202111615e-05, + "loss": 0.4756, + "step": 1223 + }, + { + "epoch": 0.46153846153846156, + "grad_norm": 1.25754383071287, + "learning_rate": 7.384615384615386e-05, + "loss": 0.4953, + "step": 1224 + }, + { + "epoch": 0.46191553544494723, + "grad_norm": 0.9126874336226573, + "learning_rate": 7.390648567119156e-05, + "loss": 0.4872, + "step": 1225 + }, + { + "epoch": 0.4622926093514329, + "grad_norm": 0.7841983210822663, + "learning_rate": 7.396681749622927e-05, + "loss": 0.4995, + "step": 1226 + }, + { + "epoch": 0.4626696832579186, + "grad_norm": 0.7431903935078802, + "learning_rate": 7.402714932126698e-05, + "loss": 0.4823, + "step": 1227 + }, + { + "epoch": 0.46304675716440424, + "grad_norm": 0.7995006149198922, + "learning_rate": 7.408748114630468e-05, + "loss": 0.4887, + "step": 1228 + }, + { + "epoch": 0.4634238310708899, + "grad_norm": 0.8333685599551648, + "learning_rate": 7.414781297134239e-05, + "loss": 0.4706, + "step": 1229 + }, + { + "epoch": 0.4638009049773756, + "grad_norm": 0.7830951120849894, + "learning_rate": 7.42081447963801e-05, + "loss": 0.4746, + "step": 1230 + }, + { + "epoch": 0.46417797888386125, + "grad_norm": 0.704468183705522, + "learning_rate": 7.42684766214178e-05, + "loss": 0.4884, + "step": 1231 + }, + { + "epoch": 0.4645550527903469, + "grad_norm": 0.6864429492493237, + "learning_rate": 7.432880844645552e-05, + "loss": 0.4868, + "step": 1232 + }, + { + "epoch": 0.4649321266968326, + "grad_norm": 0.7461176920886038, + "learning_rate": 7.438914027149321e-05, + "loss": 0.4787, + "step": 1233 + }, + { + "epoch": 0.46530920060331826, + "grad_norm": 0.9027784201447866, + "learning_rate": 7.444947209653093e-05, + "loss": 0.4823, + "step": 1234 + }, + { + "epoch": 0.46568627450980393, + "grad_norm": 1.1031687999292594, + "learning_rate": 7.450980392156864e-05, + "loss": 0.4781, + "step": 1235 + }, + { + "epoch": 0.4660633484162896, + "grad_norm": 0.9255548646877445, + "learning_rate": 7.457013574660634e-05, + "loss": 0.4813, + "step": 1236 + }, + { + "epoch": 0.46644042232277527, + "grad_norm": 0.8146964277705513, + "learning_rate": 7.463046757164405e-05, + "loss": 0.4813, + "step": 1237 + }, + { + "epoch": 0.46681749622926094, + "grad_norm": 0.5761751931315577, + "learning_rate": 7.469079939668176e-05, + "loss": 0.4852, + "step": 1238 + }, + { + "epoch": 0.4671945701357466, + "grad_norm": 0.42470543096042707, + "learning_rate": 7.475113122171946e-05, + "loss": 0.4922, + "step": 1239 + }, + { + "epoch": 0.4675716440422323, + "grad_norm": 0.5472130082965743, + "learning_rate": 7.481146304675717e-05, + "loss": 0.4716, + "step": 1240 + }, + { + "epoch": 0.46794871794871795, + "grad_norm": 0.6203708642654333, + "learning_rate": 7.487179487179487e-05, + "loss": 0.496, + "step": 1241 + }, + { + "epoch": 0.4683257918552036, + "grad_norm": 0.5981409521909029, + "learning_rate": 7.493212669683258e-05, + "loss": 0.4695, + "step": 1242 + }, + { + "epoch": 0.4687028657616893, + "grad_norm": 0.548817817551536, + "learning_rate": 7.49924585218703e-05, + "loss": 0.4907, + "step": 1243 + }, + { + "epoch": 0.46907993966817496, + "grad_norm": 0.5571633719226837, + "learning_rate": 7.5052790346908e-05, + "loss": 0.4746, + "step": 1244 + }, + { + "epoch": 0.46945701357466063, + "grad_norm": 0.6865093669656979, + "learning_rate": 7.511312217194571e-05, + "loss": 0.4928, + "step": 1245 + }, + { + "epoch": 0.4698340874811463, + "grad_norm": 0.8042246128888136, + "learning_rate": 7.517345399698342e-05, + "loss": 0.5028, + "step": 1246 + }, + { + "epoch": 0.47021116138763197, + "grad_norm": 0.8414818050881523, + "learning_rate": 7.523378582202112e-05, + "loss": 0.4886, + "step": 1247 + }, + { + "epoch": 0.47058823529411764, + "grad_norm": 1.036194024785996, + "learning_rate": 7.529411764705883e-05, + "loss": 0.496, + "step": 1248 + }, + { + "epoch": 0.4709653092006033, + "grad_norm": 1.1826035166158377, + "learning_rate": 7.535444947209653e-05, + "loss": 0.4913, + "step": 1249 + }, + { + "epoch": 0.471342383107089, + "grad_norm": 0.7201907416502408, + "learning_rate": 7.541478129713424e-05, + "loss": 0.4883, + "step": 1250 + }, + { + "epoch": 0.47171945701357465, + "grad_norm": 0.4969093625272548, + "learning_rate": 7.547511312217195e-05, + "loss": 0.479, + "step": 1251 + }, + { + "epoch": 0.4720965309200603, + "grad_norm": 0.6139683943050789, + "learning_rate": 7.553544494720965e-05, + "loss": 0.5052, + "step": 1252 + }, + { + "epoch": 0.472473604826546, + "grad_norm": 0.7785015674193956, + "learning_rate": 7.559577677224736e-05, + "loss": 0.4745, + "step": 1253 + }, + { + "epoch": 0.47285067873303166, + "grad_norm": 0.8161970186421533, + "learning_rate": 7.565610859728508e-05, + "loss": 0.488, + "step": 1254 + }, + { + "epoch": 0.47322775263951733, + "grad_norm": 0.6923534838553921, + "learning_rate": 7.571644042232278e-05, + "loss": 0.4721, + "step": 1255 + }, + { + "epoch": 0.473604826546003, + "grad_norm": 0.6617101166783863, + "learning_rate": 7.577677224736049e-05, + "loss": 0.4852, + "step": 1256 + }, + { + "epoch": 0.47398190045248867, + "grad_norm": 0.8532585161345871, + "learning_rate": 7.583710407239819e-05, + "loss": 0.4921, + "step": 1257 + }, + { + "epoch": 0.47435897435897434, + "grad_norm": 1.0361145083269958, + "learning_rate": 7.58974358974359e-05, + "loss": 0.4909, + "step": 1258 + }, + { + "epoch": 0.47473604826546, + "grad_norm": 0.9229921648757791, + "learning_rate": 7.595776772247361e-05, + "loss": 0.4728, + "step": 1259 + }, + { + "epoch": 0.4751131221719457, + "grad_norm": 0.8347976830737971, + "learning_rate": 7.601809954751131e-05, + "loss": 0.4896, + "step": 1260 + }, + { + "epoch": 0.47549019607843135, + "grad_norm": 0.9204188244484274, + "learning_rate": 7.607843137254902e-05, + "loss": 0.4937, + "step": 1261 + }, + { + "epoch": 0.475867269984917, + "grad_norm": 0.8466498421819117, + "learning_rate": 7.613876319758673e-05, + "loss": 0.4876, + "step": 1262 + }, + { + "epoch": 0.4762443438914027, + "grad_norm": 0.7253108412466519, + "learning_rate": 7.619909502262443e-05, + "loss": 0.4758, + "step": 1263 + }, + { + "epoch": 0.4766214177978884, + "grad_norm": 0.7926511836218335, + "learning_rate": 7.625942684766216e-05, + "loss": 0.4831, + "step": 1264 + }, + { + "epoch": 0.4769984917043741, + "grad_norm": 1.0246204122005718, + "learning_rate": 7.631975867269986e-05, + "loss": 0.4795, + "step": 1265 + }, + { + "epoch": 0.47737556561085975, + "grad_norm": 1.046302632823969, + "learning_rate": 7.638009049773757e-05, + "loss": 0.4958, + "step": 1266 + }, + { + "epoch": 0.4777526395173454, + "grad_norm": 0.6794001712088812, + "learning_rate": 7.644042232277527e-05, + "loss": 0.4737, + "step": 1267 + }, + { + "epoch": 0.4781297134238311, + "grad_norm": 0.5889240852293465, + "learning_rate": 7.650075414781298e-05, + "loss": 0.4813, + "step": 1268 + }, + { + "epoch": 0.47850678733031676, + "grad_norm": 0.6823674150595173, + "learning_rate": 7.656108597285069e-05, + "loss": 0.48, + "step": 1269 + }, + { + "epoch": 0.47888386123680243, + "grad_norm": 0.6777286048991504, + "learning_rate": 7.662141779788839e-05, + "loss": 0.4767, + "step": 1270 + }, + { + "epoch": 0.4792609351432881, + "grad_norm": 0.44587730942336157, + "learning_rate": 7.66817496229261e-05, + "loss": 0.5022, + "step": 1271 + }, + { + "epoch": 0.4796380090497738, + "grad_norm": 0.4868914636974709, + "learning_rate": 7.674208144796382e-05, + "loss": 0.4709, + "step": 1272 + }, + { + "epoch": 0.48001508295625944, + "grad_norm": 0.5920554569186129, + "learning_rate": 7.680241327300151e-05, + "loss": 0.4821, + "step": 1273 + }, + { + "epoch": 0.4803921568627451, + "grad_norm": 0.5815200714601815, + "learning_rate": 7.686274509803923e-05, + "loss": 0.4872, + "step": 1274 + }, + { + "epoch": 0.4807692307692308, + "grad_norm": 0.5528444063264479, + "learning_rate": 7.692307692307693e-05, + "loss": 0.4908, + "step": 1275 + }, + { + "epoch": 0.48114630467571645, + "grad_norm": 0.653619164373874, + "learning_rate": 7.698340874811464e-05, + "loss": 0.48, + "step": 1276 + }, + { + "epoch": 0.4815233785822021, + "grad_norm": 0.7537255753934807, + "learning_rate": 7.704374057315235e-05, + "loss": 0.4721, + "step": 1277 + }, + { + "epoch": 0.4819004524886878, + "grad_norm": 0.8447082038508779, + "learning_rate": 7.710407239819005e-05, + "loss": 0.4918, + "step": 1278 + }, + { + "epoch": 0.48227752639517346, + "grad_norm": 0.921047365761689, + "learning_rate": 7.716440422322776e-05, + "loss": 0.4825, + "step": 1279 + }, + { + "epoch": 0.48265460030165913, + "grad_norm": 0.9491678629183359, + "learning_rate": 7.722473604826547e-05, + "loss": 0.4913, + "step": 1280 + }, + { + "epoch": 0.4830316742081448, + "grad_norm": 0.9813651924924695, + "learning_rate": 7.728506787330317e-05, + "loss": 0.4827, + "step": 1281 + }, + { + "epoch": 0.48340874811463047, + "grad_norm": 0.9682482495735847, + "learning_rate": 7.734539969834088e-05, + "loss": 0.4812, + "step": 1282 + }, + { + "epoch": 0.48378582202111614, + "grad_norm": 1.192560117716014, + "learning_rate": 7.740573152337858e-05, + "loss": 0.4924, + "step": 1283 + }, + { + "epoch": 0.4841628959276018, + "grad_norm": 0.9595136549229178, + "learning_rate": 7.74660633484163e-05, + "loss": 0.4829, + "step": 1284 + }, + { + "epoch": 0.4845399698340875, + "grad_norm": 1.0322261091901805, + "learning_rate": 7.752639517345401e-05, + "loss": 0.4682, + "step": 1285 + }, + { + "epoch": 0.48491704374057315, + "grad_norm": 1.0694928499388108, + "learning_rate": 7.75867269984917e-05, + "loss": 0.4781, + "step": 1286 + }, + { + "epoch": 0.4852941176470588, + "grad_norm": 0.8051366535165937, + "learning_rate": 7.764705882352942e-05, + "loss": 0.4862, + "step": 1287 + }, + { + "epoch": 0.4856711915535445, + "grad_norm": 0.8138603738782293, + "learning_rate": 7.770739064856713e-05, + "loss": 0.4858, + "step": 1288 + }, + { + "epoch": 0.48604826546003016, + "grad_norm": 0.7983285215567861, + "learning_rate": 7.776772247360483e-05, + "loss": 0.4889, + "step": 1289 + }, + { + "epoch": 0.48642533936651583, + "grad_norm": 0.6401403469810867, + "learning_rate": 7.782805429864254e-05, + "loss": 0.4898, + "step": 1290 + }, + { + "epoch": 0.4868024132730015, + "grad_norm": 0.4743367009172047, + "learning_rate": 7.788838612368024e-05, + "loss": 0.4728, + "step": 1291 + }, + { + "epoch": 0.48717948717948717, + "grad_norm": 0.5312047827802805, + "learning_rate": 7.794871794871795e-05, + "loss": 0.4837, + "step": 1292 + }, + { + "epoch": 0.48755656108597284, + "grad_norm": 0.5282332561356533, + "learning_rate": 7.800904977375566e-05, + "loss": 0.4832, + "step": 1293 + }, + { + "epoch": 0.4879336349924585, + "grad_norm": 0.48501931588746183, + "learning_rate": 7.806938159879336e-05, + "loss": 0.4821, + "step": 1294 + }, + { + "epoch": 0.4883107088989442, + "grad_norm": 0.5292635562476903, + "learning_rate": 7.812971342383108e-05, + "loss": 0.4816, + "step": 1295 + }, + { + "epoch": 0.48868778280542985, + "grad_norm": 0.5582788576334926, + "learning_rate": 7.819004524886879e-05, + "loss": 0.4948, + "step": 1296 + }, + { + "epoch": 0.4890648567119155, + "grad_norm": 0.4557587834204229, + "learning_rate": 7.825037707390649e-05, + "loss": 0.4856, + "step": 1297 + }, + { + "epoch": 0.4894419306184012, + "grad_norm": 0.49651613123192373, + "learning_rate": 7.83107088989442e-05, + "loss": 0.4874, + "step": 1298 + }, + { + "epoch": 0.48981900452488686, + "grad_norm": 0.6626065751456269, + "learning_rate": 7.83710407239819e-05, + "loss": 0.4755, + "step": 1299 + }, + { + "epoch": 0.49019607843137253, + "grad_norm": 0.8720400901133023, + "learning_rate": 7.843137254901961e-05, + "loss": 0.4854, + "step": 1300 + }, + { + "epoch": 0.4905731523378582, + "grad_norm": 1.0688375947234887, + "learning_rate": 7.849170437405732e-05, + "loss": 0.4798, + "step": 1301 + }, + { + "epoch": 0.49095022624434387, + "grad_norm": 1.0472773280183079, + "learning_rate": 7.855203619909502e-05, + "loss": 0.4804, + "step": 1302 + }, + { + "epoch": 0.49132730015082954, + "grad_norm": 0.9138389971541989, + "learning_rate": 7.861236802413273e-05, + "loss": 0.4922, + "step": 1303 + }, + { + "epoch": 0.4917043740573152, + "grad_norm": 0.7975669198371583, + "learning_rate": 7.867269984917045e-05, + "loss": 0.479, + "step": 1304 + }, + { + "epoch": 0.4920814479638009, + "grad_norm": 0.7748844648899499, + "learning_rate": 7.873303167420814e-05, + "loss": 0.4847, + "step": 1305 + }, + { + "epoch": 0.4924585218702866, + "grad_norm": 0.8023111330904342, + "learning_rate": 7.879336349924586e-05, + "loss": 0.4773, + "step": 1306 + }, + { + "epoch": 0.4928355957767723, + "grad_norm": 0.746243796920654, + "learning_rate": 7.885369532428357e-05, + "loss": 0.4736, + "step": 1307 + }, + { + "epoch": 0.49321266968325794, + "grad_norm": 0.6508310491272455, + "learning_rate": 7.891402714932128e-05, + "loss": 0.4861, + "step": 1308 + }, + { + "epoch": 0.4935897435897436, + "grad_norm": 0.6200662197219722, + "learning_rate": 7.897435897435898e-05, + "loss": 0.4964, + "step": 1309 + }, + { + "epoch": 0.4939668174962293, + "grad_norm": 0.607812198634358, + "learning_rate": 7.903469079939669e-05, + "loss": 0.4718, + "step": 1310 + }, + { + "epoch": 0.49434389140271495, + "grad_norm": 0.49209819174427777, + "learning_rate": 7.90950226244344e-05, + "loss": 0.4704, + "step": 1311 + }, + { + "epoch": 0.4947209653092006, + "grad_norm": 0.6788289411983148, + "learning_rate": 7.91553544494721e-05, + "loss": 0.4787, + "step": 1312 + }, + { + "epoch": 0.4950980392156863, + "grad_norm": 0.8644455543651869, + "learning_rate": 7.921568627450982e-05, + "loss": 0.4912, + "step": 1313 + }, + { + "epoch": 0.49547511312217196, + "grad_norm": 0.7804958221690482, + "learning_rate": 7.927601809954751e-05, + "loss": 0.4731, + "step": 1314 + }, + { + "epoch": 0.49585218702865763, + "grad_norm": 0.7707966653078305, + "learning_rate": 7.933634992458523e-05, + "loss": 0.479, + "step": 1315 + }, + { + "epoch": 0.4962292609351433, + "grad_norm": 0.8893376085649707, + "learning_rate": 7.939668174962294e-05, + "loss": 0.4801, + "step": 1316 + }, + { + "epoch": 0.49660633484162897, + "grad_norm": 0.9600504112396079, + "learning_rate": 7.945701357466064e-05, + "loss": 0.4972, + "step": 1317 + }, + { + "epoch": 0.49698340874811464, + "grad_norm": 0.9948742936893606, + "learning_rate": 7.951734539969835e-05, + "loss": 0.4844, + "step": 1318 + }, + { + "epoch": 0.4973604826546003, + "grad_norm": 1.0011354058334383, + "learning_rate": 7.957767722473606e-05, + "loss": 0.4936, + "step": 1319 + }, + { + "epoch": 0.497737556561086, + "grad_norm": 0.8136705444671204, + "learning_rate": 7.963800904977376e-05, + "loss": 0.4844, + "step": 1320 + }, + { + "epoch": 0.49811463046757165, + "grad_norm": 0.522053033709792, + "learning_rate": 7.969834087481147e-05, + "loss": 0.4844, + "step": 1321 + }, + { + "epoch": 0.4984917043740573, + "grad_norm": 0.500129596296482, + "learning_rate": 7.975867269984917e-05, + "loss": 0.4899, + "step": 1322 + }, + { + "epoch": 0.498868778280543, + "grad_norm": 0.5480904921897977, + "learning_rate": 7.981900452488688e-05, + "loss": 0.4748, + "step": 1323 + }, + { + "epoch": 0.49924585218702866, + "grad_norm": 0.7475866246930819, + "learning_rate": 7.98793363499246e-05, + "loss": 0.4781, + "step": 1324 + }, + { + "epoch": 0.49962292609351433, + "grad_norm": 0.8645277609616923, + "learning_rate": 7.99396681749623e-05, + "loss": 0.4805, + "step": 1325 + }, + { + "epoch": 0.5, + "grad_norm": 0.696119835534159, + "learning_rate": 8e-05, + "loss": 0.4861, + "step": 1326 + }, + { + "epoch": 0.5003770739064857, + "grad_norm": 0.5608702827282387, + "learning_rate": 7.999999861401775e-05, + "loss": 0.4823, + "step": 1327 + }, + { + "epoch": 0.5007541478129713, + "grad_norm": 0.7830721009083337, + "learning_rate": 7.999999445607107e-05, + "loss": 0.476, + "step": 1328 + }, + { + "epoch": 0.501131221719457, + "grad_norm": 0.9415425589219529, + "learning_rate": 7.999998752616024e-05, + "loss": 0.4971, + "step": 1329 + }, + { + "epoch": 0.5015082956259427, + "grad_norm": 0.8059986745463619, + "learning_rate": 7.999997782428577e-05, + "loss": 0.4815, + "step": 1330 + }, + { + "epoch": 0.5018853695324283, + "grad_norm": 0.6187147195282672, + "learning_rate": 7.99999653504483e-05, + "loss": 0.482, + "step": 1331 + }, + { + "epoch": 0.502262443438914, + "grad_norm": 0.6027207596754713, + "learning_rate": 7.999995010464873e-05, + "loss": 0.4876, + "step": 1332 + }, + { + "epoch": 0.5026395173453997, + "grad_norm": 0.7450276463442068, + "learning_rate": 7.999993208688809e-05, + "loss": 0.4789, + "step": 1333 + }, + { + "epoch": 0.5030165912518854, + "grad_norm": 1.019766051550027, + "learning_rate": 7.999991129716764e-05, + "loss": 0.4875, + "step": 1334 + }, + { + "epoch": 0.503393665158371, + "grad_norm": 1.071658309767482, + "learning_rate": 7.99998877354888e-05, + "loss": 0.4843, + "step": 1335 + }, + { + "epoch": 0.5037707390648567, + "grad_norm": 0.7067595152934408, + "learning_rate": 7.999986140185324e-05, + "loss": 0.4812, + "step": 1336 + }, + { + "epoch": 0.5041478129713424, + "grad_norm": 0.4688709734375655, + "learning_rate": 7.999983229626276e-05, + "loss": 0.4763, + "step": 1337 + }, + { + "epoch": 0.504524886877828, + "grad_norm": 0.5042226432796098, + "learning_rate": 7.999980041871939e-05, + "loss": 0.4817, + "step": 1338 + }, + { + "epoch": 0.5049019607843137, + "grad_norm": 0.689413294137103, + "learning_rate": 7.999976576922531e-05, + "loss": 0.4691, + "step": 1339 + }, + { + "epoch": 0.5052790346907994, + "grad_norm": 0.7532059681358009, + "learning_rate": 7.999972834778294e-05, + "loss": 0.4816, + "step": 1340 + }, + { + "epoch": 0.505656108597285, + "grad_norm": 0.5688174405624358, + "learning_rate": 7.99996881543949e-05, + "loss": 0.4891, + "step": 1341 + }, + { + "epoch": 0.5060331825037707, + "grad_norm": 0.5042382301173044, + "learning_rate": 7.999964518906395e-05, + "loss": 0.483, + "step": 1342 + }, + { + "epoch": 0.5064102564102564, + "grad_norm": 0.6620922694138383, + "learning_rate": 7.999959945179305e-05, + "loss": 0.499, + "step": 1343 + }, + { + "epoch": 0.5067873303167421, + "grad_norm": 0.6735441685614606, + "learning_rate": 7.99995509425854e-05, + "loss": 0.4745, + "step": 1344 + }, + { + "epoch": 0.5071644042232277, + "grad_norm": 0.6421269824978663, + "learning_rate": 7.999949966144433e-05, + "loss": 0.4769, + "step": 1345 + }, + { + "epoch": 0.5075414781297134, + "grad_norm": 0.6504840807259804, + "learning_rate": 7.999944560837343e-05, + "loss": 0.4773, + "step": 1346 + }, + { + "epoch": 0.5079185520361991, + "grad_norm": 0.5909972077170543, + "learning_rate": 7.999938878337641e-05, + "loss": 0.4812, + "step": 1347 + }, + { + "epoch": 0.5082956259426847, + "grad_norm": 0.47883479750769836, + "learning_rate": 7.999932918645725e-05, + "loss": 0.4895, + "step": 1348 + }, + { + "epoch": 0.5086726998491704, + "grad_norm": 0.6148178195176226, + "learning_rate": 7.999926681762005e-05, + "loss": 0.4745, + "step": 1349 + }, + { + "epoch": 0.5090497737556561, + "grad_norm": 0.8298833723071022, + "learning_rate": 7.999920167686914e-05, + "loss": 0.4757, + "step": 1350 + }, + { + "epoch": 0.5094268476621417, + "grad_norm": 0.9074826427364199, + "learning_rate": 7.999913376420904e-05, + "loss": 0.4804, + "step": 1351 + }, + { + "epoch": 0.5098039215686274, + "grad_norm": 0.8577010692182815, + "learning_rate": 7.999906307964443e-05, + "loss": 0.4705, + "step": 1352 + }, + { + "epoch": 0.5101809954751131, + "grad_norm": 0.8769864024533849, + "learning_rate": 7.999898962318023e-05, + "loss": 0.4884, + "step": 1353 + }, + { + "epoch": 0.5105580693815988, + "grad_norm": 1.1078795680010776, + "learning_rate": 7.999891339482154e-05, + "loss": 0.4785, + "step": 1354 + }, + { + "epoch": 0.5109351432880844, + "grad_norm": 0.954708075524873, + "learning_rate": 7.999883439457362e-05, + "loss": 0.4839, + "step": 1355 + }, + { + "epoch": 0.5113122171945701, + "grad_norm": 0.8675187200707929, + "learning_rate": 7.999875262244196e-05, + "loss": 0.4733, + "step": 1356 + }, + { + "epoch": 0.5116892911010558, + "grad_norm": 0.7312819690058122, + "learning_rate": 7.999866807843223e-05, + "loss": 0.4814, + "step": 1357 + }, + { + "epoch": 0.5120663650075414, + "grad_norm": 0.5403702908033543, + "learning_rate": 7.999858076255028e-05, + "loss": 0.4804, + "step": 1358 + }, + { + "epoch": 0.5124434389140271, + "grad_norm": 0.4016829515242336, + "learning_rate": 7.999849067480216e-05, + "loss": 0.4746, + "step": 1359 + }, + { + "epoch": 0.5128205128205128, + "grad_norm": 0.5418036050345632, + "learning_rate": 7.999839781519409e-05, + "loss": 0.4848, + "step": 1360 + }, + { + "epoch": 0.5131975867269984, + "grad_norm": 0.7195500886604959, + "learning_rate": 7.999830218373256e-05, + "loss": 0.4758, + "step": 1361 + }, + { + "epoch": 0.5135746606334841, + "grad_norm": 0.7660749060232053, + "learning_rate": 7.999820378042415e-05, + "loss": 0.4854, + "step": 1362 + }, + { + "epoch": 0.5139517345399698, + "grad_norm": 0.7091563046319972, + "learning_rate": 7.999810260527572e-05, + "loss": 0.4846, + "step": 1363 + }, + { + "epoch": 0.5143288084464555, + "grad_norm": 0.6150864749323581, + "learning_rate": 7.999799865829424e-05, + "loss": 0.4773, + "step": 1364 + }, + { + "epoch": 0.5147058823529411, + "grad_norm": 0.4954289379198706, + "learning_rate": 7.999789193948694e-05, + "loss": 0.4632, + "step": 1365 + }, + { + "epoch": 0.5150829562594268, + "grad_norm": 0.38622562715710307, + "learning_rate": 7.99977824488612e-05, + "loss": 0.4862, + "step": 1366 + }, + { + "epoch": 0.5154600301659125, + "grad_norm": 0.3977292524347993, + "learning_rate": 7.999767018642461e-05, + "loss": 0.4724, + "step": 1367 + }, + { + "epoch": 0.5158371040723982, + "grad_norm": 0.32088198787958355, + "learning_rate": 7.999755515218495e-05, + "loss": 0.4662, + "step": 1368 + }, + { + "epoch": 0.5162141779788839, + "grad_norm": 0.4012019991011479, + "learning_rate": 7.999743734615021e-05, + "loss": 0.4818, + "step": 1369 + }, + { + "epoch": 0.5165912518853696, + "grad_norm": 0.45730030003034783, + "learning_rate": 7.999731676832853e-05, + "loss": 0.4841, + "step": 1370 + }, + { + "epoch": 0.5169683257918553, + "grad_norm": 0.42598718178951017, + "learning_rate": 7.999719341872829e-05, + "loss": 0.4816, + "step": 1371 + }, + { + "epoch": 0.5173453996983409, + "grad_norm": 0.4596094578615345, + "learning_rate": 7.999706729735801e-05, + "loss": 0.4801, + "step": 1372 + }, + { + "epoch": 0.5177224736048266, + "grad_norm": 0.5773531336844397, + "learning_rate": 7.999693840422643e-05, + "loss": 0.462, + "step": 1373 + }, + { + "epoch": 0.5180995475113123, + "grad_norm": 0.7387530274175477, + "learning_rate": 7.999680673934252e-05, + "loss": 0.4784, + "step": 1374 + }, + { + "epoch": 0.5184766214177979, + "grad_norm": 0.9163431147448179, + "learning_rate": 7.999667230271537e-05, + "loss": 0.4794, + "step": 1375 + }, + { + "epoch": 0.5188536953242836, + "grad_norm": 1.184918280301583, + "learning_rate": 7.999653509435432e-05, + "loss": 0.4788, + "step": 1376 + }, + { + "epoch": 0.5192307692307693, + "grad_norm": 0.7773231593273155, + "learning_rate": 7.999639511426884e-05, + "loss": 0.4851, + "step": 1377 + }, + { + "epoch": 0.5196078431372549, + "grad_norm": 0.6551806582816871, + "learning_rate": 7.999625236246868e-05, + "loss": 0.4818, + "step": 1378 + }, + { + "epoch": 0.5199849170437406, + "grad_norm": 0.7636798496532168, + "learning_rate": 7.999610683896369e-05, + "loss": 0.48, + "step": 1379 + }, + { + "epoch": 0.5203619909502263, + "grad_norm": 0.9470051147125793, + "learning_rate": 7.999595854376398e-05, + "loss": 0.4756, + "step": 1380 + }, + { + "epoch": 0.520739064856712, + "grad_norm": 1.081156593042739, + "learning_rate": 7.999580747687982e-05, + "loss": 0.4787, + "step": 1381 + }, + { + "epoch": 0.5211161387631976, + "grad_norm": 0.7495375227939104, + "learning_rate": 7.999565363832167e-05, + "loss": 0.4892, + "step": 1382 + }, + { + "epoch": 0.5214932126696833, + "grad_norm": 0.521253577418602, + "learning_rate": 7.999549702810022e-05, + "loss": 0.4729, + "step": 1383 + }, + { + "epoch": 0.521870286576169, + "grad_norm": 0.6616048312554487, + "learning_rate": 7.999533764622628e-05, + "loss": 0.4819, + "step": 1384 + }, + { + "epoch": 0.5222473604826546, + "grad_norm": 0.6839317377675527, + "learning_rate": 7.999517549271092e-05, + "loss": 0.4706, + "step": 1385 + }, + { + "epoch": 0.5226244343891403, + "grad_norm": 0.7233702930917995, + "learning_rate": 7.999501056756539e-05, + "loss": 0.4786, + "step": 1386 + }, + { + "epoch": 0.523001508295626, + "grad_norm": 0.6546028059763432, + "learning_rate": 7.999484287080109e-05, + "loss": 0.4825, + "step": 1387 + }, + { + "epoch": 0.5233785822021116, + "grad_norm": 0.5818585874210704, + "learning_rate": 7.999467240242965e-05, + "loss": 0.4732, + "step": 1388 + }, + { + "epoch": 0.5237556561085973, + "grad_norm": 0.6373910325833946, + "learning_rate": 7.999449916246289e-05, + "loss": 0.4818, + "step": 1389 + }, + { + "epoch": 0.524132730015083, + "grad_norm": 0.6323501472971758, + "learning_rate": 7.999432315091283e-05, + "loss": 0.4773, + "step": 1390 + }, + { + "epoch": 0.5245098039215687, + "grad_norm": 0.8059377555986749, + "learning_rate": 7.999414436779163e-05, + "loss": 0.4888, + "step": 1391 + }, + { + "epoch": 0.5248868778280543, + "grad_norm": 1.0109512857267695, + "learning_rate": 7.999396281311171e-05, + "loss": 0.5003, + "step": 1392 + }, + { + "epoch": 0.52526395173454, + "grad_norm": 1.023256988580555, + "learning_rate": 7.999377848688565e-05, + "loss": 0.4833, + "step": 1393 + }, + { + "epoch": 0.5256410256410257, + "grad_norm": 0.9568663959970597, + "learning_rate": 7.999359138912619e-05, + "loss": 0.4854, + "step": 1394 + }, + { + "epoch": 0.5260180995475113, + "grad_norm": 0.9466392096969107, + "learning_rate": 7.999340151984633e-05, + "loss": 0.4873, + "step": 1395 + }, + { + "epoch": 0.526395173453997, + "grad_norm": 1.0197767274638818, + "learning_rate": 7.999320887905922e-05, + "loss": 0.4806, + "step": 1396 + }, + { + "epoch": 0.5267722473604827, + "grad_norm": 0.8446773434521595, + "learning_rate": 7.999301346677822e-05, + "loss": 0.4832, + "step": 1397 + }, + { + "epoch": 0.5271493212669683, + "grad_norm": 0.5220426610048551, + "learning_rate": 7.999281528301685e-05, + "loss": 0.4717, + "step": 1398 + }, + { + "epoch": 0.527526395173454, + "grad_norm": 0.46745021571093476, + "learning_rate": 7.999261432778886e-05, + "loss": 0.4619, + "step": 1399 + }, + { + "epoch": 0.5279034690799397, + "grad_norm": 0.7014541652962314, + "learning_rate": 7.999241060110818e-05, + "loss": 0.4753, + "step": 1400 + }, + { + "epoch": 0.5282805429864253, + "grad_norm": 0.7711311114743212, + "learning_rate": 7.99922041029889e-05, + "loss": 0.4596, + "step": 1401 + }, + { + "epoch": 0.528657616892911, + "grad_norm": 0.6901096756124728, + "learning_rate": 7.999199483344536e-05, + "loss": 0.4883, + "step": 1402 + }, + { + "epoch": 0.5290346907993967, + "grad_norm": 0.519903693136476, + "learning_rate": 7.999178279249204e-05, + "loss": 0.4855, + "step": 1403 + }, + { + "epoch": 0.5294117647058824, + "grad_norm": 0.3521186614361872, + "learning_rate": 7.999156798014366e-05, + "loss": 0.488, + "step": 1404 + }, + { + "epoch": 0.529788838612368, + "grad_norm": 0.5020711838695538, + "learning_rate": 7.999135039641509e-05, + "loss": 0.4703, + "step": 1405 + }, + { + "epoch": 0.5301659125188537, + "grad_norm": 0.6458421013886018, + "learning_rate": 7.999113004132141e-05, + "loss": 0.4841, + "step": 1406 + }, + { + "epoch": 0.5305429864253394, + "grad_norm": 0.7649090545015041, + "learning_rate": 7.999090691487787e-05, + "loss": 0.491, + "step": 1407 + }, + { + "epoch": 0.530920060331825, + "grad_norm": 0.8302839793793849, + "learning_rate": 7.999068101709997e-05, + "loss": 0.4869, + "step": 1408 + }, + { + "epoch": 0.5312971342383107, + "grad_norm": 0.7998723316280588, + "learning_rate": 7.999045234800333e-05, + "loss": 0.4731, + "step": 1409 + }, + { + "epoch": 0.5316742081447964, + "grad_norm": 0.7456396731261798, + "learning_rate": 7.999022090760383e-05, + "loss": 0.4822, + "step": 1410 + }, + { + "epoch": 0.532051282051282, + "grad_norm": 0.7816603727110372, + "learning_rate": 7.99899866959175e-05, + "loss": 0.4737, + "step": 1411 + }, + { + "epoch": 0.5324283559577677, + "grad_norm": 0.8918249009589232, + "learning_rate": 7.998974971296055e-05, + "loss": 0.4765, + "step": 1412 + }, + { + "epoch": 0.5328054298642534, + "grad_norm": 0.8737995203120164, + "learning_rate": 7.99895099587494e-05, + "loss": 0.4809, + "step": 1413 + }, + { + "epoch": 0.5331825037707391, + "grad_norm": 0.8968525535896656, + "learning_rate": 7.998926743330071e-05, + "loss": 0.4972, + "step": 1414 + }, + { + "epoch": 0.5335595776772247, + "grad_norm": 0.9185940248051112, + "learning_rate": 7.998902213663126e-05, + "loss": 0.4832, + "step": 1415 + }, + { + "epoch": 0.5339366515837104, + "grad_norm": 0.8177622861920024, + "learning_rate": 7.998877406875802e-05, + "loss": 0.4719, + "step": 1416 + }, + { + "epoch": 0.5343137254901961, + "grad_norm": 0.6434411850217306, + "learning_rate": 7.998852322969822e-05, + "loss": 0.4858, + "step": 1417 + }, + { + "epoch": 0.5346907993966817, + "grad_norm": 0.5962787937553028, + "learning_rate": 7.998826961946924e-05, + "loss": 0.4855, + "step": 1418 + }, + { + "epoch": 0.5350678733031674, + "grad_norm": 0.4791523632371744, + "learning_rate": 7.998801323808864e-05, + "loss": 0.4724, + "step": 1419 + }, + { + "epoch": 0.5354449472096531, + "grad_norm": 0.5337608904457118, + "learning_rate": 7.998775408557417e-05, + "loss": 0.4817, + "step": 1420 + }, + { + "epoch": 0.5358220211161387, + "grad_norm": 0.5583098177585438, + "learning_rate": 7.998749216194383e-05, + "loss": 0.4681, + "step": 1421 + }, + { + "epoch": 0.5361990950226244, + "grad_norm": 0.5694115995351904, + "learning_rate": 7.998722746721575e-05, + "loss": 0.4729, + "step": 1422 + }, + { + "epoch": 0.5365761689291101, + "grad_norm": 0.6660379471905487, + "learning_rate": 7.998696000140827e-05, + "loss": 0.4717, + "step": 1423 + }, + { + "epoch": 0.5369532428355958, + "grad_norm": 0.6176965271369911, + "learning_rate": 7.998668976453995e-05, + "loss": 0.4874, + "step": 1424 + }, + { + "epoch": 0.5373303167420814, + "grad_norm": 0.5846467074723352, + "learning_rate": 7.998641675662948e-05, + "loss": 0.4763, + "step": 1425 + }, + { + "epoch": 0.5377073906485671, + "grad_norm": 0.6202514991555441, + "learning_rate": 7.998614097769579e-05, + "loss": 0.48, + "step": 1426 + }, + { + "epoch": 0.5380844645550528, + "grad_norm": 0.5471403346976473, + "learning_rate": 7.9985862427758e-05, + "loss": 0.4673, + "step": 1427 + }, + { + "epoch": 0.5384615384615384, + "grad_norm": 0.5117820100925716, + "learning_rate": 7.998558110683542e-05, + "loss": 0.4798, + "step": 1428 + }, + { + "epoch": 0.5388386123680241, + "grad_norm": 0.5172021650228098, + "learning_rate": 7.998529701494754e-05, + "loss": 0.4813, + "step": 1429 + }, + { + "epoch": 0.5392156862745098, + "grad_norm": 0.33725439850036243, + "learning_rate": 7.998501015211404e-05, + "loss": 0.4743, + "step": 1430 + }, + { + "epoch": 0.5395927601809954, + "grad_norm": 0.47244501998384564, + "learning_rate": 7.99847205183548e-05, + "loss": 0.4755, + "step": 1431 + }, + { + "epoch": 0.5399698340874811, + "grad_norm": 0.5754021127376935, + "learning_rate": 7.99844281136899e-05, + "loss": 0.4765, + "step": 1432 + }, + { + "epoch": 0.5403469079939668, + "grad_norm": 0.5782952094248428, + "learning_rate": 7.998413293813958e-05, + "loss": 0.4671, + "step": 1433 + }, + { + "epoch": 0.5407239819004525, + "grad_norm": 0.6552245469180674, + "learning_rate": 7.998383499172433e-05, + "loss": 0.4794, + "step": 1434 + }, + { + "epoch": 0.5411010558069381, + "grad_norm": 0.7412852830512923, + "learning_rate": 7.998353427446478e-05, + "loss": 0.4763, + "step": 1435 + }, + { + "epoch": 0.5414781297134238, + "grad_norm": 0.7997932938921664, + "learning_rate": 7.998323078638177e-05, + "loss": 0.4789, + "step": 1436 + }, + { + "epoch": 0.5418552036199095, + "grad_norm": 1.0388257179125702, + "learning_rate": 7.998292452749633e-05, + "loss": 0.496, + "step": 1437 + }, + { + "epoch": 0.5422322775263951, + "grad_norm": 1.0594396055721895, + "learning_rate": 7.998261549782967e-05, + "loss": 0.4857, + "step": 1438 + }, + { + "epoch": 0.5426093514328808, + "grad_norm": 0.6784564651963081, + "learning_rate": 7.998230369740325e-05, + "loss": 0.4743, + "step": 1439 + }, + { + "epoch": 0.5429864253393665, + "grad_norm": 0.5330386669724242, + "learning_rate": 7.998198912623863e-05, + "loss": 0.4755, + "step": 1440 + }, + { + "epoch": 0.5433634992458521, + "grad_norm": 0.8534090350742813, + "learning_rate": 7.998167178435763e-05, + "loss": 0.473, + "step": 1441 + }, + { + "epoch": 0.5437405731523378, + "grad_norm": 0.9367043258179681, + "learning_rate": 7.998135167178224e-05, + "loss": 0.4769, + "step": 1442 + }, + { + "epoch": 0.5441176470588235, + "grad_norm": 0.6926836096321594, + "learning_rate": 7.998102878853464e-05, + "loss": 0.4627, + "step": 1443 + }, + { + "epoch": 0.5444947209653092, + "grad_norm": 0.392602416754377, + "learning_rate": 7.998070313463721e-05, + "loss": 0.4873, + "step": 1444 + }, + { + "epoch": 0.5448717948717948, + "grad_norm": 0.5046495413124926, + "learning_rate": 7.998037471011252e-05, + "loss": 0.4776, + "step": 1445 + }, + { + "epoch": 0.5452488687782805, + "grad_norm": 0.6496881319929627, + "learning_rate": 7.998004351498331e-05, + "loss": 0.4683, + "step": 1446 + }, + { + "epoch": 0.5456259426847662, + "grad_norm": 0.5924871212601873, + "learning_rate": 7.997970954927257e-05, + "loss": 0.485, + "step": 1447 + }, + { + "epoch": 0.5460030165912518, + "grad_norm": 0.4743701514719809, + "learning_rate": 7.997937281300341e-05, + "loss": 0.4805, + "step": 1448 + }, + { + "epoch": 0.5463800904977375, + "grad_norm": 0.5123318045257796, + "learning_rate": 7.997903330619918e-05, + "loss": 0.4812, + "step": 1449 + }, + { + "epoch": 0.5467571644042232, + "grad_norm": 0.5040455230591171, + "learning_rate": 7.997869102888338e-05, + "loss": 0.467, + "step": 1450 + }, + { + "epoch": 0.547134238310709, + "grad_norm": 0.4111616476157631, + "learning_rate": 7.997834598107978e-05, + "loss": 0.4863, + "step": 1451 + }, + { + "epoch": 0.5475113122171946, + "grad_norm": 0.40907750909077656, + "learning_rate": 7.997799816281225e-05, + "loss": 0.4649, + "step": 1452 + }, + { + "epoch": 0.5478883861236803, + "grad_norm": 0.4807023921467035, + "learning_rate": 7.99776475741049e-05, + "loss": 0.4713, + "step": 1453 + }, + { + "epoch": 0.548265460030166, + "grad_norm": 0.5067396317909691, + "learning_rate": 7.997729421498206e-05, + "loss": 0.4726, + "step": 1454 + }, + { + "epoch": 0.5486425339366516, + "grad_norm": 0.5745541852217015, + "learning_rate": 7.997693808546817e-05, + "loss": 0.4637, + "step": 1455 + }, + { + "epoch": 0.5490196078431373, + "grad_norm": 0.6406043243860928, + "learning_rate": 7.997657918558794e-05, + "loss": 0.4704, + "step": 1456 + }, + { + "epoch": 0.549396681749623, + "grad_norm": 0.666746103535613, + "learning_rate": 7.997621751536623e-05, + "loss": 0.4727, + "step": 1457 + }, + { + "epoch": 0.5497737556561086, + "grad_norm": 0.8124486670429929, + "learning_rate": 7.997585307482808e-05, + "loss": 0.4747, + "step": 1458 + }, + { + "epoch": 0.5501508295625943, + "grad_norm": 1.0702963506163068, + "learning_rate": 7.99754858639988e-05, + "loss": 0.4812, + "step": 1459 + }, + { + "epoch": 0.55052790346908, + "grad_norm": 1.0800782056809866, + "learning_rate": 7.997511588290381e-05, + "loss": 0.4834, + "step": 1460 + }, + { + "epoch": 0.5509049773755657, + "grad_norm": 0.8074559246631627, + "learning_rate": 7.997474313156874e-05, + "loss": 0.487, + "step": 1461 + }, + { + "epoch": 0.5512820512820513, + "grad_norm": 0.6105239642955472, + "learning_rate": 7.997436761001941e-05, + "loss": 0.4718, + "step": 1462 + }, + { + "epoch": 0.551659125188537, + "grad_norm": 0.44554629372626764, + "learning_rate": 7.99739893182819e-05, + "loss": 0.4823, + "step": 1463 + }, + { + "epoch": 0.5520361990950227, + "grad_norm": 0.38025642381191904, + "learning_rate": 7.997360825638236e-05, + "loss": 0.4771, + "step": 1464 + }, + { + "epoch": 0.5524132730015083, + "grad_norm": 0.4664646133552946, + "learning_rate": 7.997322442434723e-05, + "loss": 0.4719, + "step": 1465 + }, + { + "epoch": 0.552790346907994, + "grad_norm": 0.5619554581229572, + "learning_rate": 7.99728378222031e-05, + "loss": 0.4753, + "step": 1466 + }, + { + "epoch": 0.5531674208144797, + "grad_norm": 0.6264981258107845, + "learning_rate": 7.997244844997676e-05, + "loss": 0.4833, + "step": 1467 + }, + { + "epoch": 0.5535444947209653, + "grad_norm": 0.6747267438803288, + "learning_rate": 7.997205630769521e-05, + "loss": 0.4711, + "step": 1468 + }, + { + "epoch": 0.553921568627451, + "grad_norm": 0.7167592812171402, + "learning_rate": 7.997166139538561e-05, + "loss": 0.4758, + "step": 1469 + }, + { + "epoch": 0.5542986425339367, + "grad_norm": 0.7332762656505893, + "learning_rate": 7.997126371307533e-05, + "loss": 0.4766, + "step": 1470 + }, + { + "epoch": 0.5546757164404224, + "grad_norm": 0.7276673534421594, + "learning_rate": 7.997086326079192e-05, + "loss": 0.4741, + "step": 1471 + }, + { + "epoch": 0.555052790346908, + "grad_norm": 0.7066557542895613, + "learning_rate": 7.997046003856316e-05, + "loss": 0.4603, + "step": 1472 + }, + { + "epoch": 0.5554298642533937, + "grad_norm": 0.686233252190749, + "learning_rate": 7.997005404641695e-05, + "loss": 0.4828, + "step": 1473 + }, + { + "epoch": 0.5558069381598794, + "grad_norm": 0.6214582325345872, + "learning_rate": 7.996964528438146e-05, + "loss": 0.4791, + "step": 1474 + }, + { + "epoch": 0.556184012066365, + "grad_norm": 0.6223192870253453, + "learning_rate": 7.996923375248501e-05, + "loss": 0.4914, + "step": 1475 + }, + { + "epoch": 0.5565610859728507, + "grad_norm": 0.6389830689200264, + "learning_rate": 7.996881945075612e-05, + "loss": 0.4874, + "step": 1476 + }, + { + "epoch": 0.5569381598793364, + "grad_norm": 0.5039894711485227, + "learning_rate": 7.996840237922346e-05, + "loss": 0.4652, + "step": 1477 + }, + { + "epoch": 0.557315233785822, + "grad_norm": 0.5056875189990131, + "learning_rate": 7.9967982537916e-05, + "loss": 0.4731, + "step": 1478 + }, + { + "epoch": 0.5576923076923077, + "grad_norm": 0.6043201876718909, + "learning_rate": 7.996755992686278e-05, + "loss": 0.4787, + "step": 1479 + }, + { + "epoch": 0.5580693815987934, + "grad_norm": 0.6331716862284106, + "learning_rate": 7.996713454609313e-05, + "loss": 0.478, + "step": 1480 + }, + { + "epoch": 0.558446455505279, + "grad_norm": 0.6104911350250087, + "learning_rate": 7.99667063956365e-05, + "loss": 0.4817, + "step": 1481 + }, + { + "epoch": 0.5588235294117647, + "grad_norm": 0.5675485718119494, + "learning_rate": 7.996627547552256e-05, + "loss": 0.475, + "step": 1482 + }, + { + "epoch": 0.5592006033182504, + "grad_norm": 0.5560160797394696, + "learning_rate": 7.996584178578119e-05, + "loss": 0.4842, + "step": 1483 + }, + { + "epoch": 0.5595776772247361, + "grad_norm": 0.7184394153449768, + "learning_rate": 7.996540532644242e-05, + "loss": 0.4829, + "step": 1484 + }, + { + "epoch": 0.5599547511312217, + "grad_norm": 0.7835208025764682, + "learning_rate": 7.996496609753652e-05, + "loss": 0.4906, + "step": 1485 + }, + { + "epoch": 0.5603318250377074, + "grad_norm": 0.7198942856729424, + "learning_rate": 7.996452409909391e-05, + "loss": 0.4701, + "step": 1486 + }, + { + "epoch": 0.5607088989441931, + "grad_norm": 0.6446173976111265, + "learning_rate": 7.996407933114523e-05, + "loss": 0.4708, + "step": 1487 + }, + { + "epoch": 0.5610859728506787, + "grad_norm": 0.5940246731891792, + "learning_rate": 7.996363179372131e-05, + "loss": 0.4731, + "step": 1488 + }, + { + "epoch": 0.5614630467571644, + "grad_norm": 0.6569114342327754, + "learning_rate": 7.996318148685315e-05, + "loss": 0.4695, + "step": 1489 + }, + { + "epoch": 0.5618401206636501, + "grad_norm": 0.8950260308894038, + "learning_rate": 7.996272841057195e-05, + "loss": 0.4651, + "step": 1490 + }, + { + "epoch": 0.5622171945701357, + "grad_norm": 1.0259227710826644, + "learning_rate": 7.996227256490914e-05, + "loss": 0.4764, + "step": 1491 + }, + { + "epoch": 0.5625942684766214, + "grad_norm": 0.9792453882110047, + "learning_rate": 7.996181394989627e-05, + "loss": 0.4736, + "step": 1492 + }, + { + "epoch": 0.5629713423831071, + "grad_norm": 0.9086355189586287, + "learning_rate": 7.996135256556514e-05, + "loss": 0.475, + "step": 1493 + }, + { + "epoch": 0.5633484162895928, + "grad_norm": 0.7177052691541125, + "learning_rate": 7.996088841194774e-05, + "loss": 0.4736, + "step": 1494 + }, + { + "epoch": 0.5637254901960784, + "grad_norm": 0.4362313531991958, + "learning_rate": 7.99604214890762e-05, + "loss": 0.4757, + "step": 1495 + }, + { + "epoch": 0.5641025641025641, + "grad_norm": 0.48234043372955876, + "learning_rate": 7.99599517969829e-05, + "loss": 0.4816, + "step": 1496 + }, + { + "epoch": 0.5644796380090498, + "grad_norm": 0.5456293330051059, + "learning_rate": 7.995947933570038e-05, + "loss": 0.4752, + "step": 1497 + }, + { + "epoch": 0.5648567119155354, + "grad_norm": 0.6533220107044616, + "learning_rate": 7.995900410526139e-05, + "loss": 0.4735, + "step": 1498 + }, + { + "epoch": 0.5652337858220211, + "grad_norm": 0.707106088974176, + "learning_rate": 7.995852610569886e-05, + "loss": 0.4887, + "step": 1499 + }, + { + "epoch": 0.5656108597285068, + "grad_norm": 0.6867497003727033, + "learning_rate": 7.995804533704592e-05, + "loss": 0.4853, + "step": 1500 + }, + { + "epoch": 0.5659879336349924, + "grad_norm": 0.642123226630722, + "learning_rate": 7.995756179933587e-05, + "loss": 0.4712, + "step": 1501 + }, + { + "epoch": 0.5663650075414781, + "grad_norm": 0.5541029902097432, + "learning_rate": 7.995707549260223e-05, + "loss": 0.4805, + "step": 1502 + }, + { + "epoch": 0.5667420814479638, + "grad_norm": 0.5183593723283207, + "learning_rate": 7.995658641687872e-05, + "loss": 0.4895, + "step": 1503 + }, + { + "epoch": 0.5671191553544495, + "grad_norm": 0.5077787035059389, + "learning_rate": 7.99560945721992e-05, + "loss": 0.4766, + "step": 1504 + }, + { + "epoch": 0.5674962292609351, + "grad_norm": 0.5849989595557675, + "learning_rate": 7.995559995859775e-05, + "loss": 0.4865, + "step": 1505 + }, + { + "epoch": 0.5678733031674208, + "grad_norm": 0.616877441468394, + "learning_rate": 7.995510257610869e-05, + "loss": 0.484, + "step": 1506 + }, + { + "epoch": 0.5682503770739065, + "grad_norm": 0.5636145331492206, + "learning_rate": 7.995460242476646e-05, + "loss": 0.479, + "step": 1507 + }, + { + "epoch": 0.5686274509803921, + "grad_norm": 0.5307995283515434, + "learning_rate": 7.99540995046057e-05, + "loss": 0.4725, + "step": 1508 + }, + { + "epoch": 0.5690045248868778, + "grad_norm": 0.5933063432081148, + "learning_rate": 7.995359381566129e-05, + "loss": 0.4694, + "step": 1509 + }, + { + "epoch": 0.5693815987933635, + "grad_norm": 0.7613835076385642, + "learning_rate": 7.995308535796828e-05, + "loss": 0.4856, + "step": 1510 + }, + { + "epoch": 0.5697586726998491, + "grad_norm": 0.9598572582205825, + "learning_rate": 7.995257413156188e-05, + "loss": 0.4973, + "step": 1511 + }, + { + "epoch": 0.5701357466063348, + "grad_norm": 1.0193478245581733, + "learning_rate": 7.995206013647754e-05, + "loss": 0.4799, + "step": 1512 + }, + { + "epoch": 0.5705128205128205, + "grad_norm": 0.8339407613920861, + "learning_rate": 7.995154337275087e-05, + "loss": 0.4859, + "step": 1513 + }, + { + "epoch": 0.5708898944193062, + "grad_norm": 0.6993626018797737, + "learning_rate": 7.995102384041767e-05, + "loss": 0.4819, + "step": 1514 + }, + { + "epoch": 0.5712669683257918, + "grad_norm": 0.7027512663149251, + "learning_rate": 7.995050153951397e-05, + "loss": 0.4825, + "step": 1515 + }, + { + "epoch": 0.5716440422322775, + "grad_norm": 0.7209572028692107, + "learning_rate": 7.994997647007594e-05, + "loss": 0.47, + "step": 1516 + }, + { + "epoch": 0.5720211161387632, + "grad_norm": 0.6672646620769189, + "learning_rate": 7.994944863213999e-05, + "loss": 0.4692, + "step": 1517 + }, + { + "epoch": 0.5723981900452488, + "grad_norm": 0.6849826349481714, + "learning_rate": 7.994891802574266e-05, + "loss": 0.4781, + "step": 1518 + }, + { + "epoch": 0.5727752639517345, + "grad_norm": 0.6560680840620954, + "learning_rate": 7.994838465092076e-05, + "loss": 0.4717, + "step": 1519 + }, + { + "epoch": 0.5731523378582202, + "grad_norm": 0.5980383606481835, + "learning_rate": 7.994784850771124e-05, + "loss": 0.4728, + "step": 1520 + }, + { + "epoch": 0.5735294117647058, + "grad_norm": 0.6139106030945265, + "learning_rate": 7.994730959615125e-05, + "loss": 0.4591, + "step": 1521 + }, + { + "epoch": 0.5739064856711915, + "grad_norm": 0.6754754225056999, + "learning_rate": 7.994676791627813e-05, + "loss": 0.4663, + "step": 1522 + }, + { + "epoch": 0.5742835595776772, + "grad_norm": 0.7168270107764781, + "learning_rate": 7.994622346812943e-05, + "loss": 0.4712, + "step": 1523 + }, + { + "epoch": 0.5746606334841629, + "grad_norm": 0.7066047321382108, + "learning_rate": 7.994567625174288e-05, + "loss": 0.4693, + "step": 1524 + }, + { + "epoch": 0.5750377073906485, + "grad_norm": 0.7796009265729842, + "learning_rate": 7.99451262671564e-05, + "loss": 0.4715, + "step": 1525 + }, + { + "epoch": 0.5754147812971342, + "grad_norm": 0.8954341261050641, + "learning_rate": 7.99445735144081e-05, + "loss": 0.4644, + "step": 1526 + }, + { + "epoch": 0.5757918552036199, + "grad_norm": 0.8709196219803621, + "learning_rate": 7.994401799353629e-05, + "loss": 0.4561, + "step": 1527 + }, + { + "epoch": 0.5761689291101055, + "grad_norm": 0.6381815228417985, + "learning_rate": 7.994345970457945e-05, + "loss": 0.4762, + "step": 1528 + }, + { + "epoch": 0.5765460030165912, + "grad_norm": 0.48559785983760967, + "learning_rate": 7.994289864757628e-05, + "loss": 0.4792, + "step": 1529 + }, + { + "epoch": 0.5769230769230769, + "grad_norm": 0.5954101542587852, + "learning_rate": 7.994233482256567e-05, + "loss": 0.4718, + "step": 1530 + }, + { + "epoch": 0.5773001508295625, + "grad_norm": 0.5509711644238957, + "learning_rate": 7.99417682295867e-05, + "loss": 0.4878, + "step": 1531 + }, + { + "epoch": 0.5776772247360482, + "grad_norm": 0.509079566006673, + "learning_rate": 7.994119886867862e-05, + "loss": 0.4714, + "step": 1532 + }, + { + "epoch": 0.5780542986425339, + "grad_norm": 0.43171068754472425, + "learning_rate": 7.994062673988086e-05, + "loss": 0.4786, + "step": 1533 + }, + { + "epoch": 0.5784313725490197, + "grad_norm": 0.4856335372569009, + "learning_rate": 7.99400518432331e-05, + "loss": 0.4768, + "step": 1534 + }, + { + "epoch": 0.5788084464555053, + "grad_norm": 0.4947168073093983, + "learning_rate": 7.993947417877518e-05, + "loss": 0.4703, + "step": 1535 + }, + { + "epoch": 0.579185520361991, + "grad_norm": 0.5341967106253849, + "learning_rate": 7.993889374654714e-05, + "loss": 0.4629, + "step": 1536 + }, + { + "epoch": 0.5795625942684767, + "grad_norm": 0.5526607868012797, + "learning_rate": 7.993831054658917e-05, + "loss": 0.4645, + "step": 1537 + }, + { + "epoch": 0.5799396681749623, + "grad_norm": 0.6446835621449969, + "learning_rate": 7.99377245789417e-05, + "loss": 0.4686, + "step": 1538 + }, + { + "epoch": 0.580316742081448, + "grad_norm": 0.6967654712640877, + "learning_rate": 7.993713584364536e-05, + "loss": 0.4806, + "step": 1539 + }, + { + "epoch": 0.5806938159879337, + "grad_norm": 0.6482133599653286, + "learning_rate": 7.993654434074093e-05, + "loss": 0.4886, + "step": 1540 + }, + { + "epoch": 0.5810708898944194, + "grad_norm": 0.5959217573190718, + "learning_rate": 7.993595007026939e-05, + "loss": 0.4611, + "step": 1541 + }, + { + "epoch": 0.581447963800905, + "grad_norm": 0.6376890462822675, + "learning_rate": 7.993535303227195e-05, + "loss": 0.4703, + "step": 1542 + }, + { + "epoch": 0.5818250377073907, + "grad_norm": 0.6707881829574597, + "learning_rate": 7.993475322678995e-05, + "loss": 0.4691, + "step": 1543 + }, + { + "epoch": 0.5822021116138764, + "grad_norm": 0.6767529324799344, + "learning_rate": 7.993415065386498e-05, + "loss": 0.4727, + "step": 1544 + }, + { + "epoch": 0.582579185520362, + "grad_norm": 0.6394941921405465, + "learning_rate": 7.993354531353879e-05, + "loss": 0.4677, + "step": 1545 + }, + { + "epoch": 0.5829562594268477, + "grad_norm": 0.601546254317899, + "learning_rate": 7.993293720585332e-05, + "loss": 0.4695, + "step": 1546 + }, + { + "epoch": 0.5833333333333334, + "grad_norm": 0.6044251673384743, + "learning_rate": 7.993232633085074e-05, + "loss": 0.4804, + "step": 1547 + }, + { + "epoch": 0.583710407239819, + "grad_norm": 0.5617407159635676, + "learning_rate": 7.993171268857335e-05, + "loss": 0.4783, + "step": 1548 + }, + { + "epoch": 0.5840874811463047, + "grad_norm": 0.5949806477399493, + "learning_rate": 7.993109627906369e-05, + "loss": 0.4656, + "step": 1549 + }, + { + "epoch": 0.5844645550527904, + "grad_norm": 0.691448056227597, + "learning_rate": 7.99304771023645e-05, + "loss": 0.4763, + "step": 1550 + }, + { + "epoch": 0.584841628959276, + "grad_norm": 0.7581399947155643, + "learning_rate": 7.992985515851864e-05, + "loss": 0.4679, + "step": 1551 + }, + { + "epoch": 0.5852187028657617, + "grad_norm": 0.81193167370962, + "learning_rate": 7.992923044756926e-05, + "loss": 0.4668, + "step": 1552 + }, + { + "epoch": 0.5855957767722474, + "grad_norm": 0.728120668060209, + "learning_rate": 7.992860296955959e-05, + "loss": 0.4595, + "step": 1553 + }, + { + "epoch": 0.5859728506787331, + "grad_norm": 0.5777149881257772, + "learning_rate": 7.992797272453319e-05, + "loss": 0.463, + "step": 1554 + }, + { + "epoch": 0.5863499245852187, + "grad_norm": 0.6059388336739102, + "learning_rate": 7.992733971253368e-05, + "loss": 0.4861, + "step": 1555 + }, + { + "epoch": 0.5867269984917044, + "grad_norm": 0.6724958362356178, + "learning_rate": 7.992670393360494e-05, + "loss": 0.4643, + "step": 1556 + }, + { + "epoch": 0.5871040723981901, + "grad_norm": 0.5985504690012736, + "learning_rate": 7.992606538779103e-05, + "loss": 0.4763, + "step": 1557 + }, + { + "epoch": 0.5874811463046757, + "grad_norm": 0.40731314993248635, + "learning_rate": 7.99254240751362e-05, + "loss": 0.4749, + "step": 1558 + }, + { + "epoch": 0.5878582202111614, + "grad_norm": 0.3831273618712776, + "learning_rate": 7.99247799956849e-05, + "loss": 0.4629, + "step": 1559 + }, + { + "epoch": 0.5882352941176471, + "grad_norm": 0.44938574874754983, + "learning_rate": 7.992413314948177e-05, + "loss": 0.4625, + "step": 1560 + }, + { + "epoch": 0.5886123680241327, + "grad_norm": 0.5490760746049397, + "learning_rate": 7.992348353657162e-05, + "loss": 0.4835, + "step": 1561 + }, + { + "epoch": 0.5889894419306184, + "grad_norm": 0.5103002153407, + "learning_rate": 7.992283115699947e-05, + "loss": 0.4758, + "step": 1562 + }, + { + "epoch": 0.5893665158371041, + "grad_norm": 0.510934989427715, + "learning_rate": 7.992217601081054e-05, + "loss": 0.4738, + "step": 1563 + }, + { + "epoch": 0.5897435897435898, + "grad_norm": 0.5892166917991811, + "learning_rate": 7.99215180980502e-05, + "loss": 0.4643, + "step": 1564 + }, + { + "epoch": 0.5901206636500754, + "grad_norm": 0.6382723642017842, + "learning_rate": 7.992085741876409e-05, + "loss": 0.4627, + "step": 1565 + }, + { + "epoch": 0.5904977375565611, + "grad_norm": 0.7893142179581003, + "learning_rate": 7.992019397299796e-05, + "loss": 0.4638, + "step": 1566 + }, + { + "epoch": 0.5908748114630468, + "grad_norm": 0.9842882852110918, + "learning_rate": 7.99195277607978e-05, + "loss": 0.4777, + "step": 1567 + }, + { + "epoch": 0.5912518853695324, + "grad_norm": 1.0847267938677188, + "learning_rate": 7.991885878220978e-05, + "loss": 0.4706, + "step": 1568 + }, + { + "epoch": 0.5916289592760181, + "grad_norm": 0.815749704294401, + "learning_rate": 7.991818703728024e-05, + "loss": 0.4797, + "step": 1569 + }, + { + "epoch": 0.5920060331825038, + "grad_norm": 0.6100250227481144, + "learning_rate": 7.991751252605576e-05, + "loss": 0.472, + "step": 1570 + }, + { + "epoch": 0.5923831070889894, + "grad_norm": 0.5800276925917115, + "learning_rate": 7.991683524858306e-05, + "loss": 0.4616, + "step": 1571 + }, + { + "epoch": 0.5927601809954751, + "grad_norm": 0.5263258689852558, + "learning_rate": 7.991615520490909e-05, + "loss": 0.4738, + "step": 1572 + }, + { + "epoch": 0.5931372549019608, + "grad_norm": 0.4272720701311445, + "learning_rate": 7.991547239508098e-05, + "loss": 0.469, + "step": 1573 + }, + { + "epoch": 0.5935143288084465, + "grad_norm": 0.3843526852576446, + "learning_rate": 7.991478681914602e-05, + "loss": 0.4675, + "step": 1574 + }, + { + "epoch": 0.5938914027149321, + "grad_norm": 0.3971964976530352, + "learning_rate": 7.991409847715175e-05, + "loss": 0.4734, + "step": 1575 + }, + { + "epoch": 0.5942684766214178, + "grad_norm": 0.36298703154100526, + "learning_rate": 7.991340736914584e-05, + "loss": 0.4801, + "step": 1576 + }, + { + "epoch": 0.5946455505279035, + "grad_norm": 0.38278410408311087, + "learning_rate": 7.991271349517622e-05, + "loss": 0.4711, + "step": 1577 + }, + { + "epoch": 0.5950226244343891, + "grad_norm": 0.43965011094744183, + "learning_rate": 7.991201685529094e-05, + "loss": 0.4747, + "step": 1578 + }, + { + "epoch": 0.5953996983408748, + "grad_norm": 0.5142580779628413, + "learning_rate": 7.991131744953832e-05, + "loss": 0.4756, + "step": 1579 + }, + { + "epoch": 0.5957767722473605, + "grad_norm": 0.5408278334235851, + "learning_rate": 7.991061527796679e-05, + "loss": 0.4688, + "step": 1580 + }, + { + "epoch": 0.5961538461538461, + "grad_norm": 0.48449026923570576, + "learning_rate": 7.990991034062501e-05, + "loss": 0.4634, + "step": 1581 + }, + { + "epoch": 0.5965309200603318, + "grad_norm": 0.47257990991099785, + "learning_rate": 7.990920263756185e-05, + "loss": 0.4717, + "step": 1582 + }, + { + "epoch": 0.5969079939668175, + "grad_norm": 0.507819403225788, + "learning_rate": 7.990849216882636e-05, + "loss": 0.4668, + "step": 1583 + }, + { + "epoch": 0.5972850678733032, + "grad_norm": 0.6416995896790122, + "learning_rate": 7.990777893446774e-05, + "loss": 0.4617, + "step": 1584 + }, + { + "epoch": 0.5976621417797888, + "grad_norm": 0.8512879867210096, + "learning_rate": 7.990706293453546e-05, + "loss": 0.487, + "step": 1585 + }, + { + "epoch": 0.5980392156862745, + "grad_norm": 1.0356426363949296, + "learning_rate": 7.99063441690791e-05, + "loss": 0.4758, + "step": 1586 + }, + { + "epoch": 0.5984162895927602, + "grad_norm": 1.0658559634324163, + "learning_rate": 7.990562263814849e-05, + "loss": 0.4815, + "step": 1587 + }, + { + "epoch": 0.5987933634992458, + "grad_norm": 0.9209398647635096, + "learning_rate": 7.990489834179363e-05, + "loss": 0.4805, + "step": 1588 + }, + { + "epoch": 0.5991704374057315, + "grad_norm": 0.8678001857795585, + "learning_rate": 7.99041712800647e-05, + "loss": 0.4624, + "step": 1589 + }, + { + "epoch": 0.5995475113122172, + "grad_norm": 0.9356097051514184, + "learning_rate": 7.990344145301212e-05, + "loss": 0.4729, + "step": 1590 + }, + { + "epoch": 0.5999245852187028, + "grad_norm": 0.9554989427585765, + "learning_rate": 7.990270886068643e-05, + "loss": 0.4631, + "step": 1591 + }, + { + "epoch": 0.6003016591251885, + "grad_norm": 0.7665856511599173, + "learning_rate": 7.990197350313839e-05, + "loss": 0.4832, + "step": 1592 + }, + { + "epoch": 0.6006787330316742, + "grad_norm": 0.5615981197502721, + "learning_rate": 7.9901235380419e-05, + "loss": 0.475, + "step": 1593 + }, + { + "epoch": 0.6010558069381599, + "grad_norm": 0.46073729025617355, + "learning_rate": 7.990049449257939e-05, + "loss": 0.4686, + "step": 1594 + }, + { + "epoch": 0.6014328808446455, + "grad_norm": 0.6187851311532422, + "learning_rate": 7.989975083967089e-05, + "loss": 0.474, + "step": 1595 + }, + { + "epoch": 0.6018099547511312, + "grad_norm": 0.7047257827159044, + "learning_rate": 7.989900442174506e-05, + "loss": 0.4744, + "step": 1596 + }, + { + "epoch": 0.6021870286576169, + "grad_norm": 0.4696790207845636, + "learning_rate": 7.989825523885363e-05, + "loss": 0.4732, + "step": 1597 + }, + { + "epoch": 0.6025641025641025, + "grad_norm": 0.39994093230247096, + "learning_rate": 7.989750329104847e-05, + "loss": 0.4761, + "step": 1598 + }, + { + "epoch": 0.6029411764705882, + "grad_norm": 0.5785554424456008, + "learning_rate": 7.989674857838173e-05, + "loss": 0.4691, + "step": 1599 + }, + { + "epoch": 0.6033182503770739, + "grad_norm": 0.6247920614478666, + "learning_rate": 7.98959911009057e-05, + "loss": 0.4647, + "step": 1600 + }, + { + "epoch": 0.6036953242835595, + "grad_norm": 0.45167832414784176, + "learning_rate": 7.98952308586729e-05, + "loss": 0.4687, + "step": 1601 + }, + { + "epoch": 0.6040723981900452, + "grad_norm": 0.4850464891532421, + "learning_rate": 7.989446785173595e-05, + "loss": 0.4701, + "step": 1602 + }, + { + "epoch": 0.6044494720965309, + "grad_norm": 0.5256618380255811, + "learning_rate": 7.989370208014779e-05, + "loss": 0.4699, + "step": 1603 + }, + { + "epoch": 0.6048265460030166, + "grad_norm": 0.38471846476224303, + "learning_rate": 7.989293354396145e-05, + "loss": 0.4776, + "step": 1604 + }, + { + "epoch": 0.6052036199095022, + "grad_norm": 0.3736326838747458, + "learning_rate": 7.98921622432302e-05, + "loss": 0.4685, + "step": 1605 + }, + { + "epoch": 0.6055806938159879, + "grad_norm": 0.5864551157240054, + "learning_rate": 7.98913881780075e-05, + "loss": 0.4663, + "step": 1606 + }, + { + "epoch": 0.6059577677224736, + "grad_norm": 0.6239475439330517, + "learning_rate": 7.989061134834698e-05, + "loss": 0.4667, + "step": 1607 + }, + { + "epoch": 0.6063348416289592, + "grad_norm": 0.6128069289927538, + "learning_rate": 7.988983175430248e-05, + "loss": 0.4702, + "step": 1608 + }, + { + "epoch": 0.6067119155354449, + "grad_norm": 0.7568864171335671, + "learning_rate": 7.9889049395928e-05, + "loss": 0.4618, + "step": 1609 + }, + { + "epoch": 0.6070889894419306, + "grad_norm": 0.8167041489034103, + "learning_rate": 7.98882642732778e-05, + "loss": 0.4707, + "step": 1610 + }, + { + "epoch": 0.6074660633484162, + "grad_norm": 0.8401104389990204, + "learning_rate": 7.988747638640628e-05, + "loss": 0.4805, + "step": 1611 + }, + { + "epoch": 0.6078431372549019, + "grad_norm": 0.8916293845312628, + "learning_rate": 7.9886685735368e-05, + "loss": 0.4735, + "step": 1612 + }, + { + "epoch": 0.6082202111613876, + "grad_norm": 0.8852705178073296, + "learning_rate": 7.988589232021778e-05, + "loss": 0.4756, + "step": 1613 + }, + { + "epoch": 0.6085972850678733, + "grad_norm": 0.8531803780479141, + "learning_rate": 7.988509614101061e-05, + "loss": 0.4765, + "step": 1614 + }, + { + "epoch": 0.6089743589743589, + "grad_norm": 0.8081685271056336, + "learning_rate": 7.988429719780166e-05, + "loss": 0.4698, + "step": 1615 + }, + { + "epoch": 0.6093514328808446, + "grad_norm": 0.615709870939707, + "learning_rate": 7.988349549064627e-05, + "loss": 0.4794, + "step": 1616 + }, + { + "epoch": 0.6097285067873304, + "grad_norm": 0.4923229558773908, + "learning_rate": 7.988269101960004e-05, + "loss": 0.4765, + "step": 1617 + }, + { + "epoch": 0.610105580693816, + "grad_norm": 0.6228548037839621, + "learning_rate": 7.988188378471868e-05, + "loss": 0.4748, + "step": 1618 + }, + { + "epoch": 0.6104826546003017, + "grad_norm": 0.5879312312942846, + "learning_rate": 7.988107378605817e-05, + "loss": 0.4669, + "step": 1619 + }, + { + "epoch": 0.6108597285067874, + "grad_norm": 0.4431953784747496, + "learning_rate": 7.98802610236746e-05, + "loss": 0.4723, + "step": 1620 + }, + { + "epoch": 0.611236802413273, + "grad_norm": 0.49605330998997205, + "learning_rate": 7.987944549762434e-05, + "loss": 0.4697, + "step": 1621 + }, + { + "epoch": 0.6116138763197587, + "grad_norm": 0.530015621809967, + "learning_rate": 7.987862720796387e-05, + "loss": 0.4749, + "step": 1622 + }, + { + "epoch": 0.6119909502262444, + "grad_norm": 0.533210586667907, + "learning_rate": 7.98778061547499e-05, + "loss": 0.4866, + "step": 1623 + }, + { + "epoch": 0.6123680241327301, + "grad_norm": 0.5692587287140808, + "learning_rate": 7.987698233803933e-05, + "loss": 0.4611, + "step": 1624 + }, + { + "epoch": 0.6127450980392157, + "grad_norm": 0.6505914310243933, + "learning_rate": 7.987615575788926e-05, + "loss": 0.4663, + "step": 1625 + }, + { + "epoch": 0.6131221719457014, + "grad_norm": 0.7397780892730262, + "learning_rate": 7.987532641435697e-05, + "loss": 0.4832, + "step": 1626 + }, + { + "epoch": 0.6134992458521871, + "grad_norm": 0.7781307836232441, + "learning_rate": 7.987449430749994e-05, + "loss": 0.4739, + "step": 1627 + }, + { + "epoch": 0.6138763197586727, + "grad_norm": 0.7562260195527164, + "learning_rate": 7.98736594373758e-05, + "loss": 0.4653, + "step": 1628 + }, + { + "epoch": 0.6142533936651584, + "grad_norm": 0.7444012654946375, + "learning_rate": 7.987282180404244e-05, + "loss": 0.4736, + "step": 1629 + }, + { + "epoch": 0.6146304675716441, + "grad_norm": 0.6221441093937823, + "learning_rate": 7.98719814075579e-05, + "loss": 0.4701, + "step": 1630 + }, + { + "epoch": 0.6150075414781297, + "grad_norm": 0.4567848278769926, + "learning_rate": 7.987113824798041e-05, + "loss": 0.4842, + "step": 1631 + }, + { + "epoch": 0.6153846153846154, + "grad_norm": 0.5061067831342124, + "learning_rate": 7.987029232536841e-05, + "loss": 0.4652, + "step": 1632 + }, + { + "epoch": 0.6157616892911011, + "grad_norm": 0.5143045677826533, + "learning_rate": 7.986944363978051e-05, + "loss": 0.4706, + "step": 1633 + }, + { + "epoch": 0.6161387631975868, + "grad_norm": 0.41754333824709244, + "learning_rate": 7.986859219127554e-05, + "loss": 0.4642, + "step": 1634 + }, + { + "epoch": 0.6165158371040724, + "grad_norm": 0.35609208890745003, + "learning_rate": 7.986773797991248e-05, + "loss": 0.4674, + "step": 1635 + }, + { + "epoch": 0.6168929110105581, + "grad_norm": 0.4407963309986366, + "learning_rate": 7.986688100575054e-05, + "loss": 0.4761, + "step": 1636 + }, + { + "epoch": 0.6172699849170438, + "grad_norm": 0.5509059332378756, + "learning_rate": 7.986602126884912e-05, + "loss": 0.4757, + "step": 1637 + }, + { + "epoch": 0.6176470588235294, + "grad_norm": 0.508447539698664, + "learning_rate": 7.986515876926777e-05, + "loss": 0.4785, + "step": 1638 + }, + { + "epoch": 0.6180241327300151, + "grad_norm": 0.4138787086945286, + "learning_rate": 7.98642935070663e-05, + "loss": 0.4668, + "step": 1639 + }, + { + "epoch": 0.6184012066365008, + "grad_norm": 0.40775542461077624, + "learning_rate": 7.986342548230465e-05, + "loss": 0.4676, + "step": 1640 + }, + { + "epoch": 0.6187782805429864, + "grad_norm": 0.48522294556696616, + "learning_rate": 7.986255469504295e-05, + "loss": 0.4568, + "step": 1641 + }, + { + "epoch": 0.6191553544494721, + "grad_norm": 0.5620324061135203, + "learning_rate": 7.986168114534159e-05, + "loss": 0.4764, + "step": 1642 + }, + { + "epoch": 0.6195324283559578, + "grad_norm": 0.6681492594836523, + "learning_rate": 7.986080483326108e-05, + "loss": 0.4755, + "step": 1643 + }, + { + "epoch": 0.6199095022624435, + "grad_norm": 0.7526575683440435, + "learning_rate": 7.985992575886213e-05, + "loss": 0.4733, + "step": 1644 + }, + { + "epoch": 0.6202865761689291, + "grad_norm": 0.916287143576456, + "learning_rate": 7.985904392220571e-05, + "loss": 0.4776, + "step": 1645 + }, + { + "epoch": 0.6206636500754148, + "grad_norm": 1.081235419861703, + "learning_rate": 7.985815932335288e-05, + "loss": 0.4735, + "step": 1646 + }, + { + "epoch": 0.6210407239819005, + "grad_norm": 0.8891874642710653, + "learning_rate": 7.985727196236497e-05, + "loss": 0.4719, + "step": 1647 + }, + { + "epoch": 0.6214177978883861, + "grad_norm": 0.6494769599341678, + "learning_rate": 7.985638183930346e-05, + "loss": 0.4669, + "step": 1648 + }, + { + "epoch": 0.6217948717948718, + "grad_norm": 0.48424394067898135, + "learning_rate": 7.985548895423006e-05, + "loss": 0.4737, + "step": 1649 + }, + { + "epoch": 0.6221719457013575, + "grad_norm": 0.406625508832543, + "learning_rate": 7.985459330720661e-05, + "loss": 0.4624, + "step": 1650 + }, + { + "epoch": 0.6225490196078431, + "grad_norm": 0.5322764776487544, + "learning_rate": 7.98536948982952e-05, + "loss": 0.4807, + "step": 1651 + }, + { + "epoch": 0.6229260935143288, + "grad_norm": 0.6547154242901908, + "learning_rate": 7.985279372755809e-05, + "loss": 0.4795, + "step": 1652 + }, + { + "epoch": 0.6233031674208145, + "grad_norm": 0.6210104370896493, + "learning_rate": 7.985188979505772e-05, + "loss": 0.4666, + "step": 1653 + }, + { + "epoch": 0.6236802413273002, + "grad_norm": 0.6789710963139233, + "learning_rate": 7.985098310085675e-05, + "loss": 0.4732, + "step": 1654 + }, + { + "epoch": 0.6240573152337858, + "grad_norm": 0.7303250293675343, + "learning_rate": 7.985007364501797e-05, + "loss": 0.4692, + "step": 1655 + }, + { + "epoch": 0.6244343891402715, + "grad_norm": 0.6783631964063743, + "learning_rate": 7.984916142760444e-05, + "loss": 0.4708, + "step": 1656 + }, + { + "epoch": 0.6248114630467572, + "grad_norm": 0.6514016785786284, + "learning_rate": 7.984824644867937e-05, + "loss": 0.4797, + "step": 1657 + }, + { + "epoch": 0.6251885369532428, + "grad_norm": 0.6136997880840325, + "learning_rate": 7.984732870830618e-05, + "loss": 0.4544, + "step": 1658 + }, + { + "epoch": 0.6255656108597285, + "grad_norm": 0.5741735205837548, + "learning_rate": 7.984640820654845e-05, + "loss": 0.4807, + "step": 1659 + }, + { + "epoch": 0.6259426847662142, + "grad_norm": 0.5513490871474441, + "learning_rate": 7.984548494346998e-05, + "loss": 0.4711, + "step": 1660 + }, + { + "epoch": 0.6263197586726998, + "grad_norm": 0.5916911003911975, + "learning_rate": 7.984455891913473e-05, + "loss": 0.4567, + "step": 1661 + }, + { + "epoch": 0.6266968325791855, + "grad_norm": 0.5880070685003534, + "learning_rate": 7.984363013360689e-05, + "loss": 0.4758, + "step": 1662 + }, + { + "epoch": 0.6270739064856712, + "grad_norm": 0.46592360765497665, + "learning_rate": 7.984269858695085e-05, + "loss": 0.4657, + "step": 1663 + }, + { + "epoch": 0.6274509803921569, + "grad_norm": 0.42593262846317675, + "learning_rate": 7.984176427923111e-05, + "loss": 0.4779, + "step": 1664 + }, + { + "epoch": 0.6278280542986425, + "grad_norm": 0.46655765540187083, + "learning_rate": 7.984082721051244e-05, + "loss": 0.4751, + "step": 1665 + }, + { + "epoch": 0.6282051282051282, + "grad_norm": 0.5437657272442815, + "learning_rate": 7.983988738085981e-05, + "loss": 0.4699, + "step": 1666 + }, + { + "epoch": 0.6285822021116139, + "grad_norm": 0.6490533822819712, + "learning_rate": 7.983894479033831e-05, + "loss": 0.4731, + "step": 1667 + }, + { + "epoch": 0.6289592760180995, + "grad_norm": 0.7778988942143306, + "learning_rate": 7.983799943901327e-05, + "loss": 0.4792, + "step": 1668 + }, + { + "epoch": 0.6293363499245852, + "grad_norm": 0.9005630964679507, + "learning_rate": 7.98370513269502e-05, + "loss": 0.4688, + "step": 1669 + }, + { + "epoch": 0.6297134238310709, + "grad_norm": 0.9688914750512077, + "learning_rate": 7.983610045421482e-05, + "loss": 0.4815, + "step": 1670 + }, + { + "epoch": 0.6300904977375565, + "grad_norm": 1.040119247405087, + "learning_rate": 7.983514682087302e-05, + "loss": 0.4675, + "step": 1671 + }, + { + "epoch": 0.6304675716440422, + "grad_norm": 0.9355683531877486, + "learning_rate": 7.983419042699086e-05, + "loss": 0.4718, + "step": 1672 + }, + { + "epoch": 0.6308446455505279, + "grad_norm": 0.8542809778689594, + "learning_rate": 7.983323127263465e-05, + "loss": 0.4753, + "step": 1673 + }, + { + "epoch": 0.6312217194570136, + "grad_norm": 0.6995080808595437, + "learning_rate": 7.983226935787084e-05, + "loss": 0.4647, + "step": 1674 + }, + { + "epoch": 0.6315987933634992, + "grad_norm": 0.4089614336756231, + "learning_rate": 7.983130468276608e-05, + "loss": 0.4601, + "step": 1675 + }, + { + "epoch": 0.6319758672699849, + "grad_norm": 0.5051161868891796, + "learning_rate": 7.983033724738726e-05, + "loss": 0.4675, + "step": 1676 + }, + { + "epoch": 0.6323529411764706, + "grad_norm": 0.6109308984390054, + "learning_rate": 7.982936705180139e-05, + "loss": 0.4647, + "step": 1677 + }, + { + "epoch": 0.6327300150829562, + "grad_norm": 0.4995924180014721, + "learning_rate": 7.98283940960757e-05, + "loss": 0.4674, + "step": 1678 + }, + { + "epoch": 0.6331070889894419, + "grad_norm": 0.39217795091342283, + "learning_rate": 7.982741838027764e-05, + "loss": 0.4727, + "step": 1679 + }, + { + "epoch": 0.6334841628959276, + "grad_norm": 0.4479940613272079, + "learning_rate": 7.982643990447482e-05, + "loss": 0.4622, + "step": 1680 + }, + { + "epoch": 0.6338612368024132, + "grad_norm": 0.6469632755757067, + "learning_rate": 7.982545866873502e-05, + "loss": 0.4717, + "step": 1681 + }, + { + "epoch": 0.6342383107088989, + "grad_norm": 0.6537303525214619, + "learning_rate": 7.982447467312627e-05, + "loss": 0.4517, + "step": 1682 + }, + { + "epoch": 0.6346153846153846, + "grad_norm": 0.49526045755515474, + "learning_rate": 7.982348791771675e-05, + "loss": 0.4856, + "step": 1683 + }, + { + "epoch": 0.6349924585218703, + "grad_norm": 0.5164268895330459, + "learning_rate": 7.982249840257485e-05, + "loss": 0.4823, + "step": 1684 + }, + { + "epoch": 0.6353695324283559, + "grad_norm": 0.7213403717472001, + "learning_rate": 7.982150612776911e-05, + "loss": 0.4843, + "step": 1685 + }, + { + "epoch": 0.6357466063348416, + "grad_norm": 0.6544660366311085, + "learning_rate": 7.982051109336834e-05, + "loss": 0.4639, + "step": 1686 + }, + { + "epoch": 0.6361236802413273, + "grad_norm": 0.50271653799416, + "learning_rate": 7.981951329944146e-05, + "loss": 0.4566, + "step": 1687 + }, + { + "epoch": 0.6365007541478129, + "grad_norm": 0.6589029662016969, + "learning_rate": 7.981851274605764e-05, + "loss": 0.4804, + "step": 1688 + }, + { + "epoch": 0.6368778280542986, + "grad_norm": 0.803996250667392, + "learning_rate": 7.981750943328619e-05, + "loss": 0.4698, + "step": 1689 + }, + { + "epoch": 0.6372549019607843, + "grad_norm": 0.8717872630217057, + "learning_rate": 7.981650336119666e-05, + "loss": 0.4533, + "step": 1690 + }, + { + "epoch": 0.6376319758672699, + "grad_norm": 0.9008648812314654, + "learning_rate": 7.981549452985876e-05, + "loss": 0.4839, + "step": 1691 + }, + { + "epoch": 0.6380090497737556, + "grad_norm": 0.883949082716059, + "learning_rate": 7.981448293934241e-05, + "loss": 0.4664, + "step": 1692 + }, + { + "epoch": 0.6383861236802413, + "grad_norm": 0.7956744865458131, + "learning_rate": 7.981346858971772e-05, + "loss": 0.4788, + "step": 1693 + }, + { + "epoch": 0.638763197586727, + "grad_norm": 0.6333523121726786, + "learning_rate": 7.981245148105495e-05, + "loss": 0.476, + "step": 1694 + }, + { + "epoch": 0.6391402714932126, + "grad_norm": 0.5966937504304413, + "learning_rate": 7.981143161342462e-05, + "loss": 0.4757, + "step": 1695 + }, + { + "epoch": 0.6395173453996983, + "grad_norm": 0.7977029192190008, + "learning_rate": 7.981040898689739e-05, + "loss": 0.4799, + "step": 1696 + }, + { + "epoch": 0.639894419306184, + "grad_norm": 0.9533707739924534, + "learning_rate": 7.980938360154412e-05, + "loss": 0.476, + "step": 1697 + }, + { + "epoch": 0.6402714932126696, + "grad_norm": 0.9030353174986863, + "learning_rate": 7.980835545743588e-05, + "loss": 0.4633, + "step": 1698 + }, + { + "epoch": 0.6406485671191554, + "grad_norm": 0.6132028440443207, + "learning_rate": 7.980732455464392e-05, + "loss": 0.4699, + "step": 1699 + }, + { + "epoch": 0.6410256410256411, + "grad_norm": 0.3850960311124732, + "learning_rate": 7.980629089323967e-05, + "loss": 0.4713, + "step": 1700 + }, + { + "epoch": 0.6414027149321267, + "grad_norm": 0.47365238662252007, + "learning_rate": 7.980525447329478e-05, + "loss": 0.4604, + "step": 1701 + }, + { + "epoch": 0.6417797888386124, + "grad_norm": 0.5824113977157259, + "learning_rate": 7.980421529488107e-05, + "loss": 0.4893, + "step": 1702 + }, + { + "epoch": 0.6421568627450981, + "grad_norm": 0.5623694614762492, + "learning_rate": 7.980317335807053e-05, + "loss": 0.4734, + "step": 1703 + }, + { + "epoch": 0.6425339366515838, + "grad_norm": 0.47300109506172316, + "learning_rate": 7.980212866293539e-05, + "loss": 0.4823, + "step": 1704 + }, + { + "epoch": 0.6429110105580694, + "grad_norm": 0.41799484483460003, + "learning_rate": 7.980108120954804e-05, + "loss": 0.4629, + "step": 1705 + }, + { + "epoch": 0.6432880844645551, + "grad_norm": 0.48135180626313, + "learning_rate": 7.980003099798105e-05, + "loss": 0.4756, + "step": 1706 + }, + { + "epoch": 0.6436651583710408, + "grad_norm": 0.41863582230440277, + "learning_rate": 7.979897802830722e-05, + "loss": 0.4729, + "step": 1707 + }, + { + "epoch": 0.6440422322775264, + "grad_norm": 0.4519912278663015, + "learning_rate": 7.979792230059952e-05, + "loss": 0.4746, + "step": 1708 + }, + { + "epoch": 0.6444193061840121, + "grad_norm": 0.46740341380875433, + "learning_rate": 7.979686381493112e-05, + "loss": 0.4706, + "step": 1709 + }, + { + "epoch": 0.6447963800904978, + "grad_norm": 0.5206325126303001, + "learning_rate": 7.979580257137534e-05, + "loss": 0.4647, + "step": 1710 + }, + { + "epoch": 0.6451734539969834, + "grad_norm": 0.5621532114651356, + "learning_rate": 7.979473857000573e-05, + "loss": 0.4773, + "step": 1711 + }, + { + "epoch": 0.6455505279034691, + "grad_norm": 0.521292935553593, + "learning_rate": 7.979367181089606e-05, + "loss": 0.4719, + "step": 1712 + }, + { + "epoch": 0.6459276018099548, + "grad_norm": 0.5370354203428528, + "learning_rate": 7.979260229412022e-05, + "loss": 0.46, + "step": 1713 + }, + { + "epoch": 0.6463046757164405, + "grad_norm": 0.6120949043701805, + "learning_rate": 7.979153001975233e-05, + "loss": 0.4777, + "step": 1714 + }, + { + "epoch": 0.6466817496229261, + "grad_norm": 0.6271116724057488, + "learning_rate": 7.979045498786671e-05, + "loss": 0.4722, + "step": 1715 + }, + { + "epoch": 0.6470588235294118, + "grad_norm": 0.5941292744446942, + "learning_rate": 7.978937719853786e-05, + "loss": 0.4688, + "step": 1716 + }, + { + "epoch": 0.6474358974358975, + "grad_norm": 0.6014362189122825, + "learning_rate": 7.978829665184045e-05, + "loss": 0.4701, + "step": 1717 + }, + { + "epoch": 0.6478129713423831, + "grad_norm": 0.6435279933838117, + "learning_rate": 7.978721334784938e-05, + "loss": 0.4803, + "step": 1718 + }, + { + "epoch": 0.6481900452488688, + "grad_norm": 0.73238879438093, + "learning_rate": 7.978612728663972e-05, + "loss": 0.4776, + "step": 1719 + }, + { + "epoch": 0.6485671191553545, + "grad_norm": 0.7414680374746806, + "learning_rate": 7.978503846828673e-05, + "loss": 0.4686, + "step": 1720 + }, + { + "epoch": 0.6489441930618401, + "grad_norm": 0.6448907078190602, + "learning_rate": 7.978394689286585e-05, + "loss": 0.477, + "step": 1721 + }, + { + "epoch": 0.6493212669683258, + "grad_norm": 0.5041953153274954, + "learning_rate": 7.978285256045275e-05, + "loss": 0.4732, + "step": 1722 + }, + { + "epoch": 0.6496983408748115, + "grad_norm": 0.46900858321382577, + "learning_rate": 7.978175547112325e-05, + "loss": 0.4683, + "step": 1723 + }, + { + "epoch": 0.6500754147812972, + "grad_norm": 0.5741686028058832, + "learning_rate": 7.978065562495339e-05, + "loss": 0.4817, + "step": 1724 + }, + { + "epoch": 0.6504524886877828, + "grad_norm": 0.6219828976394874, + "learning_rate": 7.977955302201938e-05, + "loss": 0.4723, + "step": 1725 + }, + { + "epoch": 0.6508295625942685, + "grad_norm": 0.5334545142267403, + "learning_rate": 7.977844766239764e-05, + "loss": 0.4668, + "step": 1726 + }, + { + "epoch": 0.6512066365007542, + "grad_norm": 0.36582737133351356, + "learning_rate": 7.977733954616474e-05, + "loss": 0.4605, + "step": 1727 + }, + { + "epoch": 0.6515837104072398, + "grad_norm": 0.40188662618941157, + "learning_rate": 7.97762286733975e-05, + "loss": 0.4736, + "step": 1728 + }, + { + "epoch": 0.6519607843137255, + "grad_norm": 0.5101553667875885, + "learning_rate": 7.977511504417289e-05, + "loss": 0.4784, + "step": 1729 + }, + { + "epoch": 0.6523378582202112, + "grad_norm": 0.4903601796294811, + "learning_rate": 7.977399865856808e-05, + "loss": 0.4578, + "step": 1730 + }, + { + "epoch": 0.6527149321266968, + "grad_norm": 0.45132130076774013, + "learning_rate": 7.977287951666045e-05, + "loss": 0.4556, + "step": 1731 + }, + { + "epoch": 0.6530920060331825, + "grad_norm": 0.5057742431180058, + "learning_rate": 7.977175761852757e-05, + "loss": 0.4534, + "step": 1732 + }, + { + "epoch": 0.6534690799396682, + "grad_norm": 0.5783752018608799, + "learning_rate": 7.977063296424714e-05, + "loss": 0.4698, + "step": 1733 + }, + { + "epoch": 0.6538461538461539, + "grad_norm": 0.5576159981628789, + "learning_rate": 7.976950555389713e-05, + "loss": 0.4618, + "step": 1734 + }, + { + "epoch": 0.6542232277526395, + "grad_norm": 0.4514528833146725, + "learning_rate": 7.976837538755565e-05, + "loss": 0.4725, + "step": 1735 + }, + { + "epoch": 0.6546003016591252, + "grad_norm": 0.37354664155086387, + "learning_rate": 7.976724246530105e-05, + "loss": 0.4641, + "step": 1736 + }, + { + "epoch": 0.6549773755656109, + "grad_norm": 0.4578425328975166, + "learning_rate": 7.97661067872118e-05, + "loss": 0.4749, + "step": 1737 + }, + { + "epoch": 0.6553544494720965, + "grad_norm": 0.581386402323544, + "learning_rate": 7.976496835336664e-05, + "loss": 0.465, + "step": 1738 + }, + { + "epoch": 0.6557315233785822, + "grad_norm": 0.6376174108722718, + "learning_rate": 7.976382716384443e-05, + "loss": 0.4718, + "step": 1739 + }, + { + "epoch": 0.6561085972850679, + "grad_norm": 0.61754728225019, + "learning_rate": 7.976268321872427e-05, + "loss": 0.4824, + "step": 1740 + }, + { + "epoch": 0.6564856711915535, + "grad_norm": 0.6681596314158349, + "learning_rate": 7.976153651808544e-05, + "loss": 0.4714, + "step": 1741 + }, + { + "epoch": 0.6568627450980392, + "grad_norm": 0.6911162739240015, + "learning_rate": 7.97603870620074e-05, + "loss": 0.4634, + "step": 1742 + }, + { + "epoch": 0.6572398190045249, + "grad_norm": 0.6837842403190242, + "learning_rate": 7.97592348505698e-05, + "loss": 0.4703, + "step": 1743 + }, + { + "epoch": 0.6576168929110106, + "grad_norm": 0.6218805706618844, + "learning_rate": 7.975807988385248e-05, + "loss": 0.4717, + "step": 1744 + }, + { + "epoch": 0.6579939668174962, + "grad_norm": 0.6634882141981258, + "learning_rate": 7.975692216193549e-05, + "loss": 0.4795, + "step": 1745 + }, + { + "epoch": 0.6583710407239819, + "grad_norm": 0.7338753393889097, + "learning_rate": 7.975576168489907e-05, + "loss": 0.4719, + "step": 1746 + }, + { + "epoch": 0.6587481146304676, + "grad_norm": 0.7775606538094558, + "learning_rate": 7.975459845282363e-05, + "loss": 0.4645, + "step": 1747 + }, + { + "epoch": 0.6591251885369532, + "grad_norm": 0.8461741574043966, + "learning_rate": 7.975343246578978e-05, + "loss": 0.4648, + "step": 1748 + }, + { + "epoch": 0.6595022624434389, + "grad_norm": 0.7686183767240669, + "learning_rate": 7.975226372387831e-05, + "loss": 0.4796, + "step": 1749 + }, + { + "epoch": 0.6598793363499246, + "grad_norm": 0.5664582176710488, + "learning_rate": 7.975109222717024e-05, + "loss": 0.4702, + "step": 1750 + }, + { + "epoch": 0.6602564102564102, + "grad_norm": 0.4999051744476222, + "learning_rate": 7.974991797574673e-05, + "loss": 0.474, + "step": 1751 + }, + { + "epoch": 0.6606334841628959, + "grad_norm": 0.5644301669714904, + "learning_rate": 7.974874096968917e-05, + "loss": 0.4677, + "step": 1752 + }, + { + "epoch": 0.6610105580693816, + "grad_norm": 0.5934642864082165, + "learning_rate": 7.97475612090791e-05, + "loss": 0.4625, + "step": 1753 + }, + { + "epoch": 0.6613876319758673, + "grad_norm": 0.5257504171342314, + "learning_rate": 7.974637869399831e-05, + "loss": 0.4689, + "step": 1754 + }, + { + "epoch": 0.6617647058823529, + "grad_norm": 0.5051667658453908, + "learning_rate": 7.974519342452872e-05, + "loss": 0.4599, + "step": 1755 + }, + { + "epoch": 0.6621417797888386, + "grad_norm": 0.5344532139814178, + "learning_rate": 7.974400540075248e-05, + "loss": 0.4639, + "step": 1756 + }, + { + "epoch": 0.6625188536953243, + "grad_norm": 0.5504000764738659, + "learning_rate": 7.974281462275193e-05, + "loss": 0.4618, + "step": 1757 + }, + { + "epoch": 0.6628959276018099, + "grad_norm": 0.5350317458265031, + "learning_rate": 7.974162109060956e-05, + "loss": 0.4859, + "step": 1758 + }, + { + "epoch": 0.6632730015082956, + "grad_norm": 0.5561594317796988, + "learning_rate": 7.974042480440813e-05, + "loss": 0.4697, + "step": 1759 + }, + { + "epoch": 0.6636500754147813, + "grad_norm": 0.6669275369501289, + "learning_rate": 7.973922576423047e-05, + "loss": 0.4765, + "step": 1760 + }, + { + "epoch": 0.6640271493212669, + "grad_norm": 0.7800938832147839, + "learning_rate": 7.973802397015973e-05, + "loss": 0.4747, + "step": 1761 + }, + { + "epoch": 0.6644042232277526, + "grad_norm": 0.7910272354791773, + "learning_rate": 7.97368194222792e-05, + "loss": 0.4665, + "step": 1762 + }, + { + "epoch": 0.6647812971342383, + "grad_norm": 0.6714445384915009, + "learning_rate": 7.97356121206723e-05, + "loss": 0.449, + "step": 1763 + }, + { + "epoch": 0.665158371040724, + "grad_norm": 0.5974473158147484, + "learning_rate": 7.973440206542274e-05, + "loss": 0.4722, + "step": 1764 + }, + { + "epoch": 0.6655354449472096, + "grad_norm": 0.4735746927367605, + "learning_rate": 7.973318925661437e-05, + "loss": 0.4717, + "step": 1765 + }, + { + "epoch": 0.6659125188536953, + "grad_norm": 0.4154200081707537, + "learning_rate": 7.97319736943312e-05, + "loss": 0.4707, + "step": 1766 + }, + { + "epoch": 0.666289592760181, + "grad_norm": 0.45517806310128966, + "learning_rate": 7.973075537865752e-05, + "loss": 0.453, + "step": 1767 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 0.4851408723095296, + "learning_rate": 7.972953430967773e-05, + "loss": 0.4644, + "step": 1768 + }, + { + "epoch": 0.6670437405731523, + "grad_norm": 0.5292379355655021, + "learning_rate": 7.972831048747645e-05, + "loss": 0.4665, + "step": 1769 + }, + { + "epoch": 0.667420814479638, + "grad_norm": 0.4842865419701662, + "learning_rate": 7.972708391213849e-05, + "loss": 0.4663, + "step": 1770 + }, + { + "epoch": 0.6677978883861236, + "grad_norm": 0.44808926604044413, + "learning_rate": 7.972585458374885e-05, + "loss": 0.4724, + "step": 1771 + }, + { + "epoch": 0.6681749622926093, + "grad_norm": 0.41652271351534015, + "learning_rate": 7.972462250239272e-05, + "loss": 0.4763, + "step": 1772 + }, + { + "epoch": 0.668552036199095, + "grad_norm": 0.5295390297379777, + "learning_rate": 7.97233876681555e-05, + "loss": 0.4828, + "step": 1773 + }, + { + "epoch": 0.6689291101055806, + "grad_norm": 0.5644517306825547, + "learning_rate": 7.972215008112273e-05, + "loss": 0.4737, + "step": 1774 + }, + { + "epoch": 0.6693061840120663, + "grad_norm": 0.4645948969258071, + "learning_rate": 7.972090974138021e-05, + "loss": 0.469, + "step": 1775 + }, + { + "epoch": 0.669683257918552, + "grad_norm": 0.5810076845287568, + "learning_rate": 7.971966664901388e-05, + "loss": 0.4674, + "step": 1776 + }, + { + "epoch": 0.6700603318250377, + "grad_norm": 0.7410898537257561, + "learning_rate": 7.971842080410987e-05, + "loss": 0.4764, + "step": 1777 + }, + { + "epoch": 0.6704374057315233, + "grad_norm": 0.8083508299056943, + "learning_rate": 7.971717220675454e-05, + "loss": 0.4675, + "step": 1778 + }, + { + "epoch": 0.670814479638009, + "grad_norm": 0.8704532691787041, + "learning_rate": 7.971592085703439e-05, + "loss": 0.4747, + "step": 1779 + }, + { + "epoch": 0.6711915535444947, + "grad_norm": 0.794209409588924, + "learning_rate": 7.971466675503615e-05, + "loss": 0.4643, + "step": 1780 + }, + { + "epoch": 0.6715686274509803, + "grad_norm": 0.6350694597470552, + "learning_rate": 7.971340990084676e-05, + "loss": 0.4646, + "step": 1781 + }, + { + "epoch": 0.6719457013574661, + "grad_norm": 0.6314580657365298, + "learning_rate": 7.971215029455326e-05, + "loss": 0.4672, + "step": 1782 + }, + { + "epoch": 0.6723227752639518, + "grad_norm": 0.5892172505701725, + "learning_rate": 7.971088793624299e-05, + "loss": 0.4738, + "step": 1783 + }, + { + "epoch": 0.6726998491704375, + "grad_norm": 0.47256337075194377, + "learning_rate": 7.97096228260034e-05, + "loss": 0.4734, + "step": 1784 + }, + { + "epoch": 0.6730769230769231, + "grad_norm": 0.470304580247822, + "learning_rate": 7.970835496392216e-05, + "loss": 0.4699, + "step": 1785 + }, + { + "epoch": 0.6734539969834088, + "grad_norm": 0.5416630749374767, + "learning_rate": 7.970708435008716e-05, + "loss": 0.488, + "step": 1786 + }, + { + "epoch": 0.6738310708898945, + "grad_norm": 0.6553519881167967, + "learning_rate": 7.970581098458642e-05, + "loss": 0.468, + "step": 1787 + }, + { + "epoch": 0.6742081447963801, + "grad_norm": 0.6747019582616063, + "learning_rate": 7.97045348675082e-05, + "loss": 0.472, + "step": 1788 + }, + { + "epoch": 0.6745852187028658, + "grad_norm": 0.5221100243648936, + "learning_rate": 7.970325599894095e-05, + "loss": 0.4619, + "step": 1789 + }, + { + "epoch": 0.6749622926093515, + "grad_norm": 0.41962867173807383, + "learning_rate": 7.970197437897327e-05, + "loss": 0.4844, + "step": 1790 + }, + { + "epoch": 0.6753393665158371, + "grad_norm": 0.47830820744311875, + "learning_rate": 7.970069000769397e-05, + "loss": 0.4679, + "step": 1791 + }, + { + "epoch": 0.6757164404223228, + "grad_norm": 0.5477147724056062, + "learning_rate": 7.969940288519207e-05, + "loss": 0.4767, + "step": 1792 + }, + { + "epoch": 0.6760935143288085, + "grad_norm": 0.587286476270804, + "learning_rate": 7.969811301155676e-05, + "loss": 0.4626, + "step": 1793 + }, + { + "epoch": 0.6764705882352942, + "grad_norm": 0.4233495151746367, + "learning_rate": 7.969682038687744e-05, + "loss": 0.4792, + "step": 1794 + }, + { + "epoch": 0.6768476621417798, + "grad_norm": 0.344997267281994, + "learning_rate": 7.969552501124367e-05, + "loss": 0.4596, + "step": 1795 + }, + { + "epoch": 0.6772247360482655, + "grad_norm": 0.44062961830573616, + "learning_rate": 7.969422688474523e-05, + "loss": 0.4709, + "step": 1796 + }, + { + "epoch": 0.6776018099547512, + "grad_norm": 0.5163988147419217, + "learning_rate": 7.969292600747207e-05, + "loss": 0.4696, + "step": 1797 + }, + { + "epoch": 0.6779788838612368, + "grad_norm": 0.43184199671586293, + "learning_rate": 7.969162237951436e-05, + "loss": 0.4612, + "step": 1798 + }, + { + "epoch": 0.6783559577677225, + "grad_norm": 0.3586493980566755, + "learning_rate": 7.969031600096241e-05, + "loss": 0.4523, + "step": 1799 + }, + { + "epoch": 0.6787330316742082, + "grad_norm": 0.36528916052902644, + "learning_rate": 7.968900687190679e-05, + "loss": 0.4752, + "step": 1800 + }, + { + "epoch": 0.6791101055806938, + "grad_norm": 0.34670894233268024, + "learning_rate": 7.968769499243816e-05, + "loss": 0.4654, + "step": 1801 + }, + { + "epoch": 0.6794871794871795, + "grad_norm": 0.37024067425162954, + "learning_rate": 7.968638036264749e-05, + "loss": 0.4808, + "step": 1802 + }, + { + "epoch": 0.6798642533936652, + "grad_norm": 0.4307224095778274, + "learning_rate": 7.968506298262586e-05, + "loss": 0.4605, + "step": 1803 + }, + { + "epoch": 0.6802413273001509, + "grad_norm": 0.541366731608238, + "learning_rate": 7.968374285246458e-05, + "loss": 0.4723, + "step": 1804 + }, + { + "epoch": 0.6806184012066365, + "grad_norm": 0.7684325327794461, + "learning_rate": 7.96824199722551e-05, + "loss": 0.474, + "step": 1805 + }, + { + "epoch": 0.6809954751131222, + "grad_norm": 1.0013175529499536, + "learning_rate": 7.968109434208912e-05, + "loss": 0.4719, + "step": 1806 + }, + { + "epoch": 0.6813725490196079, + "grad_norm": 1.1610966656722435, + "learning_rate": 7.967976596205852e-05, + "loss": 0.476, + "step": 1807 + }, + { + "epoch": 0.6817496229260935, + "grad_norm": 0.6501744188313742, + "learning_rate": 7.96784348322553e-05, + "loss": 0.4572, + "step": 1808 + }, + { + "epoch": 0.6821266968325792, + "grad_norm": 0.42587683801235304, + "learning_rate": 7.967710095277177e-05, + "loss": 0.4683, + "step": 1809 + }, + { + "epoch": 0.6825037707390649, + "grad_norm": 0.4985875354166986, + "learning_rate": 7.967576432370032e-05, + "loss": 0.4735, + "step": 1810 + }, + { + "epoch": 0.6828808446455505, + "grad_norm": 0.6416012603370084, + "learning_rate": 7.96744249451336e-05, + "loss": 0.4701, + "step": 1811 + }, + { + "epoch": 0.6832579185520362, + "grad_norm": 0.7718884734415766, + "learning_rate": 7.967308281716442e-05, + "loss": 0.4719, + "step": 1812 + }, + { + "epoch": 0.6836349924585219, + "grad_norm": 0.6700593872772177, + "learning_rate": 7.96717379398858e-05, + "loss": 0.458, + "step": 1813 + }, + { + "epoch": 0.6840120663650076, + "grad_norm": 0.4767240962325249, + "learning_rate": 7.967039031339092e-05, + "loss": 0.4708, + "step": 1814 + }, + { + "epoch": 0.6843891402714932, + "grad_norm": 0.4290724355525338, + "learning_rate": 7.966903993777318e-05, + "loss": 0.4687, + "step": 1815 + }, + { + "epoch": 0.6847662141779789, + "grad_norm": 0.5103803094198504, + "learning_rate": 7.966768681312616e-05, + "loss": 0.477, + "step": 1816 + }, + { + "epoch": 0.6851432880844646, + "grad_norm": 0.624691411425059, + "learning_rate": 7.966633093954362e-05, + "loss": 0.472, + "step": 1817 + }, + { + "epoch": 0.6855203619909502, + "grad_norm": 0.626876915613645, + "learning_rate": 7.966497231711956e-05, + "loss": 0.4695, + "step": 1818 + }, + { + "epoch": 0.6858974358974359, + "grad_norm": 0.6299187051622035, + "learning_rate": 7.966361094594807e-05, + "loss": 0.4762, + "step": 1819 + }, + { + "epoch": 0.6862745098039216, + "grad_norm": 0.6440760630210277, + "learning_rate": 7.966224682612353e-05, + "loss": 0.4584, + "step": 1820 + }, + { + "epoch": 0.6866515837104072, + "grad_norm": 0.638026170366678, + "learning_rate": 7.966087995774047e-05, + "loss": 0.4683, + "step": 1821 + }, + { + "epoch": 0.6870286576168929, + "grad_norm": 0.5451860705954865, + "learning_rate": 7.965951034089361e-05, + "loss": 0.4641, + "step": 1822 + }, + { + "epoch": 0.6874057315233786, + "grad_norm": 0.3979526162759599, + "learning_rate": 7.965813797567785e-05, + "loss": 0.4542, + "step": 1823 + }, + { + "epoch": 0.6877828054298643, + "grad_norm": 0.3903430270456936, + "learning_rate": 7.965676286218832e-05, + "loss": 0.4582, + "step": 1824 + }, + { + "epoch": 0.6881598793363499, + "grad_norm": 0.42067970049648007, + "learning_rate": 7.96553850005203e-05, + "loss": 0.4743, + "step": 1825 + }, + { + "epoch": 0.6885369532428356, + "grad_norm": 0.4320626174035961, + "learning_rate": 7.965400439076925e-05, + "loss": 0.4675, + "step": 1826 + }, + { + "epoch": 0.6889140271493213, + "grad_norm": 0.4741018924226315, + "learning_rate": 7.96526210330309e-05, + "loss": 0.4581, + "step": 1827 + }, + { + "epoch": 0.6892911010558069, + "grad_norm": 0.4914210694666556, + "learning_rate": 7.965123492740106e-05, + "loss": 0.4696, + "step": 1828 + }, + { + "epoch": 0.6896681749622926, + "grad_norm": 0.50678137122553, + "learning_rate": 7.964984607397584e-05, + "loss": 0.4706, + "step": 1829 + }, + { + "epoch": 0.6900452488687783, + "grad_norm": 0.40582544331561904, + "learning_rate": 7.964845447285141e-05, + "loss": 0.4695, + "step": 1830 + }, + { + "epoch": 0.6904223227752639, + "grad_norm": 0.3794695655961308, + "learning_rate": 7.964706012412429e-05, + "loss": 0.4552, + "step": 1831 + }, + { + "epoch": 0.6907993966817496, + "grad_norm": 0.48288448010365587, + "learning_rate": 7.964566302789105e-05, + "loss": 0.4624, + "step": 1832 + }, + { + "epoch": 0.6911764705882353, + "grad_norm": 0.5393578840871542, + "learning_rate": 7.964426318424855e-05, + "loss": 0.4619, + "step": 1833 + }, + { + "epoch": 0.691553544494721, + "grad_norm": 0.569120825028049, + "learning_rate": 7.964286059329375e-05, + "loss": 0.4743, + "step": 1834 + }, + { + "epoch": 0.6919306184012066, + "grad_norm": 0.6664333228922481, + "learning_rate": 7.96414552551239e-05, + "loss": 0.4683, + "step": 1835 + }, + { + "epoch": 0.6923076923076923, + "grad_norm": 0.7453406059700911, + "learning_rate": 7.964004716983635e-05, + "loss": 0.4717, + "step": 1836 + }, + { + "epoch": 0.692684766214178, + "grad_norm": 0.8472292864754452, + "learning_rate": 7.963863633752869e-05, + "loss": 0.4626, + "step": 1837 + }, + { + "epoch": 0.6930618401206636, + "grad_norm": 1.0304193754328193, + "learning_rate": 7.963722275829869e-05, + "loss": 0.4777, + "step": 1838 + }, + { + "epoch": 0.6934389140271493, + "grad_norm": 1.0665434364774073, + "learning_rate": 7.963580643224431e-05, + "loss": 0.468, + "step": 1839 + }, + { + "epoch": 0.693815987933635, + "grad_norm": 0.8291356630291715, + "learning_rate": 7.963438735946371e-05, + "loss": 0.46, + "step": 1840 + }, + { + "epoch": 0.6941930618401206, + "grad_norm": 0.7488175587434523, + "learning_rate": 7.963296554005522e-05, + "loss": 0.4575, + "step": 1841 + }, + { + "epoch": 0.6945701357466063, + "grad_norm": 0.7036447544559583, + "learning_rate": 7.963154097411736e-05, + "loss": 0.4669, + "step": 1842 + }, + { + "epoch": 0.694947209653092, + "grad_norm": 0.5901333964586878, + "learning_rate": 7.963011366174889e-05, + "loss": 0.4581, + "step": 1843 + }, + { + "epoch": 0.6953242835595776, + "grad_norm": 0.5658564148103064, + "learning_rate": 7.962868360304866e-05, + "loss": 0.4738, + "step": 1844 + }, + { + "epoch": 0.6957013574660633, + "grad_norm": 0.5391627683062032, + "learning_rate": 7.962725079811582e-05, + "loss": 0.4733, + "step": 1845 + }, + { + "epoch": 0.696078431372549, + "grad_norm": 0.44852635176934375, + "learning_rate": 7.962581524704966e-05, + "loss": 0.4751, + "step": 1846 + }, + { + "epoch": 0.6964555052790347, + "grad_norm": 0.35586293638693206, + "learning_rate": 7.962437694994964e-05, + "loss": 0.4565, + "step": 1847 + }, + { + "epoch": 0.6968325791855203, + "grad_norm": 0.3548936047277173, + "learning_rate": 7.962293590691545e-05, + "loss": 0.4733, + "step": 1848 + }, + { + "epoch": 0.697209653092006, + "grad_norm": 0.4739324907069103, + "learning_rate": 7.962149211804696e-05, + "loss": 0.4694, + "step": 1849 + }, + { + "epoch": 0.6975867269984917, + "grad_norm": 0.47843717614440556, + "learning_rate": 7.962004558344419e-05, + "loss": 0.4773, + "step": 1850 + }, + { + "epoch": 0.6979638009049773, + "grad_norm": 0.4625469633552629, + "learning_rate": 7.961859630320741e-05, + "loss": 0.4727, + "step": 1851 + }, + { + "epoch": 0.698340874811463, + "grad_norm": 0.5347603684340348, + "learning_rate": 7.961714427743705e-05, + "loss": 0.461, + "step": 1852 + }, + { + "epoch": 0.6987179487179487, + "grad_norm": 0.5070209685910844, + "learning_rate": 7.961568950623373e-05, + "loss": 0.4671, + "step": 1853 + }, + { + "epoch": 0.6990950226244343, + "grad_norm": 0.4262018076176518, + "learning_rate": 7.961423198969826e-05, + "loss": 0.4556, + "step": 1854 + }, + { + "epoch": 0.69947209653092, + "grad_norm": 0.4384650330302264, + "learning_rate": 7.961277172793166e-05, + "loss": 0.4639, + "step": 1855 + }, + { + "epoch": 0.6998491704374057, + "grad_norm": 0.43268485536481577, + "learning_rate": 7.961130872103511e-05, + "loss": 0.4522, + "step": 1856 + }, + { + "epoch": 0.7002262443438914, + "grad_norm": 0.4724237894773669, + "learning_rate": 7.960984296911e-05, + "loss": 0.4467, + "step": 1857 + }, + { + "epoch": 0.700603318250377, + "grad_norm": 0.5354115443930416, + "learning_rate": 7.960837447225793e-05, + "loss": 0.4599, + "step": 1858 + }, + { + "epoch": 0.7009803921568627, + "grad_norm": 0.6192210640592866, + "learning_rate": 7.960690323058062e-05, + "loss": 0.4766, + "step": 1859 + }, + { + "epoch": 0.7013574660633484, + "grad_norm": 0.7049951531924624, + "learning_rate": 7.960542924418004e-05, + "loss": 0.4701, + "step": 1860 + }, + { + "epoch": 0.701734539969834, + "grad_norm": 0.855581186755313, + "learning_rate": 7.960395251315837e-05, + "loss": 0.4698, + "step": 1861 + }, + { + "epoch": 0.7021116138763197, + "grad_norm": 0.9482578403560418, + "learning_rate": 7.960247303761788e-05, + "loss": 0.4676, + "step": 1862 + }, + { + "epoch": 0.7024886877828054, + "grad_norm": 0.9319029327117074, + "learning_rate": 7.960099081766116e-05, + "loss": 0.4526, + "step": 1863 + }, + { + "epoch": 0.702865761689291, + "grad_norm": 1.023058786764789, + "learning_rate": 7.959950585339091e-05, + "loss": 0.4771, + "step": 1864 + }, + { + "epoch": 0.7032428355957768, + "grad_norm": 0.9434501396285059, + "learning_rate": 7.959801814491002e-05, + "loss": 0.4594, + "step": 1865 + }, + { + "epoch": 0.7036199095022625, + "grad_norm": 0.7532680181282893, + "learning_rate": 7.95965276923216e-05, + "loss": 0.4625, + "step": 1866 + }, + { + "epoch": 0.7039969834087482, + "grad_norm": 0.5306656342109692, + "learning_rate": 7.959503449572892e-05, + "loss": 0.4728, + "step": 1867 + }, + { + "epoch": 0.7043740573152338, + "grad_norm": 0.3525112424255266, + "learning_rate": 7.959353855523548e-05, + "loss": 0.4798, + "step": 1868 + }, + { + "epoch": 0.7047511312217195, + "grad_norm": 0.5109492075691543, + "learning_rate": 7.959203987094494e-05, + "loss": 0.4608, + "step": 1869 + }, + { + "epoch": 0.7051282051282052, + "grad_norm": 0.6732018321676997, + "learning_rate": 7.959053844296116e-05, + "loss": 0.4605, + "step": 1870 + }, + { + "epoch": 0.7055052790346908, + "grad_norm": 0.7332531802602421, + "learning_rate": 7.958903427138816e-05, + "loss": 0.4843, + "step": 1871 + }, + { + "epoch": 0.7058823529411765, + "grad_norm": 0.6241791761600198, + "learning_rate": 7.958752735633022e-05, + "loss": 0.4621, + "step": 1872 + }, + { + "epoch": 0.7062594268476622, + "grad_norm": 0.4660477819067349, + "learning_rate": 7.958601769789174e-05, + "loss": 0.4624, + "step": 1873 + }, + { + "epoch": 0.7066365007541479, + "grad_norm": 0.4150965581615293, + "learning_rate": 7.958450529617735e-05, + "loss": 0.4723, + "step": 1874 + }, + { + "epoch": 0.7070135746606335, + "grad_norm": 0.4692944098558423, + "learning_rate": 7.958299015129186e-05, + "loss": 0.4745, + "step": 1875 + }, + { + "epoch": 0.7073906485671192, + "grad_norm": 0.5620165106521479, + "learning_rate": 7.958147226334025e-05, + "loss": 0.4677, + "step": 1876 + }, + { + "epoch": 0.7077677224736049, + "grad_norm": 0.5642318700739264, + "learning_rate": 7.957995163242773e-05, + "loss": 0.4824, + "step": 1877 + }, + { + "epoch": 0.7081447963800905, + "grad_norm": 0.49780904885832106, + "learning_rate": 7.957842825865966e-05, + "loss": 0.4659, + "step": 1878 + }, + { + "epoch": 0.7085218702865762, + "grad_norm": 0.4381762147658957, + "learning_rate": 7.957690214214163e-05, + "loss": 0.4743, + "step": 1879 + }, + { + "epoch": 0.7088989441930619, + "grad_norm": 0.4412473442863316, + "learning_rate": 7.957537328297938e-05, + "loss": 0.4613, + "step": 1880 + }, + { + "epoch": 0.7092760180995475, + "grad_norm": 0.44909225905136396, + "learning_rate": 7.957384168127886e-05, + "loss": 0.4696, + "step": 1881 + }, + { + "epoch": 0.7096530920060332, + "grad_norm": 0.5411027560760422, + "learning_rate": 7.957230733714623e-05, + "loss": 0.4755, + "step": 1882 + }, + { + "epoch": 0.7100301659125189, + "grad_norm": 0.6003493531797413, + "learning_rate": 7.957077025068777e-05, + "loss": 0.4803, + "step": 1883 + }, + { + "epoch": 0.7104072398190046, + "grad_norm": 0.6038859385847807, + "learning_rate": 7.956923042201006e-05, + "loss": 0.4612, + "step": 1884 + }, + { + "epoch": 0.7107843137254902, + "grad_norm": 0.5735347351195759, + "learning_rate": 7.956768785121977e-05, + "loss": 0.4595, + "step": 1885 + }, + { + "epoch": 0.7111613876319759, + "grad_norm": 0.5585371599148299, + "learning_rate": 7.956614253842382e-05, + "loss": 0.4766, + "step": 1886 + }, + { + "epoch": 0.7115384615384616, + "grad_norm": 0.5446805402725479, + "learning_rate": 7.956459448372928e-05, + "loss": 0.4586, + "step": 1887 + }, + { + "epoch": 0.7119155354449472, + "grad_norm": 0.5602372277628969, + "learning_rate": 7.956304368724342e-05, + "loss": 0.4662, + "step": 1888 + }, + { + "epoch": 0.7122926093514329, + "grad_norm": 0.45836232773539226, + "learning_rate": 7.956149014907373e-05, + "loss": 0.4733, + "step": 1889 + }, + { + "epoch": 0.7126696832579186, + "grad_norm": 0.38006832763724363, + "learning_rate": 7.955993386932789e-05, + "loss": 0.4567, + "step": 1890 + }, + { + "epoch": 0.7130467571644042, + "grad_norm": 0.45521506297727554, + "learning_rate": 7.955837484811369e-05, + "loss": 0.4585, + "step": 1891 + }, + { + "epoch": 0.7134238310708899, + "grad_norm": 0.5792202299898633, + "learning_rate": 7.955681308553921e-05, + "loss": 0.4677, + "step": 1892 + }, + { + "epoch": 0.7138009049773756, + "grad_norm": 0.6062017326328913, + "learning_rate": 7.955524858171269e-05, + "loss": 0.4727, + "step": 1893 + }, + { + "epoch": 0.7141779788838613, + "grad_norm": 0.46760901423138446, + "learning_rate": 7.955368133674251e-05, + "loss": 0.4644, + "step": 1894 + }, + { + "epoch": 0.7145550527903469, + "grad_norm": 0.4400777576897557, + "learning_rate": 7.95521113507373e-05, + "loss": 0.47, + "step": 1895 + }, + { + "epoch": 0.7149321266968326, + "grad_norm": 0.4454787470162203, + "learning_rate": 7.955053862380586e-05, + "loss": 0.4695, + "step": 1896 + }, + { + "epoch": 0.7153092006033183, + "grad_norm": 0.5547039215009569, + "learning_rate": 7.954896315605717e-05, + "loss": 0.4645, + "step": 1897 + }, + { + "epoch": 0.7156862745098039, + "grad_norm": 0.5404548082817168, + "learning_rate": 7.954738494760041e-05, + "loss": 0.4693, + "step": 1898 + }, + { + "epoch": 0.7160633484162896, + "grad_norm": 0.45873574842573966, + "learning_rate": 7.954580399854497e-05, + "loss": 0.4616, + "step": 1899 + }, + { + "epoch": 0.7164404223227753, + "grad_norm": 0.563809071684689, + "learning_rate": 7.954422030900037e-05, + "loss": 0.4698, + "step": 1900 + }, + { + "epoch": 0.7168174962292609, + "grad_norm": 0.5668404479021832, + "learning_rate": 7.954263387907638e-05, + "loss": 0.4737, + "step": 1901 + }, + { + "epoch": 0.7171945701357466, + "grad_norm": 0.6430518695294222, + "learning_rate": 7.954104470888295e-05, + "loss": 0.4571, + "step": 1902 + }, + { + "epoch": 0.7175716440422323, + "grad_norm": 0.6775272045724203, + "learning_rate": 7.953945279853018e-05, + "loss": 0.4579, + "step": 1903 + }, + { + "epoch": 0.717948717948718, + "grad_norm": 0.7910948341940582, + "learning_rate": 7.953785814812841e-05, + "loss": 0.4622, + "step": 1904 + }, + { + "epoch": 0.7183257918552036, + "grad_norm": 0.9586512678749178, + "learning_rate": 7.953626075778814e-05, + "loss": 0.471, + "step": 1905 + }, + { + "epoch": 0.7187028657616893, + "grad_norm": 1.0726204954209364, + "learning_rate": 7.953466062762006e-05, + "loss": 0.4746, + "step": 1906 + }, + { + "epoch": 0.719079939668175, + "grad_norm": 0.7871618515109674, + "learning_rate": 7.953305775773508e-05, + "loss": 0.465, + "step": 1907 + }, + { + "epoch": 0.7194570135746606, + "grad_norm": 0.46781166058288265, + "learning_rate": 7.953145214824424e-05, + "loss": 0.4593, + "step": 1908 + }, + { + "epoch": 0.7198340874811463, + "grad_norm": 0.6121840953099396, + "learning_rate": 7.952984379925886e-05, + "loss": 0.4732, + "step": 1909 + }, + { + "epoch": 0.720211161387632, + "grad_norm": 0.6525049861408686, + "learning_rate": 7.952823271089035e-05, + "loss": 0.4527, + "step": 1910 + }, + { + "epoch": 0.7205882352941176, + "grad_norm": 0.5870853549410006, + "learning_rate": 7.952661888325038e-05, + "loss": 0.4759, + "step": 1911 + }, + { + "epoch": 0.7209653092006033, + "grad_norm": 0.4878758661835472, + "learning_rate": 7.952500231645078e-05, + "loss": 0.4685, + "step": 1912 + }, + { + "epoch": 0.721342383107089, + "grad_norm": 0.5038613212633528, + "learning_rate": 7.952338301060358e-05, + "loss": 0.465, + "step": 1913 + }, + { + "epoch": 0.7217194570135747, + "grad_norm": 0.4825209817929185, + "learning_rate": 7.9521760965821e-05, + "loss": 0.4711, + "step": 1914 + }, + { + "epoch": 0.7220965309200603, + "grad_norm": 0.47790636696431815, + "learning_rate": 7.952013618221543e-05, + "loss": 0.469, + "step": 1915 + }, + { + "epoch": 0.722473604826546, + "grad_norm": 0.5077382518062945, + "learning_rate": 7.951850865989947e-05, + "loss": 0.4753, + "step": 1916 + }, + { + "epoch": 0.7228506787330317, + "grad_norm": 0.6277386869468015, + "learning_rate": 7.951687839898594e-05, + "loss": 0.4669, + "step": 1917 + }, + { + "epoch": 0.7232277526395173, + "grad_norm": 0.6028688381517565, + "learning_rate": 7.951524539958776e-05, + "loss": 0.4575, + "step": 1918 + }, + { + "epoch": 0.723604826546003, + "grad_norm": 0.4725981758136127, + "learning_rate": 7.951360966181814e-05, + "loss": 0.4574, + "step": 1919 + }, + { + "epoch": 0.7239819004524887, + "grad_norm": 0.45683129748425266, + "learning_rate": 7.951197118579042e-05, + "loss": 0.4591, + "step": 1920 + }, + { + "epoch": 0.7243589743589743, + "grad_norm": 0.5400892917120779, + "learning_rate": 7.951032997161813e-05, + "loss": 0.4652, + "step": 1921 + }, + { + "epoch": 0.72473604826546, + "grad_norm": 0.5669909781894079, + "learning_rate": 7.950868601941503e-05, + "loss": 0.4693, + "step": 1922 + }, + { + "epoch": 0.7251131221719457, + "grad_norm": 0.5588748169506493, + "learning_rate": 7.950703932929503e-05, + "loss": 0.4639, + "step": 1923 + }, + { + "epoch": 0.7254901960784313, + "grad_norm": 0.6134663843612755, + "learning_rate": 7.950538990137225e-05, + "loss": 0.4715, + "step": 1924 + }, + { + "epoch": 0.725867269984917, + "grad_norm": 0.6999566712159058, + "learning_rate": 7.950373773576099e-05, + "loss": 0.4587, + "step": 1925 + }, + { + "epoch": 0.7262443438914027, + "grad_norm": 0.7428293362757588, + "learning_rate": 7.950208283257573e-05, + "loss": 0.4678, + "step": 1926 + }, + { + "epoch": 0.7266214177978884, + "grad_norm": 0.7050739533104382, + "learning_rate": 7.950042519193119e-05, + "loss": 0.4726, + "step": 1927 + }, + { + "epoch": 0.726998491704374, + "grad_norm": 0.5957553737917737, + "learning_rate": 7.94987648139422e-05, + "loss": 0.4712, + "step": 1928 + }, + { + "epoch": 0.7273755656108597, + "grad_norm": 0.546422621590952, + "learning_rate": 7.949710169872387e-05, + "loss": 0.4468, + "step": 1929 + }, + { + "epoch": 0.7277526395173454, + "grad_norm": 0.6003222526946465, + "learning_rate": 7.94954358463914e-05, + "loss": 0.4628, + "step": 1930 + }, + { + "epoch": 0.728129713423831, + "grad_norm": 0.6704071155443961, + "learning_rate": 7.949376725706027e-05, + "loss": 0.4582, + "step": 1931 + }, + { + "epoch": 0.7285067873303167, + "grad_norm": 0.6293464468311817, + "learning_rate": 7.949209593084612e-05, + "loss": 0.4806, + "step": 1932 + }, + { + "epoch": 0.7288838612368024, + "grad_norm": 0.5179562491466876, + "learning_rate": 7.949042186786472e-05, + "loss": 0.4567, + "step": 1933 + }, + { + "epoch": 0.729260935143288, + "grad_norm": 0.3744575320402108, + "learning_rate": 7.948874506823212e-05, + "loss": 0.461, + "step": 1934 + }, + { + "epoch": 0.7296380090497737, + "grad_norm": 0.45212714665749604, + "learning_rate": 7.948706553206451e-05, + "loss": 0.4646, + "step": 1935 + }, + { + "epoch": 0.7300150829562594, + "grad_norm": 0.5421381568034673, + "learning_rate": 7.94853832594783e-05, + "loss": 0.4577, + "step": 1936 + }, + { + "epoch": 0.7303921568627451, + "grad_norm": 0.5347532976690473, + "learning_rate": 7.948369825059005e-05, + "loss": 0.4652, + "step": 1937 + }, + { + "epoch": 0.7307692307692307, + "grad_norm": 0.5562261528824566, + "learning_rate": 7.948201050551651e-05, + "loss": 0.454, + "step": 1938 + }, + { + "epoch": 0.7311463046757164, + "grad_norm": 0.5891376166813845, + "learning_rate": 7.94803200243747e-05, + "loss": 0.4538, + "step": 1939 + }, + { + "epoch": 0.7315233785822021, + "grad_norm": 0.5975497002135858, + "learning_rate": 7.94786268072817e-05, + "loss": 0.4708, + "step": 1940 + }, + { + "epoch": 0.7319004524886877, + "grad_norm": 0.6157848215646881, + "learning_rate": 7.94769308543549e-05, + "loss": 0.472, + "step": 1941 + }, + { + "epoch": 0.7322775263951734, + "grad_norm": 0.5670728790878106, + "learning_rate": 7.947523216571179e-05, + "loss": 0.4574, + "step": 1942 + }, + { + "epoch": 0.7326546003016591, + "grad_norm": 0.49806275653035176, + "learning_rate": 7.947353074147013e-05, + "loss": 0.4684, + "step": 1943 + }, + { + "epoch": 0.7330316742081447, + "grad_norm": 0.4970497130021244, + "learning_rate": 7.947182658174778e-05, + "loss": 0.4612, + "step": 1944 + }, + { + "epoch": 0.7334087481146304, + "grad_norm": 0.48300053264586484, + "learning_rate": 7.947011968666285e-05, + "loss": 0.4654, + "step": 1945 + }, + { + "epoch": 0.7337858220211161, + "grad_norm": 0.46285512663988354, + "learning_rate": 7.946841005633366e-05, + "loss": 0.47, + "step": 1946 + }, + { + "epoch": 0.7341628959276018, + "grad_norm": 0.352372487527536, + "learning_rate": 7.946669769087866e-05, + "loss": 0.4664, + "step": 1947 + }, + { + "epoch": 0.7345399698340875, + "grad_norm": 0.3941003778306099, + "learning_rate": 7.946498259041651e-05, + "loss": 0.4746, + "step": 1948 + }, + { + "epoch": 0.7349170437405732, + "grad_norm": 0.46891676850222724, + "learning_rate": 7.946326475506606e-05, + "loss": 0.4707, + "step": 1949 + }, + { + "epoch": 0.7352941176470589, + "grad_norm": 0.5115759626402087, + "learning_rate": 7.946154418494639e-05, + "loss": 0.4743, + "step": 1950 + }, + { + "epoch": 0.7356711915535445, + "grad_norm": 0.5159784354679234, + "learning_rate": 7.94598208801767e-05, + "loss": 0.4651, + "step": 1951 + }, + { + "epoch": 0.7360482654600302, + "grad_norm": 0.5484275194976055, + "learning_rate": 7.945809484087643e-05, + "loss": 0.4552, + "step": 1952 + }, + { + "epoch": 0.7364253393665159, + "grad_norm": 0.5672924463467541, + "learning_rate": 7.945636606716517e-05, + "loss": 0.4666, + "step": 1953 + }, + { + "epoch": 0.7368024132730016, + "grad_norm": 0.4758235520540251, + "learning_rate": 7.945463455916275e-05, + "loss": 0.4626, + "step": 1954 + }, + { + "epoch": 0.7371794871794872, + "grad_norm": 0.49829592565664405, + "learning_rate": 7.945290031698914e-05, + "loss": 0.4798, + "step": 1955 + }, + { + "epoch": 0.7375565610859729, + "grad_norm": 0.7132804171273617, + "learning_rate": 7.945116334076455e-05, + "loss": 0.467, + "step": 1956 + }, + { + "epoch": 0.7379336349924586, + "grad_norm": 0.8723669448222299, + "learning_rate": 7.944942363060933e-05, + "loss": 0.4669, + "step": 1957 + }, + { + "epoch": 0.7383107088989442, + "grad_norm": 0.8943957263632657, + "learning_rate": 7.944768118664404e-05, + "loss": 0.4581, + "step": 1958 + }, + { + "epoch": 0.7386877828054299, + "grad_norm": 0.853535798167205, + "learning_rate": 7.944593600898943e-05, + "loss": 0.4534, + "step": 1959 + }, + { + "epoch": 0.7390648567119156, + "grad_norm": 0.8083042136490884, + "learning_rate": 7.944418809776645e-05, + "loss": 0.4528, + "step": 1960 + }, + { + "epoch": 0.7394419306184012, + "grad_norm": 0.7077395039239203, + "learning_rate": 7.944243745309622e-05, + "loss": 0.4571, + "step": 1961 + }, + { + "epoch": 0.7398190045248869, + "grad_norm": 0.5461164989011992, + "learning_rate": 7.944068407510007e-05, + "loss": 0.45, + "step": 1962 + }, + { + "epoch": 0.7401960784313726, + "grad_norm": 0.3527713905657684, + "learning_rate": 7.943892796389947e-05, + "loss": 0.4536, + "step": 1963 + }, + { + "epoch": 0.7405731523378583, + "grad_norm": 0.3740258201318803, + "learning_rate": 7.943716911961617e-05, + "loss": 0.4678, + "step": 1964 + }, + { + "epoch": 0.7409502262443439, + "grad_norm": 0.4572801793296222, + "learning_rate": 7.943540754237202e-05, + "loss": 0.4766, + "step": 1965 + }, + { + "epoch": 0.7413273001508296, + "grad_norm": 0.4368666583627222, + "learning_rate": 7.943364323228911e-05, + "loss": 0.4627, + "step": 1966 + }, + { + "epoch": 0.7417043740573153, + "grad_norm": 0.4400583147261448, + "learning_rate": 7.943187618948969e-05, + "loss": 0.4706, + "step": 1967 + }, + { + "epoch": 0.7420814479638009, + "grad_norm": 0.4705825600103608, + "learning_rate": 7.943010641409623e-05, + "loss": 0.4781, + "step": 1968 + }, + { + "epoch": 0.7424585218702866, + "grad_norm": 0.5156636821166356, + "learning_rate": 7.942833390623139e-05, + "loss": 0.4597, + "step": 1969 + }, + { + "epoch": 0.7428355957767723, + "grad_norm": 0.512968284149845, + "learning_rate": 7.942655866601797e-05, + "loss": 0.4586, + "step": 1970 + }, + { + "epoch": 0.7432126696832579, + "grad_norm": 0.4540811136780335, + "learning_rate": 7.942478069357901e-05, + "loss": 0.471, + "step": 1971 + }, + { + "epoch": 0.7435897435897436, + "grad_norm": 0.4392311715879416, + "learning_rate": 7.942299998903771e-05, + "loss": 0.4627, + "step": 1972 + }, + { + "epoch": 0.7439668174962293, + "grad_norm": 0.5415156720508941, + "learning_rate": 7.942121655251748e-05, + "loss": 0.4611, + "step": 1973 + }, + { + "epoch": 0.744343891402715, + "grad_norm": 0.5911545171062365, + "learning_rate": 7.94194303841419e-05, + "loss": 0.4616, + "step": 1974 + }, + { + "epoch": 0.7447209653092006, + "grad_norm": 0.5616525419904006, + "learning_rate": 7.941764148403477e-05, + "loss": 0.4719, + "step": 1975 + }, + { + "epoch": 0.7450980392156863, + "grad_norm": 0.572872341378192, + "learning_rate": 7.941584985232005e-05, + "loss": 0.4674, + "step": 1976 + }, + { + "epoch": 0.745475113122172, + "grad_norm": 0.5784029269299524, + "learning_rate": 7.94140554891219e-05, + "loss": 0.4503, + "step": 1977 + }, + { + "epoch": 0.7458521870286576, + "grad_norm": 0.5876774797249145, + "learning_rate": 7.941225839456467e-05, + "loss": 0.4586, + "step": 1978 + }, + { + "epoch": 0.7462292609351433, + "grad_norm": 0.6386205171038545, + "learning_rate": 7.941045856877288e-05, + "loss": 0.4515, + "step": 1979 + }, + { + "epoch": 0.746606334841629, + "grad_norm": 0.6830183863572631, + "learning_rate": 7.940865601187127e-05, + "loss": 0.4751, + "step": 1980 + }, + { + "epoch": 0.7469834087481146, + "grad_norm": 0.5565850832152539, + "learning_rate": 7.940685072398475e-05, + "loss": 0.4642, + "step": 1981 + }, + { + "epoch": 0.7473604826546003, + "grad_norm": 0.4905518932168944, + "learning_rate": 7.940504270523844e-05, + "loss": 0.4625, + "step": 1982 + }, + { + "epoch": 0.747737556561086, + "grad_norm": 0.552090991191018, + "learning_rate": 7.940323195575762e-05, + "loss": 0.4562, + "step": 1983 + }, + { + "epoch": 0.7481146304675717, + "grad_norm": 0.6214598954758601, + "learning_rate": 7.940141847566777e-05, + "loss": 0.4565, + "step": 1984 + }, + { + "epoch": 0.7484917043740573, + "grad_norm": 0.630091493113887, + "learning_rate": 7.939960226509457e-05, + "loss": 0.4693, + "step": 1985 + }, + { + "epoch": 0.748868778280543, + "grad_norm": 0.6440415453626175, + "learning_rate": 7.939778332416387e-05, + "loss": 0.4737, + "step": 1986 + }, + { + "epoch": 0.7492458521870287, + "grad_norm": 0.7792671530096045, + "learning_rate": 7.939596165300174e-05, + "loss": 0.4566, + "step": 1987 + }, + { + "epoch": 0.7496229260935143, + "grad_norm": 0.7840040778893033, + "learning_rate": 7.939413725173442e-05, + "loss": 0.4439, + "step": 1988 + }, + { + "epoch": 0.75, + "grad_norm": 0.6446811310122565, + "learning_rate": 7.939231012048833e-05, + "loss": 0.4618, + "step": 1989 + }, + { + "epoch": 0.7503770739064857, + "grad_norm": 0.4366031997869782, + "learning_rate": 7.939048025939009e-05, + "loss": 0.4658, + "step": 1990 + }, + { + "epoch": 0.7507541478129713, + "grad_norm": 0.31832556852213534, + "learning_rate": 7.93886476685665e-05, + "loss": 0.4656, + "step": 1991 + }, + { + "epoch": 0.751131221719457, + "grad_norm": 0.4220851180583822, + "learning_rate": 7.938681234814456e-05, + "loss": 0.4743, + "step": 1992 + }, + { + "epoch": 0.7515082956259427, + "grad_norm": 0.4784975894034486, + "learning_rate": 7.938497429825146e-05, + "loss": 0.4555, + "step": 1993 + }, + { + "epoch": 0.7518853695324283, + "grad_norm": 0.4472556216471383, + "learning_rate": 7.938313351901459e-05, + "loss": 0.4655, + "step": 1994 + }, + { + "epoch": 0.752262443438914, + "grad_norm": 0.3333475277423364, + "learning_rate": 7.93812900105615e-05, + "loss": 0.45, + "step": 1995 + }, + { + "epoch": 0.7526395173453997, + "grad_norm": 0.34851595593163276, + "learning_rate": 7.937944377301993e-05, + "loss": 0.4742, + "step": 1996 + }, + { + "epoch": 0.7530165912518854, + "grad_norm": 0.4711557043054166, + "learning_rate": 7.937759480651784e-05, + "loss": 0.4677, + "step": 1997 + }, + { + "epoch": 0.753393665158371, + "grad_norm": 0.46369038315362643, + "learning_rate": 7.937574311118336e-05, + "loss": 0.4623, + "step": 1998 + }, + { + "epoch": 0.7537707390648567, + "grad_norm": 0.41602393614829314, + "learning_rate": 7.937388868714482e-05, + "loss": 0.4577, + "step": 1999 + }, + { + "epoch": 0.7541478129713424, + "grad_norm": 0.4628099774376263, + "learning_rate": 7.93720315345307e-05, + "loss": 0.458, + "step": 2000 + }, + { + "epoch": 0.754524886877828, + "grad_norm": 0.5398127434394624, + "learning_rate": 7.937017165346973e-05, + "loss": 0.4642, + "step": 2001 + }, + { + "epoch": 0.7549019607843137, + "grad_norm": 0.539563503259105, + "learning_rate": 7.936830904409076e-05, + "loss": 0.4732, + "step": 2002 + }, + { + "epoch": 0.7552790346907994, + "grad_norm": 0.5187078960518746, + "learning_rate": 7.936644370652292e-05, + "loss": 0.4617, + "step": 2003 + }, + { + "epoch": 0.755656108597285, + "grad_norm": 0.5355131758799087, + "learning_rate": 7.936457564089545e-05, + "loss": 0.4838, + "step": 2004 + }, + { + "epoch": 0.7560331825037707, + "grad_norm": 0.5661295078924811, + "learning_rate": 7.93627048473378e-05, + "loss": 0.4764, + "step": 2005 + }, + { + "epoch": 0.7564102564102564, + "grad_norm": 0.5614849463257705, + "learning_rate": 7.936083132597961e-05, + "loss": 0.4658, + "step": 2006 + }, + { + "epoch": 0.7567873303167421, + "grad_norm": 0.5814786757680512, + "learning_rate": 7.935895507695072e-05, + "loss": 0.4794, + "step": 2007 + }, + { + "epoch": 0.7571644042232277, + "grad_norm": 0.6307096226018888, + "learning_rate": 7.935707610038117e-05, + "loss": 0.4713, + "step": 2008 + }, + { + "epoch": 0.7575414781297134, + "grad_norm": 0.8792909027722161, + "learning_rate": 7.935519439640115e-05, + "loss": 0.4608, + "step": 2009 + }, + { + "epoch": 0.7579185520361991, + "grad_norm": 1.0550562258190435, + "learning_rate": 7.935330996514106e-05, + "loss": 0.465, + "step": 2010 + }, + { + "epoch": 0.7582956259426847, + "grad_norm": 0.725756582135744, + "learning_rate": 7.93514228067315e-05, + "loss": 0.4599, + "step": 2011 + }, + { + "epoch": 0.7586726998491704, + "grad_norm": 0.39724176332468797, + "learning_rate": 7.934953292130324e-05, + "loss": 0.4637, + "step": 2012 + }, + { + "epoch": 0.7590497737556561, + "grad_norm": 0.47290459737271734, + "learning_rate": 7.934764030898725e-05, + "loss": 0.471, + "step": 2013 + }, + { + "epoch": 0.7594268476621417, + "grad_norm": 0.7427493547905546, + "learning_rate": 7.93457449699147e-05, + "loss": 0.4664, + "step": 2014 + }, + { + "epoch": 0.7598039215686274, + "grad_norm": 0.768729039408056, + "learning_rate": 7.934384690421691e-05, + "loss": 0.4829, + "step": 2015 + }, + { + "epoch": 0.7601809954751131, + "grad_norm": 0.6694775336069767, + "learning_rate": 7.934194611202545e-05, + "loss": 0.4526, + "step": 2016 + }, + { + "epoch": 0.7605580693815988, + "grad_norm": 0.6590830511415078, + "learning_rate": 7.9340042593472e-05, + "loss": 0.4604, + "step": 2017 + }, + { + "epoch": 0.7609351432880844, + "grad_norm": 0.6751094499175414, + "learning_rate": 7.93381363486885e-05, + "loss": 0.4653, + "step": 2018 + }, + { + "epoch": 0.7613122171945701, + "grad_norm": 0.5797286201343327, + "learning_rate": 7.933622737780704e-05, + "loss": 0.4597, + "step": 2019 + }, + { + "epoch": 0.7616892911010558, + "grad_norm": 0.5024208788268818, + "learning_rate": 7.933431568095993e-05, + "loss": 0.4678, + "step": 2020 + }, + { + "epoch": 0.7620663650075414, + "grad_norm": 0.4856252658249792, + "learning_rate": 7.93324012582796e-05, + "loss": 0.4643, + "step": 2021 + }, + { + "epoch": 0.7624434389140271, + "grad_norm": 0.5310224757523093, + "learning_rate": 7.933048410989876e-05, + "loss": 0.4653, + "step": 2022 + }, + { + "epoch": 0.7628205128205128, + "grad_norm": 0.5494234412659831, + "learning_rate": 7.932856423595028e-05, + "loss": 0.4577, + "step": 2023 + }, + { + "epoch": 0.7631975867269984, + "grad_norm": 0.5249143224356269, + "learning_rate": 7.932664163656717e-05, + "loss": 0.4669, + "step": 2024 + }, + { + "epoch": 0.7635746606334841, + "grad_norm": 0.49813981162756243, + "learning_rate": 7.932471631188268e-05, + "loss": 0.4644, + "step": 2025 + }, + { + "epoch": 0.7639517345399698, + "grad_norm": 0.3617992594553467, + "learning_rate": 7.932278826203022e-05, + "loss": 0.463, + "step": 2026 + }, + { + "epoch": 0.7643288084464555, + "grad_norm": 0.32545408995947106, + "learning_rate": 7.932085748714343e-05, + "loss": 0.4574, + "step": 2027 + }, + { + "epoch": 0.7647058823529411, + "grad_norm": 0.4612105158100972, + "learning_rate": 7.931892398735608e-05, + "loss": 0.4607, + "step": 2028 + }, + { + "epoch": 0.7650829562594268, + "grad_norm": 0.5559460246569317, + "learning_rate": 7.931698776280217e-05, + "loss": 0.4568, + "step": 2029 + }, + { + "epoch": 0.7654600301659125, + "grad_norm": 0.4938524933015102, + "learning_rate": 7.931504881361589e-05, + "loss": 0.4529, + "step": 2030 + }, + { + "epoch": 0.7658371040723982, + "grad_norm": 0.5085378595501573, + "learning_rate": 7.93131071399316e-05, + "loss": 0.4711, + "step": 2031 + }, + { + "epoch": 0.7662141779788839, + "grad_norm": 0.6266369314804758, + "learning_rate": 7.931116274188385e-05, + "loss": 0.4614, + "step": 2032 + }, + { + "epoch": 0.7665912518853696, + "grad_norm": 0.7064363726681805, + "learning_rate": 7.930921561960741e-05, + "loss": 0.4664, + "step": 2033 + }, + { + "epoch": 0.7669683257918553, + "grad_norm": 0.7299909593311533, + "learning_rate": 7.930726577323718e-05, + "loss": 0.4701, + "step": 2034 + }, + { + "epoch": 0.7673453996983409, + "grad_norm": 0.7831240088245967, + "learning_rate": 7.93053132029083e-05, + "loss": 0.4604, + "step": 2035 + }, + { + "epoch": 0.7677224736048266, + "grad_norm": 0.8232874164716165, + "learning_rate": 7.930335790875607e-05, + "loss": 0.4696, + "step": 2036 + }, + { + "epoch": 0.7680995475113123, + "grad_norm": 0.6986086395525712, + "learning_rate": 7.930139989091601e-05, + "loss": 0.4461, + "step": 2037 + }, + { + "epoch": 0.7684766214177979, + "grad_norm": 0.528150648133401, + "learning_rate": 7.92994391495238e-05, + "loss": 0.4566, + "step": 2038 + }, + { + "epoch": 0.7688536953242836, + "grad_norm": 0.5507126672643068, + "learning_rate": 7.92974756847153e-05, + "loss": 0.4578, + "step": 2039 + }, + { + "epoch": 0.7692307692307693, + "grad_norm": 0.5651366813687866, + "learning_rate": 7.929550949662659e-05, + "loss": 0.4608, + "step": 2040 + }, + { + "epoch": 0.7696078431372549, + "grad_norm": 0.5333214909682272, + "learning_rate": 7.929354058539394e-05, + "loss": 0.4617, + "step": 2041 + }, + { + "epoch": 0.7699849170437406, + "grad_norm": 0.47380483400342194, + "learning_rate": 7.929156895115379e-05, + "loss": 0.4518, + "step": 2042 + }, + { + "epoch": 0.7703619909502263, + "grad_norm": 0.4048503564345619, + "learning_rate": 7.928959459404274e-05, + "loss": 0.4433, + "step": 2043 + }, + { + "epoch": 0.770739064856712, + "grad_norm": 0.38343425319073593, + "learning_rate": 7.928761751419766e-05, + "loss": 0.4549, + "step": 2044 + }, + { + "epoch": 0.7711161387631976, + "grad_norm": 0.4352669333070765, + "learning_rate": 7.928563771175551e-05, + "loss": 0.4532, + "step": 2045 + }, + { + "epoch": 0.7714932126696833, + "grad_norm": 0.49734858660973136, + "learning_rate": 7.928365518685352e-05, + "loss": 0.4605, + "step": 2046 + }, + { + "epoch": 0.771870286576169, + "grad_norm": 0.5054885901032271, + "learning_rate": 7.928166993962908e-05, + "loss": 0.4643, + "step": 2047 + }, + { + "epoch": 0.7722473604826546, + "grad_norm": 0.4762539856952944, + "learning_rate": 7.927968197021974e-05, + "loss": 0.4541, + "step": 2048 + }, + { + "epoch": 0.7726244343891403, + "grad_norm": 0.4890321164181861, + "learning_rate": 7.92776912787633e-05, + "loss": 0.4585, + "step": 2049 + }, + { + "epoch": 0.773001508295626, + "grad_norm": 0.48212722433601773, + "learning_rate": 7.927569786539769e-05, + "loss": 0.4526, + "step": 2050 + }, + { + "epoch": 0.7733785822021116, + "grad_norm": 0.5264401768002099, + "learning_rate": 7.927370173026104e-05, + "loss": 0.4834, + "step": 2051 + }, + { + "epoch": 0.7737556561085973, + "grad_norm": 0.47464439090543525, + "learning_rate": 7.927170287349172e-05, + "loss": 0.4606, + "step": 2052 + }, + { + "epoch": 0.774132730015083, + "grad_norm": 0.5099441734118199, + "learning_rate": 7.92697012952282e-05, + "loss": 0.4603, + "step": 2053 + }, + { + "epoch": 0.7745098039215687, + "grad_norm": 0.6402007822656793, + "learning_rate": 7.926769699560922e-05, + "loss": 0.466, + "step": 2054 + }, + { + "epoch": 0.7748868778280543, + "grad_norm": 0.7616888006253619, + "learning_rate": 7.926568997477367e-05, + "loss": 0.4643, + "step": 2055 + }, + { + "epoch": 0.77526395173454, + "grad_norm": 0.7925457615952876, + "learning_rate": 7.926368023286063e-05, + "loss": 0.4675, + "step": 2056 + }, + { + "epoch": 0.7756410256410257, + "grad_norm": 0.70905217838524, + "learning_rate": 7.926166777000939e-05, + "loss": 0.4635, + "step": 2057 + }, + { + "epoch": 0.7760180995475113, + "grad_norm": 0.6812739280860207, + "learning_rate": 7.925965258635939e-05, + "loss": 0.4601, + "step": 2058 + }, + { + "epoch": 0.776395173453997, + "grad_norm": 0.7840318919122666, + "learning_rate": 7.925763468205029e-05, + "loss": 0.4496, + "step": 2059 + }, + { + "epoch": 0.7767722473604827, + "grad_norm": 0.7849942090980935, + "learning_rate": 7.925561405722193e-05, + "loss": 0.4675, + "step": 2060 + }, + { + "epoch": 0.7771493212669683, + "grad_norm": 0.6667605579153818, + "learning_rate": 7.925359071201432e-05, + "loss": 0.4632, + "step": 2061 + }, + { + "epoch": 0.777526395173454, + "grad_norm": 0.5951875026392702, + "learning_rate": 7.92515646465677e-05, + "loss": 0.4585, + "step": 2062 + }, + { + "epoch": 0.7779034690799397, + "grad_norm": 0.6002677173998577, + "learning_rate": 7.924953586102248e-05, + "loss": 0.4598, + "step": 2063 + }, + { + "epoch": 0.7782805429864253, + "grad_norm": 0.7112544574165608, + "learning_rate": 7.924750435551923e-05, + "loss": 0.4763, + "step": 2064 + }, + { + "epoch": 0.778657616892911, + "grad_norm": 0.7892723561226833, + "learning_rate": 7.924547013019875e-05, + "loss": 0.4604, + "step": 2065 + }, + { + "epoch": 0.7790346907993967, + "grad_norm": 0.7461567431651722, + "learning_rate": 7.924343318520198e-05, + "loss": 0.4783, + "step": 2066 + }, + { + "epoch": 0.7794117647058824, + "grad_norm": 0.6446023166424683, + "learning_rate": 7.92413935206701e-05, + "loss": 0.4742, + "step": 2067 + }, + { + "epoch": 0.779788838612368, + "grad_norm": 0.6034943136462864, + "learning_rate": 7.923935113674447e-05, + "loss": 0.4666, + "step": 2068 + }, + { + "epoch": 0.7801659125188537, + "grad_norm": 0.5045788311283957, + "learning_rate": 7.923730603356661e-05, + "loss": 0.4589, + "step": 2069 + }, + { + "epoch": 0.7805429864253394, + "grad_norm": 0.46362261456222703, + "learning_rate": 7.923525821127824e-05, + "loss": 0.4567, + "step": 2070 + }, + { + "epoch": 0.780920060331825, + "grad_norm": 0.5279277038849359, + "learning_rate": 7.923320767002128e-05, + "loss": 0.4595, + "step": 2071 + }, + { + "epoch": 0.7812971342383107, + "grad_norm": 0.6159079021138688, + "learning_rate": 7.923115440993784e-05, + "loss": 0.4727, + "step": 2072 + }, + { + "epoch": 0.7816742081447964, + "grad_norm": 0.5010442999193937, + "learning_rate": 7.922909843117018e-05, + "loss": 0.4529, + "step": 2073 + }, + { + "epoch": 0.782051282051282, + "grad_norm": 0.3316518147830029, + "learning_rate": 7.92270397338608e-05, + "loss": 0.4592, + "step": 2074 + }, + { + "epoch": 0.7824283559577677, + "grad_norm": 0.5439295686578682, + "learning_rate": 7.922497831815236e-05, + "loss": 0.4715, + "step": 2075 + }, + { + "epoch": 0.7828054298642534, + "grad_norm": 0.5829150876376727, + "learning_rate": 7.922291418418772e-05, + "loss": 0.4594, + "step": 2076 + }, + { + "epoch": 0.7831825037707391, + "grad_norm": 0.4806488629652838, + "learning_rate": 7.922084733210992e-05, + "loss": 0.4605, + "step": 2077 + }, + { + "epoch": 0.7835595776772247, + "grad_norm": 0.4257235598237733, + "learning_rate": 7.921877776206216e-05, + "loss": 0.4771, + "step": 2078 + }, + { + "epoch": 0.7839366515837104, + "grad_norm": 0.4113694563987875, + "learning_rate": 7.921670547418792e-05, + "loss": 0.4661, + "step": 2079 + }, + { + "epoch": 0.7843137254901961, + "grad_norm": 0.4544700927423602, + "learning_rate": 7.921463046863077e-05, + "loss": 0.4527, + "step": 2080 + }, + { + "epoch": 0.7846907993966817, + "grad_norm": 0.5189615392917963, + "learning_rate": 7.921255274553451e-05, + "loss": 0.4733, + "step": 2081 + }, + { + "epoch": 0.7850678733031674, + "grad_norm": 0.5677426847427335, + "learning_rate": 7.921047230504313e-05, + "loss": 0.4654, + "step": 2082 + }, + { + "epoch": 0.7854449472096531, + "grad_norm": 0.6368232721303704, + "learning_rate": 7.920838914730079e-05, + "loss": 0.4563, + "step": 2083 + }, + { + "epoch": 0.7858220211161387, + "grad_norm": 0.661345096876884, + "learning_rate": 7.920630327245186e-05, + "loss": 0.4509, + "step": 2084 + }, + { + "epoch": 0.7861990950226244, + "grad_norm": 0.7073935044107469, + "learning_rate": 7.920421468064088e-05, + "loss": 0.4696, + "step": 2085 + }, + { + "epoch": 0.7865761689291101, + "grad_norm": 0.7828914042052272, + "learning_rate": 7.920212337201263e-05, + "loss": 0.4683, + "step": 2086 + }, + { + "epoch": 0.7869532428355958, + "grad_norm": 0.8287488700222719, + "learning_rate": 7.920002934671198e-05, + "loss": 0.4586, + "step": 2087 + }, + { + "epoch": 0.7873303167420814, + "grad_norm": 0.8157732934493979, + "learning_rate": 7.919793260488407e-05, + "loss": 0.4606, + "step": 2088 + }, + { + "epoch": 0.7877073906485671, + "grad_norm": 0.8615012988596968, + "learning_rate": 7.91958331466742e-05, + "loss": 0.4762, + "step": 2089 + }, + { + "epoch": 0.7880844645550528, + "grad_norm": 0.7373881746168521, + "learning_rate": 7.919373097222785e-05, + "loss": 0.4642, + "step": 2090 + }, + { + "epoch": 0.7884615384615384, + "grad_norm": 0.435002279370283, + "learning_rate": 7.919162608169072e-05, + "loss": 0.4646, + "step": 2091 + }, + { + "epoch": 0.7888386123680241, + "grad_norm": 0.5719345021241309, + "learning_rate": 7.918951847520865e-05, + "loss": 0.4651, + "step": 2092 + }, + { + "epoch": 0.7892156862745098, + "grad_norm": 0.795997915910881, + "learning_rate": 7.918740815292773e-05, + "loss": 0.465, + "step": 2093 + }, + { + "epoch": 0.7895927601809954, + "grad_norm": 0.7433958500259796, + "learning_rate": 7.918529511499418e-05, + "loss": 0.4622, + "step": 2094 + }, + { + "epoch": 0.7899698340874811, + "grad_norm": 0.5249893186530009, + "learning_rate": 7.918317936155442e-05, + "loss": 0.4525, + "step": 2095 + }, + { + "epoch": 0.7903469079939668, + "grad_norm": 0.4936069530191135, + "learning_rate": 7.91810608927551e-05, + "loss": 0.4547, + "step": 2096 + }, + { + "epoch": 0.7907239819004525, + "grad_norm": 0.6043432735042655, + "learning_rate": 7.917893970874302e-05, + "loss": 0.464, + "step": 2097 + }, + { + "epoch": 0.7911010558069381, + "grad_norm": 0.563930641643093, + "learning_rate": 7.917681580966516e-05, + "loss": 0.4698, + "step": 2098 + }, + { + "epoch": 0.7914781297134238, + "grad_norm": 0.46494178755975424, + "learning_rate": 7.917468919566871e-05, + "loss": 0.459, + "step": 2099 + }, + { + "epoch": 0.7918552036199095, + "grad_norm": 0.5019288694819658, + "learning_rate": 7.917255986690106e-05, + "loss": 0.4564, + "step": 2100 + }, + { + "epoch": 0.7922322775263951, + "grad_norm": 0.47587390202520835, + "learning_rate": 7.917042782350975e-05, + "loss": 0.4613, + "step": 2101 + }, + { + "epoch": 0.7926093514328808, + "grad_norm": 0.5424111370413562, + "learning_rate": 7.916829306564253e-05, + "loss": 0.4695, + "step": 2102 + }, + { + "epoch": 0.7929864253393665, + "grad_norm": 0.7308688486883405, + "learning_rate": 7.916615559344735e-05, + "loss": 0.4652, + "step": 2103 + }, + { + "epoch": 0.7933634992458521, + "grad_norm": 0.7491023545620091, + "learning_rate": 7.916401540707232e-05, + "loss": 0.459, + "step": 2104 + }, + { + "epoch": 0.7937405731523378, + "grad_norm": 0.7966147200678769, + "learning_rate": 7.916187250666576e-05, + "loss": 0.4624, + "step": 2105 + }, + { + "epoch": 0.7941176470588235, + "grad_norm": 0.7810717320254101, + "learning_rate": 7.915972689237618e-05, + "loss": 0.4501, + "step": 2106 + }, + { + "epoch": 0.7944947209653092, + "grad_norm": 0.6333798453663121, + "learning_rate": 7.915757856435227e-05, + "loss": 0.4666, + "step": 2107 + }, + { + "epoch": 0.7948717948717948, + "grad_norm": 0.4871849978181178, + "learning_rate": 7.915542752274289e-05, + "loss": 0.464, + "step": 2108 + }, + { + "epoch": 0.7952488687782805, + "grad_norm": 0.4211442950143248, + "learning_rate": 7.91532737676971e-05, + "loss": 0.4584, + "step": 2109 + }, + { + "epoch": 0.7956259426847662, + "grad_norm": 0.4133360950088245, + "learning_rate": 7.915111729936418e-05, + "loss": 0.4522, + "step": 2110 + }, + { + "epoch": 0.7960030165912518, + "grad_norm": 0.3994533439049555, + "learning_rate": 7.914895811789356e-05, + "loss": 0.4696, + "step": 2111 + }, + { + "epoch": 0.7963800904977375, + "grad_norm": 0.3354999134095492, + "learning_rate": 7.914679622343486e-05, + "loss": 0.4449, + "step": 2112 + }, + { + "epoch": 0.7967571644042232, + "grad_norm": 0.2818656271385539, + "learning_rate": 7.91446316161379e-05, + "loss": 0.4534, + "step": 2113 + }, + { + "epoch": 0.797134238310709, + "grad_norm": 0.34548703436210315, + "learning_rate": 7.914246429615271e-05, + "loss": 0.4651, + "step": 2114 + }, + { + "epoch": 0.7975113122171946, + "grad_norm": 0.4911574809731901, + "learning_rate": 7.914029426362945e-05, + "loss": 0.4736, + "step": 2115 + }, + { + "epoch": 0.7978883861236803, + "grad_norm": 0.5444397928749916, + "learning_rate": 7.913812151871852e-05, + "loss": 0.4668, + "step": 2116 + }, + { + "epoch": 0.798265460030166, + "grad_norm": 0.5566503176332582, + "learning_rate": 7.913594606157048e-05, + "loss": 0.4692, + "step": 2117 + }, + { + "epoch": 0.7986425339366516, + "grad_norm": 0.6279501476398094, + "learning_rate": 7.913376789233609e-05, + "loss": 0.4612, + "step": 2118 + }, + { + "epoch": 0.7990196078431373, + "grad_norm": 0.7260369732119129, + "learning_rate": 7.91315870111663e-05, + "loss": 0.4701, + "step": 2119 + }, + { + "epoch": 0.799396681749623, + "grad_norm": 0.8248632526547335, + "learning_rate": 7.912940341821224e-05, + "loss": 0.4482, + "step": 2120 + }, + { + "epoch": 0.7997737556561086, + "grad_norm": 0.943250066886947, + "learning_rate": 7.912721711362523e-05, + "loss": 0.4658, + "step": 2121 + }, + { + "epoch": 0.8001508295625943, + "grad_norm": 1.0141702587409198, + "learning_rate": 7.912502809755678e-05, + "loss": 0.4762, + "step": 2122 + }, + { + "epoch": 0.80052790346908, + "grad_norm": 0.9384422183530581, + "learning_rate": 7.912283637015859e-05, + "loss": 0.4658, + "step": 2123 + }, + { + "epoch": 0.8009049773755657, + "grad_norm": 0.7195495566811249, + "learning_rate": 7.912064193158253e-05, + "loss": 0.479, + "step": 2124 + }, + { + "epoch": 0.8012820512820513, + "grad_norm": 0.46935841718765114, + "learning_rate": 7.91184447819807e-05, + "loss": 0.4736, + "step": 2125 + }, + { + "epoch": 0.801659125188537, + "grad_norm": 0.40185867501721606, + "learning_rate": 7.911624492150534e-05, + "loss": 0.4588, + "step": 2126 + }, + { + "epoch": 0.8020361990950227, + "grad_norm": 0.5253901291031929, + "learning_rate": 7.91140423503089e-05, + "loss": 0.4633, + "step": 2127 + }, + { + "epoch": 0.8024132730015083, + "grad_norm": 0.5740474463360998, + "learning_rate": 7.911183706854402e-05, + "loss": 0.4654, + "step": 2128 + }, + { + "epoch": 0.802790346907994, + "grad_norm": 0.5483191039784865, + "learning_rate": 7.910962907636352e-05, + "loss": 0.4745, + "step": 2129 + }, + { + "epoch": 0.8031674208144797, + "grad_norm": 0.4159915246836675, + "learning_rate": 7.910741837392043e-05, + "loss": 0.4691, + "step": 2130 + }, + { + "epoch": 0.8035444947209653, + "grad_norm": 0.39489736404661185, + "learning_rate": 7.910520496136791e-05, + "loss": 0.4648, + "step": 2131 + }, + { + "epoch": 0.803921568627451, + "grad_norm": 0.4333646781210713, + "learning_rate": 7.910298883885939e-05, + "loss": 0.468, + "step": 2132 + }, + { + "epoch": 0.8042986425339367, + "grad_norm": 0.41926507840905325, + "learning_rate": 7.910077000654842e-05, + "loss": 0.4537, + "step": 2133 + }, + { + "epoch": 0.8046757164404224, + "grad_norm": 0.4069473887742344, + "learning_rate": 7.909854846458878e-05, + "loss": 0.468, + "step": 2134 + }, + { + "epoch": 0.805052790346908, + "grad_norm": 0.390771082972518, + "learning_rate": 7.90963242131344e-05, + "loss": 0.4644, + "step": 2135 + }, + { + "epoch": 0.8054298642533937, + "grad_norm": 0.3732649888786411, + "learning_rate": 7.909409725233943e-05, + "loss": 0.4583, + "step": 2136 + }, + { + "epoch": 0.8058069381598794, + "grad_norm": 0.5108278760239857, + "learning_rate": 7.90918675823582e-05, + "loss": 0.4563, + "step": 2137 + }, + { + "epoch": 0.806184012066365, + "grad_norm": 0.5999569919665764, + "learning_rate": 7.908963520334523e-05, + "loss": 0.4619, + "step": 2138 + }, + { + "epoch": 0.8065610859728507, + "grad_norm": 0.5558207623663161, + "learning_rate": 7.908740011545521e-05, + "loss": 0.4635, + "step": 2139 + }, + { + "epoch": 0.8069381598793364, + "grad_norm": 0.46758160512688784, + "learning_rate": 7.908516231884304e-05, + "loss": 0.471, + "step": 2140 + }, + { + "epoch": 0.807315233785822, + "grad_norm": 0.48962966764539206, + "learning_rate": 7.908292181366376e-05, + "loss": 0.4782, + "step": 2141 + }, + { + "epoch": 0.8076923076923077, + "grad_norm": 0.5155196215440516, + "learning_rate": 7.908067860007268e-05, + "loss": 0.465, + "step": 2142 + }, + { + "epoch": 0.8080693815987934, + "grad_norm": 0.496320025896346, + "learning_rate": 7.907843267822526e-05, + "loss": 0.4716, + "step": 2143 + }, + { + "epoch": 0.808446455505279, + "grad_norm": 0.4668189608322313, + "learning_rate": 7.90761840482771e-05, + "loss": 0.4513, + "step": 2144 + }, + { + "epoch": 0.8088235294117647, + "grad_norm": 0.40439110688049545, + "learning_rate": 7.907393271038403e-05, + "loss": 0.4599, + "step": 2145 + }, + { + "epoch": 0.8092006033182504, + "grad_norm": 0.46679750655655927, + "learning_rate": 7.90716786647021e-05, + "loss": 0.4616, + "step": 2146 + }, + { + "epoch": 0.8095776772247361, + "grad_norm": 0.5743398588870426, + "learning_rate": 7.906942191138748e-05, + "loss": 0.4547, + "step": 2147 + }, + { + "epoch": 0.8099547511312217, + "grad_norm": 0.7041379275256482, + "learning_rate": 7.90671624505966e-05, + "loss": 0.4623, + "step": 2148 + }, + { + "epoch": 0.8103318250377074, + "grad_norm": 0.7963336627803664, + "learning_rate": 7.9064900282486e-05, + "loss": 0.4605, + "step": 2149 + }, + { + "epoch": 0.8107088989441931, + "grad_norm": 0.7773406958069577, + "learning_rate": 7.906263540721246e-05, + "loss": 0.4608, + "step": 2150 + }, + { + "epoch": 0.8110859728506787, + "grad_norm": 0.7429198240552445, + "learning_rate": 7.906036782493294e-05, + "loss": 0.4491, + "step": 2151 + }, + { + "epoch": 0.8114630467571644, + "grad_norm": 0.7281937803424637, + "learning_rate": 7.905809753580459e-05, + "loss": 0.4624, + "step": 2152 + }, + { + "epoch": 0.8118401206636501, + "grad_norm": 0.6354012322877869, + "learning_rate": 7.90558245399847e-05, + "loss": 0.4677, + "step": 2153 + }, + { + "epoch": 0.8122171945701357, + "grad_norm": 0.45943393830452994, + "learning_rate": 7.905354883763082e-05, + "loss": 0.4619, + "step": 2154 + }, + { + "epoch": 0.8125942684766214, + "grad_norm": 0.4017588526997071, + "learning_rate": 7.905127042890065e-05, + "loss": 0.4638, + "step": 2155 + }, + { + "epoch": 0.8129713423831071, + "grad_norm": 0.48859681144154626, + "learning_rate": 7.904898931395208e-05, + "loss": 0.4533, + "step": 2156 + }, + { + "epoch": 0.8133484162895928, + "grad_norm": 0.5319931643411169, + "learning_rate": 7.904670549294319e-05, + "loss": 0.4631, + "step": 2157 + }, + { + "epoch": 0.8137254901960784, + "grad_norm": 0.44431644578189416, + "learning_rate": 7.904441896603223e-05, + "loss": 0.4682, + "step": 2158 + }, + { + "epoch": 0.8141025641025641, + "grad_norm": 0.43016072451568577, + "learning_rate": 7.904212973337768e-05, + "loss": 0.4802, + "step": 2159 + }, + { + "epoch": 0.8144796380090498, + "grad_norm": 0.5165211942655732, + "learning_rate": 7.903983779513816e-05, + "loss": 0.4696, + "step": 2160 + }, + { + "epoch": 0.8148567119155354, + "grad_norm": 0.5613321285901771, + "learning_rate": 7.903754315147251e-05, + "loss": 0.4563, + "step": 2161 + }, + { + "epoch": 0.8152337858220211, + "grad_norm": 0.493431985345958, + "learning_rate": 7.903524580253974e-05, + "loss": 0.4525, + "step": 2162 + }, + { + "epoch": 0.8156108597285068, + "grad_norm": 0.42644655693613154, + "learning_rate": 7.903294574849907e-05, + "loss": 0.4576, + "step": 2163 + }, + { + "epoch": 0.8159879336349924, + "grad_norm": 0.51046721360647, + "learning_rate": 7.903064298950988e-05, + "loss": 0.4587, + "step": 2164 + }, + { + "epoch": 0.8163650075414781, + "grad_norm": 0.5693549409185105, + "learning_rate": 7.902833752573176e-05, + "loss": 0.4586, + "step": 2165 + }, + { + "epoch": 0.8167420814479638, + "grad_norm": 0.46334563591136435, + "learning_rate": 7.902602935732446e-05, + "loss": 0.4713, + "step": 2166 + }, + { + "epoch": 0.8171191553544495, + "grad_norm": 0.4713863877157801, + "learning_rate": 7.902371848444794e-05, + "loss": 0.4541, + "step": 2167 + }, + { + "epoch": 0.8174962292609351, + "grad_norm": 0.5989805931563865, + "learning_rate": 7.902140490726234e-05, + "loss": 0.4753, + "step": 2168 + }, + { + "epoch": 0.8178733031674208, + "grad_norm": 0.7554866847008653, + "learning_rate": 7.9019088625928e-05, + "loss": 0.4626, + "step": 2169 + }, + { + "epoch": 0.8182503770739065, + "grad_norm": 0.9006045090387536, + "learning_rate": 7.901676964060543e-05, + "loss": 0.4778, + "step": 2170 + }, + { + "epoch": 0.8186274509803921, + "grad_norm": 0.9568524194007781, + "learning_rate": 7.901444795145532e-05, + "loss": 0.4549, + "step": 2171 + }, + { + "epoch": 0.8190045248868778, + "grad_norm": 0.9245771116880399, + "learning_rate": 7.901212355863858e-05, + "loss": 0.4479, + "step": 2172 + }, + { + "epoch": 0.8193815987933635, + "grad_norm": 0.7843238870144473, + "learning_rate": 7.900979646231628e-05, + "loss": 0.4621, + "step": 2173 + }, + { + "epoch": 0.8197586726998491, + "grad_norm": 0.6225234628218923, + "learning_rate": 7.90074666626497e-05, + "loss": 0.4632, + "step": 2174 + }, + { + "epoch": 0.8201357466063348, + "grad_norm": 0.40841359253861165, + "learning_rate": 7.900513415980026e-05, + "loss": 0.4796, + "step": 2175 + }, + { + "epoch": 0.8205128205128205, + "grad_norm": 0.38427828073770354, + "learning_rate": 7.900279895392961e-05, + "loss": 0.4605, + "step": 2176 + }, + { + "epoch": 0.8208898944193062, + "grad_norm": 0.460671590760494, + "learning_rate": 7.900046104519961e-05, + "loss": 0.4607, + "step": 2177 + }, + { + "epoch": 0.8212669683257918, + "grad_norm": 0.4893856287085096, + "learning_rate": 7.899812043377225e-05, + "loss": 0.4545, + "step": 2178 + }, + { + "epoch": 0.8216440422322775, + "grad_norm": 0.4532317990951028, + "learning_rate": 7.899577711980973e-05, + "loss": 0.4602, + "step": 2179 + }, + { + "epoch": 0.8220211161387632, + "grad_norm": 0.35113631937374007, + "learning_rate": 7.899343110347446e-05, + "loss": 0.4627, + "step": 2180 + }, + { + "epoch": 0.8223981900452488, + "grad_norm": 0.4032494338526516, + "learning_rate": 7.899108238492899e-05, + "loss": 0.4502, + "step": 2181 + }, + { + "epoch": 0.8227752639517345, + "grad_norm": 0.5374837533198421, + "learning_rate": 7.898873096433608e-05, + "loss": 0.4655, + "step": 2182 + }, + { + "epoch": 0.8231523378582202, + "grad_norm": 0.5342723113464377, + "learning_rate": 7.898637684185872e-05, + "loss": 0.459, + "step": 2183 + }, + { + "epoch": 0.8235294117647058, + "grad_norm": 0.5336531360141132, + "learning_rate": 7.898402001766002e-05, + "loss": 0.4536, + "step": 2184 + }, + { + "epoch": 0.8239064856711915, + "grad_norm": 0.6317488112539265, + "learning_rate": 7.89816604919033e-05, + "loss": 0.4531, + "step": 2185 + }, + { + "epoch": 0.8242835595776772, + "grad_norm": 0.6163128731844791, + "learning_rate": 7.897929826475211e-05, + "loss": 0.4569, + "step": 2186 + }, + { + "epoch": 0.8246606334841629, + "grad_norm": 0.4844560905043238, + "learning_rate": 7.89769333363701e-05, + "loss": 0.4736, + "step": 2187 + }, + { + "epoch": 0.8250377073906485, + "grad_norm": 0.4013587746079773, + "learning_rate": 7.897456570692119e-05, + "loss": 0.4713, + "step": 2188 + }, + { + "epoch": 0.8254147812971342, + "grad_norm": 0.5476692516220868, + "learning_rate": 7.897219537656946e-05, + "loss": 0.4587, + "step": 2189 + }, + { + "epoch": 0.8257918552036199, + "grad_norm": 0.7355198726103527, + "learning_rate": 7.896982234547914e-05, + "loss": 0.4721, + "step": 2190 + }, + { + "epoch": 0.8261689291101055, + "grad_norm": 0.7592927972860928, + "learning_rate": 7.896744661381469e-05, + "loss": 0.4602, + "step": 2191 + }, + { + "epoch": 0.8265460030165912, + "grad_norm": 0.6953221403878195, + "learning_rate": 7.896506818174077e-05, + "loss": 0.4496, + "step": 2192 + }, + { + "epoch": 0.8269230769230769, + "grad_norm": 0.5953255557743831, + "learning_rate": 7.896268704942219e-05, + "loss": 0.4521, + "step": 2193 + }, + { + "epoch": 0.8273001508295625, + "grad_norm": 0.5023508392077511, + "learning_rate": 7.896030321702395e-05, + "loss": 0.4747, + "step": 2194 + }, + { + "epoch": 0.8276772247360482, + "grad_norm": 0.5058226794673732, + "learning_rate": 7.895791668471126e-05, + "loss": 0.4534, + "step": 2195 + }, + { + "epoch": 0.8280542986425339, + "grad_norm": 0.5781322991579745, + "learning_rate": 7.895552745264949e-05, + "loss": 0.4689, + "step": 2196 + }, + { + "epoch": 0.8284313725490197, + "grad_norm": 0.56525546373648, + "learning_rate": 7.895313552100424e-05, + "loss": 0.453, + "step": 2197 + }, + { + "epoch": 0.8288084464555053, + "grad_norm": 0.4872288648723252, + "learning_rate": 7.895074088994123e-05, + "loss": 0.4713, + "step": 2198 + }, + { + "epoch": 0.829185520361991, + "grad_norm": 0.3923496724726736, + "learning_rate": 7.894834355962643e-05, + "loss": 0.4485, + "step": 2199 + }, + { + "epoch": 0.8295625942684767, + "grad_norm": 0.3287140041290878, + "learning_rate": 7.894594353022598e-05, + "loss": 0.459, + "step": 2200 + }, + { + "epoch": 0.8299396681749623, + "grad_norm": 0.27675661570668486, + "learning_rate": 7.894354080190618e-05, + "loss": 0.4869, + "step": 2201 + }, + { + "epoch": 0.830316742081448, + "grad_norm": 0.3440823905377867, + "learning_rate": 7.894113537483354e-05, + "loss": 0.4674, + "step": 2202 + }, + { + "epoch": 0.8306938159879337, + "grad_norm": 0.3749729893306855, + "learning_rate": 7.893872724917476e-05, + "loss": 0.4658, + "step": 2203 + }, + { + "epoch": 0.8310708898944194, + "grad_norm": 0.3873023495756802, + "learning_rate": 7.893631642509674e-05, + "loss": 0.4683, + "step": 2204 + }, + { + "epoch": 0.831447963800905, + "grad_norm": 0.3359055174467217, + "learning_rate": 7.893390290276651e-05, + "loss": 0.4662, + "step": 2205 + }, + { + "epoch": 0.8318250377073907, + "grad_norm": 0.29158980744372726, + "learning_rate": 7.893148668235134e-05, + "loss": 0.4536, + "step": 2206 + }, + { + "epoch": 0.8322021116138764, + "grad_norm": 0.291163854287167, + "learning_rate": 7.892906776401868e-05, + "loss": 0.473, + "step": 2207 + }, + { + "epoch": 0.832579185520362, + "grad_norm": 0.3238305920979451, + "learning_rate": 7.892664614793615e-05, + "loss": 0.4573, + "step": 2208 + }, + { + "epoch": 0.8329562594268477, + "grad_norm": 0.37280595129302513, + "learning_rate": 7.892422183427158e-05, + "loss": 0.4595, + "step": 2209 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 0.43812312602864084, + "learning_rate": 7.892179482319297e-05, + "loss": 0.4554, + "step": 2210 + }, + { + "epoch": 0.833710407239819, + "grad_norm": 0.628700305899036, + "learning_rate": 7.891936511486848e-05, + "loss": 0.4555, + "step": 2211 + }, + { + "epoch": 0.8340874811463047, + "grad_norm": 0.7874492705530838, + "learning_rate": 7.891693270946652e-05, + "loss": 0.4604, + "step": 2212 + }, + { + "epoch": 0.8344645550527904, + "grad_norm": 0.8848223429645162, + "learning_rate": 7.891449760715563e-05, + "loss": 0.4737, + "step": 2213 + }, + { + "epoch": 0.834841628959276, + "grad_norm": 0.9295919721045026, + "learning_rate": 7.89120598081046e-05, + "loss": 0.4627, + "step": 2214 + }, + { + "epoch": 0.8352187028657617, + "grad_norm": 1.0159587485256563, + "learning_rate": 7.890961931248232e-05, + "loss": 0.4644, + "step": 2215 + }, + { + "epoch": 0.8355957767722474, + "grad_norm": 0.9434682201813416, + "learning_rate": 7.890717612045793e-05, + "loss": 0.4616, + "step": 2216 + }, + { + "epoch": 0.8359728506787331, + "grad_norm": 0.8001069950500044, + "learning_rate": 7.890473023220076e-05, + "loss": 0.4643, + "step": 2217 + }, + { + "epoch": 0.8363499245852187, + "grad_norm": 0.6676002308741859, + "learning_rate": 7.89022816478803e-05, + "loss": 0.4648, + "step": 2218 + }, + { + "epoch": 0.8367269984917044, + "grad_norm": 0.6367474907555267, + "learning_rate": 7.88998303676662e-05, + "loss": 0.4616, + "step": 2219 + }, + { + "epoch": 0.8371040723981901, + "grad_norm": 0.6641111590888641, + "learning_rate": 7.889737639172839e-05, + "loss": 0.4601, + "step": 2220 + }, + { + "epoch": 0.8374811463046757, + "grad_norm": 0.5890508777537113, + "learning_rate": 7.889491972023688e-05, + "loss": 0.4704, + "step": 2221 + }, + { + "epoch": 0.8378582202111614, + "grad_norm": 0.3963059081759537, + "learning_rate": 7.889246035336193e-05, + "loss": 0.455, + "step": 2222 + }, + { + "epoch": 0.8382352941176471, + "grad_norm": 0.39459535188047107, + "learning_rate": 7.888999829127398e-05, + "loss": 0.4476, + "step": 2223 + }, + { + "epoch": 0.8386123680241327, + "grad_norm": 0.5803861169178918, + "learning_rate": 7.888753353414364e-05, + "loss": 0.4577, + "step": 2224 + }, + { + "epoch": 0.8389894419306184, + "grad_norm": 0.5809085311219315, + "learning_rate": 7.888506608214174e-05, + "loss": 0.468, + "step": 2225 + }, + { + "epoch": 0.8393665158371041, + "grad_norm": 0.4477744025341835, + "learning_rate": 7.888259593543924e-05, + "loss": 0.4579, + "step": 2226 + }, + { + "epoch": 0.8397435897435898, + "grad_norm": 0.3860661250116361, + "learning_rate": 7.888012309420732e-05, + "loss": 0.4602, + "step": 2227 + }, + { + "epoch": 0.8401206636500754, + "grad_norm": 0.37106491761318633, + "learning_rate": 7.887764755861736e-05, + "loss": 0.4512, + "step": 2228 + }, + { + "epoch": 0.8404977375565611, + "grad_norm": 0.4343168992687313, + "learning_rate": 7.887516932884093e-05, + "loss": 0.4597, + "step": 2229 + }, + { + "epoch": 0.8408748114630468, + "grad_norm": 0.5007669947619559, + "learning_rate": 7.887268840504972e-05, + "loss": 0.4546, + "step": 2230 + }, + { + "epoch": 0.8412518853695324, + "grad_norm": 0.49185469464722814, + "learning_rate": 7.88702047874157e-05, + "loss": 0.4603, + "step": 2231 + }, + { + "epoch": 0.8416289592760181, + "grad_norm": 0.5056473897099316, + "learning_rate": 7.886771847611094e-05, + "loss": 0.4456, + "step": 2232 + }, + { + "epoch": 0.8420060331825038, + "grad_norm": 0.48102599110126265, + "learning_rate": 7.88652294713078e-05, + "loss": 0.4575, + "step": 2233 + }, + { + "epoch": 0.8423831070889894, + "grad_norm": 0.34963984204576193, + "learning_rate": 7.886273777317871e-05, + "loss": 0.4536, + "step": 2234 + }, + { + "epoch": 0.8427601809954751, + "grad_norm": 0.41441116664472694, + "learning_rate": 7.886024338189636e-05, + "loss": 0.4659, + "step": 2235 + }, + { + "epoch": 0.8431372549019608, + "grad_norm": 0.4689303751193918, + "learning_rate": 7.885774629763363e-05, + "loss": 0.4659, + "step": 2236 + }, + { + "epoch": 0.8435143288084465, + "grad_norm": 0.48658394957665224, + "learning_rate": 7.885524652056354e-05, + "loss": 0.461, + "step": 2237 + }, + { + "epoch": 0.8438914027149321, + "grad_norm": 0.4764256820821457, + "learning_rate": 7.885274405085932e-05, + "loss": 0.4703, + "step": 2238 + }, + { + "epoch": 0.8442684766214178, + "grad_norm": 0.4713347542701761, + "learning_rate": 7.885023888869441e-05, + "loss": 0.4643, + "step": 2239 + }, + { + "epoch": 0.8446455505279035, + "grad_norm": 0.3995081241627039, + "learning_rate": 7.884773103424242e-05, + "loss": 0.4698, + "step": 2240 + }, + { + "epoch": 0.8450226244343891, + "grad_norm": 0.3355900590355575, + "learning_rate": 7.884522048767712e-05, + "loss": 0.4587, + "step": 2241 + }, + { + "epoch": 0.8453996983408748, + "grad_norm": 0.40285395011140634, + "learning_rate": 7.884270724917248e-05, + "loss": 0.453, + "step": 2242 + }, + { + "epoch": 0.8457767722473605, + "grad_norm": 0.4142663200511909, + "learning_rate": 7.884019131890268e-05, + "loss": 0.4535, + "step": 2243 + }, + { + "epoch": 0.8461538461538461, + "grad_norm": 0.40667780888090627, + "learning_rate": 7.883767269704209e-05, + "loss": 0.4609, + "step": 2244 + }, + { + "epoch": 0.8465309200603318, + "grad_norm": 0.4527719548166259, + "learning_rate": 7.883515138376522e-05, + "loss": 0.4556, + "step": 2245 + }, + { + "epoch": 0.8469079939668175, + "grad_norm": 0.554470331306824, + "learning_rate": 7.88326273792468e-05, + "loss": 0.4636, + "step": 2246 + }, + { + "epoch": 0.8472850678733032, + "grad_norm": 0.6632114581755701, + "learning_rate": 7.883010068366175e-05, + "loss": 0.4546, + "step": 2247 + }, + { + "epoch": 0.8476621417797888, + "grad_norm": 0.7863475366684481, + "learning_rate": 7.882757129718519e-05, + "loss": 0.4538, + "step": 2248 + }, + { + "epoch": 0.8480392156862745, + "grad_norm": 0.925356131766317, + "learning_rate": 7.882503921999237e-05, + "loss": 0.46, + "step": 2249 + }, + { + "epoch": 0.8484162895927602, + "grad_norm": 0.9130005800174423, + "learning_rate": 7.882250445225876e-05, + "loss": 0.4643, + "step": 2250 + }, + { + "epoch": 0.8487933634992458, + "grad_norm": 0.7389968827519982, + "learning_rate": 7.881996699416002e-05, + "loss": 0.4591, + "step": 2251 + }, + { + "epoch": 0.8491704374057315, + "grad_norm": 0.5206267990167278, + "learning_rate": 7.8817426845872e-05, + "loss": 0.4602, + "step": 2252 + }, + { + "epoch": 0.8495475113122172, + "grad_norm": 0.48385509471125665, + "learning_rate": 7.881488400757074e-05, + "loss": 0.4525, + "step": 2253 + }, + { + "epoch": 0.8499245852187028, + "grad_norm": 0.5915138954549053, + "learning_rate": 7.881233847943245e-05, + "loss": 0.4508, + "step": 2254 + }, + { + "epoch": 0.8503016591251885, + "grad_norm": 0.5300013630615287, + "learning_rate": 7.880979026163353e-05, + "loss": 0.4652, + "step": 2255 + }, + { + "epoch": 0.8506787330316742, + "grad_norm": 0.41719816707604834, + "learning_rate": 7.880723935435055e-05, + "loss": 0.4627, + "step": 2256 + }, + { + "epoch": 0.8510558069381599, + "grad_norm": 0.35559296007692826, + "learning_rate": 7.880468575776032e-05, + "loss": 0.4676, + "step": 2257 + }, + { + "epoch": 0.8514328808446455, + "grad_norm": 0.4784572642162056, + "learning_rate": 7.880212947203979e-05, + "loss": 0.4616, + "step": 2258 + }, + { + "epoch": 0.8518099547511312, + "grad_norm": 0.5977772616174469, + "learning_rate": 7.87995704973661e-05, + "loss": 0.4735, + "step": 2259 + }, + { + "epoch": 0.8521870286576169, + "grad_norm": 0.6218003084962648, + "learning_rate": 7.879700883391659e-05, + "loss": 0.4546, + "step": 2260 + }, + { + "epoch": 0.8525641025641025, + "grad_norm": 0.5242717984428593, + "learning_rate": 7.879444448186878e-05, + "loss": 0.4773, + "step": 2261 + }, + { + "epoch": 0.8529411764705882, + "grad_norm": 0.4213846077557083, + "learning_rate": 7.879187744140039e-05, + "loss": 0.4442, + "step": 2262 + }, + { + "epoch": 0.8533182503770739, + "grad_norm": 0.44683103335875185, + "learning_rate": 7.87893077126893e-05, + "loss": 0.4718, + "step": 2263 + }, + { + "epoch": 0.8536953242835595, + "grad_norm": 0.49798688493072507, + "learning_rate": 7.878673529591357e-05, + "loss": 0.4593, + "step": 2264 + }, + { + "epoch": 0.8540723981900452, + "grad_norm": 0.391873754334553, + "learning_rate": 7.87841601912515e-05, + "loss": 0.4665, + "step": 2265 + }, + { + "epoch": 0.8544494720965309, + "grad_norm": 0.44749644967769814, + "learning_rate": 7.878158239888154e-05, + "loss": 0.454, + "step": 2266 + }, + { + "epoch": 0.8548265460030166, + "grad_norm": 0.5302879372597463, + "learning_rate": 7.877900191898229e-05, + "loss": 0.4658, + "step": 2267 + }, + { + "epoch": 0.8552036199095022, + "grad_norm": 0.4496941117147298, + "learning_rate": 7.877641875173264e-05, + "loss": 0.4595, + "step": 2268 + }, + { + "epoch": 0.8555806938159879, + "grad_norm": 0.4385197608551854, + "learning_rate": 7.877383289731154e-05, + "loss": 0.4547, + "step": 2269 + }, + { + "epoch": 0.8559577677224736, + "grad_norm": 0.5030697508418475, + "learning_rate": 7.877124435589822e-05, + "loss": 0.453, + "step": 2270 + }, + { + "epoch": 0.8563348416289592, + "grad_norm": 0.5437857024101127, + "learning_rate": 7.876865312767206e-05, + "loss": 0.4629, + "step": 2271 + }, + { + "epoch": 0.8567119155354449, + "grad_norm": 0.5178403447289991, + "learning_rate": 7.876605921281263e-05, + "loss": 0.4591, + "step": 2272 + }, + { + "epoch": 0.8570889894419306, + "grad_norm": 0.4815009972523698, + "learning_rate": 7.876346261149968e-05, + "loss": 0.4708, + "step": 2273 + }, + { + "epoch": 0.8574660633484162, + "grad_norm": 0.6015874495707233, + "learning_rate": 7.876086332391314e-05, + "loss": 0.4518, + "step": 2274 + }, + { + "epoch": 0.8578431372549019, + "grad_norm": 0.5941538930282906, + "learning_rate": 7.875826135023316e-05, + "loss": 0.4497, + "step": 2275 + }, + { + "epoch": 0.8582202111613876, + "grad_norm": 0.4116214363500213, + "learning_rate": 7.875565669064005e-05, + "loss": 0.459, + "step": 2276 + }, + { + "epoch": 0.8585972850678733, + "grad_norm": 0.374143454154581, + "learning_rate": 7.87530493453143e-05, + "loss": 0.4483, + "step": 2277 + }, + { + "epoch": 0.8589743589743589, + "grad_norm": 0.3758131303753649, + "learning_rate": 7.875043931443663e-05, + "loss": 0.465, + "step": 2278 + }, + { + "epoch": 0.8593514328808446, + "grad_norm": 0.395670453858731, + "learning_rate": 7.874782659818786e-05, + "loss": 0.4647, + "step": 2279 + }, + { + "epoch": 0.8597285067873304, + "grad_norm": 0.42162755631820537, + "learning_rate": 7.874521119674909e-05, + "loss": 0.4572, + "step": 2280 + }, + { + "epoch": 0.860105580693816, + "grad_norm": 0.39537689930685055, + "learning_rate": 7.874259311030154e-05, + "loss": 0.4634, + "step": 2281 + }, + { + "epoch": 0.8604826546003017, + "grad_norm": 0.32000809953460235, + "learning_rate": 7.873997233902666e-05, + "loss": 0.4611, + "step": 2282 + }, + { + "epoch": 0.8608597285067874, + "grad_norm": 0.3852264954046526, + "learning_rate": 7.873734888310605e-05, + "loss": 0.459, + "step": 2283 + }, + { + "epoch": 0.861236802413273, + "grad_norm": 0.5446781568813421, + "learning_rate": 7.873472274272153e-05, + "loss": 0.4687, + "step": 2284 + }, + { + "epoch": 0.8616138763197587, + "grad_norm": 0.646463119283529, + "learning_rate": 7.873209391805507e-05, + "loss": 0.4569, + "step": 2285 + }, + { + "epoch": 0.8619909502262444, + "grad_norm": 0.5779771126241356, + "learning_rate": 7.872946240928888e-05, + "loss": 0.4521, + "step": 2286 + }, + { + "epoch": 0.8623680241327301, + "grad_norm": 0.46069082180200965, + "learning_rate": 7.872682821660527e-05, + "loss": 0.4557, + "step": 2287 + }, + { + "epoch": 0.8627450980392157, + "grad_norm": 0.5512105711229978, + "learning_rate": 7.872419134018683e-05, + "loss": 0.4594, + "step": 2288 + }, + { + "epoch": 0.8631221719457014, + "grad_norm": 0.635459665368003, + "learning_rate": 7.872155178021629e-05, + "loss": 0.4664, + "step": 2289 + }, + { + "epoch": 0.8634992458521871, + "grad_norm": 0.6893693782334428, + "learning_rate": 7.871890953687655e-05, + "loss": 0.4643, + "step": 2290 + }, + { + "epoch": 0.8638763197586727, + "grad_norm": 0.6431503138329531, + "learning_rate": 7.871626461035072e-05, + "loss": 0.4638, + "step": 2291 + }, + { + "epoch": 0.8642533936651584, + "grad_norm": 0.5546949273512062, + "learning_rate": 7.871361700082209e-05, + "loss": 0.452, + "step": 2292 + }, + { + "epoch": 0.8646304675716441, + "grad_norm": 0.596509010145194, + "learning_rate": 7.871096670847415e-05, + "loss": 0.4517, + "step": 2293 + }, + { + "epoch": 0.8650075414781297, + "grad_norm": 0.6575621953945664, + "learning_rate": 7.870831373349054e-05, + "loss": 0.466, + "step": 2294 + }, + { + "epoch": 0.8653846153846154, + "grad_norm": 0.6511574378153028, + "learning_rate": 7.870565807605513e-05, + "loss": 0.4611, + "step": 2295 + }, + { + "epoch": 0.8657616892911011, + "grad_norm": 0.6081181715619959, + "learning_rate": 7.870299973635195e-05, + "loss": 0.4569, + "step": 2296 + }, + { + "epoch": 0.8661387631975868, + "grad_norm": 0.5328455709694206, + "learning_rate": 7.870033871456521e-05, + "loss": 0.4562, + "step": 2297 + }, + { + "epoch": 0.8665158371040724, + "grad_norm": 0.4153627437471054, + "learning_rate": 7.869767501087934e-05, + "loss": 0.4611, + "step": 2298 + }, + { + "epoch": 0.8668929110105581, + "grad_norm": 0.37397218000669746, + "learning_rate": 7.86950086254789e-05, + "loss": 0.4811, + "step": 2299 + }, + { + "epoch": 0.8672699849170438, + "grad_norm": 0.3701114835754512, + "learning_rate": 7.869233955854869e-05, + "loss": 0.4565, + "step": 2300 + }, + { + "epoch": 0.8676470588235294, + "grad_norm": 0.3887981793026982, + "learning_rate": 7.868966781027367e-05, + "loss": 0.4631, + "step": 2301 + }, + { + "epoch": 0.8680241327300151, + "grad_norm": 0.39615960635245734, + "learning_rate": 7.868699338083898e-05, + "loss": 0.4466, + "step": 2302 + }, + { + "epoch": 0.8684012066365008, + "grad_norm": 0.5458700805350426, + "learning_rate": 7.868431627042998e-05, + "loss": 0.4492, + "step": 2303 + }, + { + "epoch": 0.8687782805429864, + "grad_norm": 0.5877144290136597, + "learning_rate": 7.868163647923217e-05, + "loss": 0.4592, + "step": 2304 + }, + { + "epoch": 0.8691553544494721, + "grad_norm": 0.44362653998911494, + "learning_rate": 7.867895400743125e-05, + "loss": 0.4618, + "step": 2305 + }, + { + "epoch": 0.8695324283559578, + "grad_norm": 0.3356219814337405, + "learning_rate": 7.867626885521315e-05, + "loss": 0.457, + "step": 2306 + }, + { + "epoch": 0.8699095022624435, + "grad_norm": 0.494696122379597, + "learning_rate": 7.86735810227639e-05, + "loss": 0.4562, + "step": 2307 + }, + { + "epoch": 0.8702865761689291, + "grad_norm": 0.6537308103887629, + "learning_rate": 7.86708905102698e-05, + "loss": 0.4521, + "step": 2308 + }, + { + "epoch": 0.8706636500754148, + "grad_norm": 0.6705227748275269, + "learning_rate": 7.866819731791728e-05, + "loss": 0.4503, + "step": 2309 + }, + { + "epoch": 0.8710407239819005, + "grad_norm": 0.6745855462793455, + "learning_rate": 7.8665501445893e-05, + "loss": 0.4383, + "step": 2310 + }, + { + "epoch": 0.8714177978883861, + "grad_norm": 0.7312165739417816, + "learning_rate": 7.866280289438376e-05, + "loss": 0.4679, + "step": 2311 + }, + { + "epoch": 0.8717948717948718, + "grad_norm": 0.7569135740986688, + "learning_rate": 7.866010166357656e-05, + "loss": 0.4635, + "step": 2312 + }, + { + "epoch": 0.8721719457013575, + "grad_norm": 0.7932692179232969, + "learning_rate": 7.865739775365863e-05, + "loss": 0.455, + "step": 2313 + }, + { + "epoch": 0.8725490196078431, + "grad_norm": 0.849592433633073, + "learning_rate": 7.865469116481731e-05, + "loss": 0.4554, + "step": 2314 + }, + { + "epoch": 0.8729260935143288, + "grad_norm": 0.8551827558095709, + "learning_rate": 7.865198189724017e-05, + "loss": 0.4627, + "step": 2315 + }, + { + "epoch": 0.8733031674208145, + "grad_norm": 0.8211911458119362, + "learning_rate": 7.864926995111498e-05, + "loss": 0.4545, + "step": 2316 + }, + { + "epoch": 0.8736802413273002, + "grad_norm": 0.6630325891089343, + "learning_rate": 7.864655532662966e-05, + "loss": 0.4485, + "step": 2317 + }, + { + "epoch": 0.8740573152337858, + "grad_norm": 0.4765778267723367, + "learning_rate": 7.864383802397235e-05, + "loss": 0.4579, + "step": 2318 + }, + { + "epoch": 0.8744343891402715, + "grad_norm": 0.47819678854251213, + "learning_rate": 7.864111804333132e-05, + "loss": 0.4447, + "step": 2319 + }, + { + "epoch": 0.8748114630467572, + "grad_norm": 0.6145547952297102, + "learning_rate": 7.86383953848951e-05, + "loss": 0.4488, + "step": 2320 + }, + { + "epoch": 0.8751885369532428, + "grad_norm": 0.7079730482016183, + "learning_rate": 7.863567004885234e-05, + "loss": 0.4578, + "step": 2321 + }, + { + "epoch": 0.8755656108597285, + "grad_norm": 0.6904835932222998, + "learning_rate": 7.863294203539193e-05, + "loss": 0.4553, + "step": 2322 + }, + { + "epoch": 0.8759426847662142, + "grad_norm": 0.5245011190543785, + "learning_rate": 7.86302113447029e-05, + "loss": 0.4539, + "step": 2323 + }, + { + "epoch": 0.8763197586726998, + "grad_norm": 0.41168775194815355, + "learning_rate": 7.862747797697448e-05, + "loss": 0.4625, + "step": 2324 + }, + { + "epoch": 0.8766968325791855, + "grad_norm": 0.5437541559522523, + "learning_rate": 7.86247419323961e-05, + "loss": 0.4584, + "step": 2325 + }, + { + "epoch": 0.8770739064856712, + "grad_norm": 0.6455234810254277, + "learning_rate": 7.862200321115737e-05, + "loss": 0.4665, + "step": 2326 + }, + { + "epoch": 0.8774509803921569, + "grad_norm": 0.5898188083463254, + "learning_rate": 7.861926181344807e-05, + "loss": 0.458, + "step": 2327 + }, + { + "epoch": 0.8778280542986425, + "grad_norm": 0.4056185735940619, + "learning_rate": 7.861651773945818e-05, + "loss": 0.4835, + "step": 2328 + }, + { + "epoch": 0.8782051282051282, + "grad_norm": 0.3728449009294552, + "learning_rate": 7.861377098937788e-05, + "loss": 0.4505, + "step": 2329 + }, + { + "epoch": 0.8785822021116139, + "grad_norm": 0.4271318241031113, + "learning_rate": 7.861102156339748e-05, + "loss": 0.4661, + "step": 2330 + }, + { + "epoch": 0.8789592760180995, + "grad_norm": 0.4884780318604645, + "learning_rate": 7.860826946170755e-05, + "loss": 0.4624, + "step": 2331 + }, + { + "epoch": 0.8793363499245852, + "grad_norm": 0.4532077312011205, + "learning_rate": 7.86055146844988e-05, + "loss": 0.4618, + "step": 2332 + }, + { + "epoch": 0.8797134238310709, + "grad_norm": 0.4309176983917419, + "learning_rate": 7.860275723196209e-05, + "loss": 0.4617, + "step": 2333 + }, + { + "epoch": 0.8800904977375565, + "grad_norm": 0.5393322039461019, + "learning_rate": 7.859999710428856e-05, + "loss": 0.4707, + "step": 2334 + }, + { + "epoch": 0.8804675716440422, + "grad_norm": 0.7105748530210938, + "learning_rate": 7.859723430166949e-05, + "loss": 0.4603, + "step": 2335 + }, + { + "epoch": 0.8808446455505279, + "grad_norm": 0.6699515995897791, + "learning_rate": 7.859446882429629e-05, + "loss": 0.456, + "step": 2336 + }, + { + "epoch": 0.8812217194570136, + "grad_norm": 0.5032590642625869, + "learning_rate": 7.859170067236065e-05, + "loss": 0.4547, + "step": 2337 + }, + { + "epoch": 0.8815987933634992, + "grad_norm": 0.45750770635110705, + "learning_rate": 7.858892984605439e-05, + "loss": 0.4625, + "step": 2338 + }, + { + "epoch": 0.8819758672699849, + "grad_norm": 0.5606854719911016, + "learning_rate": 7.85861563455695e-05, + "loss": 0.4544, + "step": 2339 + }, + { + "epoch": 0.8823529411764706, + "grad_norm": 0.5535111797427748, + "learning_rate": 7.858338017109822e-05, + "loss": 0.4409, + "step": 2340 + }, + { + "epoch": 0.8827300150829562, + "grad_norm": 0.4809324244994769, + "learning_rate": 7.85806013228329e-05, + "loss": 0.46, + "step": 2341 + }, + { + "epoch": 0.8831070889894419, + "grad_norm": 0.4794923994555877, + "learning_rate": 7.857781980096613e-05, + "loss": 0.4704, + "step": 2342 + }, + { + "epoch": 0.8834841628959276, + "grad_norm": 0.5837584309396989, + "learning_rate": 7.857503560569069e-05, + "loss": 0.4738, + "step": 2343 + }, + { + "epoch": 0.8838612368024132, + "grad_norm": 0.7266067300739205, + "learning_rate": 7.857224873719948e-05, + "loss": 0.4623, + "step": 2344 + }, + { + "epoch": 0.8842383107088989, + "grad_norm": 0.7741725891368612, + "learning_rate": 7.856945919568564e-05, + "loss": 0.4601, + "step": 2345 + }, + { + "epoch": 0.8846153846153846, + "grad_norm": 0.7471199311577178, + "learning_rate": 7.85666669813425e-05, + "loss": 0.4557, + "step": 2346 + }, + { + "epoch": 0.8849924585218703, + "grad_norm": 0.6828403636839858, + "learning_rate": 7.856387209436355e-05, + "loss": 0.4536, + "step": 2347 + }, + { + "epoch": 0.8853695324283559, + "grad_norm": 0.5668192270706496, + "learning_rate": 7.856107453494246e-05, + "loss": 0.4723, + "step": 2348 + }, + { + "epoch": 0.8857466063348416, + "grad_norm": 0.46679805170319105, + "learning_rate": 7.855827430327312e-05, + "loss": 0.4546, + "step": 2349 + }, + { + "epoch": 0.8861236802413273, + "grad_norm": 0.39563161456284124, + "learning_rate": 7.855547139954956e-05, + "loss": 0.4529, + "step": 2350 + }, + { + "epoch": 0.8865007541478129, + "grad_norm": 0.4037488079934052, + "learning_rate": 7.855266582396603e-05, + "loss": 0.4533, + "step": 2351 + }, + { + "epoch": 0.8868778280542986, + "grad_norm": 0.45846685287420696, + "learning_rate": 7.854985757671697e-05, + "loss": 0.458, + "step": 2352 + }, + { + "epoch": 0.8872549019607843, + "grad_norm": 0.47449812322261165, + "learning_rate": 7.854704665799696e-05, + "loss": 0.46, + "step": 2353 + }, + { + "epoch": 0.8876319758672699, + "grad_norm": 0.41869080054814534, + "learning_rate": 7.854423306800081e-05, + "loss": 0.459, + "step": 2354 + }, + { + "epoch": 0.8880090497737556, + "grad_norm": 0.42478882896687137, + "learning_rate": 7.85414168069235e-05, + "loss": 0.4472, + "step": 2355 + }, + { + "epoch": 0.8883861236802413, + "grad_norm": 0.4867920265715907, + "learning_rate": 7.853859787496019e-05, + "loss": 0.4402, + "step": 2356 + }, + { + "epoch": 0.888763197586727, + "grad_norm": 0.5227934982430485, + "learning_rate": 7.853577627230623e-05, + "loss": 0.4671, + "step": 2357 + }, + { + "epoch": 0.8891402714932126, + "grad_norm": 0.5477416467442977, + "learning_rate": 7.853295199915716e-05, + "loss": 0.4431, + "step": 2358 + }, + { + "epoch": 0.8895173453996983, + "grad_norm": 0.5769949828687656, + "learning_rate": 7.853012505570869e-05, + "loss": 0.4463, + "step": 2359 + }, + { + "epoch": 0.889894419306184, + "grad_norm": 0.7205592487382947, + "learning_rate": 7.852729544215672e-05, + "loss": 0.4716, + "step": 2360 + }, + { + "epoch": 0.8902714932126696, + "grad_norm": 0.83005249446074, + "learning_rate": 7.852446315869737e-05, + "loss": 0.4704, + "step": 2361 + }, + { + "epoch": 0.8906485671191554, + "grad_norm": 0.8145064549302102, + "learning_rate": 7.852162820552688e-05, + "loss": 0.4684, + "step": 2362 + }, + { + "epoch": 0.8910256410256411, + "grad_norm": 0.7535627567769007, + "learning_rate": 7.851879058284173e-05, + "loss": 0.4625, + "step": 2363 + }, + { + "epoch": 0.8914027149321267, + "grad_norm": 0.6508427992868157, + "learning_rate": 7.851595029083856e-05, + "loss": 0.4535, + "step": 2364 + }, + { + "epoch": 0.8917797888386124, + "grad_norm": 0.4949565965680191, + "learning_rate": 7.851310732971419e-05, + "loss": 0.461, + "step": 2365 + }, + { + "epoch": 0.8921568627450981, + "grad_norm": 0.49069134768214095, + "learning_rate": 7.851026169966565e-05, + "loss": 0.4694, + "step": 2366 + }, + { + "epoch": 0.8925339366515838, + "grad_norm": 0.5507552007799265, + "learning_rate": 7.850741340089014e-05, + "loss": 0.4644, + "step": 2367 + }, + { + "epoch": 0.8929110105580694, + "grad_norm": 0.5550107182919863, + "learning_rate": 7.850456243358504e-05, + "loss": 0.4697, + "step": 2368 + }, + { + "epoch": 0.8932880844645551, + "grad_norm": 0.48895123105839905, + "learning_rate": 7.850170879794789e-05, + "loss": 0.4528, + "step": 2369 + }, + { + "epoch": 0.8936651583710408, + "grad_norm": 0.32137585140100855, + "learning_rate": 7.849885249417648e-05, + "loss": 0.4615, + "step": 2370 + }, + { + "epoch": 0.8940422322775264, + "grad_norm": 0.30587267259475, + "learning_rate": 7.849599352246875e-05, + "loss": 0.4422, + "step": 2371 + }, + { + "epoch": 0.8944193061840121, + "grad_norm": 0.3279300956883759, + "learning_rate": 7.849313188302281e-05, + "loss": 0.4582, + "step": 2372 + }, + { + "epoch": 0.8947963800904978, + "grad_norm": 0.38145051119354606, + "learning_rate": 7.849026757603699e-05, + "loss": 0.457, + "step": 2373 + }, + { + "epoch": 0.8951734539969834, + "grad_norm": 0.4552825779455011, + "learning_rate": 7.848740060170976e-05, + "loss": 0.4543, + "step": 2374 + }, + { + "epoch": 0.8955505279034691, + "grad_norm": 0.473385862189418, + "learning_rate": 7.84845309602398e-05, + "loss": 0.455, + "step": 2375 + }, + { + "epoch": 0.8959276018099548, + "grad_norm": 0.39971757061152036, + "learning_rate": 7.848165865182597e-05, + "loss": 0.458, + "step": 2376 + }, + { + "epoch": 0.8963046757164405, + "grad_norm": 0.34093354108842927, + "learning_rate": 7.847878367666735e-05, + "loss": 0.457, + "step": 2377 + }, + { + "epoch": 0.8966817496229261, + "grad_norm": 0.3657173962777399, + "learning_rate": 7.847590603496313e-05, + "loss": 0.4392, + "step": 2378 + }, + { + "epoch": 0.8970588235294118, + "grad_norm": 0.44984865741081304, + "learning_rate": 7.847302572691277e-05, + "loss": 0.4575, + "step": 2379 + }, + { + "epoch": 0.8974358974358975, + "grad_norm": 0.5239453747485677, + "learning_rate": 7.847014275271584e-05, + "loss": 0.4603, + "step": 2380 + }, + { + "epoch": 0.8978129713423831, + "grad_norm": 0.5781843166379373, + "learning_rate": 7.846725711257215e-05, + "loss": 0.4605, + "step": 2381 + }, + { + "epoch": 0.8981900452488688, + "grad_norm": 0.5644779485994659, + "learning_rate": 7.846436880668166e-05, + "loss": 0.4643, + "step": 2382 + }, + { + "epoch": 0.8985671191553545, + "grad_norm": 0.48540878047885927, + "learning_rate": 7.846147783524452e-05, + "loss": 0.4602, + "step": 2383 + }, + { + "epoch": 0.8989441930618401, + "grad_norm": 0.46020975739367914, + "learning_rate": 7.845858419846109e-05, + "loss": 0.4544, + "step": 2384 + }, + { + "epoch": 0.8993212669683258, + "grad_norm": 0.6621602892959404, + "learning_rate": 7.845568789653189e-05, + "loss": 0.4655, + "step": 2385 + }, + { + "epoch": 0.8996983408748115, + "grad_norm": 0.8336183108583648, + "learning_rate": 7.845278892965762e-05, + "loss": 0.4608, + "step": 2386 + }, + { + "epoch": 0.9000754147812972, + "grad_norm": 0.8469093466831817, + "learning_rate": 7.84498872980392e-05, + "loss": 0.4497, + "step": 2387 + }, + { + "epoch": 0.9004524886877828, + "grad_norm": 0.6703526886418256, + "learning_rate": 7.844698300187768e-05, + "loss": 0.4554, + "step": 2388 + }, + { + "epoch": 0.9008295625942685, + "grad_norm": 0.5075474607424338, + "learning_rate": 7.844407604137434e-05, + "loss": 0.4563, + "step": 2389 + }, + { + "epoch": 0.9012066365007542, + "grad_norm": 0.46831264549984436, + "learning_rate": 7.844116641673063e-05, + "loss": 0.4507, + "step": 2390 + }, + { + "epoch": 0.9015837104072398, + "grad_norm": 0.43206505917992394, + "learning_rate": 7.84382541281482e-05, + "loss": 0.4648, + "step": 2391 + }, + { + "epoch": 0.9019607843137255, + "grad_norm": 0.41043059596719333, + "learning_rate": 7.843533917582884e-05, + "loss": 0.4481, + "step": 2392 + }, + { + "epoch": 0.9023378582202112, + "grad_norm": 0.40417888487319764, + "learning_rate": 7.843242155997457e-05, + "loss": 0.4418, + "step": 2393 + }, + { + "epoch": 0.9027149321266968, + "grad_norm": 0.45029960527565677, + "learning_rate": 7.842950128078757e-05, + "loss": 0.4611, + "step": 2394 + }, + { + "epoch": 0.9030920060331825, + "grad_norm": 0.4735865920446896, + "learning_rate": 7.842657833847024e-05, + "loss": 0.4504, + "step": 2395 + }, + { + "epoch": 0.9034690799396682, + "grad_norm": 0.40663634214469463, + "learning_rate": 7.842365273322509e-05, + "loss": 0.4521, + "step": 2396 + }, + { + "epoch": 0.9038461538461539, + "grad_norm": 0.37403557527852543, + "learning_rate": 7.84207244652549e-05, + "loss": 0.4473, + "step": 2397 + }, + { + "epoch": 0.9042232277526395, + "grad_norm": 0.38319741807744007, + "learning_rate": 7.841779353476258e-05, + "loss": 0.461, + "step": 2398 + }, + { + "epoch": 0.9046003016591252, + "grad_norm": 0.43407474622533987, + "learning_rate": 7.841485994195125e-05, + "loss": 0.4557, + "step": 2399 + }, + { + "epoch": 0.9049773755656109, + "grad_norm": 0.5652818287004823, + "learning_rate": 7.841192368702419e-05, + "loss": 0.4689, + "step": 2400 + }, + { + "epoch": 0.9053544494720965, + "grad_norm": 0.6099435181870148, + "learning_rate": 7.84089847701849e-05, + "loss": 0.4589, + "step": 2401 + }, + { + "epoch": 0.9057315233785822, + "grad_norm": 0.6407480485444202, + "learning_rate": 7.840604319163703e-05, + "loss": 0.4673, + "step": 2402 + }, + { + "epoch": 0.9061085972850679, + "grad_norm": 0.5473719851241616, + "learning_rate": 7.840309895158442e-05, + "loss": 0.4616, + "step": 2403 + }, + { + "epoch": 0.9064856711915535, + "grad_norm": 0.49065349200374014, + "learning_rate": 7.840015205023112e-05, + "loss": 0.4575, + "step": 2404 + }, + { + "epoch": 0.9068627450980392, + "grad_norm": 0.4340380917392645, + "learning_rate": 7.839720248778136e-05, + "loss": 0.453, + "step": 2405 + }, + { + "epoch": 0.9072398190045249, + "grad_norm": 0.4649848827255399, + "learning_rate": 7.839425026443952e-05, + "loss": 0.4602, + "step": 2406 + }, + { + "epoch": 0.9076168929110106, + "grad_norm": 0.4351481836467266, + "learning_rate": 7.83912953804102e-05, + "loss": 0.4677, + "step": 2407 + }, + { + "epoch": 0.9079939668174962, + "grad_norm": 0.34205575292547663, + "learning_rate": 7.838833783589815e-05, + "loss": 0.4554, + "step": 2408 + }, + { + "epoch": 0.9083710407239819, + "grad_norm": 0.4307042882048875, + "learning_rate": 7.838537763110834e-05, + "loss": 0.4567, + "step": 2409 + }, + { + "epoch": 0.9087481146304676, + "grad_norm": 0.46895557531748266, + "learning_rate": 7.838241476624591e-05, + "loss": 0.4636, + "step": 2410 + }, + { + "epoch": 0.9091251885369532, + "grad_norm": 0.4990367440878599, + "learning_rate": 7.837944924151618e-05, + "loss": 0.4492, + "step": 2411 + }, + { + "epoch": 0.9095022624434389, + "grad_norm": 0.49254163618438607, + "learning_rate": 7.837648105712467e-05, + "loss": 0.4483, + "step": 2412 + }, + { + "epoch": 0.9098793363499246, + "grad_norm": 0.6132720674470622, + "learning_rate": 7.837351021327704e-05, + "loss": 0.4691, + "step": 2413 + }, + { + "epoch": 0.9102564102564102, + "grad_norm": 0.7784609898888267, + "learning_rate": 7.837053671017921e-05, + "loss": 0.4446, + "step": 2414 + }, + { + "epoch": 0.9106334841628959, + "grad_norm": 0.7209085838368429, + "learning_rate": 7.836756054803723e-05, + "loss": 0.4662, + "step": 2415 + }, + { + "epoch": 0.9110105580693816, + "grad_norm": 0.6581710741214253, + "learning_rate": 7.836458172705731e-05, + "loss": 0.4592, + "step": 2416 + }, + { + "epoch": 0.9113876319758673, + "grad_norm": 0.6070678102640751, + "learning_rate": 7.836160024744592e-05, + "loss": 0.4533, + "step": 2417 + }, + { + "epoch": 0.9117647058823529, + "grad_norm": 0.536273887104888, + "learning_rate": 7.835861610940965e-05, + "loss": 0.4555, + "step": 2418 + }, + { + "epoch": 0.9121417797888386, + "grad_norm": 0.43438201311643626, + "learning_rate": 7.835562931315532e-05, + "loss": 0.4582, + "step": 2419 + }, + { + "epoch": 0.9125188536953243, + "grad_norm": 0.34090505175181157, + "learning_rate": 7.83526398588899e-05, + "loss": 0.4561, + "step": 2420 + }, + { + "epoch": 0.9128959276018099, + "grad_norm": 0.37286567294066286, + "learning_rate": 7.834964774682054e-05, + "loss": 0.4455, + "step": 2421 + }, + { + "epoch": 0.9132730015082956, + "grad_norm": 0.5061505702474683, + "learning_rate": 7.834665297715462e-05, + "loss": 0.4519, + "step": 2422 + }, + { + "epoch": 0.9136500754147813, + "grad_norm": 0.6346629187919427, + "learning_rate": 7.834365555009966e-05, + "loss": 0.4574, + "step": 2423 + }, + { + "epoch": 0.9140271493212669, + "grad_norm": 0.648776675063349, + "learning_rate": 7.834065546586339e-05, + "loss": 0.4566, + "step": 2424 + }, + { + "epoch": 0.9144042232277526, + "grad_norm": 0.6660447319385485, + "learning_rate": 7.833765272465368e-05, + "loss": 0.4753, + "step": 2425 + }, + { + "epoch": 0.9147812971342383, + "grad_norm": 0.7096198082783398, + "learning_rate": 7.833464732667866e-05, + "loss": 0.4528, + "step": 2426 + }, + { + "epoch": 0.915158371040724, + "grad_norm": 0.7704770471874413, + "learning_rate": 7.833163927214659e-05, + "loss": 0.4502, + "step": 2427 + }, + { + "epoch": 0.9155354449472096, + "grad_norm": 0.8109288716824973, + "learning_rate": 7.832862856126589e-05, + "loss": 0.4566, + "step": 2428 + }, + { + "epoch": 0.9159125188536953, + "grad_norm": 0.7463149133421982, + "learning_rate": 7.832561519424526e-05, + "loss": 0.4478, + "step": 2429 + }, + { + "epoch": 0.916289592760181, + "grad_norm": 0.7361515374468393, + "learning_rate": 7.832259917129347e-05, + "loss": 0.465, + "step": 2430 + }, + { + "epoch": 0.9166666666666666, + "grad_norm": 0.6589761376223425, + "learning_rate": 7.831958049261956e-05, + "loss": 0.4408, + "step": 2431 + }, + { + "epoch": 0.9170437405731523, + "grad_norm": 0.5650216370175573, + "learning_rate": 7.831655915843272e-05, + "loss": 0.4554, + "step": 2432 + }, + { + "epoch": 0.917420814479638, + "grad_norm": 0.5162775985240002, + "learning_rate": 7.831353516894229e-05, + "loss": 0.46, + "step": 2433 + }, + { + "epoch": 0.9177978883861236, + "grad_norm": 0.39510015936323384, + "learning_rate": 7.831050852435788e-05, + "loss": 0.4427, + "step": 2434 + }, + { + "epoch": 0.9181749622926093, + "grad_norm": 0.5218837816295826, + "learning_rate": 7.83074792248892e-05, + "loss": 0.4578, + "step": 2435 + }, + { + "epoch": 0.918552036199095, + "grad_norm": 0.6521470052571586, + "learning_rate": 7.83044472707462e-05, + "loss": 0.4759, + "step": 2436 + }, + { + "epoch": 0.9189291101055806, + "grad_norm": 0.7137185906653102, + "learning_rate": 7.830141266213897e-05, + "loss": 0.4628, + "step": 2437 + }, + { + "epoch": 0.9193061840120663, + "grad_norm": 0.5939959980507289, + "learning_rate": 7.829837539927781e-05, + "loss": 0.4475, + "step": 2438 + }, + { + "epoch": 0.919683257918552, + "grad_norm": 0.48727153439437426, + "learning_rate": 7.829533548237323e-05, + "loss": 0.4502, + "step": 2439 + }, + { + "epoch": 0.9200603318250377, + "grad_norm": 0.44022737775917387, + "learning_rate": 7.829229291163584e-05, + "loss": 0.4517, + "step": 2440 + }, + { + "epoch": 0.9204374057315233, + "grad_norm": 0.5218904024625849, + "learning_rate": 7.828924768727654e-05, + "loss": 0.4474, + "step": 2441 + }, + { + "epoch": 0.920814479638009, + "grad_norm": 0.6185125987632292, + "learning_rate": 7.828619980950632e-05, + "loss": 0.4725, + "step": 2442 + }, + { + "epoch": 0.9211915535444947, + "grad_norm": 0.6388856594304254, + "learning_rate": 7.828314927853643e-05, + "loss": 0.456, + "step": 2443 + }, + { + "epoch": 0.9215686274509803, + "grad_norm": 0.5865978349359486, + "learning_rate": 7.828009609457823e-05, + "loss": 0.4698, + "step": 2444 + }, + { + "epoch": 0.9219457013574661, + "grad_norm": 0.5508080832721717, + "learning_rate": 7.827704025784335e-05, + "loss": 0.4576, + "step": 2445 + }, + { + "epoch": 0.9223227752639518, + "grad_norm": 0.48098377349880395, + "learning_rate": 7.827398176854351e-05, + "loss": 0.4568, + "step": 2446 + }, + { + "epoch": 0.9226998491704375, + "grad_norm": 0.4166722464578625, + "learning_rate": 7.827092062689069e-05, + "loss": 0.4539, + "step": 2447 + }, + { + "epoch": 0.9230769230769231, + "grad_norm": 0.5562076328667078, + "learning_rate": 7.826785683309702e-05, + "loss": 0.4655, + "step": 2448 + }, + { + "epoch": 0.9234539969834088, + "grad_norm": 0.6889416321224704, + "learning_rate": 7.826479038737481e-05, + "loss": 0.4541, + "step": 2449 + }, + { + "epoch": 0.9238310708898945, + "grad_norm": 0.6526693291611412, + "learning_rate": 7.826172128993657e-05, + "loss": 0.475, + "step": 2450 + }, + { + "epoch": 0.9242081447963801, + "grad_norm": 0.5091741523576696, + "learning_rate": 7.825864954099497e-05, + "loss": 0.4585, + "step": 2451 + }, + { + "epoch": 0.9245852187028658, + "grad_norm": 0.443569702561046, + "learning_rate": 7.825557514076291e-05, + "loss": 0.4506, + "step": 2452 + }, + { + "epoch": 0.9249622926093515, + "grad_norm": 0.5670860253126732, + "learning_rate": 7.825249808945343e-05, + "loss": 0.4614, + "step": 2453 + }, + { + "epoch": 0.9253393665158371, + "grad_norm": 0.5831375281602698, + "learning_rate": 7.824941838727975e-05, + "loss": 0.4539, + "step": 2454 + }, + { + "epoch": 0.9257164404223228, + "grad_norm": 0.5241037169868463, + "learning_rate": 7.82463360344553e-05, + "loss": 0.4532, + "step": 2455 + }, + { + "epoch": 0.9260935143288085, + "grad_norm": 0.507337456591544, + "learning_rate": 7.824325103119369e-05, + "loss": 0.4623, + "step": 2456 + }, + { + "epoch": 0.9264705882352942, + "grad_norm": 0.4668332436869916, + "learning_rate": 7.824016337770872e-05, + "loss": 0.4504, + "step": 2457 + }, + { + "epoch": 0.9268476621417798, + "grad_norm": 0.40770028449014684, + "learning_rate": 7.823707307421433e-05, + "loss": 0.4679, + "step": 2458 + }, + { + "epoch": 0.9272247360482655, + "grad_norm": 0.4868377892041612, + "learning_rate": 7.82339801209247e-05, + "loss": 0.4529, + "step": 2459 + }, + { + "epoch": 0.9276018099547512, + "grad_norm": 0.6632361208347001, + "learning_rate": 7.823088451805415e-05, + "loss": 0.4681, + "step": 2460 + }, + { + "epoch": 0.9279788838612368, + "grad_norm": 0.7136086361571885, + "learning_rate": 7.822778626581723e-05, + "loss": 0.4466, + "step": 2461 + }, + { + "epoch": 0.9283559577677225, + "grad_norm": 0.5244727876324483, + "learning_rate": 7.822468536442861e-05, + "loss": 0.4514, + "step": 2462 + }, + { + "epoch": 0.9287330316742082, + "grad_norm": 0.34675648797123737, + "learning_rate": 7.822158181410322e-05, + "loss": 0.4478, + "step": 2463 + }, + { + "epoch": 0.9291101055806938, + "grad_norm": 0.4264673289212982, + "learning_rate": 7.82184756150561e-05, + "loss": 0.4531, + "step": 2464 + }, + { + "epoch": 0.9294871794871795, + "grad_norm": 0.5338203669360504, + "learning_rate": 7.821536676750253e-05, + "loss": 0.4486, + "step": 2465 + }, + { + "epoch": 0.9298642533936652, + "grad_norm": 0.4631151058930683, + "learning_rate": 7.821225527165793e-05, + "loss": 0.4715, + "step": 2466 + }, + { + "epoch": 0.9302413273001509, + "grad_norm": 0.3851973521429347, + "learning_rate": 7.820914112773794e-05, + "loss": 0.4569, + "step": 2467 + }, + { + "epoch": 0.9306184012066365, + "grad_norm": 0.41875467718682446, + "learning_rate": 7.820602433595837e-05, + "loss": 0.4645, + "step": 2468 + }, + { + "epoch": 0.9309954751131222, + "grad_norm": 0.4666053916661331, + "learning_rate": 7.820290489653518e-05, + "loss": 0.4519, + "step": 2469 + }, + { + "epoch": 0.9313725490196079, + "grad_norm": 0.4558721638560448, + "learning_rate": 7.819978280968457e-05, + "loss": 0.4803, + "step": 2470 + }, + { + "epoch": 0.9317496229260935, + "grad_norm": 0.4109857536917159, + "learning_rate": 7.81966580756229e-05, + "loss": 0.4409, + "step": 2471 + }, + { + "epoch": 0.9321266968325792, + "grad_norm": 0.3743645115273935, + "learning_rate": 7.819353069456672e-05, + "loss": 0.4593, + "step": 2472 + }, + { + "epoch": 0.9325037707390649, + "grad_norm": 0.47418325760087554, + "learning_rate": 7.819040066673274e-05, + "loss": 0.4582, + "step": 2473 + }, + { + "epoch": 0.9328808446455505, + "grad_norm": 0.6402359060264488, + "learning_rate": 7.818726799233784e-05, + "loss": 0.4566, + "step": 2474 + }, + { + "epoch": 0.9332579185520362, + "grad_norm": 0.7587697845495379, + "learning_rate": 7.818413267159915e-05, + "loss": 0.4463, + "step": 2475 + }, + { + "epoch": 0.9336349924585219, + "grad_norm": 0.7271056386604156, + "learning_rate": 7.818099470473395e-05, + "loss": 0.4475, + "step": 2476 + }, + { + "epoch": 0.9340120663650076, + "grad_norm": 0.7361560572116016, + "learning_rate": 7.817785409195967e-05, + "loss": 0.4717, + "step": 2477 + }, + { + "epoch": 0.9343891402714932, + "grad_norm": 0.7376752895909492, + "learning_rate": 7.817471083349398e-05, + "loss": 0.4624, + "step": 2478 + }, + { + "epoch": 0.9347662141779789, + "grad_norm": 0.6581519268429337, + "learning_rate": 7.817156492955468e-05, + "loss": 0.4586, + "step": 2479 + }, + { + "epoch": 0.9351432880844646, + "grad_norm": 0.4987979480601088, + "learning_rate": 7.816841638035978e-05, + "loss": 0.4623, + "step": 2480 + }, + { + "epoch": 0.9355203619909502, + "grad_norm": 0.4778047373369799, + "learning_rate": 7.81652651861275e-05, + "loss": 0.4587, + "step": 2481 + }, + { + "epoch": 0.9358974358974359, + "grad_norm": 0.5482245497265705, + "learning_rate": 7.816211134707619e-05, + "loss": 0.4573, + "step": 2482 + }, + { + "epoch": 0.9362745098039216, + "grad_norm": 0.503012744791461, + "learning_rate": 7.81589548634244e-05, + "loss": 0.4552, + "step": 2483 + }, + { + "epoch": 0.9366515837104072, + "grad_norm": 0.4550038216551682, + "learning_rate": 7.815579573539091e-05, + "loss": 0.4587, + "step": 2484 + }, + { + "epoch": 0.9370286576168929, + "grad_norm": 0.4560703774232684, + "learning_rate": 7.81526339631946e-05, + "loss": 0.4562, + "step": 2485 + }, + { + "epoch": 0.9374057315233786, + "grad_norm": 0.41580139001853944, + "learning_rate": 7.81494695470546e-05, + "loss": 0.4578, + "step": 2486 + }, + { + "epoch": 0.9377828054298643, + "grad_norm": 0.3737583907277238, + "learning_rate": 7.814630248719018e-05, + "loss": 0.4536, + "step": 2487 + }, + { + "epoch": 0.9381598793363499, + "grad_norm": 0.4708258499179867, + "learning_rate": 7.814313278382086e-05, + "loss": 0.4588, + "step": 2488 + }, + { + "epoch": 0.9385369532428356, + "grad_norm": 0.5748652567190145, + "learning_rate": 7.813996043716627e-05, + "loss": 0.4521, + "step": 2489 + }, + { + "epoch": 0.9389140271493213, + "grad_norm": 0.5733971177768343, + "learning_rate": 7.813678544744624e-05, + "loss": 0.4549, + "step": 2490 + }, + { + "epoch": 0.9392911010558069, + "grad_norm": 0.5226175521741881, + "learning_rate": 7.813360781488082e-05, + "loss": 0.4646, + "step": 2491 + }, + { + "epoch": 0.9396681749622926, + "grad_norm": 0.5382345811814075, + "learning_rate": 7.813042753969019e-05, + "loss": 0.4728, + "step": 2492 + }, + { + "epoch": 0.9400452488687783, + "grad_norm": 0.5486903182482851, + "learning_rate": 7.812724462209476e-05, + "loss": 0.4581, + "step": 2493 + }, + { + "epoch": 0.9404223227752639, + "grad_norm": 0.4891999915211238, + "learning_rate": 7.81240590623151e-05, + "loss": 0.4522, + "step": 2494 + }, + { + "epoch": 0.9407993966817496, + "grad_norm": 0.3786789339856116, + "learning_rate": 7.812087086057195e-05, + "loss": 0.4601, + "step": 2495 + }, + { + "epoch": 0.9411764705882353, + "grad_norm": 0.38178328742473217, + "learning_rate": 7.811768001708627e-05, + "loss": 0.4576, + "step": 2496 + }, + { + "epoch": 0.941553544494721, + "grad_norm": 0.4117986547207329, + "learning_rate": 7.811448653207917e-05, + "loss": 0.4479, + "step": 2497 + }, + { + "epoch": 0.9419306184012066, + "grad_norm": 0.3227071307534147, + "learning_rate": 7.811129040577196e-05, + "loss": 0.4485, + "step": 2498 + }, + { + "epoch": 0.9423076923076923, + "grad_norm": 0.3097059846165193, + "learning_rate": 7.810809163838615e-05, + "loss": 0.4492, + "step": 2499 + }, + { + "epoch": 0.942684766214178, + "grad_norm": 0.3961405479276893, + "learning_rate": 7.810489023014337e-05, + "loss": 0.4716, + "step": 2500 + }, + { + "epoch": 0.9430618401206636, + "grad_norm": 0.42735271900846594, + "learning_rate": 7.810168618126551e-05, + "loss": 0.4601, + "step": 2501 + }, + { + "epoch": 0.9434389140271493, + "grad_norm": 0.34091341269858133, + "learning_rate": 7.809847949197459e-05, + "loss": 0.4627, + "step": 2502 + }, + { + "epoch": 0.943815987933635, + "grad_norm": 0.32628673150155724, + "learning_rate": 7.809527016249283e-05, + "loss": 0.4517, + "step": 2503 + }, + { + "epoch": 0.9441930618401206, + "grad_norm": 0.41634780373782615, + "learning_rate": 7.809205819304263e-05, + "loss": 0.4607, + "step": 2504 + }, + { + "epoch": 0.9445701357466063, + "grad_norm": 0.4441731096338133, + "learning_rate": 7.808884358384659e-05, + "loss": 0.4627, + "step": 2505 + }, + { + "epoch": 0.944947209653092, + "grad_norm": 0.49640362623136847, + "learning_rate": 7.808562633512749e-05, + "loss": 0.4501, + "step": 2506 + }, + { + "epoch": 0.9453242835595776, + "grad_norm": 0.5450780898257838, + "learning_rate": 7.808240644710826e-05, + "loss": 0.4449, + "step": 2507 + }, + { + "epoch": 0.9457013574660633, + "grad_norm": 0.5589497958477028, + "learning_rate": 7.807918392001203e-05, + "loss": 0.4599, + "step": 2508 + }, + { + "epoch": 0.946078431372549, + "grad_norm": 0.5833297438413336, + "learning_rate": 7.807595875406213e-05, + "loss": 0.4585, + "step": 2509 + }, + { + "epoch": 0.9464555052790347, + "grad_norm": 0.6144693628842968, + "learning_rate": 7.807273094948206e-05, + "loss": 0.4601, + "step": 2510 + }, + { + "epoch": 0.9468325791855203, + "grad_norm": 0.6556003419993336, + "learning_rate": 7.806950050649553e-05, + "loss": 0.4588, + "step": 2511 + }, + { + "epoch": 0.947209653092006, + "grad_norm": 0.6399747542441208, + "learning_rate": 7.806626742532637e-05, + "loss": 0.4557, + "step": 2512 + }, + { + "epoch": 0.9475867269984917, + "grad_norm": 0.575167573280602, + "learning_rate": 7.806303170619864e-05, + "loss": 0.4727, + "step": 2513 + }, + { + "epoch": 0.9479638009049773, + "grad_norm": 0.4562346461850267, + "learning_rate": 7.805979334933658e-05, + "loss": 0.4563, + "step": 2514 + }, + { + "epoch": 0.948340874811463, + "grad_norm": 0.40111730677026436, + "learning_rate": 7.805655235496458e-05, + "loss": 0.4679, + "step": 2515 + }, + { + "epoch": 0.9487179487179487, + "grad_norm": 0.40256799583862285, + "learning_rate": 7.805330872330727e-05, + "loss": 0.4408, + "step": 2516 + }, + { + "epoch": 0.9490950226244343, + "grad_norm": 0.41490025941688125, + "learning_rate": 7.805006245458943e-05, + "loss": 0.4555, + "step": 2517 + }, + { + "epoch": 0.94947209653092, + "grad_norm": 0.4254351650023242, + "learning_rate": 7.8046813549036e-05, + "loss": 0.4667, + "step": 2518 + }, + { + "epoch": 0.9498491704374057, + "grad_norm": 0.41100202276241743, + "learning_rate": 7.804356200687215e-05, + "loss": 0.4409, + "step": 2519 + }, + { + "epoch": 0.9502262443438914, + "grad_norm": 0.44245062569218324, + "learning_rate": 7.80403078283232e-05, + "loss": 0.4706, + "step": 2520 + }, + { + "epoch": 0.950603318250377, + "grad_norm": 0.4401015835517484, + "learning_rate": 7.803705101361462e-05, + "loss": 0.4607, + "step": 2521 + }, + { + "epoch": 0.9509803921568627, + "grad_norm": 0.5191558463082769, + "learning_rate": 7.80337915629722e-05, + "loss": 0.446, + "step": 2522 + }, + { + "epoch": 0.9513574660633484, + "grad_norm": 0.6185293378359356, + "learning_rate": 7.803052947662172e-05, + "loss": 0.4552, + "step": 2523 + }, + { + "epoch": 0.951734539969834, + "grad_norm": 0.680556625357452, + "learning_rate": 7.802726475478929e-05, + "loss": 0.477, + "step": 2524 + }, + { + "epoch": 0.9521116138763197, + "grad_norm": 0.7325677510473969, + "learning_rate": 7.802399739770115e-05, + "loss": 0.4561, + "step": 2525 + }, + { + "epoch": 0.9524886877828054, + "grad_norm": 0.7350513212689214, + "learning_rate": 7.802072740558371e-05, + "loss": 0.4659, + "step": 2526 + }, + { + "epoch": 0.952865761689291, + "grad_norm": 0.7236699103931825, + "learning_rate": 7.801745477866359e-05, + "loss": 0.4684, + "step": 2527 + }, + { + "epoch": 0.9532428355957768, + "grad_norm": 0.6918951834019635, + "learning_rate": 7.801417951716757e-05, + "loss": 0.4732, + "step": 2528 + }, + { + "epoch": 0.9536199095022625, + "grad_norm": 0.5782097521475957, + "learning_rate": 7.801090162132264e-05, + "loss": 0.4625, + "step": 2529 + }, + { + "epoch": 0.9539969834087482, + "grad_norm": 0.41985406870418734, + "learning_rate": 7.800762109135593e-05, + "loss": 0.4455, + "step": 2530 + }, + { + "epoch": 0.9543740573152338, + "grad_norm": 0.3824227734003607, + "learning_rate": 7.80043379274948e-05, + "loss": 0.4534, + "step": 2531 + }, + { + "epoch": 0.9547511312217195, + "grad_norm": 0.4680691042552022, + "learning_rate": 7.800105212996675e-05, + "loss": 0.4611, + "step": 2532 + }, + { + "epoch": 0.9551282051282052, + "grad_norm": 0.5294647923863369, + "learning_rate": 7.799776369899951e-05, + "loss": 0.4546, + "step": 2533 + }, + { + "epoch": 0.9555052790346908, + "grad_norm": 0.4912154994019678, + "learning_rate": 7.799447263482094e-05, + "loss": 0.4522, + "step": 2534 + }, + { + "epoch": 0.9558823529411765, + "grad_norm": 0.452303785533309, + "learning_rate": 7.799117893765913e-05, + "loss": 0.464, + "step": 2535 + }, + { + "epoch": 0.9562594268476622, + "grad_norm": 0.45428760713811067, + "learning_rate": 7.79878826077423e-05, + "loss": 0.4588, + "step": 2536 + }, + { + "epoch": 0.9566365007541479, + "grad_norm": 0.45179635732801854, + "learning_rate": 7.79845836452989e-05, + "loss": 0.4597, + "step": 2537 + }, + { + "epoch": 0.9570135746606335, + "grad_norm": 0.41780174813144505, + "learning_rate": 7.798128205055756e-05, + "loss": 0.4554, + "step": 2538 + }, + { + "epoch": 0.9573906485671192, + "grad_norm": 0.40672179729149033, + "learning_rate": 7.797797782374705e-05, + "loss": 0.4516, + "step": 2539 + }, + { + "epoch": 0.9577677224736049, + "grad_norm": 0.4625604612441199, + "learning_rate": 7.797467096509637e-05, + "loss": 0.4565, + "step": 2540 + }, + { + "epoch": 0.9581447963800905, + "grad_norm": 0.577157552275001, + "learning_rate": 7.797136147483467e-05, + "loss": 0.4478, + "step": 2541 + }, + { + "epoch": 0.9585218702865762, + "grad_norm": 0.5745114033486434, + "learning_rate": 7.796804935319132e-05, + "loss": 0.4569, + "step": 2542 + }, + { + "epoch": 0.9588989441930619, + "grad_norm": 0.4570286270414864, + "learning_rate": 7.796473460039581e-05, + "loss": 0.4601, + "step": 2543 + }, + { + "epoch": 0.9592760180995475, + "grad_norm": 0.43259128092564436, + "learning_rate": 7.796141721667787e-05, + "loss": 0.4654, + "step": 2544 + }, + { + "epoch": 0.9596530920060332, + "grad_norm": 0.5031535210549228, + "learning_rate": 7.795809720226738e-05, + "loss": 0.4596, + "step": 2545 + }, + { + "epoch": 0.9600301659125189, + "grad_norm": 0.5603159524584264, + "learning_rate": 7.795477455739444e-05, + "loss": 0.4653, + "step": 2546 + }, + { + "epoch": 0.9604072398190046, + "grad_norm": 0.5523679028539962, + "learning_rate": 7.795144928228928e-05, + "loss": 0.4437, + "step": 2547 + }, + { + "epoch": 0.9607843137254902, + "grad_norm": 0.5479155012741221, + "learning_rate": 7.794812137718235e-05, + "loss": 0.4524, + "step": 2548 + }, + { + "epoch": 0.9611613876319759, + "grad_norm": 0.6712355125481678, + "learning_rate": 7.794479084230425e-05, + "loss": 0.4612, + "step": 2549 + }, + { + "epoch": 0.9615384615384616, + "grad_norm": 0.7459987994962591, + "learning_rate": 7.794145767788582e-05, + "loss": 0.4557, + "step": 2550 + }, + { + "epoch": 0.9619155354449472, + "grad_norm": 0.7494935760203404, + "learning_rate": 7.793812188415804e-05, + "loss": 0.4528, + "step": 2551 + }, + { + "epoch": 0.9622926093514329, + "grad_norm": 0.6994871280776321, + "learning_rate": 7.793478346135204e-05, + "loss": 0.4536, + "step": 2552 + }, + { + "epoch": 0.9626696832579186, + "grad_norm": 0.6279353577686287, + "learning_rate": 7.79314424096992e-05, + "loss": 0.4576, + "step": 2553 + }, + { + "epoch": 0.9630467571644042, + "grad_norm": 0.5846257279925337, + "learning_rate": 7.792809872943105e-05, + "loss": 0.4594, + "step": 2554 + }, + { + "epoch": 0.9634238310708899, + "grad_norm": 0.47551741707483813, + "learning_rate": 7.792475242077931e-05, + "loss": 0.4507, + "step": 2555 + }, + { + "epoch": 0.9638009049773756, + "grad_norm": 0.3595402342178627, + "learning_rate": 7.792140348397585e-05, + "loss": 0.4608, + "step": 2556 + }, + { + "epoch": 0.9641779788838613, + "grad_norm": 0.31085412394098155, + "learning_rate": 7.791805191925277e-05, + "loss": 0.456, + "step": 2557 + }, + { + "epoch": 0.9645550527903469, + "grad_norm": 0.4160548363123637, + "learning_rate": 7.791469772684233e-05, + "loss": 0.4637, + "step": 2558 + }, + { + "epoch": 0.9649321266968326, + "grad_norm": 0.4388561240611685, + "learning_rate": 7.791134090697697e-05, + "loss": 0.4634, + "step": 2559 + }, + { + "epoch": 0.9653092006033183, + "grad_norm": 0.38517809558335947, + "learning_rate": 7.790798145988931e-05, + "loss": 0.449, + "step": 2560 + }, + { + "epoch": 0.9656862745098039, + "grad_norm": 0.4421910254171229, + "learning_rate": 7.790461938581215e-05, + "loss": 0.4445, + "step": 2561 + }, + { + "epoch": 0.9660633484162896, + "grad_norm": 0.5915033247997871, + "learning_rate": 7.79012546849785e-05, + "loss": 0.4599, + "step": 2562 + }, + { + "epoch": 0.9664404223227753, + "grad_norm": 0.642648252386967, + "learning_rate": 7.789788735762151e-05, + "loss": 0.4586, + "step": 2563 + }, + { + "epoch": 0.9668174962292609, + "grad_norm": 0.6171729863319062, + "learning_rate": 7.789451740397453e-05, + "loss": 0.4535, + "step": 2564 + }, + { + "epoch": 0.9671945701357466, + "grad_norm": 0.5614439750552926, + "learning_rate": 7.789114482427112e-05, + "loss": 0.4437, + "step": 2565 + }, + { + "epoch": 0.9675716440422323, + "grad_norm": 0.5100388070174567, + "learning_rate": 7.788776961874499e-05, + "loss": 0.45, + "step": 2566 + }, + { + "epoch": 0.967948717948718, + "grad_norm": 0.46921630723201735, + "learning_rate": 7.788439178763002e-05, + "loss": 0.4622, + "step": 2567 + }, + { + "epoch": 0.9683257918552036, + "grad_norm": 0.5221540494680961, + "learning_rate": 7.788101133116032e-05, + "loss": 0.4468, + "step": 2568 + }, + { + "epoch": 0.9687028657616893, + "grad_norm": 0.5071552821743068, + "learning_rate": 7.787762824957011e-05, + "loss": 0.4578, + "step": 2569 + }, + { + "epoch": 0.969079939668175, + "grad_norm": 0.47637228115040803, + "learning_rate": 7.787424254309388e-05, + "loss": 0.4756, + "step": 2570 + }, + { + "epoch": 0.9694570135746606, + "grad_norm": 0.5410116104258856, + "learning_rate": 7.787085421196623e-05, + "loss": 0.4595, + "step": 2571 + }, + { + "epoch": 0.9698340874811463, + "grad_norm": 0.5391743628153276, + "learning_rate": 7.786746325642197e-05, + "loss": 0.4406, + "step": 2572 + }, + { + "epoch": 0.970211161387632, + "grad_norm": 0.4620164212566707, + "learning_rate": 7.78640696766961e-05, + "loss": 0.4594, + "step": 2573 + }, + { + "epoch": 0.9705882352941176, + "grad_norm": 0.36969933471536487, + "learning_rate": 7.786067347302379e-05, + "loss": 0.464, + "step": 2574 + }, + { + "epoch": 0.9709653092006033, + "grad_norm": 0.30885850777812296, + "learning_rate": 7.785727464564039e-05, + "loss": 0.4628, + "step": 2575 + }, + { + "epoch": 0.971342383107089, + "grad_norm": 0.2893618120947177, + "learning_rate": 7.785387319478143e-05, + "loss": 0.4643, + "step": 2576 + }, + { + "epoch": 0.9717194570135747, + "grad_norm": 0.3084391323582833, + "learning_rate": 7.785046912068263e-05, + "loss": 0.4473, + "step": 2577 + }, + { + "epoch": 0.9720965309200603, + "grad_norm": 0.34720406261095227, + "learning_rate": 7.78470624235799e-05, + "loss": 0.4611, + "step": 2578 + }, + { + "epoch": 0.972473604826546, + "grad_norm": 0.36191452422364845, + "learning_rate": 7.78436531037093e-05, + "loss": 0.4617, + "step": 2579 + }, + { + "epoch": 0.9728506787330317, + "grad_norm": 0.3356476592496949, + "learning_rate": 7.784024116130714e-05, + "loss": 0.4537, + "step": 2580 + }, + { + "epoch": 0.9732277526395173, + "grad_norm": 0.34895183949050745, + "learning_rate": 7.783682659660981e-05, + "loss": 0.443, + "step": 2581 + }, + { + "epoch": 0.973604826546003, + "grad_norm": 0.3581508678234303, + "learning_rate": 7.783340940985396e-05, + "loss": 0.4558, + "step": 2582 + }, + { + "epoch": 0.9739819004524887, + "grad_norm": 0.42225597377495905, + "learning_rate": 7.782998960127639e-05, + "loss": 0.4626, + "step": 2583 + }, + { + "epoch": 0.9743589743589743, + "grad_norm": 0.46581555327452695, + "learning_rate": 7.782656717111411e-05, + "loss": 0.4555, + "step": 2584 + }, + { + "epoch": 0.97473604826546, + "grad_norm": 0.44318974613583434, + "learning_rate": 7.782314211960426e-05, + "loss": 0.4531, + "step": 2585 + }, + { + "epoch": 0.9751131221719457, + "grad_norm": 0.47175416577346163, + "learning_rate": 7.781971444698423e-05, + "loss": 0.4606, + "step": 2586 + }, + { + "epoch": 0.9754901960784313, + "grad_norm": 0.5882152736788316, + "learning_rate": 7.781628415349152e-05, + "loss": 0.4652, + "step": 2587 + }, + { + "epoch": 0.975867269984917, + "grad_norm": 0.6555551392705135, + "learning_rate": 7.781285123936386e-05, + "loss": 0.4622, + "step": 2588 + }, + { + "epoch": 0.9762443438914027, + "grad_norm": 0.6636782620882953, + "learning_rate": 7.780941570483916e-05, + "loss": 0.4562, + "step": 2589 + }, + { + "epoch": 0.9766214177978884, + "grad_norm": 0.556412422317267, + "learning_rate": 7.780597755015549e-05, + "loss": 0.4422, + "step": 2590 + }, + { + "epoch": 0.976998491704374, + "grad_norm": 0.45278587000160525, + "learning_rate": 7.78025367755511e-05, + "loss": 0.4588, + "step": 2591 + }, + { + "epoch": 0.9773755656108597, + "grad_norm": 0.5364545528583924, + "learning_rate": 7.779909338126445e-05, + "loss": 0.4583, + "step": 2592 + }, + { + "epoch": 0.9777526395173454, + "grad_norm": 0.5576145789161077, + "learning_rate": 7.779564736753415e-05, + "loss": 0.4544, + "step": 2593 + }, + { + "epoch": 0.978129713423831, + "grad_norm": 0.5279584033419209, + "learning_rate": 7.779219873459901e-05, + "loss": 0.4671, + "step": 2594 + }, + { + "epoch": 0.9785067873303167, + "grad_norm": 0.49790186365841255, + "learning_rate": 7.778874748269802e-05, + "loss": 0.4555, + "step": 2595 + }, + { + "epoch": 0.9788838612368024, + "grad_norm": 0.5106878955761082, + "learning_rate": 7.778529361207035e-05, + "loss": 0.4546, + "step": 2596 + }, + { + "epoch": 0.979260935143288, + "grad_norm": 0.534827779217347, + "learning_rate": 7.778183712295535e-05, + "loss": 0.4535, + "step": 2597 + }, + { + "epoch": 0.9796380090497737, + "grad_norm": 0.5645104046382136, + "learning_rate": 7.777837801559256e-05, + "loss": 0.456, + "step": 2598 + }, + { + "epoch": 0.9800150829562594, + "grad_norm": 0.5920560056206776, + "learning_rate": 7.777491629022167e-05, + "loss": 0.4495, + "step": 2599 + }, + { + "epoch": 0.9803921568627451, + "grad_norm": 0.5444873883914445, + "learning_rate": 7.777145194708258e-05, + "loss": 0.4509, + "step": 2600 + }, + { + "epoch": 0.9807692307692307, + "grad_norm": 0.5000454479269602, + "learning_rate": 7.776798498641539e-05, + "loss": 0.4573, + "step": 2601 + }, + { + "epoch": 0.9811463046757164, + "grad_norm": 0.4286454383041004, + "learning_rate": 7.776451540846033e-05, + "loss": 0.4616, + "step": 2602 + }, + { + "epoch": 0.9815233785822021, + "grad_norm": 0.48121061274170457, + "learning_rate": 7.776104321345785e-05, + "loss": 0.456, + "step": 2603 + }, + { + "epoch": 0.9819004524886877, + "grad_norm": 0.5212937335546614, + "learning_rate": 7.775756840164858e-05, + "loss": 0.4656, + "step": 2604 + }, + { + "epoch": 0.9822775263951734, + "grad_norm": 0.5741862630013432, + "learning_rate": 7.775409097327331e-05, + "loss": 0.4558, + "step": 2605 + }, + { + "epoch": 0.9826546003016591, + "grad_norm": 0.634590985960841, + "learning_rate": 7.775061092857301e-05, + "loss": 0.4688, + "step": 2606 + }, + { + "epoch": 0.9830316742081447, + "grad_norm": 0.5423530562536157, + "learning_rate": 7.774712826778886e-05, + "loss": 0.4484, + "step": 2607 + }, + { + "epoch": 0.9834087481146304, + "grad_norm": 0.4989170018136076, + "learning_rate": 7.77436429911622e-05, + "loss": 0.4605, + "step": 2608 + }, + { + "epoch": 0.9837858220211161, + "grad_norm": 0.4975887857144609, + "learning_rate": 7.774015509893457e-05, + "loss": 0.4547, + "step": 2609 + }, + { + "epoch": 0.9841628959276018, + "grad_norm": 0.43396172584331905, + "learning_rate": 7.773666459134767e-05, + "loss": 0.453, + "step": 2610 + }, + { + "epoch": 0.9845399698340875, + "grad_norm": 0.37642401056472297, + "learning_rate": 7.773317146864337e-05, + "loss": 0.4615, + "step": 2611 + }, + { + "epoch": 0.9849170437405732, + "grad_norm": 0.3506751395358835, + "learning_rate": 7.772967573106377e-05, + "loss": 0.4378, + "step": 2612 + }, + { + "epoch": 0.9852941176470589, + "grad_norm": 0.36023850886867015, + "learning_rate": 7.77261773788511e-05, + "loss": 0.4649, + "step": 2613 + }, + { + "epoch": 0.9856711915535445, + "grad_norm": 0.488629752355657, + "learning_rate": 7.772267641224781e-05, + "loss": 0.4663, + "step": 2614 + }, + { + "epoch": 0.9860482654600302, + "grad_norm": 0.6245929969965341, + "learning_rate": 7.77191728314965e-05, + "loss": 0.4552, + "step": 2615 + }, + { + "epoch": 0.9864253393665159, + "grad_norm": 0.7536315250364505, + "learning_rate": 7.771566663683998e-05, + "loss": 0.4467, + "step": 2616 + }, + { + "epoch": 0.9868024132730016, + "grad_norm": 0.776394534123457, + "learning_rate": 7.77121578285212e-05, + "loss": 0.4569, + "step": 2617 + }, + { + "epoch": 0.9871794871794872, + "grad_norm": 0.6596197173192979, + "learning_rate": 7.770864640678333e-05, + "loss": 0.4641, + "step": 2618 + }, + { + "epoch": 0.9875565610859729, + "grad_norm": 0.5553834638238073, + "learning_rate": 7.770513237186972e-05, + "loss": 0.436, + "step": 2619 + }, + { + "epoch": 0.9879336349924586, + "grad_norm": 0.5161432887173532, + "learning_rate": 7.770161572402387e-05, + "loss": 0.4562, + "step": 2620 + }, + { + "epoch": 0.9883107088989442, + "grad_norm": 0.497021958316261, + "learning_rate": 7.76980964634895e-05, + "loss": 0.4603, + "step": 2621 + }, + { + "epoch": 0.9886877828054299, + "grad_norm": 0.4515456583815306, + "learning_rate": 7.76945745905105e-05, + "loss": 0.4498, + "step": 2622 + }, + { + "epoch": 0.9890648567119156, + "grad_norm": 0.35004210314227596, + "learning_rate": 7.76910501053309e-05, + "loss": 0.4472, + "step": 2623 + }, + { + "epoch": 0.9894419306184012, + "grad_norm": 0.36431023764751236, + "learning_rate": 7.768752300819496e-05, + "loss": 0.4577, + "step": 2624 + }, + { + "epoch": 0.9898190045248869, + "grad_norm": 0.3831862814737958, + "learning_rate": 7.76839932993471e-05, + "loss": 0.4489, + "step": 2625 + }, + { + "epoch": 0.9901960784313726, + "grad_norm": 0.37687670075928753, + "learning_rate": 7.768046097903194e-05, + "loss": 0.452, + "step": 2626 + }, + { + "epoch": 0.9905731523378583, + "grad_norm": 0.3402447775147092, + "learning_rate": 7.767692604749424e-05, + "loss": 0.4626, + "step": 2627 + }, + { + "epoch": 0.9909502262443439, + "grad_norm": 0.28802760803765554, + "learning_rate": 7.767338850497901e-05, + "loss": 0.4505, + "step": 2628 + }, + { + "epoch": 0.9913273001508296, + "grad_norm": 0.39195849161955365, + "learning_rate": 7.766984835173136e-05, + "loss": 0.4508, + "step": 2629 + }, + { + "epoch": 0.9917043740573153, + "grad_norm": 0.493569068142465, + "learning_rate": 7.766630558799662e-05, + "loss": 0.4556, + "step": 2630 + }, + { + "epoch": 0.9920814479638009, + "grad_norm": 0.47776381614956437, + "learning_rate": 7.766276021402032e-05, + "loss": 0.4543, + "step": 2631 + }, + { + "epoch": 0.9924585218702866, + "grad_norm": 0.4389594133536926, + "learning_rate": 7.765921223004816e-05, + "loss": 0.4568, + "step": 2632 + }, + { + "epoch": 0.9928355957767723, + "grad_norm": 0.3992737196447484, + "learning_rate": 7.765566163632598e-05, + "loss": 0.4555, + "step": 2633 + }, + { + "epoch": 0.9932126696832579, + "grad_norm": 0.3758773513805594, + "learning_rate": 7.765210843309986e-05, + "loss": 0.4544, + "step": 2634 + }, + { + "epoch": 0.9935897435897436, + "grad_norm": 0.4032537127436583, + "learning_rate": 7.7648552620616e-05, + "loss": 0.4513, + "step": 2635 + }, + { + "epoch": 0.9939668174962293, + "grad_norm": 0.43061366840781196, + "learning_rate": 7.764499419912087e-05, + "loss": 0.4442, + "step": 2636 + }, + { + "epoch": 0.994343891402715, + "grad_norm": 0.4318133794528706, + "learning_rate": 7.764143316886101e-05, + "loss": 0.46, + "step": 2637 + }, + { + "epoch": 0.9947209653092006, + "grad_norm": 0.46493680024990147, + "learning_rate": 7.763786953008322e-05, + "loss": 0.4535, + "step": 2638 + }, + { + "epoch": 0.9950980392156863, + "grad_norm": 0.4590915710844103, + "learning_rate": 7.763430328303446e-05, + "loss": 0.4469, + "step": 2639 + }, + { + "epoch": 0.995475113122172, + "grad_norm": 0.44795479341461986, + "learning_rate": 7.763073442796187e-05, + "loss": 0.4601, + "step": 2640 + }, + { + "epoch": 0.9958521870286576, + "grad_norm": 0.4527171919941686, + "learning_rate": 7.762716296511275e-05, + "loss": 0.4597, + "step": 2641 + }, + { + "epoch": 0.9962292609351433, + "grad_norm": 0.41410667998774997, + "learning_rate": 7.762358889473464e-05, + "loss": 0.4424, + "step": 2642 + }, + { + "epoch": 0.996606334841629, + "grad_norm": 0.42586598410652565, + "learning_rate": 7.762001221707516e-05, + "loss": 0.455, + "step": 2643 + }, + { + "epoch": 0.9969834087481146, + "grad_norm": 0.5494765530708864, + "learning_rate": 7.761643293238223e-05, + "loss": 0.449, + "step": 2644 + }, + { + "epoch": 0.9973604826546003, + "grad_norm": 0.5529558231978502, + "learning_rate": 7.761285104090383e-05, + "loss": 0.4417, + "step": 2645 + }, + { + "epoch": 0.997737556561086, + "grad_norm": 0.46263109076955067, + "learning_rate": 7.760926654288825e-05, + "loss": 0.4404, + "step": 2646 + }, + { + "epoch": 0.9981146304675717, + "grad_norm": 0.42306641792236305, + "learning_rate": 7.760567943858383e-05, + "loss": 0.454, + "step": 2647 + }, + { + "epoch": 0.9984917043740573, + "grad_norm": 0.41320674293379017, + "learning_rate": 7.760208972823921e-05, + "loss": 0.4463, + "step": 2648 + }, + { + "epoch": 0.998868778280543, + "grad_norm": 0.46829889624382903, + "learning_rate": 7.75984974121031e-05, + "loss": 0.4472, + "step": 2649 + }, + { + "epoch": 0.9992458521870287, + "grad_norm": 0.4605281520969338, + "learning_rate": 7.759490249042448e-05, + "loss": 0.4618, + "step": 2650 + }, + { + "epoch": 0.9996229260935143, + "grad_norm": 0.47481479620669015, + "learning_rate": 7.759130496345245e-05, + "loss": 0.4643, + "step": 2651 + }, + { + "epoch": 1.0, + "grad_norm": 0.45187655209416167, + "learning_rate": 7.758770483143634e-05, + "loss": 0.4491, + "step": 2652 + }, + { + "epoch": 1.0003770739064857, + "grad_norm": 0.41826636176949905, + "learning_rate": 7.758410209462564e-05, + "loss": 0.4504, + "step": 2653 + }, + { + "epoch": 1.0007541478129713, + "grad_norm": 0.45108851537455175, + "learning_rate": 7.758049675326997e-05, + "loss": 0.4448, + "step": 2654 + }, + { + "epoch": 1.001131221719457, + "grad_norm": 0.4510230590885474, + "learning_rate": 7.757688880761922e-05, + "loss": 0.4442, + "step": 2655 + }, + { + "epoch": 1.0015082956259427, + "grad_norm": 0.3965757048531137, + "learning_rate": 7.757327825792341e-05, + "loss": 0.4469, + "step": 2656 + }, + { + "epoch": 1.0018853695324283, + "grad_norm": 0.3551346064292115, + "learning_rate": 7.756966510443277e-05, + "loss": 0.4395, + "step": 2657 + }, + { + "epoch": 1.002262443438914, + "grad_norm": 0.4225580482772868, + "learning_rate": 7.756604934739762e-05, + "loss": 0.4482, + "step": 2658 + }, + { + "epoch": 1.0026395173453997, + "grad_norm": 0.46870955666192293, + "learning_rate": 7.75624309870686e-05, + "loss": 0.4464, + "step": 2659 + }, + { + "epoch": 1.0030165912518854, + "grad_norm": 0.49764412362319443, + "learning_rate": 7.755881002369643e-05, + "loss": 0.4466, + "step": 2660 + }, + { + "epoch": 1.003393665158371, + "grad_norm": 0.5733514083976906, + "learning_rate": 7.755518645753203e-05, + "loss": 0.4394, + "step": 2661 + }, + { + "epoch": 1.0037707390648567, + "grad_norm": 0.5898628243857371, + "learning_rate": 7.755156028882651e-05, + "loss": 0.4448, + "step": 2662 + }, + { + "epoch": 1.0041478129713424, + "grad_norm": 0.6752885323396658, + "learning_rate": 7.754793151783118e-05, + "loss": 0.449, + "step": 2663 + }, + { + "epoch": 1.004524886877828, + "grad_norm": 0.8098786901263236, + "learning_rate": 7.754430014479751e-05, + "loss": 0.4387, + "step": 2664 + }, + { + "epoch": 1.0049019607843137, + "grad_norm": 0.7477382623050911, + "learning_rate": 7.754066616997713e-05, + "loss": 0.4425, + "step": 2665 + }, + { + "epoch": 1.0052790346907994, + "grad_norm": 0.61794948295399, + "learning_rate": 7.753702959362189e-05, + "loss": 0.4525, + "step": 2666 + }, + { + "epoch": 1.005656108597285, + "grad_norm": 0.5991769231953279, + "learning_rate": 7.75333904159838e-05, + "loss": 0.4567, + "step": 2667 + }, + { + "epoch": 1.0060331825037707, + "grad_norm": 0.5899533865195682, + "learning_rate": 7.752974863731503e-05, + "loss": 0.4415, + "step": 2668 + }, + { + "epoch": 1.0064102564102564, + "grad_norm": 0.43571590822796286, + "learning_rate": 7.752610425786798e-05, + "loss": 0.4352, + "step": 2669 + }, + { + "epoch": 1.006787330316742, + "grad_norm": 0.363165501534787, + "learning_rate": 7.752245727789519e-05, + "loss": 0.4321, + "step": 2670 + }, + { + "epoch": 1.0071644042232277, + "grad_norm": 0.4375343414120086, + "learning_rate": 7.751880769764938e-05, + "loss": 0.4496, + "step": 2671 + }, + { + "epoch": 1.0075414781297134, + "grad_norm": 0.4039149105776388, + "learning_rate": 7.751515551738348e-05, + "loss": 0.4394, + "step": 2672 + }, + { + "epoch": 1.007918552036199, + "grad_norm": 0.3006017819444832, + "learning_rate": 7.751150073735059e-05, + "loss": 0.4433, + "step": 2673 + }, + { + "epoch": 1.0082956259426847, + "grad_norm": 0.3364795045070861, + "learning_rate": 7.750784335780395e-05, + "loss": 0.4349, + "step": 2674 + }, + { + "epoch": 1.0086726998491704, + "grad_norm": 0.4194037935514924, + "learning_rate": 7.750418337899707e-05, + "loss": 0.4339, + "step": 2675 + }, + { + "epoch": 1.009049773755656, + "grad_norm": 0.390212788213026, + "learning_rate": 7.750052080118352e-05, + "loss": 0.453, + "step": 2676 + }, + { + "epoch": 1.0094268476621417, + "grad_norm": 0.3463091170946586, + "learning_rate": 7.749685562461714e-05, + "loss": 0.4346, + "step": 2677 + }, + { + "epoch": 1.0098039215686274, + "grad_norm": 0.3999244333164006, + "learning_rate": 7.749318784955195e-05, + "loss": 0.4261, + "step": 2678 + }, + { + "epoch": 1.010180995475113, + "grad_norm": 0.38681378132647826, + "learning_rate": 7.748951747624208e-05, + "loss": 0.4435, + "step": 2679 + }, + { + "epoch": 1.0105580693815988, + "grad_norm": 0.4907360840889778, + "learning_rate": 7.74858445049419e-05, + "loss": 0.4368, + "step": 2680 + }, + { + "epoch": 1.0109351432880844, + "grad_norm": 0.6492602655558145, + "learning_rate": 7.748216893590596e-05, + "loss": 0.4561, + "step": 2681 + }, + { + "epoch": 1.01131221719457, + "grad_norm": 0.7834700403813045, + "learning_rate": 7.747849076938893e-05, + "loss": 0.4439, + "step": 2682 + }, + { + "epoch": 1.0116892911010558, + "grad_norm": 0.8811777511970565, + "learning_rate": 7.747481000564577e-05, + "loss": 0.4374, + "step": 2683 + }, + { + "epoch": 1.0120663650075414, + "grad_norm": 0.8568353138930986, + "learning_rate": 7.747112664493148e-05, + "loss": 0.4514, + "step": 2684 + }, + { + "epoch": 1.012443438914027, + "grad_norm": 0.7176686320942236, + "learning_rate": 7.746744068750138e-05, + "loss": 0.4517, + "step": 2685 + }, + { + "epoch": 1.0128205128205128, + "grad_norm": 0.5139035897077856, + "learning_rate": 7.746375213361086e-05, + "loss": 0.447, + "step": 2686 + }, + { + "epoch": 1.0131975867269984, + "grad_norm": 0.44139213254123383, + "learning_rate": 7.746006098351557e-05, + "loss": 0.4307, + "step": 2687 + }, + { + "epoch": 1.0135746606334841, + "grad_norm": 0.45266146043893557, + "learning_rate": 7.745636723747124e-05, + "loss": 0.4505, + "step": 2688 + }, + { + "epoch": 1.0139517345399698, + "grad_norm": 0.5391380040114788, + "learning_rate": 7.745267089573391e-05, + "loss": 0.4426, + "step": 2689 + }, + { + "epoch": 1.0143288084464555, + "grad_norm": 0.5357437293093, + "learning_rate": 7.74489719585597e-05, + "loss": 0.4363, + "step": 2690 + }, + { + "epoch": 1.0147058823529411, + "grad_norm": 0.511535712254849, + "learning_rate": 7.744527042620496e-05, + "loss": 0.4473, + "step": 2691 + }, + { + "epoch": 1.0150829562594268, + "grad_norm": 0.48316249989617344, + "learning_rate": 7.744156629892619e-05, + "loss": 0.4344, + "step": 2692 + }, + { + "epoch": 1.0154600301659125, + "grad_norm": 0.4440104154988776, + "learning_rate": 7.743785957698008e-05, + "loss": 0.4369, + "step": 2693 + }, + { + "epoch": 1.0158371040723981, + "grad_norm": 0.36533131226814713, + "learning_rate": 7.743415026062351e-05, + "loss": 0.442, + "step": 2694 + }, + { + "epoch": 1.0162141779788838, + "grad_norm": 0.36610872744683315, + "learning_rate": 7.743043835011354e-05, + "loss": 0.4351, + "step": 2695 + }, + { + "epoch": 1.0165912518853695, + "grad_norm": 0.3163287764725775, + "learning_rate": 7.742672384570739e-05, + "loss": 0.4432, + "step": 2696 + }, + { + "epoch": 1.0169683257918551, + "grad_norm": 0.28690694143306555, + "learning_rate": 7.742300674766247e-05, + "loss": 0.4395, + "step": 2697 + }, + { + "epoch": 1.0173453996983408, + "grad_norm": 0.31084931280469114, + "learning_rate": 7.741928705623638e-05, + "loss": 0.4339, + "step": 2698 + }, + { + "epoch": 1.0177224736048265, + "grad_norm": 0.36245128677659877, + "learning_rate": 7.741556477168689e-05, + "loss": 0.4357, + "step": 2699 + }, + { + "epoch": 1.0180995475113122, + "grad_norm": 0.4332867901741453, + "learning_rate": 7.741183989427194e-05, + "loss": 0.4167, + "step": 2700 + }, + { + "epoch": 1.0184766214177978, + "grad_norm": 0.4204367235853343, + "learning_rate": 7.740811242424967e-05, + "loss": 0.437, + "step": 2701 + }, + { + "epoch": 1.0188536953242835, + "grad_norm": 0.3893390877164388, + "learning_rate": 7.740438236187839e-05, + "loss": 0.4535, + "step": 2702 + }, + { + "epoch": 1.0192307692307692, + "grad_norm": 0.41209514321598834, + "learning_rate": 7.740064970741661e-05, + "loss": 0.4384, + "step": 2703 + }, + { + "epoch": 1.0196078431372548, + "grad_norm": 0.4595124749449053, + "learning_rate": 7.739691446112294e-05, + "loss": 0.4388, + "step": 2704 + }, + { + "epoch": 1.0199849170437405, + "grad_norm": 0.4759333461365069, + "learning_rate": 7.739317662325631e-05, + "loss": 0.4399, + "step": 2705 + }, + { + "epoch": 1.0203619909502262, + "grad_norm": 0.4909527819211695, + "learning_rate": 7.738943619407568e-05, + "loss": 0.4385, + "step": 2706 + }, + { + "epoch": 1.0207390648567118, + "grad_norm": 0.5594897946192379, + "learning_rate": 7.738569317384031e-05, + "loss": 0.4432, + "step": 2707 + }, + { + "epoch": 1.0211161387631975, + "grad_norm": 0.5641316295684021, + "learning_rate": 7.738194756280954e-05, + "loss": 0.4312, + "step": 2708 + }, + { + "epoch": 1.0214932126696832, + "grad_norm": 0.5169093389326491, + "learning_rate": 7.737819936124297e-05, + "loss": 0.4384, + "step": 2709 + }, + { + "epoch": 1.0218702865761689, + "grad_norm": 0.45756254459237516, + "learning_rate": 7.737444856940034e-05, + "loss": 0.4469, + "step": 2710 + }, + { + "epoch": 1.0222473604826545, + "grad_norm": 0.4739011891694032, + "learning_rate": 7.737069518754158e-05, + "loss": 0.4355, + "step": 2711 + }, + { + "epoch": 1.0226244343891402, + "grad_norm": 0.4489630674702445, + "learning_rate": 7.73669392159268e-05, + "loss": 0.4367, + "step": 2712 + }, + { + "epoch": 1.0230015082956259, + "grad_norm": 0.3530613456238394, + "learning_rate": 7.736318065481625e-05, + "loss": 0.4477, + "step": 2713 + }, + { + "epoch": 1.0233785822021115, + "grad_norm": 0.2877411903093052, + "learning_rate": 7.735941950447044e-05, + "loss": 0.4577, + "step": 2714 + }, + { + "epoch": 1.0237556561085972, + "grad_norm": 0.3535323348460711, + "learning_rate": 7.735565576515e-05, + "loss": 0.4481, + "step": 2715 + }, + { + "epoch": 1.0241327300150829, + "grad_norm": 0.378840751543373, + "learning_rate": 7.735188943711574e-05, + "loss": 0.4495, + "step": 2716 + }, + { + "epoch": 1.0245098039215685, + "grad_norm": 0.40020023170851815, + "learning_rate": 7.734812052062867e-05, + "loss": 0.4494, + "step": 2717 + }, + { + "epoch": 1.0248868778280542, + "grad_norm": 0.5029785174150662, + "learning_rate": 7.734434901594998e-05, + "loss": 0.4429, + "step": 2718 + }, + { + "epoch": 1.0252639517345399, + "grad_norm": 0.5005948621927988, + "learning_rate": 7.734057492334103e-05, + "loss": 0.4483, + "step": 2719 + }, + { + "epoch": 1.0256410256410255, + "grad_norm": 0.5480445763628007, + "learning_rate": 7.733679824306336e-05, + "loss": 0.4439, + "step": 2720 + }, + { + "epoch": 1.0260180995475112, + "grad_norm": 0.6380201630747682, + "learning_rate": 7.733301897537868e-05, + "loss": 0.4449, + "step": 2721 + }, + { + "epoch": 1.026395173453997, + "grad_norm": 0.670412467570155, + "learning_rate": 7.732923712054891e-05, + "loss": 0.4353, + "step": 2722 + }, + { + "epoch": 1.0267722473604826, + "grad_norm": 0.5950028857962635, + "learning_rate": 7.732545267883612e-05, + "loss": 0.4427, + "step": 2723 + }, + { + "epoch": 1.0271493212669682, + "grad_norm": 0.4058950709419382, + "learning_rate": 7.732166565050256e-05, + "loss": 0.4465, + "step": 2724 + }, + { + "epoch": 1.027526395173454, + "grad_norm": 0.31571710613550097, + "learning_rate": 7.731787603581067e-05, + "loss": 0.4271, + "step": 2725 + }, + { + "epoch": 1.0279034690799396, + "grad_norm": 0.37232617080558295, + "learning_rate": 7.731408383502308e-05, + "loss": 0.4359, + "step": 2726 + }, + { + "epoch": 1.0282805429864252, + "grad_norm": 0.45051225313837123, + "learning_rate": 7.731028904840257e-05, + "loss": 0.443, + "step": 2727 + }, + { + "epoch": 1.028657616892911, + "grad_norm": 0.5892307543624249, + "learning_rate": 7.730649167621214e-05, + "loss": 0.447, + "step": 2728 + }, + { + "epoch": 1.0290346907993966, + "grad_norm": 0.6480386963502381, + "learning_rate": 7.730269171871491e-05, + "loss": 0.4533, + "step": 2729 + }, + { + "epoch": 1.0294117647058822, + "grad_norm": 0.6442572530209886, + "learning_rate": 7.729888917617424e-05, + "loss": 0.4446, + "step": 2730 + }, + { + "epoch": 1.029788838612368, + "grad_norm": 0.5768687306290857, + "learning_rate": 7.729508404885364e-05, + "loss": 0.4635, + "step": 2731 + }, + { + "epoch": 1.0301659125188536, + "grad_norm": 0.5853693815571502, + "learning_rate": 7.729127633701679e-05, + "loss": 0.4506, + "step": 2732 + }, + { + "epoch": 1.0305429864253393, + "grad_norm": 0.6014021445592119, + "learning_rate": 7.728746604092755e-05, + "loss": 0.4292, + "step": 2733 + }, + { + "epoch": 1.030920060331825, + "grad_norm": 0.6170160595253512, + "learning_rate": 7.728365316085e-05, + "loss": 0.4438, + "step": 2734 + }, + { + "epoch": 1.0312971342383106, + "grad_norm": 0.5260913384467304, + "learning_rate": 7.727983769704837e-05, + "loss": 0.4312, + "step": 2735 + }, + { + "epoch": 1.0316742081447963, + "grad_norm": 0.3826842626360242, + "learning_rate": 7.727601964978704e-05, + "loss": 0.4408, + "step": 2736 + }, + { + "epoch": 1.032051282051282, + "grad_norm": 0.39138675316549665, + "learning_rate": 7.727219901933063e-05, + "loss": 0.4422, + "step": 2737 + }, + { + "epoch": 1.0324283559577678, + "grad_norm": 0.5339129857080382, + "learning_rate": 7.726837580594385e-05, + "loss": 0.4317, + "step": 2738 + }, + { + "epoch": 1.0328054298642535, + "grad_norm": 0.5226965170143294, + "learning_rate": 7.726455000989171e-05, + "loss": 0.4541, + "step": 2739 + }, + { + "epoch": 1.0331825037707392, + "grad_norm": 0.39034931120059235, + "learning_rate": 7.726072163143931e-05, + "loss": 0.4501, + "step": 2740 + }, + { + "epoch": 1.0335595776772248, + "grad_norm": 0.4051607517218453, + "learning_rate": 7.725689067085195e-05, + "loss": 0.4385, + "step": 2741 + }, + { + "epoch": 1.0339366515837105, + "grad_norm": 0.6009338455211946, + "learning_rate": 7.725305712839511e-05, + "loss": 0.4545, + "step": 2742 + }, + { + "epoch": 1.0343137254901962, + "grad_norm": 0.6579713224071955, + "learning_rate": 7.724922100433445e-05, + "loss": 0.4471, + "step": 2743 + }, + { + "epoch": 1.0346907993966818, + "grad_norm": 0.7227534704518747, + "learning_rate": 7.724538229893583e-05, + "loss": 0.4525, + "step": 2744 + }, + { + "epoch": 1.0350678733031675, + "grad_norm": 0.6250041538628708, + "learning_rate": 7.724154101246523e-05, + "loss": 0.4425, + "step": 2745 + }, + { + "epoch": 1.0354449472096532, + "grad_norm": 0.4081365038031126, + "learning_rate": 7.723769714518889e-05, + "loss": 0.4424, + "step": 2746 + }, + { + "epoch": 1.0358220211161389, + "grad_norm": 0.39608541793002666, + "learning_rate": 7.723385069737315e-05, + "loss": 0.4508, + "step": 2747 + }, + { + "epoch": 1.0361990950226245, + "grad_norm": 0.5495111371910583, + "learning_rate": 7.723000166928461e-05, + "loss": 0.4413, + "step": 2748 + }, + { + "epoch": 1.0365761689291102, + "grad_norm": 0.6673590249473939, + "learning_rate": 7.722615006118996e-05, + "loss": 0.4315, + "step": 2749 + }, + { + "epoch": 1.0369532428355959, + "grad_norm": 0.5500806076611555, + "learning_rate": 7.722229587335613e-05, + "loss": 0.4383, + "step": 2750 + }, + { + "epoch": 1.0373303167420815, + "grad_norm": 0.45259663230780406, + "learning_rate": 7.72184391060502e-05, + "loss": 0.4413, + "step": 2751 + }, + { + "epoch": 1.0377073906485672, + "grad_norm": 0.42923986415443643, + "learning_rate": 7.721457975953949e-05, + "loss": 0.4281, + "step": 2752 + }, + { + "epoch": 1.0380844645550529, + "grad_norm": 0.41584574931890483, + "learning_rate": 7.721071783409138e-05, + "loss": 0.4476, + "step": 2753 + }, + { + "epoch": 1.0384615384615385, + "grad_norm": 0.3410735144490017, + "learning_rate": 7.720685332997354e-05, + "loss": 0.4472, + "step": 2754 + }, + { + "epoch": 1.0388386123680242, + "grad_norm": 0.30148027338257277, + "learning_rate": 7.720298624745376e-05, + "loss": 0.435, + "step": 2755 + }, + { + "epoch": 1.0392156862745099, + "grad_norm": 0.3638116325276319, + "learning_rate": 7.719911658680005e-05, + "loss": 0.4462, + "step": 2756 + }, + { + "epoch": 1.0395927601809956, + "grad_norm": 0.3566600077491078, + "learning_rate": 7.719524434828054e-05, + "loss": 0.4492, + "step": 2757 + }, + { + "epoch": 1.0399698340874812, + "grad_norm": 0.32414191844850143, + "learning_rate": 7.719136953216359e-05, + "loss": 0.4483, + "step": 2758 + }, + { + "epoch": 1.040346907993967, + "grad_norm": 0.36981025280099317, + "learning_rate": 7.718749213871771e-05, + "loss": 0.4273, + "step": 2759 + }, + { + "epoch": 1.0407239819004526, + "grad_norm": 0.4491812613205279, + "learning_rate": 7.718361216821163e-05, + "loss": 0.4473, + "step": 2760 + }, + { + "epoch": 1.0411010558069382, + "grad_norm": 0.434474286498073, + "learning_rate": 7.71797296209142e-05, + "loss": 0.4581, + "step": 2761 + }, + { + "epoch": 1.041478129713424, + "grad_norm": 0.42319768604576585, + "learning_rate": 7.717584449709449e-05, + "loss": 0.4446, + "step": 2762 + }, + { + "epoch": 1.0418552036199096, + "grad_norm": 0.4090248346112003, + "learning_rate": 7.717195679702172e-05, + "loss": 0.4519, + "step": 2763 + }, + { + "epoch": 1.0422322775263952, + "grad_norm": 0.35577129997584916, + "learning_rate": 7.716806652096534e-05, + "loss": 0.4412, + "step": 2764 + }, + { + "epoch": 1.042609351432881, + "grad_norm": 0.3263649418659946, + "learning_rate": 7.716417366919489e-05, + "loss": 0.4466, + "step": 2765 + }, + { + "epoch": 1.0429864253393666, + "grad_norm": 0.4142378428741153, + "learning_rate": 7.716027824198017e-05, + "loss": 0.4482, + "step": 2766 + }, + { + "epoch": 1.0433634992458523, + "grad_norm": 0.5162837152228126, + "learning_rate": 7.715638023959115e-05, + "loss": 0.4416, + "step": 2767 + }, + { + "epoch": 1.043740573152338, + "grad_norm": 0.5737592136253435, + "learning_rate": 7.715247966229794e-05, + "loss": 0.4368, + "step": 2768 + }, + { + "epoch": 1.0441176470588236, + "grad_norm": 0.5592396260971948, + "learning_rate": 7.714857651037081e-05, + "loss": 0.4415, + "step": 2769 + }, + { + "epoch": 1.0444947209653093, + "grad_norm": 0.463655304352472, + "learning_rate": 7.71446707840803e-05, + "loss": 0.4272, + "step": 2770 + }, + { + "epoch": 1.044871794871795, + "grad_norm": 0.42665238250489856, + "learning_rate": 7.714076248369707e-05, + "loss": 0.4472, + "step": 2771 + }, + { + "epoch": 1.0452488687782806, + "grad_norm": 0.4612077059929547, + "learning_rate": 7.713685160949191e-05, + "loss": 0.4397, + "step": 2772 + }, + { + "epoch": 1.0456259426847663, + "grad_norm": 0.4568130212482222, + "learning_rate": 7.713293816173588e-05, + "loss": 0.447, + "step": 2773 + }, + { + "epoch": 1.046003016591252, + "grad_norm": 0.4832402647292717, + "learning_rate": 7.712902214070019e-05, + "loss": 0.4489, + "step": 2774 + }, + { + "epoch": 1.0463800904977376, + "grad_norm": 0.5086604251023653, + "learning_rate": 7.71251035466562e-05, + "loss": 0.45, + "step": 2775 + }, + { + "epoch": 1.0467571644042233, + "grad_norm": 0.6078086102744419, + "learning_rate": 7.712118237987545e-05, + "loss": 0.4365, + "step": 2776 + }, + { + "epoch": 1.047134238310709, + "grad_norm": 0.6162430122649188, + "learning_rate": 7.711725864062969e-05, + "loss": 0.4289, + "step": 2777 + }, + { + "epoch": 1.0475113122171946, + "grad_norm": 0.5295049982154493, + "learning_rate": 7.711333232919084e-05, + "loss": 0.4437, + "step": 2778 + }, + { + "epoch": 1.0478883861236803, + "grad_norm": 0.4329258924945145, + "learning_rate": 7.710940344583097e-05, + "loss": 0.4456, + "step": 2779 + }, + { + "epoch": 1.048265460030166, + "grad_norm": 0.39053082615713, + "learning_rate": 7.710547199082234e-05, + "loss": 0.4487, + "step": 2780 + }, + { + "epoch": 1.0486425339366516, + "grad_norm": 0.37873152298929524, + "learning_rate": 7.710153796443743e-05, + "loss": 0.4504, + "step": 2781 + }, + { + "epoch": 1.0490196078431373, + "grad_norm": 0.42648728693607774, + "learning_rate": 7.709760136694886e-05, + "loss": 0.4371, + "step": 2782 + }, + { + "epoch": 1.049396681749623, + "grad_norm": 0.47434820714692544, + "learning_rate": 7.70936621986294e-05, + "loss": 0.4326, + "step": 2783 + }, + { + "epoch": 1.0497737556561086, + "grad_norm": 0.510150800511455, + "learning_rate": 7.708972045975207e-05, + "loss": 0.4462, + "step": 2784 + }, + { + "epoch": 1.0501508295625943, + "grad_norm": 0.5469556848338892, + "learning_rate": 7.708577615059e-05, + "loss": 0.4362, + "step": 2785 + }, + { + "epoch": 1.05052790346908, + "grad_norm": 0.6053875138178022, + "learning_rate": 7.708182927141654e-05, + "loss": 0.4509, + "step": 2786 + }, + { + "epoch": 1.0509049773755657, + "grad_norm": 0.6880172509088254, + "learning_rate": 7.70778798225052e-05, + "loss": 0.4461, + "step": 2787 + }, + { + "epoch": 1.0512820512820513, + "grad_norm": 0.7357799008032001, + "learning_rate": 7.707392780412968e-05, + "loss": 0.432, + "step": 2788 + }, + { + "epoch": 1.051659125188537, + "grad_norm": 0.6227493733347262, + "learning_rate": 7.706997321656385e-05, + "loss": 0.4514, + "step": 2789 + }, + { + "epoch": 1.0520361990950227, + "grad_norm": 0.4499922222789563, + "learning_rate": 7.706601606008176e-05, + "loss": 0.4468, + "step": 2790 + }, + { + "epoch": 1.0524132730015083, + "grad_norm": 0.3906769354479975, + "learning_rate": 7.706205633495762e-05, + "loss": 0.4388, + "step": 2791 + }, + { + "epoch": 1.052790346907994, + "grad_norm": 0.36346769588130884, + "learning_rate": 7.705809404146585e-05, + "loss": 0.4387, + "step": 2792 + }, + { + "epoch": 1.0531674208144797, + "grad_norm": 0.355567907416694, + "learning_rate": 7.705412917988105e-05, + "loss": 0.4565, + "step": 2793 + }, + { + "epoch": 1.0535444947209653, + "grad_norm": 0.36111956228781295, + "learning_rate": 7.705016175047795e-05, + "loss": 0.4365, + "step": 2794 + }, + { + "epoch": 1.053921568627451, + "grad_norm": 0.3942389485358135, + "learning_rate": 7.704619175353152e-05, + "loss": 0.4391, + "step": 2795 + }, + { + "epoch": 1.0542986425339367, + "grad_norm": 0.4077270620920054, + "learning_rate": 7.704221918931684e-05, + "loss": 0.4358, + "step": 2796 + }, + { + "epoch": 1.0546757164404224, + "grad_norm": 0.38358679283221275, + "learning_rate": 7.703824405810924e-05, + "loss": 0.4511, + "step": 2797 + }, + { + "epoch": 1.055052790346908, + "grad_norm": 0.318579665882504, + "learning_rate": 7.703426636018419e-05, + "loss": 0.43, + "step": 2798 + }, + { + "epoch": 1.0554298642533937, + "grad_norm": 0.3439574383649951, + "learning_rate": 7.703028609581731e-05, + "loss": 0.435, + "step": 2799 + }, + { + "epoch": 1.0558069381598794, + "grad_norm": 0.38717354022513484, + "learning_rate": 7.702630326528446e-05, + "loss": 0.433, + "step": 2800 + }, + { + "epoch": 1.056184012066365, + "grad_norm": 0.42074569499438785, + "learning_rate": 7.702231786886164e-05, + "loss": 0.4399, + "step": 2801 + }, + { + "epoch": 1.0565610859728507, + "grad_norm": 0.42807021842518966, + "learning_rate": 7.701832990682502e-05, + "loss": 0.4416, + "step": 2802 + }, + { + "epoch": 1.0569381598793364, + "grad_norm": 0.43754657992341617, + "learning_rate": 7.701433937945098e-05, + "loss": 0.4443, + "step": 2803 + }, + { + "epoch": 1.057315233785822, + "grad_norm": 0.42216463478072663, + "learning_rate": 7.701034628701605e-05, + "loss": 0.4336, + "step": 2804 + }, + { + "epoch": 1.0576923076923077, + "grad_norm": 0.4031066297495303, + "learning_rate": 7.700635062979695e-05, + "loss": 0.4349, + "step": 2805 + }, + { + "epoch": 1.0580693815987934, + "grad_norm": 0.4359120339157377, + "learning_rate": 7.700235240807057e-05, + "loss": 0.4551, + "step": 2806 + }, + { + "epoch": 1.058446455505279, + "grad_norm": 0.47542168855880396, + "learning_rate": 7.6998351622114e-05, + "loss": 0.4493, + "step": 2807 + }, + { + "epoch": 1.0588235294117647, + "grad_norm": 0.5761201879507254, + "learning_rate": 7.699434827220448e-05, + "loss": 0.4459, + "step": 2808 + }, + { + "epoch": 1.0592006033182504, + "grad_norm": 0.616056014232236, + "learning_rate": 7.699034235861943e-05, + "loss": 0.4362, + "step": 2809 + }, + { + "epoch": 1.059577677224736, + "grad_norm": 0.6325474065411265, + "learning_rate": 7.698633388163647e-05, + "loss": 0.4395, + "step": 2810 + }, + { + "epoch": 1.0599547511312217, + "grad_norm": 0.5876656385770604, + "learning_rate": 7.698232284153338e-05, + "loss": 0.4474, + "step": 2811 + }, + { + "epoch": 1.0603318250377074, + "grad_norm": 0.5542061057704593, + "learning_rate": 7.697830923858813e-05, + "loss": 0.4366, + "step": 2812 + }, + { + "epoch": 1.060708898944193, + "grad_norm": 0.5128903039607103, + "learning_rate": 7.697429307307882e-05, + "loss": 0.4404, + "step": 2813 + }, + { + "epoch": 1.0610859728506787, + "grad_norm": 0.4220875853137095, + "learning_rate": 7.697027434528383e-05, + "loss": 0.4276, + "step": 2814 + }, + { + "epoch": 1.0614630467571644, + "grad_norm": 0.3846737503151704, + "learning_rate": 7.69662530554816e-05, + "loss": 0.4377, + "step": 2815 + }, + { + "epoch": 1.06184012066365, + "grad_norm": 0.33530781665014026, + "learning_rate": 7.696222920395082e-05, + "loss": 0.4374, + "step": 2816 + }, + { + "epoch": 1.0622171945701357, + "grad_norm": 0.3486122128685782, + "learning_rate": 7.695820279097035e-05, + "loss": 0.4373, + "step": 2817 + }, + { + "epoch": 1.0625942684766214, + "grad_norm": 0.38016443162956925, + "learning_rate": 7.695417381681921e-05, + "loss": 0.4469, + "step": 2818 + }, + { + "epoch": 1.062971342383107, + "grad_norm": 0.3891591182130235, + "learning_rate": 7.69501422817766e-05, + "loss": 0.4267, + "step": 2819 + }, + { + "epoch": 1.0633484162895928, + "grad_norm": 0.4400369270837802, + "learning_rate": 7.69461081861219e-05, + "loss": 0.4484, + "step": 2820 + }, + { + "epoch": 1.0637254901960784, + "grad_norm": 0.4735358750823295, + "learning_rate": 7.694207153013468e-05, + "loss": 0.4632, + "step": 2821 + }, + { + "epoch": 1.064102564102564, + "grad_norm": 0.48783194951289016, + "learning_rate": 7.693803231409468e-05, + "loss": 0.4478, + "step": 2822 + }, + { + "epoch": 1.0644796380090498, + "grad_norm": 0.48525896987206085, + "learning_rate": 7.693399053828179e-05, + "loss": 0.4398, + "step": 2823 + }, + { + "epoch": 1.0648567119155354, + "grad_norm": 0.5050341537157015, + "learning_rate": 7.692994620297611e-05, + "loss": 0.4316, + "step": 2824 + }, + { + "epoch": 1.065233785822021, + "grad_norm": 0.6071229487827794, + "learning_rate": 7.692589930845792e-05, + "loss": 0.4462, + "step": 2825 + }, + { + "epoch": 1.0656108597285068, + "grad_norm": 0.6971832112513713, + "learning_rate": 7.692184985500767e-05, + "loss": 0.4473, + "step": 2826 + }, + { + "epoch": 1.0659879336349924, + "grad_norm": 0.7621506431850956, + "learning_rate": 7.691779784290598e-05, + "loss": 0.4366, + "step": 2827 + }, + { + "epoch": 1.0663650075414781, + "grad_norm": 0.758222404084359, + "learning_rate": 7.691374327243363e-05, + "loss": 0.4281, + "step": 2828 + }, + { + "epoch": 1.0667420814479638, + "grad_norm": 0.5644986602526576, + "learning_rate": 7.690968614387162e-05, + "loss": 0.4333, + "step": 2829 + }, + { + "epoch": 1.0671191553544495, + "grad_norm": 0.3541013282009001, + "learning_rate": 7.69056264575011e-05, + "loss": 0.4458, + "step": 2830 + }, + { + "epoch": 1.0674962292609351, + "grad_norm": 0.3294391937639758, + "learning_rate": 7.69015642136034e-05, + "loss": 0.4333, + "step": 2831 + }, + { + "epoch": 1.0678733031674208, + "grad_norm": 0.3583187850901964, + "learning_rate": 7.689749941246004e-05, + "loss": 0.4407, + "step": 2832 + }, + { + "epoch": 1.0682503770739065, + "grad_norm": 0.4555838742372709, + "learning_rate": 7.689343205435269e-05, + "loss": 0.4403, + "step": 2833 + }, + { + "epoch": 1.0686274509803921, + "grad_norm": 0.45669842733972504, + "learning_rate": 7.688936213956321e-05, + "loss": 0.4282, + "step": 2834 + }, + { + "epoch": 1.0690045248868778, + "grad_norm": 0.3989186003364033, + "learning_rate": 7.688528966837369e-05, + "loss": 0.4362, + "step": 2835 + }, + { + "epoch": 1.0693815987933635, + "grad_norm": 0.4069994037530831, + "learning_rate": 7.688121464106627e-05, + "loss": 0.4324, + "step": 2836 + }, + { + "epoch": 1.0697586726998491, + "grad_norm": 0.356027588336116, + "learning_rate": 7.687713705792342e-05, + "loss": 0.4561, + "step": 2837 + }, + { + "epoch": 1.0701357466063348, + "grad_norm": 0.31249405006910685, + "learning_rate": 7.687305691922767e-05, + "loss": 0.4439, + "step": 2838 + }, + { + "epoch": 1.0705128205128205, + "grad_norm": 0.3253491315723791, + "learning_rate": 7.686897422526178e-05, + "loss": 0.4404, + "step": 2839 + }, + { + "epoch": 1.0708898944193062, + "grad_norm": 0.36796184904823337, + "learning_rate": 7.686488897630869e-05, + "loss": 0.4409, + "step": 2840 + }, + { + "epoch": 1.0712669683257918, + "grad_norm": 0.34595682191881505, + "learning_rate": 7.686080117265147e-05, + "loss": 0.4447, + "step": 2841 + }, + { + "epoch": 1.0716440422322775, + "grad_norm": 0.3266034167604996, + "learning_rate": 7.685671081457346e-05, + "loss": 0.4527, + "step": 2842 + }, + { + "epoch": 1.0720211161387632, + "grad_norm": 0.34537910233709657, + "learning_rate": 7.685261790235805e-05, + "loss": 0.4433, + "step": 2843 + }, + { + "epoch": 1.0723981900452488, + "grad_norm": 0.38783821657015793, + "learning_rate": 7.684852243628893e-05, + "loss": 0.4318, + "step": 2844 + }, + { + "epoch": 1.0727752639517345, + "grad_norm": 0.48905132697113196, + "learning_rate": 7.684442441664988e-05, + "loss": 0.4336, + "step": 2845 + }, + { + "epoch": 1.0731523378582202, + "grad_norm": 0.4965784917066625, + "learning_rate": 7.68403238437249e-05, + "loss": 0.4448, + "step": 2846 + }, + { + "epoch": 1.0735294117647058, + "grad_norm": 0.5366711854712453, + "learning_rate": 7.683622071779816e-05, + "loss": 0.4393, + "step": 2847 + }, + { + "epoch": 1.0739064856711915, + "grad_norm": 0.6376942470631008, + "learning_rate": 7.683211503915398e-05, + "loss": 0.4499, + "step": 2848 + }, + { + "epoch": 1.0742835595776772, + "grad_norm": 0.6828017997548351, + "learning_rate": 7.682800680807692e-05, + "loss": 0.4373, + "step": 2849 + }, + { + "epoch": 1.0746606334841629, + "grad_norm": 0.6968801952220715, + "learning_rate": 7.682389602485164e-05, + "loss": 0.4494, + "step": 2850 + }, + { + "epoch": 1.0750377073906485, + "grad_norm": 0.688372641155137, + "learning_rate": 7.681978268976303e-05, + "loss": 0.4415, + "step": 2851 + }, + { + "epoch": 1.0754147812971342, + "grad_norm": 0.5846333020877416, + "learning_rate": 7.681566680309614e-05, + "loss": 0.4365, + "step": 2852 + }, + { + "epoch": 1.0757918552036199, + "grad_norm": 0.4467804719198124, + "learning_rate": 7.68115483651362e-05, + "loss": 0.4388, + "step": 2853 + }, + { + "epoch": 1.0761689291101055, + "grad_norm": 0.3764037982225008, + "learning_rate": 7.68074273761686e-05, + "loss": 0.4429, + "step": 2854 + }, + { + "epoch": 1.0765460030165912, + "grad_norm": 0.3155125459586304, + "learning_rate": 7.680330383647894e-05, + "loss": 0.4411, + "step": 2855 + }, + { + "epoch": 1.0769230769230769, + "grad_norm": 0.3552539924606071, + "learning_rate": 7.679917774635298e-05, + "loss": 0.4374, + "step": 2856 + }, + { + "epoch": 1.0773001508295625, + "grad_norm": 0.43609339080536275, + "learning_rate": 7.679504910607663e-05, + "loss": 0.4488, + "step": 2857 + }, + { + "epoch": 1.0776772247360482, + "grad_norm": 0.46386117282546613, + "learning_rate": 7.679091791593601e-05, + "loss": 0.4557, + "step": 2858 + }, + { + "epoch": 1.0780542986425339, + "grad_norm": 0.4577561799713789, + "learning_rate": 7.678678417621742e-05, + "loss": 0.4474, + "step": 2859 + }, + { + "epoch": 1.0784313725490196, + "grad_norm": 0.4031390016206854, + "learning_rate": 7.67826478872073e-05, + "loss": 0.4513, + "step": 2860 + }, + { + "epoch": 1.0788084464555052, + "grad_norm": 0.37554063740070676, + "learning_rate": 7.677850904919233e-05, + "loss": 0.4299, + "step": 2861 + }, + { + "epoch": 1.079185520361991, + "grad_norm": 0.4032338776508124, + "learning_rate": 7.67743676624593e-05, + "loss": 0.4302, + "step": 2862 + }, + { + "epoch": 1.0795625942684766, + "grad_norm": 0.43874637398926786, + "learning_rate": 7.677022372729519e-05, + "loss": 0.4347, + "step": 2863 + }, + { + "epoch": 1.0799396681749622, + "grad_norm": 0.4805157552580412, + "learning_rate": 7.676607724398722e-05, + "loss": 0.4621, + "step": 2864 + }, + { + "epoch": 1.080316742081448, + "grad_norm": 0.5446994311820056, + "learning_rate": 7.676192821282269e-05, + "loss": 0.4546, + "step": 2865 + }, + { + "epoch": 1.0806938159879336, + "grad_norm": 0.5773000207892947, + "learning_rate": 7.675777663408915e-05, + "loss": 0.4415, + "step": 2866 + }, + { + "epoch": 1.0810708898944192, + "grad_norm": 0.7078193743726987, + "learning_rate": 7.675362250807429e-05, + "loss": 0.4361, + "step": 2867 + }, + { + "epoch": 1.081447963800905, + "grad_norm": 0.879438699212154, + "learning_rate": 7.674946583506599e-05, + "loss": 0.4474, + "step": 2868 + }, + { + "epoch": 1.0818250377073906, + "grad_norm": 0.9890899548662427, + "learning_rate": 7.67453066153523e-05, + "loss": 0.4337, + "step": 2869 + }, + { + "epoch": 1.0822021116138762, + "grad_norm": 0.9099183639896694, + "learning_rate": 7.674114484922147e-05, + "loss": 0.4423, + "step": 2870 + }, + { + "epoch": 1.082579185520362, + "grad_norm": 0.6168906357016131, + "learning_rate": 7.673698053696188e-05, + "loss": 0.4438, + "step": 2871 + }, + { + "epoch": 1.0829562594268476, + "grad_norm": 0.3185610394713129, + "learning_rate": 7.673281367886212e-05, + "loss": 0.4379, + "step": 2872 + }, + { + "epoch": 1.0833333333333333, + "grad_norm": 0.345069183608444, + "learning_rate": 7.672864427521097e-05, + "loss": 0.448, + "step": 2873 + }, + { + "epoch": 1.083710407239819, + "grad_norm": 0.5331411837180404, + "learning_rate": 7.672447232629734e-05, + "loss": 0.4407, + "step": 2874 + }, + { + "epoch": 1.0840874811463046, + "grad_norm": 0.5371285535586221, + "learning_rate": 7.672029783241036e-05, + "loss": 0.4361, + "step": 2875 + }, + { + "epoch": 1.0844645550527903, + "grad_norm": 0.3435349922733753, + "learning_rate": 7.671612079383929e-05, + "loss": 0.4382, + "step": 2876 + }, + { + "epoch": 1.084841628959276, + "grad_norm": 0.2663981324764964, + "learning_rate": 7.671194121087363e-05, + "loss": 0.4492, + "step": 2877 + }, + { + "epoch": 1.0852187028657616, + "grad_norm": 0.409851279102452, + "learning_rate": 7.670775908380301e-05, + "loss": 0.434, + "step": 2878 + }, + { + "epoch": 1.0855957767722473, + "grad_norm": 0.5071893330201671, + "learning_rate": 7.670357441291724e-05, + "loss": 0.4374, + "step": 2879 + }, + { + "epoch": 1.085972850678733, + "grad_norm": 0.5198442961666164, + "learning_rate": 7.669938719850634e-05, + "loss": 0.4469, + "step": 2880 + }, + { + "epoch": 1.0863499245852186, + "grad_norm": 0.4865308498908082, + "learning_rate": 7.669519744086042e-05, + "loss": 0.4406, + "step": 2881 + }, + { + "epoch": 1.0867269984917043, + "grad_norm": 0.4358907846855424, + "learning_rate": 7.669100514026988e-05, + "loss": 0.4456, + "step": 2882 + }, + { + "epoch": 1.08710407239819, + "grad_norm": 0.3670012818539026, + "learning_rate": 7.668681029702526e-05, + "loss": 0.4356, + "step": 2883 + }, + { + "epoch": 1.0874811463046756, + "grad_norm": 0.34408917243854675, + "learning_rate": 7.66826129114172e-05, + "loss": 0.4494, + "step": 2884 + }, + { + "epoch": 1.0878582202111613, + "grad_norm": 0.2533564710850084, + "learning_rate": 7.66784129837366e-05, + "loss": 0.4544, + "step": 2885 + }, + { + "epoch": 1.088235294117647, + "grad_norm": 0.23844724638012535, + "learning_rate": 7.667421051427453e-05, + "loss": 0.4331, + "step": 2886 + }, + { + "epoch": 1.0886123680241326, + "grad_norm": 0.2869930538373074, + "learning_rate": 7.667000550332217e-05, + "loss": 0.4393, + "step": 2887 + }, + { + "epoch": 1.0889894419306183, + "grad_norm": 0.2775716157154722, + "learning_rate": 7.666579795117098e-05, + "loss": 0.4345, + "step": 2888 + }, + { + "epoch": 1.089366515837104, + "grad_norm": 0.3726045318514661, + "learning_rate": 7.666158785811252e-05, + "loss": 0.4428, + "step": 2889 + }, + { + "epoch": 1.0897435897435896, + "grad_norm": 0.40154459265362297, + "learning_rate": 7.665737522443852e-05, + "loss": 0.4508, + "step": 2890 + }, + { + "epoch": 1.0901206636500753, + "grad_norm": 0.4518279991480145, + "learning_rate": 7.665316005044096e-05, + "loss": 0.4434, + "step": 2891 + }, + { + "epoch": 1.090497737556561, + "grad_norm": 0.48588492196112565, + "learning_rate": 7.66489423364119e-05, + "loss": 0.4393, + "step": 2892 + }, + { + "epoch": 1.0908748114630467, + "grad_norm": 0.43769473363378025, + "learning_rate": 7.664472208264364e-05, + "loss": 0.444, + "step": 2893 + }, + { + "epoch": 1.0912518853695323, + "grad_norm": 0.4168563173294864, + "learning_rate": 7.664049928942865e-05, + "loss": 0.4338, + "step": 2894 + }, + { + "epoch": 1.091628959276018, + "grad_norm": 0.3781528078222642, + "learning_rate": 7.663627395705956e-05, + "loss": 0.4395, + "step": 2895 + }, + { + "epoch": 1.0920060331825039, + "grad_norm": 0.3749782776044312, + "learning_rate": 7.663204608582918e-05, + "loss": 0.432, + "step": 2896 + }, + { + "epoch": 1.0923831070889896, + "grad_norm": 0.38489438810687665, + "learning_rate": 7.66278156760305e-05, + "loss": 0.4272, + "step": 2897 + }, + { + "epoch": 1.0927601809954752, + "grad_norm": 0.36614598355889516, + "learning_rate": 7.662358272795668e-05, + "loss": 0.4368, + "step": 2898 + }, + { + "epoch": 1.093137254901961, + "grad_norm": 0.3516053083619188, + "learning_rate": 7.661934724190107e-05, + "loss": 0.4285, + "step": 2899 + }, + { + "epoch": 1.0935143288084466, + "grad_norm": 0.34402675217702744, + "learning_rate": 7.661510921815718e-05, + "loss": 0.4347, + "step": 2900 + }, + { + "epoch": 1.0938914027149322, + "grad_norm": 0.38954387932056733, + "learning_rate": 7.661086865701867e-05, + "loss": 0.4497, + "step": 2901 + }, + { + "epoch": 1.094268476621418, + "grad_norm": 0.38762031644191763, + "learning_rate": 7.660662555877948e-05, + "loss": 0.4481, + "step": 2902 + }, + { + "epoch": 1.0946455505279036, + "grad_norm": 0.3541802641037805, + "learning_rate": 7.660237992373357e-05, + "loss": 0.4569, + "step": 2903 + }, + { + "epoch": 1.0950226244343892, + "grad_norm": 0.38072397980807193, + "learning_rate": 7.659813175217521e-05, + "loss": 0.4406, + "step": 2904 + }, + { + "epoch": 1.095399698340875, + "grad_norm": 0.4512573006123397, + "learning_rate": 7.659388104439879e-05, + "loss": 0.4439, + "step": 2905 + }, + { + "epoch": 1.0957767722473606, + "grad_norm": 0.5038275642741754, + "learning_rate": 7.658962780069886e-05, + "loss": 0.4272, + "step": 2906 + }, + { + "epoch": 1.0961538461538463, + "grad_norm": 0.44135675493284904, + "learning_rate": 7.658537202137018e-05, + "loss": 0.4441, + "step": 2907 + }, + { + "epoch": 1.096530920060332, + "grad_norm": 0.3465056255378611, + "learning_rate": 7.658111370670768e-05, + "loss": 0.455, + "step": 2908 + }, + { + "epoch": 1.0969079939668176, + "grad_norm": 0.3385704622032131, + "learning_rate": 7.657685285700645e-05, + "loss": 0.4337, + "step": 2909 + }, + { + "epoch": 1.0972850678733033, + "grad_norm": 0.3908632236653765, + "learning_rate": 7.657258947256176e-05, + "loss": 0.4401, + "step": 2910 + }, + { + "epoch": 1.097662141779789, + "grad_norm": 0.43706436447201963, + "learning_rate": 7.656832355366904e-05, + "loss": 0.4431, + "step": 2911 + }, + { + "epoch": 1.0980392156862746, + "grad_norm": 0.43563747253619, + "learning_rate": 7.656405510062395e-05, + "loss": 0.4437, + "step": 2912 + }, + { + "epoch": 1.0984162895927603, + "grad_norm": 0.4053698411647875, + "learning_rate": 7.655978411372227e-05, + "loss": 0.4609, + "step": 2913 + }, + { + "epoch": 1.098793363499246, + "grad_norm": 0.44290256159918534, + "learning_rate": 7.655551059325998e-05, + "loss": 0.4396, + "step": 2914 + }, + { + "epoch": 1.0991704374057316, + "grad_norm": 0.4527151544719504, + "learning_rate": 7.655123453953323e-05, + "loss": 0.4517, + "step": 2915 + }, + { + "epoch": 1.0995475113122173, + "grad_norm": 0.5078481356188055, + "learning_rate": 7.654695595283834e-05, + "loss": 0.4359, + "step": 2916 + }, + { + "epoch": 1.099924585218703, + "grad_norm": 0.515423120472323, + "learning_rate": 7.654267483347185e-05, + "loss": 0.4581, + "step": 2917 + }, + { + "epoch": 1.1003016591251886, + "grad_norm": 0.5133923924893792, + "learning_rate": 7.653839118173036e-05, + "loss": 0.4467, + "step": 2918 + }, + { + "epoch": 1.1006787330316743, + "grad_norm": 0.5463138160894941, + "learning_rate": 7.65341049979108e-05, + "loss": 0.4417, + "step": 2919 + }, + { + "epoch": 1.10105580693816, + "grad_norm": 0.5268296381970509, + "learning_rate": 7.652981628231016e-05, + "loss": 0.4469, + "step": 2920 + }, + { + "epoch": 1.1014328808446456, + "grad_norm": 0.5545512817625627, + "learning_rate": 7.652552503522567e-05, + "loss": 0.4451, + "step": 2921 + }, + { + "epoch": 1.1018099547511313, + "grad_norm": 0.530223070789488, + "learning_rate": 7.652123125695469e-05, + "loss": 0.447, + "step": 2922 + }, + { + "epoch": 1.102187028657617, + "grad_norm": 0.43243201426121264, + "learning_rate": 7.651693494779476e-05, + "loss": 0.4376, + "step": 2923 + }, + { + "epoch": 1.1025641025641026, + "grad_norm": 0.42179218056063333, + "learning_rate": 7.651263610804364e-05, + "loss": 0.4413, + "step": 2924 + }, + { + "epoch": 1.1029411764705883, + "grad_norm": 0.3863053874852189, + "learning_rate": 7.650833473799922e-05, + "loss": 0.432, + "step": 2925 + }, + { + "epoch": 1.103318250377074, + "grad_norm": 0.3697253458425399, + "learning_rate": 7.650403083795958e-05, + "loss": 0.4541, + "step": 2926 + }, + { + "epoch": 1.1036953242835597, + "grad_norm": 0.4646437514853196, + "learning_rate": 7.6499724408223e-05, + "loss": 0.4464, + "step": 2927 + }, + { + "epoch": 1.1040723981900453, + "grad_norm": 0.5191601616698993, + "learning_rate": 7.649541544908788e-05, + "loss": 0.4386, + "step": 2928 + }, + { + "epoch": 1.104449472096531, + "grad_norm": 0.543532314390369, + "learning_rate": 7.649110396085285e-05, + "loss": 0.4592, + "step": 2929 + }, + { + "epoch": 1.1048265460030167, + "grad_norm": 0.5387741175358804, + "learning_rate": 7.648678994381667e-05, + "loss": 0.4555, + "step": 2930 + }, + { + "epoch": 1.1052036199095023, + "grad_norm": 0.5353664052724264, + "learning_rate": 7.648247339827832e-05, + "loss": 0.4391, + "step": 2931 + }, + { + "epoch": 1.105580693815988, + "grad_norm": 0.4771134257589829, + "learning_rate": 7.647815432453692e-05, + "loss": 0.4492, + "step": 2932 + }, + { + "epoch": 1.1059577677224737, + "grad_norm": 0.34617750651863066, + "learning_rate": 7.647383272289178e-05, + "loss": 0.4456, + "step": 2933 + }, + { + "epoch": 1.1063348416289593, + "grad_norm": 0.2931231505698022, + "learning_rate": 7.64695085936424e-05, + "loss": 0.4433, + "step": 2934 + }, + { + "epoch": 1.106711915535445, + "grad_norm": 0.4037403565493025, + "learning_rate": 7.646518193708841e-05, + "loss": 0.4372, + "step": 2935 + }, + { + "epoch": 1.1070889894419307, + "grad_norm": 0.4919092490766811, + "learning_rate": 7.646085275352965e-05, + "loss": 0.4315, + "step": 2936 + }, + { + "epoch": 1.1074660633484164, + "grad_norm": 0.3833139213186309, + "learning_rate": 7.645652104326614e-05, + "loss": 0.4349, + "step": 2937 + }, + { + "epoch": 1.107843137254902, + "grad_norm": 0.3831235012056497, + "learning_rate": 7.645218680659807e-05, + "loss": 0.4382, + "step": 2938 + }, + { + "epoch": 1.1082202111613877, + "grad_norm": 0.44458622242583395, + "learning_rate": 7.644785004382577e-05, + "loss": 0.4346, + "step": 2939 + }, + { + "epoch": 1.1085972850678734, + "grad_norm": 0.4920096612100432, + "learning_rate": 7.644351075524981e-05, + "loss": 0.4532, + "step": 2940 + }, + { + "epoch": 1.108974358974359, + "grad_norm": 0.46940224502356726, + "learning_rate": 7.643916894117086e-05, + "loss": 0.442, + "step": 2941 + }, + { + "epoch": 1.1093514328808447, + "grad_norm": 0.4291161579753593, + "learning_rate": 7.643482460188983e-05, + "loss": 0.4289, + "step": 2942 + }, + { + "epoch": 1.1097285067873304, + "grad_norm": 0.4141949263109128, + "learning_rate": 7.643047773770778e-05, + "loss": 0.4423, + "step": 2943 + }, + { + "epoch": 1.110105580693816, + "grad_norm": 0.3383552743480246, + "learning_rate": 7.642612834892593e-05, + "loss": 0.4329, + "step": 2944 + }, + { + "epoch": 1.1104826546003017, + "grad_norm": 0.37836118817524816, + "learning_rate": 7.642177643584569e-05, + "loss": 0.443, + "step": 2945 + }, + { + "epoch": 1.1108597285067874, + "grad_norm": 0.4032028432494278, + "learning_rate": 7.641742199876866e-05, + "loss": 0.4447, + "step": 2946 + }, + { + "epoch": 1.111236802413273, + "grad_norm": 0.30828759357610014, + "learning_rate": 7.641306503799658e-05, + "loss": 0.4385, + "step": 2947 + }, + { + "epoch": 1.1116138763197587, + "grad_norm": 0.32318119632459785, + "learning_rate": 7.64087055538314e-05, + "loss": 0.4438, + "step": 2948 + }, + { + "epoch": 1.1119909502262444, + "grad_norm": 0.34611679748406027, + "learning_rate": 7.640434354657522e-05, + "loss": 0.4444, + "step": 2949 + }, + { + "epoch": 1.11236802413273, + "grad_norm": 0.33010783039620684, + "learning_rate": 7.639997901653032e-05, + "loss": 0.4458, + "step": 2950 + }, + { + "epoch": 1.1127450980392157, + "grad_norm": 0.35712127952616973, + "learning_rate": 7.639561196399916e-05, + "loss": 0.4473, + "step": 2951 + }, + { + "epoch": 1.1131221719457014, + "grad_norm": 0.35821706638496276, + "learning_rate": 7.639124238928438e-05, + "loss": 0.4283, + "step": 2952 + }, + { + "epoch": 1.113499245852187, + "grad_norm": 0.3542861171248702, + "learning_rate": 7.638687029268877e-05, + "loss": 0.4362, + "step": 2953 + }, + { + "epoch": 1.1138763197586727, + "grad_norm": 0.32281405886852454, + "learning_rate": 7.638249567451532e-05, + "loss": 0.4349, + "step": 2954 + }, + { + "epoch": 1.1142533936651584, + "grad_norm": 0.33886380600065596, + "learning_rate": 7.63781185350672e-05, + "loss": 0.4529, + "step": 2955 + }, + { + "epoch": 1.114630467571644, + "grad_norm": 0.4388435899622333, + "learning_rate": 7.637373887464773e-05, + "loss": 0.4383, + "step": 2956 + }, + { + "epoch": 1.1150075414781297, + "grad_norm": 0.5822678726226735, + "learning_rate": 7.636935669356043e-05, + "loss": 0.4225, + "step": 2957 + }, + { + "epoch": 1.1153846153846154, + "grad_norm": 0.7407139256635408, + "learning_rate": 7.636497199210895e-05, + "loss": 0.4506, + "step": 2958 + }, + { + "epoch": 1.115761689291101, + "grad_norm": 0.8200516877792774, + "learning_rate": 7.636058477059717e-05, + "loss": 0.439, + "step": 2959 + }, + { + "epoch": 1.1161387631975868, + "grad_norm": 0.8090479757914449, + "learning_rate": 7.635619502932914e-05, + "loss": 0.4334, + "step": 2960 + }, + { + "epoch": 1.1165158371040724, + "grad_norm": 0.7463922862955303, + "learning_rate": 7.635180276860903e-05, + "loss": 0.4406, + "step": 2961 + }, + { + "epoch": 1.116892911010558, + "grad_norm": 0.6134765033983144, + "learning_rate": 7.634740798874123e-05, + "loss": 0.4383, + "step": 2962 + }, + { + "epoch": 1.1172699849170438, + "grad_norm": 0.5770407443086097, + "learning_rate": 7.634301069003028e-05, + "loss": 0.4474, + "step": 2963 + }, + { + "epoch": 1.1176470588235294, + "grad_norm": 0.6035154862975619, + "learning_rate": 7.633861087278095e-05, + "loss": 0.4532, + "step": 2964 + }, + { + "epoch": 1.118024132730015, + "grad_norm": 0.5251179903633241, + "learning_rate": 7.633420853729813e-05, + "loss": 0.4465, + "step": 2965 + }, + { + "epoch": 1.1184012066365008, + "grad_norm": 0.4160711396675011, + "learning_rate": 7.632980368388685e-05, + "loss": 0.4407, + "step": 2966 + }, + { + "epoch": 1.1187782805429864, + "grad_norm": 0.32550986419142575, + "learning_rate": 7.632539631285242e-05, + "loss": 0.446, + "step": 2967 + }, + { + "epoch": 1.1191553544494721, + "grad_norm": 0.2948755310951733, + "learning_rate": 7.632098642450024e-05, + "loss": 0.4569, + "step": 2968 + }, + { + "epoch": 1.1195324283559578, + "grad_norm": 0.29570851732275727, + "learning_rate": 7.631657401913593e-05, + "loss": 0.4322, + "step": 2969 + }, + { + "epoch": 1.1199095022624435, + "grad_norm": 0.29430837054498243, + "learning_rate": 7.631215909706523e-05, + "loss": 0.437, + "step": 2970 + }, + { + "epoch": 1.1202865761689291, + "grad_norm": 0.3310677422441225, + "learning_rate": 7.630774165859413e-05, + "loss": 0.4245, + "step": 2971 + }, + { + "epoch": 1.1206636500754148, + "grad_norm": 0.4146582661306805, + "learning_rate": 7.630332170402874e-05, + "loss": 0.4436, + "step": 2972 + }, + { + "epoch": 1.1210407239819005, + "grad_norm": 0.44114055227546933, + "learning_rate": 7.629889923367535e-05, + "loss": 0.4357, + "step": 2973 + }, + { + "epoch": 1.1214177978883861, + "grad_norm": 0.41948448328911214, + "learning_rate": 7.629447424784045e-05, + "loss": 0.4405, + "step": 2974 + }, + { + "epoch": 1.1217948717948718, + "grad_norm": 0.3967881827842816, + "learning_rate": 7.629004674683065e-05, + "loss": 0.4261, + "step": 2975 + }, + { + "epoch": 1.1221719457013575, + "grad_norm": 0.417872149280174, + "learning_rate": 7.628561673095283e-05, + "loss": 0.433, + "step": 2976 + }, + { + "epoch": 1.1225490196078431, + "grad_norm": 0.4650287143572178, + "learning_rate": 7.628118420051394e-05, + "loss": 0.4495, + "step": 2977 + }, + { + "epoch": 1.1229260935143288, + "grad_norm": 0.4595936323134962, + "learning_rate": 7.627674915582119e-05, + "loss": 0.4329, + "step": 2978 + }, + { + "epoch": 1.1233031674208145, + "grad_norm": 0.43035639574171736, + "learning_rate": 7.627231159718187e-05, + "loss": 0.4368, + "step": 2979 + }, + { + "epoch": 1.1236802413273002, + "grad_norm": 0.4693675190619924, + "learning_rate": 7.626787152490355e-05, + "loss": 0.4257, + "step": 2980 + }, + { + "epoch": 1.1240573152337858, + "grad_norm": 0.6124893061532263, + "learning_rate": 7.626342893929389e-05, + "loss": 0.4511, + "step": 2981 + }, + { + "epoch": 1.1244343891402715, + "grad_norm": 0.7945032199531892, + "learning_rate": 7.625898384066078e-05, + "loss": 0.4579, + "step": 2982 + }, + { + "epoch": 1.1248114630467572, + "grad_norm": 0.9111024899764619, + "learning_rate": 7.625453622931223e-05, + "loss": 0.444, + "step": 2983 + }, + { + "epoch": 1.1251885369532428, + "grad_norm": 0.9225822059575735, + "learning_rate": 7.62500861055565e-05, + "loss": 0.4333, + "step": 2984 + }, + { + "epoch": 1.1255656108597285, + "grad_norm": 0.7967918774868081, + "learning_rate": 7.624563346970196e-05, + "loss": 0.4479, + "step": 2985 + }, + { + "epoch": 1.1259426847662142, + "grad_norm": 0.5453814970667881, + "learning_rate": 7.624117832205717e-05, + "loss": 0.4377, + "step": 2986 + }, + { + "epoch": 1.1263197586726998, + "grad_norm": 0.34414576519599066, + "learning_rate": 7.623672066293084e-05, + "loss": 0.4401, + "step": 2987 + }, + { + "epoch": 1.1266968325791855, + "grad_norm": 0.4898560848589351, + "learning_rate": 7.623226049263194e-05, + "loss": 0.4341, + "step": 2988 + }, + { + "epoch": 1.1270739064856712, + "grad_norm": 0.6225387779890218, + "learning_rate": 7.622779781146952e-05, + "loss": 0.4387, + "step": 2989 + }, + { + "epoch": 1.1274509803921569, + "grad_norm": 0.6058851969270713, + "learning_rate": 7.622333261975283e-05, + "loss": 0.4361, + "step": 2990 + }, + { + "epoch": 1.1278280542986425, + "grad_norm": 0.5975884824743825, + "learning_rate": 7.621886491779134e-05, + "loss": 0.4605, + "step": 2991 + }, + { + "epoch": 1.1282051282051282, + "grad_norm": 0.5178377108170968, + "learning_rate": 7.621439470589462e-05, + "loss": 0.4515, + "step": 2992 + }, + { + "epoch": 1.1285822021116139, + "grad_norm": 0.40651320125204576, + "learning_rate": 7.620992198437248e-05, + "loss": 0.438, + "step": 2993 + }, + { + "epoch": 1.1289592760180995, + "grad_norm": 0.4360409972832924, + "learning_rate": 7.620544675353486e-05, + "loss": 0.4289, + "step": 2994 + }, + { + "epoch": 1.1293363499245852, + "grad_norm": 0.46248244438884134, + "learning_rate": 7.62009690136919e-05, + "loss": 0.4395, + "step": 2995 + }, + { + "epoch": 1.1297134238310709, + "grad_norm": 0.5703593004783021, + "learning_rate": 7.619648876515389e-05, + "loss": 0.435, + "step": 2996 + }, + { + "epoch": 1.1300904977375565, + "grad_norm": 0.5713866902946504, + "learning_rate": 7.619200600823131e-05, + "loss": 0.4377, + "step": 2997 + }, + { + "epoch": 1.1304675716440422, + "grad_norm": 0.43990108640302694, + "learning_rate": 7.618752074323482e-05, + "loss": 0.4369, + "step": 2998 + }, + { + "epoch": 1.1308446455505279, + "grad_norm": 0.35111915370064917, + "learning_rate": 7.618303297047525e-05, + "loss": 0.4485, + "step": 2999 + }, + { + "epoch": 1.1312217194570136, + "grad_norm": 0.395392959895956, + "learning_rate": 7.617854269026359e-05, + "loss": 0.4426, + "step": 3000 + }, + { + "epoch": 1.1315987933634992, + "grad_norm": 0.4950550371100359, + "learning_rate": 7.6174049902911e-05, + "loss": 0.4233, + "step": 3001 + }, + { + "epoch": 1.131975867269985, + "grad_norm": 0.5770057100652731, + "learning_rate": 7.616955460872884e-05, + "loss": 0.4513, + "step": 3002 + }, + { + "epoch": 1.1323529411764706, + "grad_norm": 0.5569348817691324, + "learning_rate": 7.616505680802863e-05, + "loss": 0.4408, + "step": 3003 + }, + { + "epoch": 1.1327300150829562, + "grad_norm": 0.45415579142761986, + "learning_rate": 7.616055650112207e-05, + "loss": 0.4435, + "step": 3004 + }, + { + "epoch": 1.133107088989442, + "grad_norm": 0.36126450372511204, + "learning_rate": 7.615605368832102e-05, + "loss": 0.4394, + "step": 3005 + }, + { + "epoch": 1.1334841628959276, + "grad_norm": 0.33566504473118186, + "learning_rate": 7.615154836993751e-05, + "loss": 0.453, + "step": 3006 + }, + { + "epoch": 1.1338612368024132, + "grad_norm": 0.3606632704701776, + "learning_rate": 7.614704054628377e-05, + "loss": 0.4414, + "step": 3007 + }, + { + "epoch": 1.134238310708899, + "grad_norm": 0.43366011466268406, + "learning_rate": 7.614253021767218e-05, + "loss": 0.4336, + "step": 3008 + }, + { + "epoch": 1.1346153846153846, + "grad_norm": 0.44678920560900653, + "learning_rate": 7.61380173844153e-05, + "loss": 0.4455, + "step": 3009 + }, + { + "epoch": 1.1349924585218703, + "grad_norm": 0.34815415852930187, + "learning_rate": 7.613350204682586e-05, + "loss": 0.444, + "step": 3010 + }, + { + "epoch": 1.135369532428356, + "grad_norm": 0.24459811147439522, + "learning_rate": 7.61289842052168e-05, + "loss": 0.4377, + "step": 3011 + }, + { + "epoch": 1.1357466063348416, + "grad_norm": 0.31076078856464273, + "learning_rate": 7.612446385990118e-05, + "loss": 0.4292, + "step": 3012 + }, + { + "epoch": 1.1361236802413273, + "grad_norm": 0.3827533163369764, + "learning_rate": 7.611994101119225e-05, + "loss": 0.4384, + "step": 3013 + }, + { + "epoch": 1.136500754147813, + "grad_norm": 0.3956328891391689, + "learning_rate": 7.611541565940343e-05, + "loss": 0.4317, + "step": 3014 + }, + { + "epoch": 1.1368778280542986, + "grad_norm": 0.38306491279059957, + "learning_rate": 7.611088780484837e-05, + "loss": 0.444, + "step": 3015 + }, + { + "epoch": 1.1372549019607843, + "grad_norm": 0.4330431514089398, + "learning_rate": 7.61063574478408e-05, + "loss": 0.4472, + "step": 3016 + }, + { + "epoch": 1.13763197586727, + "grad_norm": 0.45215374847494927, + "learning_rate": 7.610182458869467e-05, + "loss": 0.4314, + "step": 3017 + }, + { + "epoch": 1.1380090497737556, + "grad_norm": 0.42827837503295446, + "learning_rate": 7.609728922772414e-05, + "loss": 0.4537, + "step": 3018 + }, + { + "epoch": 1.1383861236802413, + "grad_norm": 0.4915207983992711, + "learning_rate": 7.609275136524348e-05, + "loss": 0.4452, + "step": 3019 + }, + { + "epoch": 1.138763197586727, + "grad_norm": 0.6617262171418098, + "learning_rate": 7.608821100156716e-05, + "loss": 0.4497, + "step": 3020 + }, + { + "epoch": 1.1391402714932126, + "grad_norm": 0.7107404006013788, + "learning_rate": 7.608366813700984e-05, + "loss": 0.4447, + "step": 3021 + }, + { + "epoch": 1.1395173453996983, + "grad_norm": 0.6894038704442369, + "learning_rate": 7.607912277188631e-05, + "loss": 0.442, + "step": 3022 + }, + { + "epoch": 1.139894419306184, + "grad_norm": 0.627663519673197, + "learning_rate": 7.607457490651157e-05, + "loss": 0.4528, + "step": 3023 + }, + { + "epoch": 1.1402714932126696, + "grad_norm": 0.5543575848008556, + "learning_rate": 7.60700245412008e-05, + "loss": 0.4451, + "step": 3024 + }, + { + "epoch": 1.1406485671191553, + "grad_norm": 0.4587046216792381, + "learning_rate": 7.606547167626933e-05, + "loss": 0.4361, + "step": 3025 + }, + { + "epoch": 1.141025641025641, + "grad_norm": 0.3368831162728567, + "learning_rate": 7.606091631203265e-05, + "loss": 0.432, + "step": 3026 + }, + { + "epoch": 1.1414027149321266, + "grad_norm": 0.34355476709774907, + "learning_rate": 7.605635844880645e-05, + "loss": 0.4283, + "step": 3027 + }, + { + "epoch": 1.1417797888386123, + "grad_norm": 0.45695516223122284, + "learning_rate": 7.60517980869066e-05, + "loss": 0.4414, + "step": 3028 + }, + { + "epoch": 1.142156862745098, + "grad_norm": 0.4398227583543785, + "learning_rate": 7.604723522664913e-05, + "loss": 0.43, + "step": 3029 + }, + { + "epoch": 1.1425339366515836, + "grad_norm": 0.358352752578214, + "learning_rate": 7.604266986835022e-05, + "loss": 0.4531, + "step": 3030 + }, + { + "epoch": 1.1429110105580693, + "grad_norm": 0.31280707812423114, + "learning_rate": 7.603810201232626e-05, + "loss": 0.4334, + "step": 3031 + }, + { + "epoch": 1.143288084464555, + "grad_norm": 0.34514444748412815, + "learning_rate": 7.60335316588938e-05, + "loss": 0.447, + "step": 3032 + }, + { + "epoch": 1.1436651583710407, + "grad_norm": 5.34520914395709, + "learning_rate": 7.602895880836956e-05, + "loss": 0.4622, + "step": 3033 + }, + { + "epoch": 1.1440422322775263, + "grad_norm": 362.0767969218401, + "learning_rate": 7.602438346107042e-05, + "loss": 2.812, + "step": 3034 + }, + { + "epoch": 1.144419306184012, + "grad_norm": 1.227301999219436, + "learning_rate": 7.601980561731345e-05, + "loss": 0.4548, + "step": 3035 + }, + { + "epoch": 1.1447963800904977, + "grad_norm": 1.4805484112583414, + "learning_rate": 7.601522527741594e-05, + "loss": 0.482, + "step": 3036 + }, + { + "epoch": 1.1451734539969833, + "grad_norm": 0.7764475439737206, + "learning_rate": 7.601064244169523e-05, + "loss": 0.4444, + "step": 3037 + }, + { + "epoch": 1.145550527903469, + "grad_norm": 1.3727699344700872, + "learning_rate": 7.600605711046895e-05, + "loss": 0.4684, + "step": 3038 + }, + { + "epoch": 1.1459276018099547, + "grad_norm": 0.8669613157362069, + "learning_rate": 7.600146928405485e-05, + "loss": 0.4598, + "step": 3039 + }, + { + "epoch": 1.1463046757164403, + "grad_norm": 1.5566642836226694, + "learning_rate": 7.599687896277086e-05, + "loss": 0.4981, + "step": 3040 + }, + { + "epoch": 1.146681749622926, + "grad_norm": 60.22681522447443, + "learning_rate": 7.599228614693509e-05, + "loss": 0.7432, + "step": 3041 + }, + { + "epoch": 1.1470588235294117, + "grad_norm": 1.5994598179195614, + "learning_rate": 7.598769083686582e-05, + "loss": 0.4936, + "step": 3042 + }, + { + "epoch": 1.1474358974358974, + "grad_norm": 0.5126533394812827, + "learning_rate": 7.59830930328815e-05, + "loss": 0.4564, + "step": 3043 + }, + { + "epoch": 1.147812971342383, + "grad_norm": 1.0379809907543478, + "learning_rate": 7.597849273530073e-05, + "loss": 0.4855, + "step": 3044 + }, + { + "epoch": 1.1481900452488687, + "grad_norm": 59.42251591367119, + "learning_rate": 7.597388994444234e-05, + "loss": 0.683, + "step": 3045 + }, + { + "epoch": 1.1485671191553544, + "grad_norm": 10.915614400690412, + "learning_rate": 7.596928466062528e-05, + "loss": 0.6165, + "step": 3046 + }, + { + "epoch": 1.14894419306184, + "grad_norm": 42.954106648457646, + "learning_rate": 7.596467688416869e-05, + "loss": 1.8807, + "step": 3047 + }, + { + "epoch": 1.1493212669683257, + "grad_norm": 71.9610093521069, + "learning_rate": 7.596006661539191e-05, + "loss": 9.1651, + "step": 3048 + }, + { + "epoch": 1.1496983408748114, + "grad_norm": 86.33372407903695, + "learning_rate": 7.595545385461441e-05, + "loss": 5.1052, + "step": 3049 + }, + { + "epoch": 1.150075414781297, + "grad_norm": 73.91763005281103, + "learning_rate": 7.595083860215583e-05, + "loss": 8.4591, + "step": 3050 + }, + { + "epoch": 1.1504524886877827, + "grad_norm": 62.88558152182579, + "learning_rate": 7.594622085833604e-05, + "loss": 9.1744, + "step": 3051 + }, + { + "epoch": 1.1508295625942684, + "grad_norm": 53.322562015928895, + "learning_rate": 7.594160062347502e-05, + "loss": 8.5093, + "step": 3052 + }, + { + "epoch": 1.151206636500754, + "grad_norm": 49.693481743436955, + "learning_rate": 7.593697789789296e-05, + "loss": 7.6585, + "step": 3053 + }, + { + "epoch": 1.1515837104072397, + "grad_norm": 27.891396903152902, + "learning_rate": 7.593235268191022e-05, + "loss": 7.2274, + "step": 3054 + }, + { + "epoch": 1.1519607843137254, + "grad_norm": 20.10717350675179, + "learning_rate": 7.592772497584729e-05, + "loss": 7.6339, + "step": 3055 + }, + { + "epoch": 1.152337858220211, + "grad_norm": 116.36083464143164, + "learning_rate": 7.59230947800249e-05, + "loss": 9.6932, + "step": 3056 + }, + { + "epoch": 1.1527149321266967, + "grad_norm": 239.5504684661475, + "learning_rate": 7.591846209476392e-05, + "loss": 9.9791, + "step": 3057 + }, + { + "epoch": 1.1530920060331824, + "grad_norm": 37.53361220760065, + "learning_rate": 7.591382692038535e-05, + "loss": 8.9642, + "step": 3058 + }, + { + "epoch": 1.153469079939668, + "grad_norm": 58.40591784286071, + "learning_rate": 7.590918925721044e-05, + "loss": 8.7886, + "step": 3059 + }, + { + "epoch": 1.1538461538461537, + "grad_norm": 226.99067677403443, + "learning_rate": 7.590454910556058e-05, + "loss": 14.2519, + "step": 3060 + }, + { + "epoch": 1.1542232277526394, + "grad_norm": 91.24142221247645, + "learning_rate": 7.589990646575729e-05, + "loss": 10.1416, + "step": 3061 + }, + { + "epoch": 1.154600301659125, + "grad_norm": 150.06620932581126, + "learning_rate": 7.589526133812233e-05, + "loss": 8.4663, + "step": 3062 + }, + { + "epoch": 1.1549773755656108, + "grad_norm": 32.056196963176035, + "learning_rate": 7.589061372297761e-05, + "loss": 7.2857, + "step": 3063 + }, + { + "epoch": 1.1553544494720964, + "grad_norm": 107.1994911346753, + "learning_rate": 7.588596362064519e-05, + "loss": 11.2361, + "step": 3064 + }, + { + "epoch": 1.155731523378582, + "grad_norm": 34.00658063723824, + "learning_rate": 7.588131103144732e-05, + "loss": 7.579, + "step": 3065 + }, + { + "epoch": 1.1561085972850678, + "grad_norm": 14.124247291541291, + "learning_rate": 7.587665595570642e-05, + "loss": 7.8921, + "step": 3066 + }, + { + "epoch": 1.1564856711915534, + "grad_norm": 61.7416496102752, + "learning_rate": 7.587199839374508e-05, + "loss": 8.778, + "step": 3067 + }, + { + "epoch": 1.156862745098039, + "grad_norm": 15.28926556131397, + "learning_rate": 7.586733834588607e-05, + "loss": 6.8869, + "step": 3068 + }, + { + "epoch": 1.1572398190045248, + "grad_norm": 8.32254423743692, + "learning_rate": 7.586267581245233e-05, + "loss": 6.7446, + "step": 3069 + }, + { + "epoch": 1.1576168929110104, + "grad_norm": 16.52534631825698, + "learning_rate": 7.585801079376698e-05, + "loss": 6.6925, + "step": 3070 + }, + { + "epoch": 1.1579939668174961, + "grad_norm": 10.221464369254253, + "learning_rate": 7.585334329015327e-05, + "loss": 6.6644, + "step": 3071 + }, + { + "epoch": 1.1583710407239818, + "grad_norm": 21.472892241812318, + "learning_rate": 7.584867330193468e-05, + "loss": 6.9711, + "step": 3072 + }, + { + "epoch": 1.1587481146304675, + "grad_norm": 17.340075187201563, + "learning_rate": 7.584400082943483e-05, + "loss": 6.6697, + "step": 3073 + }, + { + "epoch": 1.1591251885369531, + "grad_norm": 8.175382128931115, + "learning_rate": 7.583932587297752e-05, + "loss": 6.4349, + "step": 3074 + }, + { + "epoch": 1.1595022624434388, + "grad_norm": 6.623589691086457, + "learning_rate": 7.583464843288671e-05, + "loss": 6.3632, + "step": 3075 + }, + { + "epoch": 1.1598793363499245, + "grad_norm": 7.704040787388317, + "learning_rate": 7.582996850948654e-05, + "loss": 6.3787, + "step": 3076 + }, + { + "epoch": 1.1602564102564104, + "grad_norm": 6.6403709230411945, + "learning_rate": 7.582528610310135e-05, + "loss": 6.3017, + "step": 3077 + }, + { + "epoch": 1.160633484162896, + "grad_norm": 5.628010602297843, + "learning_rate": 7.582060121405561e-05, + "loss": 6.2744, + "step": 3078 + }, + { + "epoch": 1.1610105580693817, + "grad_norm": 9.811570725155708, + "learning_rate": 7.581591384267396e-05, + "loss": 6.2498, + "step": 3079 + }, + { + "epoch": 1.1613876319758674, + "grad_norm": 10.49309170079755, + "learning_rate": 7.581122398928126e-05, + "loss": 6.256, + "step": 3080 + }, + { + "epoch": 1.161764705882353, + "grad_norm": 5.612251400775109, + "learning_rate": 7.58065316542025e-05, + "loss": 6.2088, + "step": 3081 + }, + { + "epoch": 1.1621417797888387, + "grad_norm": 5.055929852755966, + "learning_rate": 7.580183683776286e-05, + "loss": 6.1123, + "step": 3082 + }, + { + "epoch": 1.1625188536953244, + "grad_norm": 5.856159256072582, + "learning_rate": 7.579713954028767e-05, + "loss": 6.1387, + "step": 3083 + }, + { + "epoch": 1.16289592760181, + "grad_norm": 4.165640765549386, + "learning_rate": 7.579243976210249e-05, + "loss": 6.0537, + "step": 3084 + }, + { + "epoch": 1.1632730015082957, + "grad_norm": 6.17131716325805, + "learning_rate": 7.578773750353296e-05, + "loss": 6.0522, + "step": 3085 + }, + { + "epoch": 1.1636500754147814, + "grad_norm": 3.901389303867557, + "learning_rate": 7.578303276490497e-05, + "loss": 6.0664, + "step": 3086 + }, + { + "epoch": 1.164027149321267, + "grad_norm": 2.8811154909377295, + "learning_rate": 7.577832554654455e-05, + "loss": 5.9114, + "step": 3087 + }, + { + "epoch": 1.1644042232277527, + "grad_norm": 7.687373546417385, + "learning_rate": 7.57736158487779e-05, + "loss": 6.1191, + "step": 3088 + }, + { + "epoch": 1.1647812971342384, + "grad_norm": 4.833645080767921, + "learning_rate": 7.576890367193142e-05, + "loss": 5.9971, + "step": 3089 + }, + { + "epoch": 1.165158371040724, + "grad_norm": 5.551474898921566, + "learning_rate": 7.576418901633161e-05, + "loss": 5.894, + "step": 3090 + }, + { + "epoch": 1.1655354449472097, + "grad_norm": 5.080534605524621, + "learning_rate": 7.575947188230526e-05, + "loss": 5.9241, + "step": 3091 + }, + { + "epoch": 1.1659125188536954, + "grad_norm": 3.209396561292437, + "learning_rate": 7.575475227017921e-05, + "loss": 5.8382, + "step": 3092 + }, + { + "epoch": 1.166289592760181, + "grad_norm": 5.294130719895125, + "learning_rate": 7.575003018028055e-05, + "loss": 5.8151, + "step": 3093 + }, + { + "epoch": 1.1666666666666667, + "grad_norm": 5.6232999846450085, + "learning_rate": 7.57453056129365e-05, + "loss": 5.8895, + "step": 3094 + }, + { + "epoch": 1.1670437405731524, + "grad_norm": 3.673523145198866, + "learning_rate": 7.574057856847448e-05, + "loss": 5.7263, + "step": 3095 + }, + { + "epoch": 1.167420814479638, + "grad_norm": 5.3142022454820586, + "learning_rate": 7.573584904722206e-05, + "loss": 5.8592, + "step": 3096 + }, + { + "epoch": 1.1677978883861238, + "grad_norm": 3.3400112447163104, + "learning_rate": 7.5731117049507e-05, + "loss": 5.6574, + "step": 3097 + }, + { + "epoch": 1.1681749622926094, + "grad_norm": 4.324823987972047, + "learning_rate": 7.572638257565723e-05, + "loss": 5.7172, + "step": 3098 + }, + { + "epoch": 1.168552036199095, + "grad_norm": 4.732553790295642, + "learning_rate": 7.572164562600084e-05, + "loss": 5.6832, + "step": 3099 + }, + { + "epoch": 1.1689291101055808, + "grad_norm": 3.5437048751281663, + "learning_rate": 7.571690620086609e-05, + "loss": 5.6235, + "step": 3100 + }, + { + "epoch": 1.1693061840120664, + "grad_norm": 4.193098375990047, + "learning_rate": 7.571216430058141e-05, + "loss": 5.6507, + "step": 3101 + }, + { + "epoch": 1.169683257918552, + "grad_norm": 3.4878864426389704, + "learning_rate": 7.570741992547544e-05, + "loss": 5.6113, + "step": 3102 + }, + { + "epoch": 1.1700603318250378, + "grad_norm": 6.081038738633236, + "learning_rate": 7.570267307587692e-05, + "loss": 5.6462, + "step": 3103 + }, + { + "epoch": 1.1704374057315234, + "grad_norm": 2.7812758378921933, + "learning_rate": 7.569792375211483e-05, + "loss": 5.5536, + "step": 3104 + }, + { + "epoch": 1.170814479638009, + "grad_norm": 3.379792459518115, + "learning_rate": 7.56931719545183e-05, + "loss": 5.5379, + "step": 3105 + }, + { + "epoch": 1.1711915535444948, + "grad_norm": 4.382894805862144, + "learning_rate": 7.56884176834166e-05, + "loss": 5.5993, + "step": 3106 + }, + { + "epoch": 1.1715686274509804, + "grad_norm": 3.2164915440855997, + "learning_rate": 7.568366093913922e-05, + "loss": 5.5158, + "step": 3107 + }, + { + "epoch": 1.1719457013574661, + "grad_norm": 2.7248512809296144, + "learning_rate": 7.567890172201578e-05, + "loss": 5.4874, + "step": 3108 + }, + { + "epoch": 1.1723227752639518, + "grad_norm": 5.6057976196028765, + "learning_rate": 7.567414003237611e-05, + "loss": 5.5121, + "step": 3109 + }, + { + "epoch": 1.1726998491704375, + "grad_norm": 3.3246466338116925, + "learning_rate": 7.566937587055016e-05, + "loss": 5.4514, + "step": 3110 + }, + { + "epoch": 1.1730769230769231, + "grad_norm": 3.89549562703663, + "learning_rate": 7.566460923686812e-05, + "loss": 5.4119, + "step": 3111 + }, + { + "epoch": 1.1734539969834088, + "grad_norm": 3.9681650883874884, + "learning_rate": 7.565984013166029e-05, + "loss": 5.3829, + "step": 3112 + }, + { + "epoch": 1.1738310708898945, + "grad_norm": 4.300595454407629, + "learning_rate": 7.565506855525715e-05, + "loss": 5.3929, + "step": 3113 + }, + { + "epoch": 1.1742081447963801, + "grad_norm": 3.4997725304016165, + "learning_rate": 7.565029450798942e-05, + "loss": 5.3817, + "step": 3114 + }, + { + "epoch": 1.1745852187028658, + "grad_norm": 3.4700035440713393, + "learning_rate": 7.564551799018788e-05, + "loss": 5.396, + "step": 3115 + }, + { + "epoch": 1.1749622926093515, + "grad_norm": 4.315663546320227, + "learning_rate": 7.564073900218356e-05, + "loss": 5.3638, + "step": 3116 + }, + { + "epoch": 1.1753393665158371, + "grad_norm": 1.9490511028555344, + "learning_rate": 7.563595754430764e-05, + "loss": 5.3326, + "step": 3117 + }, + { + "epoch": 1.1757164404223228, + "grad_norm": 3.683142744265034, + "learning_rate": 7.563117361689147e-05, + "loss": 5.3258, + "step": 3118 + }, + { + "epoch": 1.1760935143288085, + "grad_norm": 2.774794107561982, + "learning_rate": 7.562638722026659e-05, + "loss": 5.3276, + "step": 3119 + }, + { + "epoch": 1.1764705882352942, + "grad_norm": 4.390195325548224, + "learning_rate": 7.562159835476466e-05, + "loss": 5.3704, + "step": 3120 + }, + { + "epoch": 1.1768476621417798, + "grad_norm": 2.16050081587265, + "learning_rate": 7.561680702071756e-05, + "loss": 5.2412, + "step": 3121 + }, + { + "epoch": 1.1772247360482655, + "grad_norm": 4.324172548108098, + "learning_rate": 7.561201321845733e-05, + "loss": 5.4203, + "step": 3122 + }, + { + "epoch": 1.1776018099547512, + "grad_norm": 2.9095477577437348, + "learning_rate": 7.560721694831617e-05, + "loss": 5.2493, + "step": 3123 + }, + { + "epoch": 1.1779788838612368, + "grad_norm": 4.195112028565799, + "learning_rate": 7.560241821062646e-05, + "loss": 5.3109, + "step": 3124 + }, + { + "epoch": 1.1783559577677225, + "grad_norm": 3.9541949732166737, + "learning_rate": 7.559761700572075e-05, + "loss": 5.2465, + "step": 3125 + }, + { + "epoch": 1.1787330316742082, + "grad_norm": 2.662253851563683, + "learning_rate": 7.559281333393176e-05, + "loss": 5.2416, + "step": 3126 + }, + { + "epoch": 1.1791101055806938, + "grad_norm": 2.5386598170511396, + "learning_rate": 7.558800719559237e-05, + "loss": 5.2374, + "step": 3127 + }, + { + "epoch": 1.1794871794871795, + "grad_norm": 2.2116874231194292, + "learning_rate": 7.558319859103565e-05, + "loss": 5.1763, + "step": 3128 + }, + { + "epoch": 1.1798642533936652, + "grad_norm": 2.79556597239872, + "learning_rate": 7.557838752059484e-05, + "loss": 5.2612, + "step": 3129 + }, + { + "epoch": 1.1802413273001509, + "grad_norm": 1.8476845708399798, + "learning_rate": 7.557357398460331e-05, + "loss": 5.1757, + "step": 3130 + }, + { + "epoch": 1.1806184012066365, + "grad_norm": 2.3583589478431985, + "learning_rate": 7.556875798339468e-05, + "loss": 5.1271, + "step": 3131 + }, + { + "epoch": 1.1809954751131222, + "grad_norm": 2.031950500451867, + "learning_rate": 7.556393951730266e-05, + "loss": 5.136, + "step": 3132 + }, + { + "epoch": 1.1813725490196079, + "grad_norm": 2.8876694488432673, + "learning_rate": 7.555911858666116e-05, + "loss": 5.0771, + "step": 3133 + }, + { + "epoch": 1.1817496229260935, + "grad_norm": 4.339884067205675, + "learning_rate": 7.555429519180431e-05, + "loss": 5.1094, + "step": 3134 + }, + { + "epoch": 1.1821266968325792, + "grad_norm": 2.184317424697193, + "learning_rate": 7.554946933306633e-05, + "loss": 5.0952, + "step": 3135 + }, + { + "epoch": 1.1825037707390649, + "grad_norm": 4.223674824483176, + "learning_rate": 7.554464101078165e-05, + "loss": 5.0968, + "step": 3136 + }, + { + "epoch": 1.1828808446455505, + "grad_norm": 3.206993626930752, + "learning_rate": 7.553981022528489e-05, + "loss": 5.0395, + "step": 3137 + }, + { + "epoch": 1.1832579185520362, + "grad_norm": 3.493448403795824, + "learning_rate": 7.553497697691079e-05, + "loss": 5.0778, + "step": 3138 + }, + { + "epoch": 1.1836349924585219, + "grad_norm": 2.888962238227541, + "learning_rate": 7.553014126599431e-05, + "loss": 5.016, + "step": 3139 + }, + { + "epoch": 1.1840120663650076, + "grad_norm": 2.9833614277417824, + "learning_rate": 7.552530309287056e-05, + "loss": 4.9745, + "step": 3140 + }, + { + "epoch": 1.1843891402714932, + "grad_norm": 2.039447702182038, + "learning_rate": 7.552046245787482e-05, + "loss": 4.9568, + "step": 3141 + }, + { + "epoch": 1.184766214177979, + "grad_norm": 2.3652665702712476, + "learning_rate": 7.551561936134254e-05, + "loss": 4.927, + "step": 3142 + }, + { + "epoch": 1.1851432880844646, + "grad_norm": 1.936202326912784, + "learning_rate": 7.551077380360932e-05, + "loss": 4.9489, + "step": 3143 + }, + { + "epoch": 1.1855203619909502, + "grad_norm": 5.879769330643313, + "learning_rate": 7.5505925785011e-05, + "loss": 4.9757, + "step": 3144 + }, + { + "epoch": 1.185897435897436, + "grad_norm": 2.7129411185031573, + "learning_rate": 7.55010753058835e-05, + "loss": 4.9305, + "step": 3145 + }, + { + "epoch": 1.1862745098039216, + "grad_norm": 4.288983425606686, + "learning_rate": 7.549622236656299e-05, + "loss": 4.965, + "step": 3146 + }, + { + "epoch": 1.1866515837104072, + "grad_norm": 3.3039723117203277, + "learning_rate": 7.549136696738572e-05, + "loss": 4.9199, + "step": 3147 + }, + { + "epoch": 1.187028657616893, + "grad_norm": 4.77990518199856, + "learning_rate": 7.548650910868823e-05, + "loss": 4.9825, + "step": 3148 + }, + { + "epoch": 1.1874057315233786, + "grad_norm": 3.157546176523627, + "learning_rate": 7.548164879080712e-05, + "loss": 4.8635, + "step": 3149 + }, + { + "epoch": 1.1877828054298643, + "grad_norm": 4.728206648043748, + "learning_rate": 7.547678601407923e-05, + "loss": 4.9407, + "step": 3150 + }, + { + "epoch": 1.18815987933635, + "grad_norm": 3.8984306323785587, + "learning_rate": 7.547192077884153e-05, + "loss": 4.9003, + "step": 3151 + }, + { + "epoch": 1.1885369532428356, + "grad_norm": 4.239390760911319, + "learning_rate": 7.54670530854312e-05, + "loss": 4.8727, + "step": 3152 + }, + { + "epoch": 1.1889140271493213, + "grad_norm": 4.098815053640368, + "learning_rate": 7.546218293418552e-05, + "loss": 4.8674, + "step": 3153 + }, + { + "epoch": 1.189291101055807, + "grad_norm": 2.6036834316602007, + "learning_rate": 7.545731032544204e-05, + "loss": 4.8239, + "step": 3154 + }, + { + "epoch": 1.1896681749622926, + "grad_norm": 4.138622102341476, + "learning_rate": 7.545243525953841e-05, + "loss": 4.8848, + "step": 3155 + }, + { + "epoch": 1.1900452488687783, + "grad_norm": 2.7028946815052324, + "learning_rate": 7.544755773681245e-05, + "loss": 4.7882, + "step": 3156 + }, + { + "epoch": 1.190422322775264, + "grad_norm": 2.4360848974706206, + "learning_rate": 7.544267775760217e-05, + "loss": 4.7948, + "step": 3157 + }, + { + "epoch": 1.1907993966817496, + "grad_norm": 4.329611227733041, + "learning_rate": 7.543779532224578e-05, + "loss": 4.7931, + "step": 3158 + }, + { + "epoch": 1.1911764705882353, + "grad_norm": 1.9786674365128196, + "learning_rate": 7.543291043108159e-05, + "loss": 4.7604, + "step": 3159 + }, + { + "epoch": 1.191553544494721, + "grad_norm": 4.727677637706452, + "learning_rate": 7.542802308444815e-05, + "loss": 4.8084, + "step": 3160 + }, + { + "epoch": 1.1919306184012066, + "grad_norm": 3.002157611524037, + "learning_rate": 7.542313328268414e-05, + "loss": 4.7241, + "step": 3161 + }, + { + "epoch": 1.1923076923076923, + "grad_norm": 4.9842619124271685, + "learning_rate": 7.541824102612839e-05, + "loss": 4.8505, + "step": 3162 + }, + { + "epoch": 1.192684766214178, + "grad_norm": 2.8876247784440947, + "learning_rate": 7.541334631511998e-05, + "loss": 4.7234, + "step": 3163 + }, + { + "epoch": 1.1930618401206636, + "grad_norm": 4.815375671364087, + "learning_rate": 7.540844914999807e-05, + "loss": 4.7191, + "step": 3164 + }, + { + "epoch": 1.1934389140271493, + "grad_norm": 3.2830554397629084, + "learning_rate": 7.540354953110203e-05, + "loss": 4.7119, + "step": 3165 + }, + { + "epoch": 1.193815987933635, + "grad_norm": 5.48218612327807, + "learning_rate": 7.539864745877144e-05, + "loss": 4.7325, + "step": 3166 + }, + { + "epoch": 1.1941930618401206, + "grad_norm": 5.273503055925479, + "learning_rate": 7.539374293334595e-05, + "loss": 4.7393, + "step": 3167 + }, + { + "epoch": 1.1945701357466063, + "grad_norm": 2.915024507346468, + "learning_rate": 7.538883595516549e-05, + "loss": 4.741, + "step": 3168 + }, + { + "epoch": 1.194947209653092, + "grad_norm": 3.4932067676833958, + "learning_rate": 7.538392652457007e-05, + "loss": 4.621, + "step": 3169 + }, + { + "epoch": 1.1953242835595776, + "grad_norm": 3.5813409987595843, + "learning_rate": 7.537901464189994e-05, + "loss": 4.652, + "step": 3170 + }, + { + "epoch": 1.1957013574660633, + "grad_norm": 2.8955732850305314, + "learning_rate": 7.537410030749546e-05, + "loss": 4.5976, + "step": 3171 + }, + { + "epoch": 1.196078431372549, + "grad_norm": 3.8421729617138003, + "learning_rate": 7.536918352169722e-05, + "loss": 4.6361, + "step": 3172 + }, + { + "epoch": 1.1964555052790347, + "grad_norm": 3.4204522758245455, + "learning_rate": 7.536426428484594e-05, + "loss": 4.6875, + "step": 3173 + }, + { + "epoch": 1.1968325791855203, + "grad_norm": 8.217973138286695, + "learning_rate": 7.535934259728249e-05, + "loss": 4.7916, + "step": 3174 + }, + { + "epoch": 1.197209653092006, + "grad_norm": 3.124995019363496, + "learning_rate": 7.5354418459348e-05, + "loss": 4.601, + "step": 3175 + }, + { + "epoch": 1.1975867269984917, + "grad_norm": 4.055321228006656, + "learning_rate": 7.534949187138363e-05, + "loss": 4.5865, + "step": 3176 + }, + { + "epoch": 1.1979638009049773, + "grad_norm": 2.9594438923208064, + "learning_rate": 7.534456283373084e-05, + "loss": 4.5579, + "step": 3177 + }, + { + "epoch": 1.198340874811463, + "grad_norm": 3.2035966370110245, + "learning_rate": 7.53396313467312e-05, + "loss": 4.5234, + "step": 3178 + }, + { + "epoch": 1.1987179487179487, + "grad_norm": 2.8520664393872472, + "learning_rate": 7.533469741072645e-05, + "loss": 4.5824, + "step": 3179 + }, + { + "epoch": 1.1990950226244343, + "grad_norm": 2.572901396561905, + "learning_rate": 7.532976102605853e-05, + "loss": 4.5309, + "step": 3180 + }, + { + "epoch": 1.19947209653092, + "grad_norm": 2.995971477305802, + "learning_rate": 7.53248221930695e-05, + "loss": 4.5751, + "step": 3181 + }, + { + "epoch": 1.1998491704374057, + "grad_norm": 2.7852099491872915, + "learning_rate": 7.531988091210162e-05, + "loss": 4.4769, + "step": 3182 + }, + { + "epoch": 1.2002262443438914, + "grad_norm": 2.843158398765122, + "learning_rate": 7.531493718349732e-05, + "loss": 4.4545, + "step": 3183 + }, + { + "epoch": 1.200603318250377, + "grad_norm": 2.8685060392381136, + "learning_rate": 7.530999100759919e-05, + "loss": 4.4494, + "step": 3184 + }, + { + "epoch": 1.2009803921568627, + "grad_norm": 2.4486762961858988, + "learning_rate": 7.530504238475002e-05, + "loss": 4.4823, + "step": 3185 + }, + { + "epoch": 1.2013574660633484, + "grad_norm": 1.9842710915037483, + "learning_rate": 7.530009131529272e-05, + "loss": 4.3853, + "step": 3186 + }, + { + "epoch": 1.201734539969834, + "grad_norm": 4.119364436845313, + "learning_rate": 7.52951377995704e-05, + "loss": 4.4775, + "step": 3187 + }, + { + "epoch": 1.2021116138763197, + "grad_norm": 3.0539209156494027, + "learning_rate": 7.529018183792634e-05, + "loss": 4.4313, + "step": 3188 + }, + { + "epoch": 1.2024886877828054, + "grad_norm": 3.381765675729316, + "learning_rate": 7.528522343070399e-05, + "loss": 4.3961, + "step": 3189 + }, + { + "epoch": 1.202865761689291, + "grad_norm": 2.735905791738492, + "learning_rate": 7.528026257824694e-05, + "loss": 4.3716, + "step": 3190 + }, + { + "epoch": 1.2032428355957767, + "grad_norm": 3.317451251504489, + "learning_rate": 7.527529928089899e-05, + "loss": 4.386, + "step": 3191 + }, + { + "epoch": 1.2036199095022624, + "grad_norm": 2.2361081898815027, + "learning_rate": 7.527033353900409e-05, + "loss": 4.2891, + "step": 3192 + }, + { + "epoch": 1.203996983408748, + "grad_norm": 3.315273791487747, + "learning_rate": 7.526536535290635e-05, + "loss": 4.3712, + "step": 3193 + }, + { + "epoch": 1.2043740573152337, + "grad_norm": 2.254539463205753, + "learning_rate": 7.52603947229501e-05, + "loss": 4.2718, + "step": 3194 + }, + { + "epoch": 1.2047511312217194, + "grad_norm": 3.2247010029832324, + "learning_rate": 7.525542164947976e-05, + "loss": 4.2852, + "step": 3195 + }, + { + "epoch": 1.205128205128205, + "grad_norm": 2.704619539156359, + "learning_rate": 7.525044613283997e-05, + "loss": 4.2315, + "step": 3196 + }, + { + "epoch": 1.2055052790346907, + "grad_norm": 3.2345537689579493, + "learning_rate": 7.524546817337552e-05, + "loss": 4.2868, + "step": 3197 + }, + { + "epoch": 1.2058823529411764, + "grad_norm": 2.5207753741551686, + "learning_rate": 7.524048777143139e-05, + "loss": 4.276, + "step": 3198 + }, + { + "epoch": 1.206259426847662, + "grad_norm": 2.499857690845048, + "learning_rate": 7.523550492735272e-05, + "loss": 4.2472, + "step": 3199 + }, + { + "epoch": 1.2066365007541477, + "grad_norm": 2.898430214994682, + "learning_rate": 7.523051964148481e-05, + "loss": 4.2035, + "step": 3200 + }, + { + "epoch": 1.2070135746606334, + "grad_norm": 2.370291918228175, + "learning_rate": 7.522553191417313e-05, + "loss": 4.2228, + "step": 3201 + }, + { + "epoch": 1.207390648567119, + "grad_norm": 2.7307388022351846, + "learning_rate": 7.522054174576335e-05, + "loss": 4.2094, + "step": 3202 + }, + { + "epoch": 1.2077677224736048, + "grad_norm": 3.0223051389915074, + "learning_rate": 7.521554913660127e-05, + "loss": 4.231, + "step": 3203 + }, + { + "epoch": 1.2081447963800904, + "grad_norm": 3.013063050895142, + "learning_rate": 7.521055408703285e-05, + "loss": 4.1719, + "step": 3204 + }, + { + "epoch": 1.208521870286576, + "grad_norm": 1.9249597627602955, + "learning_rate": 7.520555659740428e-05, + "loss": 4.1748, + "step": 3205 + }, + { + "epoch": 1.2088989441930618, + "grad_norm": 2.4998722965478684, + "learning_rate": 7.520055666806186e-05, + "loss": 4.1658, + "step": 3206 + }, + { + "epoch": 1.2092760180995474, + "grad_norm": 2.8686432876812393, + "learning_rate": 7.519555429935209e-05, + "loss": 4.1558, + "step": 3207 + }, + { + "epoch": 1.209653092006033, + "grad_norm": 3.778943893540285, + "learning_rate": 7.519054949162163e-05, + "loss": 4.1536, + "step": 3208 + }, + { + "epoch": 1.2100301659125188, + "grad_norm": 2.6825991051171814, + "learning_rate": 7.51855422452173e-05, + "loss": 4.1076, + "step": 3209 + }, + { + "epoch": 1.2104072398190044, + "grad_norm": 4.16059579696688, + "learning_rate": 7.518053256048609e-05, + "loss": 4.0955, + "step": 3210 + }, + { + "epoch": 1.2107843137254901, + "grad_norm": 2.569002450534991, + "learning_rate": 7.517552043777519e-05, + "loss": 4.06, + "step": 3211 + }, + { + "epoch": 1.2111613876319758, + "grad_norm": 2.2543815132668654, + "learning_rate": 7.517050587743193e-05, + "loss": 4.0774, + "step": 3212 + }, + { + "epoch": 1.2115384615384615, + "grad_norm": 3.5205614932086178, + "learning_rate": 7.51654888798038e-05, + "loss": 4.0641, + "step": 3213 + }, + { + "epoch": 1.2119155354449471, + "grad_norm": 2.679666489533801, + "learning_rate": 7.516046944523848e-05, + "loss": 4.0821, + "step": 3214 + }, + { + "epoch": 1.2122926093514328, + "grad_norm": 2.5649642703643813, + "learning_rate": 7.515544757408381e-05, + "loss": 4.0372, + "step": 3215 + }, + { + "epoch": 1.2126696832579185, + "grad_norm": 2.541359039175253, + "learning_rate": 7.515042326668782e-05, + "loss": 3.9916, + "step": 3216 + }, + { + "epoch": 1.2130467571644041, + "grad_norm": 2.227173712804317, + "learning_rate": 7.514539652339867e-05, + "loss": 3.9717, + "step": 3217 + }, + { + "epoch": 1.2134238310708898, + "grad_norm": 3.2694322247190453, + "learning_rate": 7.514036734456471e-05, + "loss": 4.0398, + "step": 3218 + }, + { + "epoch": 1.2138009049773755, + "grad_norm": 2.7688091513690005, + "learning_rate": 7.513533573053447e-05, + "loss": 4.0128, + "step": 3219 + }, + { + "epoch": 1.2141779788838611, + "grad_norm": 2.197163456521445, + "learning_rate": 7.513030168165662e-05, + "loss": 3.9437, + "step": 3220 + }, + { + "epoch": 1.2145550527903468, + "grad_norm": 3.672946768076899, + "learning_rate": 7.512526519828003e-05, + "loss": 3.9742, + "step": 3221 + }, + { + "epoch": 1.2149321266968327, + "grad_norm": 2.473067832827731, + "learning_rate": 7.512022628075372e-05, + "loss": 3.9772, + "step": 3222 + }, + { + "epoch": 1.2153092006033184, + "grad_norm": 2.779279696960139, + "learning_rate": 7.511518492942688e-05, + "loss": 3.919, + "step": 3223 + }, + { + "epoch": 1.215686274509804, + "grad_norm": 2.787838651117651, + "learning_rate": 7.511014114464887e-05, + "loss": 3.8785, + "step": 3224 + }, + { + "epoch": 1.2160633484162897, + "grad_norm": 2.7536967341024376, + "learning_rate": 7.510509492676922e-05, + "loss": 3.9845, + "step": 3225 + }, + { + "epoch": 1.2164404223227754, + "grad_norm": 2.828646541460513, + "learning_rate": 7.510004627613762e-05, + "loss": 3.9062, + "step": 3226 + }, + { + "epoch": 1.216817496229261, + "grad_norm": 2.5126995384027815, + "learning_rate": 7.509499519310396e-05, + "loss": 3.8673, + "step": 3227 + }, + { + "epoch": 1.2171945701357467, + "grad_norm": 2.568517500312691, + "learning_rate": 7.508994167801826e-05, + "loss": 3.906, + "step": 3228 + }, + { + "epoch": 1.2175716440422324, + "grad_norm": 2.938461775226455, + "learning_rate": 7.508488573123073e-05, + "loss": 3.8763, + "step": 3229 + }, + { + "epoch": 1.217948717948718, + "grad_norm": 3.0151155100112264, + "learning_rate": 7.507982735309175e-05, + "loss": 3.8743, + "step": 3230 + }, + { + "epoch": 1.2183257918552037, + "grad_norm": 2.5945760438027143, + "learning_rate": 7.507476654395183e-05, + "loss": 3.866, + "step": 3231 + }, + { + "epoch": 1.2187028657616894, + "grad_norm": 2.63538837919159, + "learning_rate": 7.50697033041617e-05, + "loss": 3.785, + "step": 3232 + }, + { + "epoch": 1.219079939668175, + "grad_norm": 2.650946206034339, + "learning_rate": 7.506463763407225e-05, + "loss": 3.8683, + "step": 3233 + }, + { + "epoch": 1.2194570135746607, + "grad_norm": 1.9992789603118861, + "learning_rate": 7.50595695340345e-05, + "loss": 3.771, + "step": 3234 + }, + { + "epoch": 1.2198340874811464, + "grad_norm": 3.359637721351007, + "learning_rate": 7.50544990043997e-05, + "loss": 3.7959, + "step": 3235 + }, + { + "epoch": 1.220211161387632, + "grad_norm": 2.4914119054928783, + "learning_rate": 7.504942604551918e-05, + "loss": 3.8234, + "step": 3236 + }, + { + "epoch": 1.2205882352941178, + "grad_norm": 2.6908950543996317, + "learning_rate": 7.504435065774455e-05, + "loss": 3.7922, + "step": 3237 + }, + { + "epoch": 1.2209653092006034, + "grad_norm": 3.0817457583346073, + "learning_rate": 7.50392728414275e-05, + "loss": 3.7801, + "step": 3238 + }, + { + "epoch": 1.221342383107089, + "grad_norm": 2.06401916877314, + "learning_rate": 7.503419259691991e-05, + "loss": 3.7403, + "step": 3239 + }, + { + "epoch": 1.2217194570135748, + "grad_norm": 3.5905471781217373, + "learning_rate": 7.502910992457387e-05, + "loss": 3.7673, + "step": 3240 + }, + { + "epoch": 1.2220965309200604, + "grad_norm": 3.181690259293202, + "learning_rate": 7.502402482474155e-05, + "loss": 3.7363, + "step": 3241 + }, + { + "epoch": 1.222473604826546, + "grad_norm": 2.442969721441257, + "learning_rate": 7.501893729777541e-05, + "loss": 3.728, + "step": 3242 + }, + { + "epoch": 1.2228506787330318, + "grad_norm": 3.6752870164528684, + "learning_rate": 7.501384734402796e-05, + "loss": 3.7057, + "step": 3243 + }, + { + "epoch": 1.2232277526395174, + "grad_norm": 2.898752304545921, + "learning_rate": 7.500875496385194e-05, + "loss": 3.766, + "step": 3244 + }, + { + "epoch": 1.223604826546003, + "grad_norm": 3.830467069931141, + "learning_rate": 7.500366015760027e-05, + "loss": 3.7405, + "step": 3245 + }, + { + "epoch": 1.2239819004524888, + "grad_norm": 1.8414415249322655, + "learning_rate": 7.4998562925626e-05, + "loss": 3.6658, + "step": 3246 + }, + { + "epoch": 1.2243589743589745, + "grad_norm": 5.3544401436652755, + "learning_rate": 7.499346326828234e-05, + "loss": 3.7339, + "step": 3247 + }, + { + "epoch": 1.2247360482654601, + "grad_norm": 2.9761376104728647, + "learning_rate": 7.498836118592273e-05, + "loss": 3.6487, + "step": 3248 + }, + { + "epoch": 1.2251131221719458, + "grad_norm": 3.1235261973294755, + "learning_rate": 7.498325667890073e-05, + "loss": 3.7048, + "step": 3249 + }, + { + "epoch": 1.2254901960784315, + "grad_norm": 2.9264062393005856, + "learning_rate": 7.497814974757007e-05, + "loss": 3.6424, + "step": 3250 + }, + { + "epoch": 1.2258672699849171, + "grad_norm": 3.042989380998319, + "learning_rate": 7.497304039228465e-05, + "loss": 3.651, + "step": 3251 + }, + { + "epoch": 1.2262443438914028, + "grad_norm": 1.9387922855938622, + "learning_rate": 7.496792861339857e-05, + "loss": 3.596, + "step": 3252 + }, + { + "epoch": 1.2266214177978885, + "grad_norm": 3.945162239761862, + "learning_rate": 7.496281441126602e-05, + "loss": 3.6535, + "step": 3253 + }, + { + "epoch": 1.2269984917043741, + "grad_norm": 2.094049209189415, + "learning_rate": 7.495769778624146e-05, + "loss": 3.6243, + "step": 3254 + }, + { + "epoch": 1.2273755656108598, + "grad_norm": 4.168584236371468, + "learning_rate": 7.495257873867947e-05, + "loss": 3.6357, + "step": 3255 + }, + { + "epoch": 1.2277526395173455, + "grad_norm": 3.491464302015669, + "learning_rate": 7.494745726893475e-05, + "loss": 3.5883, + "step": 3256 + }, + { + "epoch": 1.2281297134238311, + "grad_norm": 3.2456832098086497, + "learning_rate": 7.494233337736224e-05, + "loss": 3.5793, + "step": 3257 + }, + { + "epoch": 1.2285067873303168, + "grad_norm": 3.291855083794507, + "learning_rate": 7.493720706431705e-05, + "loss": 3.6304, + "step": 3258 + }, + { + "epoch": 1.2288838612368025, + "grad_norm": 1.9038418356030424, + "learning_rate": 7.493207833015438e-05, + "loss": 3.5643, + "step": 3259 + }, + { + "epoch": 1.2292609351432882, + "grad_norm": 4.4874514953861935, + "learning_rate": 7.492694717522967e-05, + "loss": 3.64, + "step": 3260 + }, + { + "epoch": 1.2296380090497738, + "grad_norm": 3.0549918386608885, + "learning_rate": 7.492181359989848e-05, + "loss": 3.5644, + "step": 3261 + }, + { + "epoch": 1.2300150829562595, + "grad_norm": 3.0848400331851606, + "learning_rate": 7.49166776045166e-05, + "loss": 3.5739, + "step": 3262 + }, + { + "epoch": 1.2303921568627452, + "grad_norm": 3.3579013126469275, + "learning_rate": 7.491153918943994e-05, + "loss": 3.6067, + "step": 3263 + }, + { + "epoch": 1.2307692307692308, + "grad_norm": 3.113807859231381, + "learning_rate": 7.490639835502458e-05, + "loss": 3.5702, + "step": 3264 + }, + { + "epoch": 1.2311463046757165, + "grad_norm": 2.098303666881095, + "learning_rate": 7.490125510162677e-05, + "loss": 3.4744, + "step": 3265 + }, + { + "epoch": 1.2315233785822022, + "grad_norm": 4.144482831269979, + "learning_rate": 7.489610942960295e-05, + "loss": 3.6148, + "step": 3266 + }, + { + "epoch": 1.2319004524886878, + "grad_norm": 2.55037595648224, + "learning_rate": 7.489096133930968e-05, + "loss": 3.529, + "step": 3267 + }, + { + "epoch": 1.2322775263951735, + "grad_norm": 3.205722694459517, + "learning_rate": 7.488581083110374e-05, + "loss": 3.5066, + "step": 3268 + }, + { + "epoch": 1.2326546003016592, + "grad_norm": 2.6762753220807074, + "learning_rate": 7.488065790534206e-05, + "loss": 3.4645, + "step": 3269 + }, + { + "epoch": 1.2330316742081449, + "grad_norm": 3.1704416100640027, + "learning_rate": 7.487550256238173e-05, + "loss": 3.4628, + "step": 3270 + }, + { + "epoch": 1.2334087481146305, + "grad_norm": 3.3899430590311193, + "learning_rate": 7.487034480258001e-05, + "loss": 3.507, + "step": 3271 + }, + { + "epoch": 1.2337858220211162, + "grad_norm": 2.89267188313314, + "learning_rate": 7.486518462629431e-05, + "loss": 3.48, + "step": 3272 + }, + { + "epoch": 1.2341628959276019, + "grad_norm": 2.8792151936260417, + "learning_rate": 7.486002203388226e-05, + "loss": 3.4653, + "step": 3273 + }, + { + "epoch": 1.2345399698340875, + "grad_norm": 2.859118289135484, + "learning_rate": 7.485485702570158e-05, + "loss": 3.4685, + "step": 3274 + }, + { + "epoch": 1.2349170437405732, + "grad_norm": 3.476405491382434, + "learning_rate": 7.484968960211023e-05, + "loss": 3.382, + "step": 3275 + }, + { + "epoch": 1.2352941176470589, + "grad_norm": 2.628967225601228, + "learning_rate": 7.48445197634663e-05, + "loss": 3.4133, + "step": 3276 + }, + { + "epoch": 1.2356711915535445, + "grad_norm": 2.681463113175572, + "learning_rate": 7.483934751012807e-05, + "loss": 3.3623, + "step": 3277 + }, + { + "epoch": 1.2360482654600302, + "grad_norm": 3.0539844808777676, + "learning_rate": 7.483417284245397e-05, + "loss": 3.4221, + "step": 3278 + }, + { + "epoch": 1.2364253393665159, + "grad_norm": 2.641650875363173, + "learning_rate": 7.482899576080257e-05, + "loss": 3.4184, + "step": 3279 + }, + { + "epoch": 1.2368024132730016, + "grad_norm": 3.4253813962886075, + "learning_rate": 7.482381626553266e-05, + "loss": 3.3734, + "step": 3280 + }, + { + "epoch": 1.2371794871794872, + "grad_norm": 2.741044171178224, + "learning_rate": 7.481863435700318e-05, + "loss": 3.3858, + "step": 3281 + }, + { + "epoch": 1.237556561085973, + "grad_norm": 3.438192991103355, + "learning_rate": 7.481345003557321e-05, + "loss": 3.3875, + "step": 3282 + }, + { + "epoch": 1.2379336349924586, + "grad_norm": 2.7980241765670986, + "learning_rate": 7.480826330160204e-05, + "loss": 3.3738, + "step": 3283 + }, + { + "epoch": 1.2383107088989442, + "grad_norm": 2.1828624268676227, + "learning_rate": 7.48030741554491e-05, + "loss": 3.3446, + "step": 3284 + }, + { + "epoch": 1.23868778280543, + "grad_norm": 3.7818438453328342, + "learning_rate": 7.479788259747398e-05, + "loss": 3.4077, + "step": 3285 + }, + { + "epoch": 1.2390648567119156, + "grad_norm": 2.9786142720077757, + "learning_rate": 7.47926886280365e-05, + "loss": 3.3086, + "step": 3286 + }, + { + "epoch": 1.2394419306184012, + "grad_norm": 3.5053687589943077, + "learning_rate": 7.478749224749651e-05, + "loss": 3.2892, + "step": 3287 + }, + { + "epoch": 1.239819004524887, + "grad_norm": 2.0714235909301015, + "learning_rate": 7.478229345621419e-05, + "loss": 3.2847, + "step": 3288 + }, + { + "epoch": 1.2401960784313726, + "grad_norm": 4.171003647361005, + "learning_rate": 7.477709225454978e-05, + "loss": 3.3552, + "step": 3289 + }, + { + "epoch": 1.2405731523378583, + "grad_norm": 3.4264157263097785, + "learning_rate": 7.477188864286371e-05, + "loss": 3.3602, + "step": 3290 + }, + { + "epoch": 1.240950226244344, + "grad_norm": 2.561719564051761, + "learning_rate": 7.476668262151663e-05, + "loss": 3.276, + "step": 3291 + }, + { + "epoch": 1.2413273001508296, + "grad_norm": 3.018867788776277, + "learning_rate": 7.476147419086925e-05, + "loss": 3.2959, + "step": 3292 + }, + { + "epoch": 1.2417043740573153, + "grad_norm": 2.9557641655005855, + "learning_rate": 7.475626335128257e-05, + "loss": 3.3136, + "step": 3293 + }, + { + "epoch": 1.242081447963801, + "grad_norm": 2.954086148079769, + "learning_rate": 7.475105010311765e-05, + "loss": 3.2565, + "step": 3294 + }, + { + "epoch": 1.2424585218702866, + "grad_norm": 2.6936416093123823, + "learning_rate": 7.474583444673579e-05, + "loss": 3.2473, + "step": 3295 + }, + { + "epoch": 1.2428355957767723, + "grad_norm": 4.155366002630867, + "learning_rate": 7.474061638249842e-05, + "loss": 3.3219, + "step": 3296 + }, + { + "epoch": 1.243212669683258, + "grad_norm": 2.097887637004017, + "learning_rate": 7.473539591076714e-05, + "loss": 3.2077, + "step": 3297 + }, + { + "epoch": 1.2435897435897436, + "grad_norm": 3.68277712343178, + "learning_rate": 7.473017303190376e-05, + "loss": 3.2365, + "step": 3298 + }, + { + "epoch": 1.2439668174962293, + "grad_norm": 2.3730085937615732, + "learning_rate": 7.472494774627016e-05, + "loss": 3.2382, + "step": 3299 + }, + { + "epoch": 1.244343891402715, + "grad_norm": 3.114489450591905, + "learning_rate": 7.471972005422851e-05, + "loss": 3.2155, + "step": 3300 + }, + { + "epoch": 1.2447209653092006, + "grad_norm": 2.357315075096837, + "learning_rate": 7.471448995614104e-05, + "loss": 3.1664, + "step": 3301 + }, + { + "epoch": 1.2450980392156863, + "grad_norm": 3.420923996857177, + "learning_rate": 7.470925745237022e-05, + "loss": 3.2364, + "step": 3302 + }, + { + "epoch": 1.245475113122172, + "grad_norm": 3.0370452139954773, + "learning_rate": 7.470402254327864e-05, + "loss": 3.2315, + "step": 3303 + }, + { + "epoch": 1.2458521870286576, + "grad_norm": 3.545363674172321, + "learning_rate": 7.469878522922907e-05, + "loss": 3.2382, + "step": 3304 + }, + { + "epoch": 1.2462292609351433, + "grad_norm": 3.361592149255364, + "learning_rate": 7.469354551058448e-05, + "loss": 3.1895, + "step": 3305 + }, + { + "epoch": 1.246606334841629, + "grad_norm": 2.873610895492555, + "learning_rate": 7.468830338770796e-05, + "loss": 3.172, + "step": 3306 + }, + { + "epoch": 1.2469834087481146, + "grad_norm": 2.073783463009952, + "learning_rate": 7.468305886096276e-05, + "loss": 3.1717, + "step": 3307 + }, + { + "epoch": 1.2473604826546003, + "grad_norm": 3.186245197340889, + "learning_rate": 7.467781193071236e-05, + "loss": 3.1144, + "step": 3308 + }, + { + "epoch": 1.247737556561086, + "grad_norm": 3.8356764173425963, + "learning_rate": 7.467256259732035e-05, + "loss": 3.1612, + "step": 3309 + }, + { + "epoch": 1.2481146304675717, + "grad_norm": 2.321482064630388, + "learning_rate": 7.466731086115051e-05, + "loss": 3.0871, + "step": 3310 + }, + { + "epoch": 1.2484917043740573, + "grad_norm": 3.562684076609389, + "learning_rate": 7.466205672256679e-05, + "loss": 3.1647, + "step": 3311 + }, + { + "epoch": 1.248868778280543, + "grad_norm": 2.8151936052597915, + "learning_rate": 7.465680018193326e-05, + "loss": 3.1288, + "step": 3312 + }, + { + "epoch": 1.2492458521870287, + "grad_norm": 3.368220809044197, + "learning_rate": 7.465154123961423e-05, + "loss": 3.1525, + "step": 3313 + }, + { + "epoch": 1.2496229260935143, + "grad_norm": 2.7836668508242295, + "learning_rate": 7.464627989597413e-05, + "loss": 3.0961, + "step": 3314 + }, + { + "epoch": 1.25, + "grad_norm": 2.7035003141480893, + "learning_rate": 7.464101615137756e-05, + "loss": 3.0791, + "step": 3315 + }, + { + "epoch": 1.2503770739064857, + "grad_norm": 2.4392028569748034, + "learning_rate": 7.463575000618928e-05, + "loss": 3.0733, + "step": 3316 + }, + { + "epoch": 1.2507541478129713, + "grad_norm": 3.217988521815518, + "learning_rate": 7.463048146077427e-05, + "loss": 3.0863, + "step": 3317 + }, + { + "epoch": 1.251131221719457, + "grad_norm": 3.2333098836794845, + "learning_rate": 7.462521051549761e-05, + "loss": 3.0326, + "step": 3318 + }, + { + "epoch": 1.2515082956259427, + "grad_norm": 2.418135962036256, + "learning_rate": 7.461993717072457e-05, + "loss": 3.011, + "step": 3319 + }, + { + "epoch": 1.2518853695324283, + "grad_norm": 3.7308191540266877, + "learning_rate": 7.461466142682058e-05, + "loss": 3.0415, + "step": 3320 + }, + { + "epoch": 1.252262443438914, + "grad_norm": 2.369487045442765, + "learning_rate": 7.460938328415127e-05, + "loss": 3.0245, + "step": 3321 + }, + { + "epoch": 1.2526395173453997, + "grad_norm": 4.635711236051148, + "learning_rate": 7.46041027430824e-05, + "loss": 3.0064, + "step": 3322 + }, + { + "epoch": 1.2530165912518854, + "grad_norm": 3.360664016143216, + "learning_rate": 7.45988198039799e-05, + "loss": 3.0169, + "step": 3323 + }, + { + "epoch": 1.253393665158371, + "grad_norm": 2.5738829393735663, + "learning_rate": 7.459353446720988e-05, + "loss": 2.9725, + "step": 3324 + }, + { + "epoch": 1.2537707390648567, + "grad_norm": 3.5852869482712757, + "learning_rate": 7.45882467331386e-05, + "loss": 2.979, + "step": 3325 + }, + { + "epoch": 1.2541478129713424, + "grad_norm": 3.629727417416808, + "learning_rate": 7.45829566021325e-05, + "loss": 2.9443, + "step": 3326 + }, + { + "epoch": 1.254524886877828, + "grad_norm": 2.2531055422954096, + "learning_rate": 7.457766407455819e-05, + "loss": 2.9522, + "step": 3327 + }, + { + "epoch": 1.2549019607843137, + "grad_norm": 3.3695654760131175, + "learning_rate": 7.457236915078241e-05, + "loss": 2.9604, + "step": 3328 + }, + { + "epoch": 1.2552790346907994, + "grad_norm": 3.710683036616324, + "learning_rate": 7.456707183117215e-05, + "loss": 2.9509, + "step": 3329 + }, + { + "epoch": 1.255656108597285, + "grad_norm": 3.691165461110462, + "learning_rate": 7.456177211609444e-05, + "loss": 2.9292, + "step": 3330 + }, + { + "epoch": 1.2560331825037707, + "grad_norm": 3.0291093794298862, + "learning_rate": 7.455647000591657e-05, + "loss": 2.8923, + "step": 3331 + }, + { + "epoch": 1.2564102564102564, + "grad_norm": 4.04546845676482, + "learning_rate": 7.455116550100602e-05, + "loss": 2.8792, + "step": 3332 + }, + { + "epoch": 1.256787330316742, + "grad_norm": 2.2287589807792396, + "learning_rate": 7.454585860173032e-05, + "loss": 2.8522, + "step": 3333 + }, + { + "epoch": 1.2571644042232277, + "grad_norm": 3.979866545144704, + "learning_rate": 7.454054930845727e-05, + "loss": 2.871, + "step": 3334 + }, + { + "epoch": 1.2575414781297134, + "grad_norm": 3.0552202762046434, + "learning_rate": 7.453523762155479e-05, + "loss": 2.8779, + "step": 3335 + }, + { + "epoch": 1.257918552036199, + "grad_norm": 3.15141756038997, + "learning_rate": 7.452992354139098e-05, + "loss": 2.8603, + "step": 3336 + }, + { + "epoch": 1.2582956259426847, + "grad_norm": 2.9025217705709743, + "learning_rate": 7.452460706833409e-05, + "loss": 2.8203, + "step": 3337 + }, + { + "epoch": 1.2586726998491704, + "grad_norm": 3.6243890994844508, + "learning_rate": 7.451928820275255e-05, + "loss": 2.7644, + "step": 3338 + }, + { + "epoch": 1.259049773755656, + "grad_norm": 2.6720255051744326, + "learning_rate": 7.451396694501497e-05, + "loss": 2.7415, + "step": 3339 + }, + { + "epoch": 1.2594268476621417, + "grad_norm": 3.637499254704771, + "learning_rate": 7.450864329549008e-05, + "loss": 2.766, + "step": 3340 + }, + { + "epoch": 1.2598039215686274, + "grad_norm": 4.382987542053495, + "learning_rate": 7.450331725454683e-05, + "loss": 2.7868, + "step": 3341 + }, + { + "epoch": 1.260180995475113, + "grad_norm": 2.7269061058014907, + "learning_rate": 7.449798882255428e-05, + "loss": 2.7229, + "step": 3342 + }, + { + "epoch": 1.2605580693815988, + "grad_norm": 3.31524500794433, + "learning_rate": 7.449265799988172e-05, + "loss": 2.7758, + "step": 3343 + }, + { + "epoch": 1.2609351432880844, + "grad_norm": 2.4206256714604657, + "learning_rate": 7.448732478689855e-05, + "loss": 2.7287, + "step": 3344 + }, + { + "epoch": 1.26131221719457, + "grad_norm": 3.4919671856238406, + "learning_rate": 7.448198918397436e-05, + "loss": 2.7008, + "step": 3345 + }, + { + "epoch": 1.2616892911010558, + "grad_norm": 3.5673346708901206, + "learning_rate": 7.447665119147891e-05, + "loss": 2.6526, + "step": 3346 + }, + { + "epoch": 1.2620663650075414, + "grad_norm": 2.65400409569768, + "learning_rate": 7.447131080978213e-05, + "loss": 2.6881, + "step": 3347 + }, + { + "epoch": 1.262443438914027, + "grad_norm": 4.477493174780845, + "learning_rate": 7.446596803925406e-05, + "loss": 2.6492, + "step": 3348 + }, + { + "epoch": 1.2628205128205128, + "grad_norm": 2.87996570294853, + "learning_rate": 7.446062288026499e-05, + "loss": 2.6839, + "step": 3349 + }, + { + "epoch": 1.2631975867269984, + "grad_norm": 2.8421314360541574, + "learning_rate": 7.445527533318532e-05, + "loss": 2.6156, + "step": 3350 + }, + { + "epoch": 1.2635746606334841, + "grad_norm": 2.362831491636535, + "learning_rate": 7.444992539838562e-05, + "loss": 2.5812, + "step": 3351 + }, + { + "epoch": 1.2639517345399698, + "grad_norm": 3.9666551606431675, + "learning_rate": 7.444457307623666e-05, + "loss": 2.6204, + "step": 3352 + }, + { + "epoch": 1.2643288084464555, + "grad_norm": 3.8928240298563472, + "learning_rate": 7.443921836710933e-05, + "loss": 2.5616, + "step": 3353 + }, + { + "epoch": 1.2647058823529411, + "grad_norm": 5.304401140423463, + "learning_rate": 7.443386127137472e-05, + "loss": 2.5261, + "step": 3354 + }, + { + "epoch": 1.2650829562594268, + "grad_norm": 1.8804896648958958, + "learning_rate": 7.442850178940407e-05, + "loss": 2.4886, + "step": 3355 + }, + { + "epoch": 1.2654600301659125, + "grad_norm": 3.6201930618641445, + "learning_rate": 7.442313992156877e-05, + "loss": 2.4944, + "step": 3356 + }, + { + "epoch": 1.2658371040723981, + "grad_norm": 3.062120066948204, + "learning_rate": 7.44177756682404e-05, + "loss": 2.5266, + "step": 3357 + }, + { + "epoch": 1.2662141779788838, + "grad_norm": 2.824920300482361, + "learning_rate": 7.441240902979074e-05, + "loss": 2.4463, + "step": 3358 + }, + { + "epoch": 1.2665912518853695, + "grad_norm": 2.83241552694284, + "learning_rate": 7.440704000659163e-05, + "loss": 2.4193, + "step": 3359 + }, + { + "epoch": 1.2669683257918551, + "grad_norm": 2.2861196670802215, + "learning_rate": 7.440166859901518e-05, + "loss": 2.4293, + "step": 3360 + }, + { + "epoch": 1.2673453996983408, + "grad_norm": 3.0089175418679623, + "learning_rate": 7.43962948074336e-05, + "loss": 2.4291, + "step": 3361 + }, + { + "epoch": 1.2677224736048265, + "grad_norm": 2.7317970415656947, + "learning_rate": 7.439091863221931e-05, + "loss": 2.395, + "step": 3362 + }, + { + "epoch": 1.2680995475113122, + "grad_norm": 3.5107930512729015, + "learning_rate": 7.438554007374485e-05, + "loss": 2.3812, + "step": 3363 + }, + { + "epoch": 1.2684766214177978, + "grad_norm": 2.6396496428310785, + "learning_rate": 7.438015913238299e-05, + "loss": 2.3451, + "step": 3364 + }, + { + "epoch": 1.2688536953242835, + "grad_norm": 3.2429956016824697, + "learning_rate": 7.437477580850656e-05, + "loss": 2.323, + "step": 3365 + }, + { + "epoch": 1.2692307692307692, + "grad_norm": 2.512742318997763, + "learning_rate": 7.436939010248868e-05, + "loss": 2.3206, + "step": 3366 + }, + { + "epoch": 1.2696078431372548, + "grad_norm": 2.3708784024153076, + "learning_rate": 7.436400201470254e-05, + "loss": 2.2857, + "step": 3367 + }, + { + "epoch": 1.2699849170437405, + "grad_norm": 3.721387654971352, + "learning_rate": 7.435861154552155e-05, + "loss": 2.269, + "step": 3368 + }, + { + "epoch": 1.2703619909502262, + "grad_norm": 2.388962266953672, + "learning_rate": 7.435321869531926e-05, + "loss": 2.2434, + "step": 3369 + }, + { + "epoch": 1.2707390648567118, + "grad_norm": 3.622192072776817, + "learning_rate": 7.434782346446938e-05, + "loss": 2.2547, + "step": 3370 + }, + { + "epoch": 1.2711161387631975, + "grad_norm": 3.0827984801201955, + "learning_rate": 7.434242585334579e-05, + "loss": 2.2424, + "step": 3371 + }, + { + "epoch": 1.2714932126696832, + "grad_norm": 2.38563417249682, + "learning_rate": 7.433702586232257e-05, + "loss": 2.1707, + "step": 3372 + }, + { + "epoch": 1.2718702865761689, + "grad_norm": 3.3899116383048926, + "learning_rate": 7.43316234917739e-05, + "loss": 2.2322, + "step": 3373 + }, + { + "epoch": 1.2722473604826545, + "grad_norm": 2.3456583733801977, + "learning_rate": 7.432621874207417e-05, + "loss": 2.1515, + "step": 3374 + }, + { + "epoch": 1.2726244343891402, + "grad_norm": 1.968731435917969, + "learning_rate": 7.432081161359794e-05, + "loss": 2.1335, + "step": 3375 + }, + { + "epoch": 1.2730015082956259, + "grad_norm": 3.3074432545059596, + "learning_rate": 7.43154021067199e-05, + "loss": 2.1236, + "step": 3376 + }, + { + "epoch": 1.2733785822021115, + "grad_norm": 1.9188533197118105, + "learning_rate": 7.430999022181494e-05, + "loss": 2.1059, + "step": 3377 + }, + { + "epoch": 1.2737556561085972, + "grad_norm": 2.9542278460140836, + "learning_rate": 7.43045759592581e-05, + "loss": 2.1238, + "step": 3378 + }, + { + "epoch": 1.2741327300150829, + "grad_norm": 2.8689857696367005, + "learning_rate": 7.429915931942456e-05, + "loss": 2.0909, + "step": 3379 + }, + { + "epoch": 1.2745098039215685, + "grad_norm": 3.3711606931166878, + "learning_rate": 7.429374030268969e-05, + "loss": 2.1723, + "step": 3380 + }, + { + "epoch": 1.2748868778280542, + "grad_norm": 2.6443605876585496, + "learning_rate": 7.428831890942906e-05, + "loss": 2.0368, + "step": 3381 + }, + { + "epoch": 1.2752639517345399, + "grad_norm": 2.0995835604359154, + "learning_rate": 7.428289514001833e-05, + "loss": 1.9919, + "step": 3382 + }, + { + "epoch": 1.2756410256410255, + "grad_norm": 2.6332562415136995, + "learning_rate": 7.427746899483337e-05, + "loss": 1.975, + "step": 3383 + }, + { + "epoch": 1.2760180995475112, + "grad_norm": 3.3040572848118006, + "learning_rate": 7.427204047425022e-05, + "loss": 1.9801, + "step": 3384 + }, + { + "epoch": 1.276395173453997, + "grad_norm": 2.2111544351850356, + "learning_rate": 7.426660957864507e-05, + "loss": 1.9179, + "step": 3385 + }, + { + "epoch": 1.2767722473604826, + "grad_norm": 3.4324940982518357, + "learning_rate": 7.426117630839427e-05, + "loss": 1.9601, + "step": 3386 + }, + { + "epoch": 1.2771493212669682, + "grad_norm": 1.7031455859669515, + "learning_rate": 7.425574066387433e-05, + "loss": 1.8951, + "step": 3387 + }, + { + "epoch": 1.277526395173454, + "grad_norm": 3.565265845921055, + "learning_rate": 7.425030264546197e-05, + "loss": 1.9206, + "step": 3388 + }, + { + "epoch": 1.2779034690799396, + "grad_norm": 2.880062691525566, + "learning_rate": 7.424486225353398e-05, + "loss": 1.8631, + "step": 3389 + }, + { + "epoch": 1.2782805429864252, + "grad_norm": 2.6930380950413317, + "learning_rate": 7.423941948846745e-05, + "loss": 1.8466, + "step": 3390 + }, + { + "epoch": 1.278657616892911, + "grad_norm": 2.6709166659443375, + "learning_rate": 7.42339743506395e-05, + "loss": 1.7993, + "step": 3391 + }, + { + "epoch": 1.2790346907993966, + "grad_norm": 2.1673257740504495, + "learning_rate": 7.42285268404275e-05, + "loss": 1.7676, + "step": 3392 + }, + { + "epoch": 1.2794117647058822, + "grad_norm": 3.1915823518094015, + "learning_rate": 7.422307695820893e-05, + "loss": 1.832, + "step": 3393 + }, + { + "epoch": 1.279788838612368, + "grad_norm": 3.9305197219729404, + "learning_rate": 7.421762470436151e-05, + "loss": 1.8783, + "step": 3394 + }, + { + "epoch": 1.2801659125188536, + "grad_norm": 2.0278068979158057, + "learning_rate": 7.421217007926302e-05, + "loss": 1.7447, + "step": 3395 + }, + { + "epoch": 1.2805429864253393, + "grad_norm": 2.6240079591149486, + "learning_rate": 7.420671308329151e-05, + "loss": 1.7592, + "step": 3396 + }, + { + "epoch": 1.280920060331825, + "grad_norm": 3.0468117217775577, + "learning_rate": 7.420125371682513e-05, + "loss": 1.742, + "step": 3397 + }, + { + "epoch": 1.2812971342383106, + "grad_norm": 2.1371374834339116, + "learning_rate": 7.419579198024219e-05, + "loss": 1.6343, + "step": 3398 + }, + { + "epoch": 1.2816742081447963, + "grad_norm": 4.240588870119116, + "learning_rate": 7.41903278739212e-05, + "loss": 1.7554, + "step": 3399 + }, + { + "epoch": 1.282051282051282, + "grad_norm": 2.7216346116650962, + "learning_rate": 7.418486139824082e-05, + "loss": 1.6396, + "step": 3400 + }, + { + "epoch": 1.2824283559577676, + "grad_norm": 3.219205071793825, + "learning_rate": 7.417939255357986e-05, + "loss": 1.6154, + "step": 3401 + }, + { + "epoch": 1.2828054298642533, + "grad_norm": 2.476799730218292, + "learning_rate": 7.417392134031732e-05, + "loss": 1.5721, + "step": 3402 + }, + { + "epoch": 1.283182503770739, + "grad_norm": 2.665759655435502, + "learning_rate": 7.416844775883234e-05, + "loss": 1.5173, + "step": 3403 + }, + { + "epoch": 1.2835595776772246, + "grad_norm": 3.958541826630898, + "learning_rate": 7.416297180950425e-05, + "loss": 1.4881, + "step": 3404 + }, + { + "epoch": 1.2839366515837103, + "grad_norm": 2.8127061731184604, + "learning_rate": 7.41574934927125e-05, + "loss": 1.4086, + "step": 3405 + }, + { + "epoch": 1.284313725490196, + "grad_norm": 2.5718328566706634, + "learning_rate": 7.415201280883675e-05, + "loss": 1.3161, + "step": 3406 + }, + { + "epoch": 1.2846907993966816, + "grad_norm": 4.612379915509242, + "learning_rate": 7.414652975825681e-05, + "loss": 1.3422, + "step": 3407 + }, + { + "epoch": 1.2850678733031673, + "grad_norm": 4.372855542549734, + "learning_rate": 7.414104434135265e-05, + "loss": 1.366, + "step": 3408 + }, + { + "epoch": 1.285444947209653, + "grad_norm": 3.865082808437602, + "learning_rate": 7.413555655850438e-05, + "loss": 1.3188, + "step": 3409 + }, + { + "epoch": 1.2858220211161386, + "grad_norm": 2.2046031161097113, + "learning_rate": 7.413006641009235e-05, + "loss": 1.2028, + "step": 3410 + }, + { + "epoch": 1.2861990950226243, + "grad_norm": 1.8874628596182292, + "learning_rate": 7.412457389649696e-05, + "loss": 1.0973, + "step": 3411 + }, + { + "epoch": 1.28657616892911, + "grad_norm": 4.207486659636661, + "learning_rate": 7.41190790180989e-05, + "loss": 1.1571, + "step": 3412 + }, + { + "epoch": 1.2869532428355956, + "grad_norm": 24.003597827741164, + "learning_rate": 7.41135817752789e-05, + "loss": 1.4307, + "step": 3413 + }, + { + "epoch": 1.2873303167420813, + "grad_norm": 9.023788318814665, + "learning_rate": 7.410808216841795e-05, + "loss": 1.5178, + "step": 3414 + }, + { + "epoch": 1.287707390648567, + "grad_norm": 3.6503874958368168, + "learning_rate": 7.410258019789716e-05, + "loss": 1.3344, + "step": 3415 + }, + { + "epoch": 1.2880844645550527, + "grad_norm": 2.848713378655892, + "learning_rate": 7.40970758640978e-05, + "loss": 1.1733, + "step": 3416 + }, + { + "epoch": 1.2884615384615383, + "grad_norm": 2.3600095324170014, + "learning_rate": 7.409156916740134e-05, + "loss": 1.0752, + "step": 3417 + }, + { + "epoch": 1.288838612368024, + "grad_norm": 1.8062694151654357, + "learning_rate": 7.408606010818938e-05, + "loss": 0.9963, + "step": 3418 + }, + { + "epoch": 1.2892156862745099, + "grad_norm": 3.5457541565600392, + "learning_rate": 7.408054868684369e-05, + "loss": 1.1141, + "step": 3419 + }, + { + "epoch": 1.2895927601809956, + "grad_norm": 1.7861206851751974, + "learning_rate": 7.40750349037462e-05, + "loss": 0.9493, + "step": 3420 + }, + { + "epoch": 1.2899698340874812, + "grad_norm": 1.1266001492435018, + "learning_rate": 7.406951875927901e-05, + "loss": 0.911, + "step": 3421 + }, + { + "epoch": 1.290346907993967, + "grad_norm": 87.26325594655773, + "learning_rate": 7.40640002538244e-05, + "loss": 1.9327, + "step": 3422 + }, + { + "epoch": 1.2907239819004526, + "grad_norm": 3.8483375833711118, + "learning_rate": 7.405847938776477e-05, + "loss": 1.1344, + "step": 3423 + }, + { + "epoch": 1.2911010558069382, + "grad_norm": 1.902000832672468, + "learning_rate": 7.405295616148276e-05, + "loss": 0.9516, + "step": 3424 + }, + { + "epoch": 1.291478129713424, + "grad_norm": 2.5054944397888046, + "learning_rate": 7.404743057536106e-05, + "loss": 0.9514, + "step": 3425 + }, + { + "epoch": 1.2918552036199096, + "grad_norm": 1.611413788474628, + "learning_rate": 7.404190262978262e-05, + "loss": 0.8851, + "step": 3426 + }, + { + "epoch": 1.2922322775263952, + "grad_norm": 1.0968134153237599, + "learning_rate": 7.403637232513055e-05, + "loss": 0.8201, + "step": 3427 + }, + { + "epoch": 1.292609351432881, + "grad_norm": 4.285881817753642, + "learning_rate": 7.403083966178805e-05, + "loss": 1.0259, + "step": 3428 + }, + { + "epoch": 1.2929864253393666, + "grad_norm": 1.595742226509383, + "learning_rate": 7.402530464013855e-05, + "loss": 0.8276, + "step": 3429 + }, + { + "epoch": 1.2933634992458523, + "grad_norm": 0.831776027694026, + "learning_rate": 7.401976726056563e-05, + "loss": 0.7635, + "step": 3430 + }, + { + "epoch": 1.293740573152338, + "grad_norm": 2.5570972617265153, + "learning_rate": 7.4014227523453e-05, + "loss": 0.8018, + "step": 3431 + }, + { + "epoch": 1.2941176470588236, + "grad_norm": 5.8162991727772955, + "learning_rate": 7.400868542918457e-05, + "loss": 1.024, + "step": 3432 + }, + { + "epoch": 1.2944947209653093, + "grad_norm": 1.4251295574519232, + "learning_rate": 7.400314097814442e-05, + "loss": 0.7864, + "step": 3433 + }, + { + "epoch": 1.294871794871795, + "grad_norm": 1.0764793768432226, + "learning_rate": 7.399759417071675e-05, + "loss": 0.7482, + "step": 3434 + }, + { + "epoch": 1.2952488687782806, + "grad_norm": 1.3512433715712724, + "learning_rate": 7.399204500728596e-05, + "loss": 0.7356, + "step": 3435 + }, + { + "epoch": 1.2956259426847663, + "grad_norm": 1.503134572513461, + "learning_rate": 7.398649348823661e-05, + "loss": 0.6973, + "step": 3436 + }, + { + "epoch": 1.296003016591252, + "grad_norm": 1.4945255533348976, + "learning_rate": 7.39809396139534e-05, + "loss": 0.7233, + "step": 3437 + }, + { + "epoch": 1.2963800904977376, + "grad_norm": 0.9990453938972537, + "learning_rate": 7.397538338482121e-05, + "loss": 0.7039, + "step": 3438 + }, + { + "epoch": 1.2967571644042233, + "grad_norm": 1.0683744149797505, + "learning_rate": 7.39698248012251e-05, + "loss": 0.6915, + "step": 3439 + }, + { + "epoch": 1.297134238310709, + "grad_norm": 0.9071263421752871, + "learning_rate": 7.396426386355027e-05, + "loss": 0.652, + "step": 3440 + }, + { + "epoch": 1.2975113122171946, + "grad_norm": 0.7535026582991622, + "learning_rate": 7.395870057218207e-05, + "loss": 0.6559, + "step": 3441 + }, + { + "epoch": 1.2978883861236803, + "grad_norm": 0.7388312947699678, + "learning_rate": 7.395313492750603e-05, + "loss": 0.6152, + "step": 3442 + }, + { + "epoch": 1.298265460030166, + "grad_norm": 0.9038219169538518, + "learning_rate": 7.394756692990787e-05, + "loss": 0.615, + "step": 3443 + }, + { + "epoch": 1.2986425339366516, + "grad_norm": 1.1189885964850224, + "learning_rate": 7.394199657977343e-05, + "loss": 0.6086, + "step": 3444 + }, + { + "epoch": 1.2990196078431373, + "grad_norm": 1.8988437761792372, + "learning_rate": 7.393642387748874e-05, + "loss": 0.6659, + "step": 3445 + }, + { + "epoch": 1.299396681749623, + "grad_norm": 1.1239399337138276, + "learning_rate": 7.393084882343997e-05, + "loss": 0.6352, + "step": 3446 + }, + { + "epoch": 1.2997737556561086, + "grad_norm": 0.8738992448287783, + "learning_rate": 7.392527141801348e-05, + "loss": 0.6253, + "step": 3447 + }, + { + "epoch": 1.3001508295625943, + "grad_norm": 0.9820117196995841, + "learning_rate": 7.391969166159575e-05, + "loss": 0.5876, + "step": 3448 + }, + { + "epoch": 1.30052790346908, + "grad_norm": 0.895720233964195, + "learning_rate": 7.39141095545735e-05, + "loss": 0.5905, + "step": 3449 + }, + { + "epoch": 1.3009049773755657, + "grad_norm": 3.893497831746255, + "learning_rate": 7.390852509733353e-05, + "loss": 0.733, + "step": 3450 + }, + { + "epoch": 1.3012820512820513, + "grad_norm": 1.3171459847327431, + "learning_rate": 7.390293829026284e-05, + "loss": 0.6101, + "step": 3451 + }, + { + "epoch": 1.301659125188537, + "grad_norm": 0.6093459163972271, + "learning_rate": 7.389734913374862e-05, + "loss": 0.6004, + "step": 3452 + }, + { + "epoch": 1.3020361990950227, + "grad_norm": 0.7449392281046087, + "learning_rate": 7.389175762817814e-05, + "loss": 0.5901, + "step": 3453 + }, + { + "epoch": 1.3024132730015083, + "grad_norm": 0.8951411222351773, + "learning_rate": 7.388616377393895e-05, + "loss": 0.5708, + "step": 3454 + }, + { + "epoch": 1.302790346907994, + "grad_norm": 0.8671075848099355, + "learning_rate": 7.388056757141865e-05, + "loss": 0.5708, + "step": 3455 + }, + { + "epoch": 1.3031674208144797, + "grad_norm": 0.6644638316604261, + "learning_rate": 7.387496902100507e-05, + "loss": 0.581, + "step": 3456 + }, + { + "epoch": 1.3035444947209653, + "grad_norm": 0.5047216121748112, + "learning_rate": 7.38693681230862e-05, + "loss": 0.5787, + "step": 3457 + }, + { + "epoch": 1.303921568627451, + "grad_norm": 0.5805998655320512, + "learning_rate": 7.386376487805013e-05, + "loss": 0.5805, + "step": 3458 + }, + { + "epoch": 1.3042986425339367, + "grad_norm": 0.5340632976981387, + "learning_rate": 7.385815928628521e-05, + "loss": 0.5501, + "step": 3459 + }, + { + "epoch": 1.3046757164404224, + "grad_norm": 0.4856532794249434, + "learning_rate": 7.385255134817989e-05, + "loss": 0.5393, + "step": 3460 + }, + { + "epoch": 1.305052790346908, + "grad_norm": 0.4427757888515942, + "learning_rate": 7.384694106412278e-05, + "loss": 0.5377, + "step": 3461 + }, + { + "epoch": 1.3054298642533937, + "grad_norm": 0.4923786532531996, + "learning_rate": 7.384132843450267e-05, + "loss": 0.5297, + "step": 3462 + }, + { + "epoch": 1.3058069381598794, + "grad_norm": 0.5816209721199326, + "learning_rate": 7.383571345970855e-05, + "loss": 0.5196, + "step": 3463 + }, + { + "epoch": 1.306184012066365, + "grad_norm": 1.029919985051297, + "learning_rate": 7.383009614012946e-05, + "loss": 0.5493, + "step": 3464 + }, + { + "epoch": 1.3065610859728507, + "grad_norm": 0.7703844090320592, + "learning_rate": 7.382447647615473e-05, + "loss": 0.5562, + "step": 3465 + }, + { + "epoch": 1.3069381598793364, + "grad_norm": 0.6239860799268636, + "learning_rate": 7.381885446817379e-05, + "loss": 0.5417, + "step": 3466 + }, + { + "epoch": 1.307315233785822, + "grad_norm": 0.5226791370555817, + "learning_rate": 7.381323011657622e-05, + "loss": 0.5374, + "step": 3467 + }, + { + "epoch": 1.3076923076923077, + "grad_norm": 0.5336645818411144, + "learning_rate": 7.38076034217518e-05, + "loss": 0.5353, + "step": 3468 + }, + { + "epoch": 1.3080693815987934, + "grad_norm": 0.6396928572090511, + "learning_rate": 7.380197438409044e-05, + "loss": 0.5283, + "step": 3469 + }, + { + "epoch": 1.308446455505279, + "grad_norm": 0.4866820344504698, + "learning_rate": 7.379634300398226e-05, + "loss": 0.5291, + "step": 3470 + }, + { + "epoch": 1.3088235294117647, + "grad_norm": 0.5748391640715327, + "learning_rate": 7.379070928181747e-05, + "loss": 0.5084, + "step": 3471 + }, + { + "epoch": 1.3092006033182504, + "grad_norm": 0.665737582055332, + "learning_rate": 7.37850732179865e-05, + "loss": 0.5172, + "step": 3472 + }, + { + "epoch": 1.309577677224736, + "grad_norm": 0.5691807657378426, + "learning_rate": 7.377943481287993e-05, + "loss": 0.5045, + "step": 3473 + }, + { + "epoch": 1.3099547511312217, + "grad_norm": 0.4928780195038956, + "learning_rate": 7.377379406688848e-05, + "loss": 0.5299, + "step": 3474 + }, + { + "epoch": 1.3103318250377074, + "grad_norm": 0.4225694260220585, + "learning_rate": 7.376815098040308e-05, + "loss": 0.4978, + "step": 3475 + }, + { + "epoch": 1.310708898944193, + "grad_norm": 0.5864891985967103, + "learning_rate": 7.376250555381474e-05, + "loss": 0.5012, + "step": 3476 + }, + { + "epoch": 1.3110859728506787, + "grad_norm": 0.3573583463488866, + "learning_rate": 7.375685778751474e-05, + "loss": 0.4992, + "step": 3477 + }, + { + "epoch": 1.3114630467571644, + "grad_norm": 0.5584638226541745, + "learning_rate": 7.375120768189442e-05, + "loss": 0.4933, + "step": 3478 + }, + { + "epoch": 1.31184012066365, + "grad_norm": 0.40059464093002956, + "learning_rate": 7.374555523734535e-05, + "loss": 0.4904, + "step": 3479 + }, + { + "epoch": 1.3122171945701357, + "grad_norm": 0.427400029223993, + "learning_rate": 7.373990045425922e-05, + "loss": 0.4891, + "step": 3480 + }, + { + "epoch": 1.3125942684766214, + "grad_norm": 0.43625524797096443, + "learning_rate": 7.373424333302794e-05, + "loss": 0.4903, + "step": 3481 + }, + { + "epoch": 1.312971342383107, + "grad_norm": 0.30999222791317343, + "learning_rate": 7.372858387404351e-05, + "loss": 0.4815, + "step": 3482 + }, + { + "epoch": 1.3133484162895928, + "grad_norm": 0.354600465619087, + "learning_rate": 7.372292207769813e-05, + "loss": 0.4875, + "step": 3483 + }, + { + "epoch": 1.3137254901960784, + "grad_norm": 0.323769079901827, + "learning_rate": 7.371725794438417e-05, + "loss": 0.5047, + "step": 3484 + }, + { + "epoch": 1.314102564102564, + "grad_norm": 0.37699320355472205, + "learning_rate": 7.371159147449414e-05, + "loss": 0.4766, + "step": 3485 + }, + { + "epoch": 1.3144796380090498, + "grad_norm": 0.4940884206045071, + "learning_rate": 7.370592266842071e-05, + "loss": 0.491, + "step": 3486 + }, + { + "epoch": 1.3148567119155354, + "grad_norm": 0.3648199903673323, + "learning_rate": 7.370025152655676e-05, + "loss": 0.4855, + "step": 3487 + }, + { + "epoch": 1.315233785822021, + "grad_norm": 0.5256157597034787, + "learning_rate": 7.369457804929525e-05, + "loss": 0.4845, + "step": 3488 + }, + { + "epoch": 1.3156108597285068, + "grad_norm": 0.7066068535431668, + "learning_rate": 7.368890223702938e-05, + "loss": 0.4862, + "step": 3489 + }, + { + "epoch": 1.3159879336349924, + "grad_norm": 0.3507981954121831, + "learning_rate": 7.368322409015246e-05, + "loss": 0.4849, + "step": 3490 + }, + { + "epoch": 1.3163650075414781, + "grad_norm": 0.542735948216231, + "learning_rate": 7.367754360905799e-05, + "loss": 0.4834, + "step": 3491 + }, + { + "epoch": 1.3167420814479638, + "grad_norm": 0.49584217057020785, + "learning_rate": 7.367186079413963e-05, + "loss": 0.4764, + "step": 3492 + }, + { + "epoch": 1.3171191553544495, + "grad_norm": 0.43286229604963783, + "learning_rate": 7.366617564579116e-05, + "loss": 0.4804, + "step": 3493 + }, + { + "epoch": 1.3174962292609351, + "grad_norm": 0.3947050131573957, + "learning_rate": 7.36604881644066e-05, + "loss": 0.4786, + "step": 3494 + }, + { + "epoch": 1.3178733031674208, + "grad_norm": 0.404676375238421, + "learning_rate": 7.365479835038006e-05, + "loss": 0.4926, + "step": 3495 + }, + { + "epoch": 1.3182503770739065, + "grad_norm": 0.38826268935348157, + "learning_rate": 7.364910620410584e-05, + "loss": 0.4707, + "step": 3496 + }, + { + "epoch": 1.3186274509803921, + "grad_norm": 0.36789972314984887, + "learning_rate": 7.364341172597842e-05, + "loss": 0.4823, + "step": 3497 + }, + { + "epoch": 1.3190045248868778, + "grad_norm": 0.43214901970803793, + "learning_rate": 7.36377149163924e-05, + "loss": 0.4808, + "step": 3498 + }, + { + "epoch": 1.3193815987933635, + "grad_norm": 0.3189200955361944, + "learning_rate": 7.363201577574257e-05, + "loss": 0.4662, + "step": 3499 + }, + { + "epoch": 1.3197586726998491, + "grad_norm": 0.4725106440293564, + "learning_rate": 7.36263143044239e-05, + "loss": 0.4669, + "step": 3500 + }, + { + "epoch": 1.3201357466063348, + "grad_norm": 0.3244629859593273, + "learning_rate": 7.362061050283145e-05, + "loss": 0.4668, + "step": 3501 + }, + { + "epoch": 1.3205128205128205, + "grad_norm": 0.4705452828543701, + "learning_rate": 7.361490437136052e-05, + "loss": 0.4727, + "step": 3502 + }, + { + "epoch": 1.3208898944193062, + "grad_norm": 0.369933281632765, + "learning_rate": 7.360919591040654e-05, + "loss": 0.4758, + "step": 3503 + }, + { + "epoch": 1.3212669683257918, + "grad_norm": 0.38572038448625734, + "learning_rate": 7.36034851203651e-05, + "loss": 0.4658, + "step": 3504 + }, + { + "epoch": 1.3216440422322775, + "grad_norm": 0.3766622763024366, + "learning_rate": 7.359777200163193e-05, + "loss": 0.4644, + "step": 3505 + }, + { + "epoch": 1.3220211161387632, + "grad_norm": 0.2983719322458127, + "learning_rate": 7.359205655460297e-05, + "loss": 0.4718, + "step": 3506 + }, + { + "epoch": 1.3223981900452488, + "grad_norm": 0.39723751167295535, + "learning_rate": 7.358633877967429e-05, + "loss": 0.4684, + "step": 3507 + }, + { + "epoch": 1.3227752639517345, + "grad_norm": 0.2873271698640366, + "learning_rate": 7.358061867724212e-05, + "loss": 0.4631, + "step": 3508 + }, + { + "epoch": 1.3231523378582202, + "grad_norm": 0.29850021634152807, + "learning_rate": 7.357489624770285e-05, + "loss": 0.4649, + "step": 3509 + }, + { + "epoch": 1.3235294117647058, + "grad_norm": 0.2964282446091471, + "learning_rate": 7.356917149145308e-05, + "loss": 0.4589, + "step": 3510 + }, + { + "epoch": 1.3239064856711915, + "grad_norm": 0.22996344110490197, + "learning_rate": 7.356344440888948e-05, + "loss": 0.4624, + "step": 3511 + }, + { + "epoch": 1.3242835595776772, + "grad_norm": 0.2416420804121888, + "learning_rate": 7.355771500040895e-05, + "loss": 0.4591, + "step": 3512 + }, + { + "epoch": 1.3246606334841629, + "grad_norm": 0.23356480182714032, + "learning_rate": 7.355198326640855e-05, + "loss": 0.4576, + "step": 3513 + }, + { + "epoch": 1.3250377073906485, + "grad_norm": 0.19519268627065742, + "learning_rate": 7.354624920728545e-05, + "loss": 0.4627, + "step": 3514 + }, + { + "epoch": 1.3254147812971342, + "grad_norm": 0.249493659906306, + "learning_rate": 7.354051282343707e-05, + "loss": 0.465, + "step": 3515 + }, + { + "epoch": 1.3257918552036199, + "grad_norm": 0.2277528493170229, + "learning_rate": 7.353477411526087e-05, + "loss": 0.4656, + "step": 3516 + }, + { + "epoch": 1.3261689291101055, + "grad_norm": 0.24377575481167546, + "learning_rate": 7.352903308315459e-05, + "loss": 0.4703, + "step": 3517 + }, + { + "epoch": 1.3265460030165912, + "grad_norm": 0.19025878771417096, + "learning_rate": 7.352328972751605e-05, + "loss": 0.4657, + "step": 3518 + }, + { + "epoch": 1.3269230769230769, + "grad_norm": 0.23964738605468594, + "learning_rate": 7.351754404874325e-05, + "loss": 0.4663, + "step": 3519 + }, + { + "epoch": 1.3273001508295625, + "grad_norm": 0.21353305049809024, + "learning_rate": 7.35117960472344e-05, + "loss": 0.4562, + "step": 3520 + }, + { + "epoch": 1.3276772247360482, + "grad_norm": 0.2128507642040722, + "learning_rate": 7.35060457233878e-05, + "loss": 0.4679, + "step": 3521 + }, + { + "epoch": 1.3280542986425339, + "grad_norm": 0.22009888715553394, + "learning_rate": 7.350029307760195e-05, + "loss": 0.449, + "step": 3522 + }, + { + "epoch": 1.3284313725490196, + "grad_norm": 0.29526851195977516, + "learning_rate": 7.349453811027551e-05, + "loss": 0.4613, + "step": 3523 + }, + { + "epoch": 1.3288084464555052, + "grad_norm": 0.3036813712856528, + "learning_rate": 7.348878082180729e-05, + "loss": 0.4637, + "step": 3524 + }, + { + "epoch": 1.329185520361991, + "grad_norm": 0.30426593149235304, + "learning_rate": 7.348302121259626e-05, + "loss": 0.4537, + "step": 3525 + }, + { + "epoch": 1.3295625942684766, + "grad_norm": 0.29263113289838477, + "learning_rate": 7.347725928304154e-05, + "loss": 0.4557, + "step": 3526 + }, + { + "epoch": 1.3299396681749622, + "grad_norm": 0.2663264661985787, + "learning_rate": 7.347149503354248e-05, + "loss": 0.4701, + "step": 3527 + }, + { + "epoch": 1.330316742081448, + "grad_norm": 0.26654682365871973, + "learning_rate": 7.346572846449848e-05, + "loss": 0.4781, + "step": 3528 + }, + { + "epoch": 1.3306938159879336, + "grad_norm": 0.2829575892968648, + "learning_rate": 7.345995957630919e-05, + "loss": 0.4663, + "step": 3529 + }, + { + "epoch": 1.3310708898944192, + "grad_norm": 0.260693356254036, + "learning_rate": 7.345418836937437e-05, + "loss": 0.4743, + "step": 3530 + }, + { + "epoch": 1.331447963800905, + "grad_norm": 0.2196474944609322, + "learning_rate": 7.344841484409399e-05, + "loss": 0.4633, + "step": 3531 + }, + { + "epoch": 1.3318250377073906, + "grad_norm": 0.2349427298946241, + "learning_rate": 7.344263900086812e-05, + "loss": 0.4677, + "step": 3532 + }, + { + "epoch": 1.3322021116138762, + "grad_norm": 0.2408467222493607, + "learning_rate": 7.343686084009703e-05, + "loss": 0.4685, + "step": 3533 + }, + { + "epoch": 1.332579185520362, + "grad_norm": 0.2356470312419382, + "learning_rate": 7.343108036218114e-05, + "loss": 0.4536, + "step": 3534 + }, + { + "epoch": 1.3329562594268476, + "grad_norm": 0.2224445983907144, + "learning_rate": 7.342529756752103e-05, + "loss": 0.464, + "step": 3535 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 0.25875044971700356, + "learning_rate": 7.341951245651747e-05, + "loss": 0.4626, + "step": 3536 + }, + { + "epoch": 1.333710407239819, + "grad_norm": 0.2835419962659994, + "learning_rate": 7.341372502957132e-05, + "loss": 0.469, + "step": 3537 + }, + { + "epoch": 1.3340874811463046, + "grad_norm": 0.3695345677757239, + "learning_rate": 7.340793528708366e-05, + "loss": 0.4687, + "step": 3538 + }, + { + "epoch": 1.3344645550527903, + "grad_norm": 0.45791398457711646, + "learning_rate": 7.340214322945572e-05, + "loss": 0.4622, + "step": 3539 + }, + { + "epoch": 1.334841628959276, + "grad_norm": 0.5513583200958098, + "learning_rate": 7.339634885708889e-05, + "loss": 0.4641, + "step": 3540 + }, + { + "epoch": 1.3352187028657616, + "grad_norm": 0.6022873775319663, + "learning_rate": 7.33905521703847e-05, + "loss": 0.4519, + "step": 3541 + }, + { + "epoch": 1.3355957767722473, + "grad_norm": 0.681406297913482, + "learning_rate": 7.338475316974487e-05, + "loss": 0.4591, + "step": 3542 + }, + { + "epoch": 1.3359728506787332, + "grad_norm": 0.686413572003419, + "learning_rate": 7.337895185557125e-05, + "loss": 0.464, + "step": 3543 + }, + { + "epoch": 1.3363499245852188, + "grad_norm": 0.5786957480158051, + "learning_rate": 7.337314822826587e-05, + "loss": 0.4487, + "step": 3544 + }, + { + "epoch": 1.3367269984917045, + "grad_norm": 0.40136833017190915, + "learning_rate": 7.336734228823094e-05, + "loss": 0.4499, + "step": 3545 + }, + { + "epoch": 1.3371040723981902, + "grad_norm": 0.2589820403082242, + "learning_rate": 7.336153403586875e-05, + "loss": 0.4636, + "step": 3546 + }, + { + "epoch": 1.3374811463046758, + "grad_norm": 0.2815357518539494, + "learning_rate": 7.335572347158188e-05, + "loss": 0.4708, + "step": 3547 + }, + { + "epoch": 1.3378582202111615, + "grad_norm": 0.34981606038838703, + "learning_rate": 7.334991059577294e-05, + "loss": 0.4631, + "step": 3548 + }, + { + "epoch": 1.3382352941176472, + "grad_norm": 0.43935521704209346, + "learning_rate": 7.334409540884479e-05, + "loss": 0.4601, + "step": 3549 + }, + { + "epoch": 1.3386123680241329, + "grad_norm": 0.4878959187199182, + "learning_rate": 7.33382779112004e-05, + "loss": 0.4576, + "step": 3550 + }, + { + "epoch": 1.3389894419306185, + "grad_norm": 0.46287877743434697, + "learning_rate": 7.33324581032429e-05, + "loss": 0.4626, + "step": 3551 + }, + { + "epoch": 1.3393665158371042, + "grad_norm": 0.34092603571712965, + "learning_rate": 7.332663598537565e-05, + "loss": 0.4578, + "step": 3552 + }, + { + "epoch": 1.3397435897435899, + "grad_norm": 0.28146262847576503, + "learning_rate": 7.332081155800207e-05, + "loss": 0.451, + "step": 3553 + }, + { + "epoch": 1.3401206636500755, + "grad_norm": 0.26810438742341963, + "learning_rate": 7.331498482152581e-05, + "loss": 0.4695, + "step": 3554 + }, + { + "epoch": 1.3404977375565612, + "grad_norm": 0.20282844151308152, + "learning_rate": 7.330915577635065e-05, + "loss": 0.451, + "step": 3555 + }, + { + "epoch": 1.3408748114630469, + "grad_norm": 0.20144618874811643, + "learning_rate": 7.330332442288054e-05, + "loss": 0.4653, + "step": 3556 + }, + { + "epoch": 1.3412518853695325, + "grad_norm": 0.2605308444958474, + "learning_rate": 7.329749076151958e-05, + "loss": 0.4618, + "step": 3557 + }, + { + "epoch": 1.3416289592760182, + "grad_norm": 0.3206101290439715, + "learning_rate": 7.329165479267206e-05, + "loss": 0.4495, + "step": 3558 + }, + { + "epoch": 1.3420060331825039, + "grad_norm": 0.25695956424315125, + "learning_rate": 7.328581651674239e-05, + "loss": 0.4628, + "step": 3559 + }, + { + "epoch": 1.3423831070889896, + "grad_norm": 0.18527376764503123, + "learning_rate": 7.327997593413516e-05, + "loss": 0.4662, + "step": 3560 + }, + { + "epoch": 1.3427601809954752, + "grad_norm": 0.20318421064807574, + "learning_rate": 7.327413304525511e-05, + "loss": 0.4511, + "step": 3561 + }, + { + "epoch": 1.343137254901961, + "grad_norm": 0.18908388170549062, + "learning_rate": 7.326828785050716e-05, + "loss": 0.4542, + "step": 3562 + }, + { + "epoch": 1.3435143288084466, + "grad_norm": 0.25311202068672706, + "learning_rate": 7.326244035029635e-05, + "loss": 0.4541, + "step": 3563 + }, + { + "epoch": 1.3438914027149322, + "grad_norm": 0.33973464594646136, + "learning_rate": 7.325659054502796e-05, + "loss": 0.4638, + "step": 3564 + }, + { + "epoch": 1.344268476621418, + "grad_norm": 0.36032325441916985, + "learning_rate": 7.325073843510732e-05, + "loss": 0.4469, + "step": 3565 + }, + { + "epoch": 1.3446455505279036, + "grad_norm": 0.36082346354537986, + "learning_rate": 7.324488402094e-05, + "loss": 0.4572, + "step": 3566 + }, + { + "epoch": 1.3450226244343892, + "grad_norm": 0.3696086427961489, + "learning_rate": 7.323902730293172e-05, + "loss": 0.4545, + "step": 3567 + }, + { + "epoch": 1.345399698340875, + "grad_norm": 0.4265275097389745, + "learning_rate": 7.323316828148833e-05, + "loss": 0.4416, + "step": 3568 + }, + { + "epoch": 1.3457767722473606, + "grad_norm": 0.4422747078187314, + "learning_rate": 7.322730695701586e-05, + "loss": 0.4546, + "step": 3569 + }, + { + "epoch": 1.3461538461538463, + "grad_norm": 0.39088133647775386, + "learning_rate": 7.322144332992047e-05, + "loss": 0.4487, + "step": 3570 + }, + { + "epoch": 1.346530920060332, + "grad_norm": 0.3557489812728841, + "learning_rate": 7.321557740060855e-05, + "loss": 0.4578, + "step": 3571 + }, + { + "epoch": 1.3469079939668176, + "grad_norm": 0.30716166202207107, + "learning_rate": 7.320970916948656e-05, + "loss": 0.4554, + "step": 3572 + }, + { + "epoch": 1.3472850678733033, + "grad_norm": 0.20395981346840164, + "learning_rate": 7.320383863696117e-05, + "loss": 0.4505, + "step": 3573 + }, + { + "epoch": 1.347662141779789, + "grad_norm": 0.2339818819061287, + "learning_rate": 7.319796580343924e-05, + "loss": 0.4411, + "step": 3574 + }, + { + "epoch": 1.3480392156862746, + "grad_norm": 0.2392463019542901, + "learning_rate": 7.319209066932772e-05, + "loss": 0.4438, + "step": 3575 + }, + { + "epoch": 1.3484162895927603, + "grad_norm": 0.3030934751505498, + "learning_rate": 7.318621323503376e-05, + "loss": 0.4517, + "step": 3576 + }, + { + "epoch": 1.348793363499246, + "grad_norm": 0.3659886913548192, + "learning_rate": 7.318033350096464e-05, + "loss": 0.4547, + "step": 3577 + }, + { + "epoch": 1.3491704374057316, + "grad_norm": 0.4339091532575386, + "learning_rate": 7.317445146752786e-05, + "loss": 0.4457, + "step": 3578 + }, + { + "epoch": 1.3495475113122173, + "grad_norm": 0.5025554797750647, + "learning_rate": 7.316856713513101e-05, + "loss": 0.4477, + "step": 3579 + }, + { + "epoch": 1.349924585218703, + "grad_norm": 0.6332955238940582, + "learning_rate": 7.316268050418188e-05, + "loss": 0.4529, + "step": 3580 + }, + { + "epoch": 1.3503016591251886, + "grad_norm": 0.7058274888766826, + "learning_rate": 7.315679157508841e-05, + "loss": 0.4526, + "step": 3581 + }, + { + "epoch": 1.3506787330316743, + "grad_norm": 0.6779423559255359, + "learning_rate": 7.315090034825869e-05, + "loss": 0.4621, + "step": 3582 + }, + { + "epoch": 1.35105580693816, + "grad_norm": 0.5593448371418995, + "learning_rate": 7.314500682410098e-05, + "loss": 0.4511, + "step": 3583 + }, + { + "epoch": 1.3514328808446456, + "grad_norm": 0.4765754553451103, + "learning_rate": 7.313911100302369e-05, + "loss": 0.4467, + "step": 3584 + }, + { + "epoch": 1.3518099547511313, + "grad_norm": 0.3434485944952262, + "learning_rate": 7.313321288543541e-05, + "loss": 0.4404, + "step": 3585 + }, + { + "epoch": 1.352187028657617, + "grad_norm": 0.29500199893243834, + "learning_rate": 7.312731247174487e-05, + "loss": 0.4494, + "step": 3586 + }, + { + "epoch": 1.3525641025641026, + "grad_norm": 0.39948030960982056, + "learning_rate": 7.312140976236094e-05, + "loss": 0.4578, + "step": 3587 + }, + { + "epoch": 1.3529411764705883, + "grad_norm": 0.41792745262641334, + "learning_rate": 7.311550475769272e-05, + "loss": 0.4478, + "step": 3588 + }, + { + "epoch": 1.353318250377074, + "grad_norm": 0.45339677540905876, + "learning_rate": 7.310959745814938e-05, + "loss": 0.4433, + "step": 3589 + }, + { + "epoch": 1.3536953242835597, + "grad_norm": 0.4783692163898853, + "learning_rate": 7.31036878641403e-05, + "loss": 0.4642, + "step": 3590 + }, + { + "epoch": 1.3540723981900453, + "grad_norm": 0.4333392070719599, + "learning_rate": 7.309777597607502e-05, + "loss": 0.453, + "step": 3591 + }, + { + "epoch": 1.354449472096531, + "grad_norm": 0.3632253152081476, + "learning_rate": 7.309186179436323e-05, + "loss": 0.4645, + "step": 3592 + }, + { + "epoch": 1.3548265460030167, + "grad_norm": 0.29493930727489964, + "learning_rate": 7.308594531941476e-05, + "loss": 0.4443, + "step": 3593 + }, + { + "epoch": 1.3552036199095023, + "grad_norm": 0.2427089841328273, + "learning_rate": 7.308002655163963e-05, + "loss": 0.4618, + "step": 3594 + }, + { + "epoch": 1.355580693815988, + "grad_norm": 0.29587877786301797, + "learning_rate": 7.3074105491448e-05, + "loss": 0.4528, + "step": 3595 + }, + { + "epoch": 1.3559577677224737, + "grad_norm": 0.33737036106846885, + "learning_rate": 7.306818213925019e-05, + "loss": 0.4504, + "step": 3596 + }, + { + "epoch": 1.3563348416289593, + "grad_norm": 0.27226975262996567, + "learning_rate": 7.30622564954567e-05, + "loss": 0.4407, + "step": 3597 + }, + { + "epoch": 1.356711915535445, + "grad_norm": 0.25182220741514766, + "learning_rate": 7.305632856047815e-05, + "loss": 0.4589, + "step": 3598 + }, + { + "epoch": 1.3570889894419307, + "grad_norm": 0.33857521763604, + "learning_rate": 7.305039833472537e-05, + "loss": 0.4412, + "step": 3599 + }, + { + "epoch": 1.3574660633484164, + "grad_norm": 0.413916504102084, + "learning_rate": 7.304446581860929e-05, + "loss": 0.4423, + "step": 3600 + }, + { + "epoch": 1.357843137254902, + "grad_norm": 0.43965919786486585, + "learning_rate": 7.303853101254103e-05, + "loss": 0.4614, + "step": 3601 + }, + { + "epoch": 1.3582202111613877, + "grad_norm": 0.4285484965254375, + "learning_rate": 7.303259391693189e-05, + "loss": 0.4525, + "step": 3602 + }, + { + "epoch": 1.3585972850678734, + "grad_norm": 0.3746689181815224, + "learning_rate": 7.302665453219329e-05, + "loss": 0.4634, + "step": 3603 + }, + { + "epoch": 1.358974358974359, + "grad_norm": 0.2836403307825816, + "learning_rate": 7.30207128587368e-05, + "loss": 0.4418, + "step": 3604 + }, + { + "epoch": 1.3593514328808447, + "grad_norm": 0.2745297608409804, + "learning_rate": 7.301476889697424e-05, + "loss": 0.451, + "step": 3605 + }, + { + "epoch": 1.3597285067873304, + "grad_norm": 0.2323738863909163, + "learning_rate": 7.300882264731746e-05, + "loss": 0.4567, + "step": 3606 + }, + { + "epoch": 1.360105580693816, + "grad_norm": 0.22947761737695832, + "learning_rate": 7.300287411017854e-05, + "loss": 0.4496, + "step": 3607 + }, + { + "epoch": 1.3604826546003017, + "grad_norm": 0.3178262059755702, + "learning_rate": 7.299692328596972e-05, + "loss": 0.4357, + "step": 3608 + }, + { + "epoch": 1.3608597285067874, + "grad_norm": 0.3828270244780204, + "learning_rate": 7.29909701751034e-05, + "loss": 0.4547, + "step": 3609 + }, + { + "epoch": 1.361236802413273, + "grad_norm": 0.37955412152591705, + "learning_rate": 7.29850147779921e-05, + "loss": 0.4546, + "step": 3610 + }, + { + "epoch": 1.3616138763197587, + "grad_norm": 0.34802872133080215, + "learning_rate": 7.297905709504853e-05, + "loss": 0.4485, + "step": 3611 + }, + { + "epoch": 1.3619909502262444, + "grad_norm": 0.28090688635418, + "learning_rate": 7.297309712668557e-05, + "loss": 0.4602, + "step": 3612 + }, + { + "epoch": 1.36236802413273, + "grad_norm": 0.22477127366099547, + "learning_rate": 7.29671348733162e-05, + "loss": 0.4533, + "step": 3613 + }, + { + "epoch": 1.3627450980392157, + "grad_norm": 0.21377304169461328, + "learning_rate": 7.296117033535365e-05, + "loss": 0.4502, + "step": 3614 + }, + { + "epoch": 1.3631221719457014, + "grad_norm": 0.2282717130510388, + "learning_rate": 7.295520351321122e-05, + "loss": 0.4456, + "step": 3615 + }, + { + "epoch": 1.363499245852187, + "grad_norm": 0.24137027387586216, + "learning_rate": 7.294923440730242e-05, + "loss": 0.4329, + "step": 3616 + }, + { + "epoch": 1.3638763197586727, + "grad_norm": 0.2741644520095255, + "learning_rate": 7.294326301804091e-05, + "loss": 0.4607, + "step": 3617 + }, + { + "epoch": 1.3642533936651584, + "grad_norm": 0.2626106667313442, + "learning_rate": 7.293728934584048e-05, + "loss": 0.4581, + "step": 3618 + }, + { + "epoch": 1.364630467571644, + "grad_norm": 0.25085067530809113, + "learning_rate": 7.293131339111511e-05, + "loss": 0.4495, + "step": 3619 + }, + { + "epoch": 1.3650075414781297, + "grad_norm": 0.27291868166940897, + "learning_rate": 7.292533515427895e-05, + "loss": 0.4554, + "step": 3620 + }, + { + "epoch": 1.3653846153846154, + "grad_norm": 0.3248509976686497, + "learning_rate": 7.291935463574626e-05, + "loss": 0.4469, + "step": 3621 + }, + { + "epoch": 1.365761689291101, + "grad_norm": 0.4015270281057459, + "learning_rate": 7.29133718359315e-05, + "loss": 0.4419, + "step": 3622 + }, + { + "epoch": 1.3661387631975868, + "grad_norm": 0.39367310018813356, + "learning_rate": 7.290738675524926e-05, + "loss": 0.4503, + "step": 3623 + }, + { + "epoch": 1.3665158371040724, + "grad_norm": 0.37536188126891434, + "learning_rate": 7.29013993941143e-05, + "loss": 0.4549, + "step": 3624 + }, + { + "epoch": 1.366892911010558, + "grad_norm": 0.37588746585661825, + "learning_rate": 7.289540975294156e-05, + "loss": 0.4348, + "step": 3625 + }, + { + "epoch": 1.3672699849170438, + "grad_norm": 0.37741385639981656, + "learning_rate": 7.288941783214609e-05, + "loss": 0.4454, + "step": 3626 + }, + { + "epoch": 1.3676470588235294, + "grad_norm": 0.36260841529830273, + "learning_rate": 7.288342363214313e-05, + "loss": 0.4409, + "step": 3627 + }, + { + "epoch": 1.368024132730015, + "grad_norm": 0.39811867886714153, + "learning_rate": 7.287742715334811e-05, + "loss": 0.4438, + "step": 3628 + }, + { + "epoch": 1.3684012066365008, + "grad_norm": 0.4570433873052253, + "learning_rate": 7.287142839617651e-05, + "loss": 0.4458, + "step": 3629 + }, + { + "epoch": 1.3687782805429864, + "grad_norm": 0.43041137319979456, + "learning_rate": 7.286542736104411e-05, + "loss": 0.4559, + "step": 3630 + }, + { + "epoch": 1.3691553544494721, + "grad_norm": 0.3704097863799893, + "learning_rate": 7.285942404836673e-05, + "loss": 0.4545, + "step": 3631 + }, + { + "epoch": 1.3695324283559578, + "grad_norm": 0.41104903734440945, + "learning_rate": 7.285341845856042e-05, + "loss": 0.458, + "step": 3632 + }, + { + "epoch": 1.3699095022624435, + "grad_norm": 0.37448764388547434, + "learning_rate": 7.284741059204135e-05, + "loss": 0.4598, + "step": 3633 + }, + { + "epoch": 1.3702865761689291, + "grad_norm": 0.24063984637114688, + "learning_rate": 7.284140044922587e-05, + "loss": 0.4474, + "step": 3634 + }, + { + "epoch": 1.3706636500754148, + "grad_norm": 0.21413260362977404, + "learning_rate": 7.283538803053045e-05, + "loss": 0.4557, + "step": 3635 + }, + { + "epoch": 1.3710407239819005, + "grad_norm": 0.2705677495665263, + "learning_rate": 7.282937333637178e-05, + "loss": 0.4556, + "step": 3636 + }, + { + "epoch": 1.3714177978883861, + "grad_norm": 0.27941972159720374, + "learning_rate": 7.282335636716664e-05, + "loss": 0.4501, + "step": 3637 + }, + { + "epoch": 1.3717948717948718, + "grad_norm": 0.24532412762383077, + "learning_rate": 7.281733712333202e-05, + "loss": 0.4481, + "step": 3638 + }, + { + "epoch": 1.3721719457013575, + "grad_norm": 0.25500268488372657, + "learning_rate": 7.281131560528506e-05, + "loss": 0.4628, + "step": 3639 + }, + { + "epoch": 1.3725490196078431, + "grad_norm": 0.3120459458740125, + "learning_rate": 7.280529181344303e-05, + "loss": 0.4492, + "step": 3640 + }, + { + "epoch": 1.3729260935143288, + "grad_norm": 0.34277822517005613, + "learning_rate": 7.279926574822337e-05, + "loss": 0.4498, + "step": 3641 + }, + { + "epoch": 1.3733031674208145, + "grad_norm": 0.4011371839747537, + "learning_rate": 7.27932374100437e-05, + "loss": 0.4589, + "step": 3642 + }, + { + "epoch": 1.3736802413273002, + "grad_norm": 0.4735130088938107, + "learning_rate": 7.278720679932175e-05, + "loss": 0.4446, + "step": 3643 + }, + { + "epoch": 1.3740573152337858, + "grad_norm": 0.48269569081563596, + "learning_rate": 7.278117391647545e-05, + "loss": 0.4485, + "step": 3644 + }, + { + "epoch": 1.3744343891402715, + "grad_norm": 0.46822588913458985, + "learning_rate": 7.277513876192288e-05, + "loss": 0.4447, + "step": 3645 + }, + { + "epoch": 1.3748114630467572, + "grad_norm": 0.4077255238186241, + "learning_rate": 7.276910133608226e-05, + "loss": 0.4671, + "step": 3646 + }, + { + "epoch": 1.3751885369532428, + "grad_norm": 0.3475111068640269, + "learning_rate": 7.276306163937199e-05, + "loss": 0.4358, + "step": 3647 + }, + { + "epoch": 1.3755656108597285, + "grad_norm": 0.2818857577008301, + "learning_rate": 7.27570196722106e-05, + "loss": 0.4641, + "step": 3648 + }, + { + "epoch": 1.3759426847662142, + "grad_norm": 0.2273808830830724, + "learning_rate": 7.275097543501682e-05, + "loss": 0.4484, + "step": 3649 + }, + { + "epoch": 1.3763197586726998, + "grad_norm": 0.23326682644991104, + "learning_rate": 7.274492892820948e-05, + "loss": 0.4497, + "step": 3650 + }, + { + "epoch": 1.3766968325791855, + "grad_norm": 0.2649929434052039, + "learning_rate": 7.273888015220762e-05, + "loss": 0.4462, + "step": 3651 + }, + { + "epoch": 1.3770739064856712, + "grad_norm": 0.27590726735258303, + "learning_rate": 7.273282910743038e-05, + "loss": 0.448, + "step": 3652 + }, + { + "epoch": 1.3774509803921569, + "grad_norm": 0.28110448856652354, + "learning_rate": 7.272677579429714e-05, + "loss": 0.4457, + "step": 3653 + }, + { + "epoch": 1.3778280542986425, + "grad_norm": 0.365996777643842, + "learning_rate": 7.272072021322736e-05, + "loss": 0.4463, + "step": 3654 + }, + { + "epoch": 1.3782051282051282, + "grad_norm": 0.42271641625147127, + "learning_rate": 7.271466236464068e-05, + "loss": 0.4337, + "step": 3655 + }, + { + "epoch": 1.3785822021116139, + "grad_norm": 0.4146021707203078, + "learning_rate": 7.270860224895694e-05, + "loss": 0.4482, + "step": 3656 + }, + { + "epoch": 1.3789592760180995, + "grad_norm": 0.48613353234101553, + "learning_rate": 7.270253986659606e-05, + "loss": 0.4522, + "step": 3657 + }, + { + "epoch": 1.3793363499245852, + "grad_norm": 0.5462061006379124, + "learning_rate": 7.269647521797819e-05, + "loss": 0.446, + "step": 3658 + }, + { + "epoch": 1.3797134238310709, + "grad_norm": 0.6059216690375794, + "learning_rate": 7.269040830352357e-05, + "loss": 0.4546, + "step": 3659 + }, + { + "epoch": 1.3800904977375565, + "grad_norm": 0.6371391944209549, + "learning_rate": 7.268433912365266e-05, + "loss": 0.4557, + "step": 3660 + }, + { + "epoch": 1.3804675716440422, + "grad_norm": 0.6382214644281684, + "learning_rate": 7.267826767878605e-05, + "loss": 0.4409, + "step": 3661 + }, + { + "epoch": 1.3808446455505279, + "grad_norm": 0.5123622979576281, + "learning_rate": 7.267219396934446e-05, + "loss": 0.4478, + "step": 3662 + }, + { + "epoch": 1.3812217194570136, + "grad_norm": 0.42283340306720524, + "learning_rate": 7.26661179957488e-05, + "loss": 0.4454, + "step": 3663 + }, + { + "epoch": 1.3815987933634992, + "grad_norm": 0.2418587112874756, + "learning_rate": 7.266003975842015e-05, + "loss": 0.456, + "step": 3664 + }, + { + "epoch": 1.381975867269985, + "grad_norm": 0.3525087102447073, + "learning_rate": 7.265395925777972e-05, + "loss": 0.4377, + "step": 3665 + }, + { + "epoch": 1.3823529411764706, + "grad_norm": 0.6266793266513276, + "learning_rate": 7.264787649424888e-05, + "loss": 0.4517, + "step": 3666 + }, + { + "epoch": 1.3827300150829562, + "grad_norm": 0.7345481781936071, + "learning_rate": 7.264179146824915e-05, + "loss": 0.4546, + "step": 3667 + }, + { + "epoch": 1.383107088989442, + "grad_norm": 0.7532219960406874, + "learning_rate": 7.263570418020221e-05, + "loss": 0.4544, + "step": 3668 + }, + { + "epoch": 1.3834841628959276, + "grad_norm": 0.5937739346430053, + "learning_rate": 7.262961463052994e-05, + "loss": 0.426, + "step": 3669 + }, + { + "epoch": 1.3838612368024132, + "grad_norm": 0.41876812306654804, + "learning_rate": 7.262352281965431e-05, + "loss": 0.4526, + "step": 3670 + }, + { + "epoch": 1.384238310708899, + "grad_norm": 0.36410751798371266, + "learning_rate": 7.261742874799747e-05, + "loss": 0.4525, + "step": 3671 + }, + { + "epoch": 1.3846153846153846, + "grad_norm": 0.5466043493922247, + "learning_rate": 7.261133241598177e-05, + "loss": 0.4513, + "step": 3672 + }, + { + "epoch": 1.3849924585218703, + "grad_norm": 0.611523606972223, + "learning_rate": 7.260523382402966e-05, + "loss": 0.4489, + "step": 3673 + }, + { + "epoch": 1.385369532428356, + "grad_norm": 0.5252777513699348, + "learning_rate": 7.259913297256376e-05, + "loss": 0.4547, + "step": 3674 + }, + { + "epoch": 1.3857466063348416, + "grad_norm": 0.5757747338478533, + "learning_rate": 7.259302986200687e-05, + "loss": 0.4544, + "step": 3675 + }, + { + "epoch": 1.3861236802413273, + "grad_norm": 0.6259838639817317, + "learning_rate": 7.258692449278192e-05, + "loss": 0.4473, + "step": 3676 + }, + { + "epoch": 1.386500754147813, + "grad_norm": 0.5033158877466665, + "learning_rate": 7.258081686531199e-05, + "loss": 0.4431, + "step": 3677 + }, + { + "epoch": 1.3868778280542986, + "grad_norm": 0.4051067659604309, + "learning_rate": 7.257470698002037e-05, + "loss": 0.4456, + "step": 3678 + }, + { + "epoch": 1.3872549019607843, + "grad_norm": 0.4705581162784912, + "learning_rate": 7.256859483733043e-05, + "loss": 0.4387, + "step": 3679 + }, + { + "epoch": 1.38763197586727, + "grad_norm": 0.4644209025221676, + "learning_rate": 7.256248043766578e-05, + "loss": 0.4504, + "step": 3680 + }, + { + "epoch": 1.3880090497737556, + "grad_norm": 0.31791304738064186, + "learning_rate": 7.25563637814501e-05, + "loss": 0.4495, + "step": 3681 + }, + { + "epoch": 1.3883861236802413, + "grad_norm": 0.2767366124867376, + "learning_rate": 7.255024486910729e-05, + "loss": 0.4484, + "step": 3682 + }, + { + "epoch": 1.388763197586727, + "grad_norm": 0.34849426844918524, + "learning_rate": 7.254412370106139e-05, + "loss": 0.4479, + "step": 3683 + }, + { + "epoch": 1.3891402714932126, + "grad_norm": 0.4098517628962141, + "learning_rate": 7.253800027773658e-05, + "loss": 0.4471, + "step": 3684 + }, + { + "epoch": 1.3895173453996983, + "grad_norm": 0.3160935509486048, + "learning_rate": 7.25318745995572e-05, + "loss": 0.4545, + "step": 3685 + }, + { + "epoch": 1.389894419306184, + "grad_norm": 0.23912521283253252, + "learning_rate": 7.252574666694778e-05, + "loss": 0.4372, + "step": 3686 + }, + { + "epoch": 1.3902714932126696, + "grad_norm": 0.2747109181390731, + "learning_rate": 7.251961648033297e-05, + "loss": 0.4454, + "step": 3687 + }, + { + "epoch": 1.3906485671191553, + "grad_norm": 0.3183192182897611, + "learning_rate": 7.251348404013758e-05, + "loss": 0.4538, + "step": 3688 + }, + { + "epoch": 1.391025641025641, + "grad_norm": 0.3427518105438008, + "learning_rate": 7.250734934678658e-05, + "loss": 0.4425, + "step": 3689 + }, + { + "epoch": 1.3914027149321266, + "grad_norm": 0.41516747551840794, + "learning_rate": 7.25012124007051e-05, + "loss": 0.4474, + "step": 3690 + }, + { + "epoch": 1.3917797888386123, + "grad_norm": 0.40657899053131175, + "learning_rate": 7.249507320231844e-05, + "loss": 0.4438, + "step": 3691 + }, + { + "epoch": 1.392156862745098, + "grad_norm": 0.3005243392574897, + "learning_rate": 7.248893175205203e-05, + "loss": 0.4386, + "step": 3692 + }, + { + "epoch": 1.3925339366515836, + "grad_norm": 0.2824015676558865, + "learning_rate": 7.248278805033145e-05, + "loss": 0.4564, + "step": 3693 + }, + { + "epoch": 1.3929110105580693, + "grad_norm": 0.37182095991876984, + "learning_rate": 7.247664209758249e-05, + "loss": 0.4513, + "step": 3694 + }, + { + "epoch": 1.393288084464555, + "grad_norm": 0.410854439919982, + "learning_rate": 7.247049389423103e-05, + "loss": 0.4497, + "step": 3695 + }, + { + "epoch": 1.3936651583710407, + "grad_norm": 0.4149820765031061, + "learning_rate": 7.246434344070315e-05, + "loss": 0.4481, + "step": 3696 + }, + { + "epoch": 1.3940422322775263, + "grad_norm": 0.48924642745216906, + "learning_rate": 7.245819073742506e-05, + "loss": 0.4492, + "step": 3697 + }, + { + "epoch": 1.394419306184012, + "grad_norm": 0.5965708408916908, + "learning_rate": 7.245203578482315e-05, + "loss": 0.4423, + "step": 3698 + }, + { + "epoch": 1.3947963800904977, + "grad_norm": 0.6269154155092947, + "learning_rate": 7.244587858332393e-05, + "loss": 0.4579, + "step": 3699 + }, + { + "epoch": 1.3951734539969833, + "grad_norm": 0.6278871697751758, + "learning_rate": 7.243971913335412e-05, + "loss": 0.4424, + "step": 3700 + }, + { + "epoch": 1.395550527903469, + "grad_norm": 0.7029594342118587, + "learning_rate": 7.243355743534054e-05, + "loss": 0.4626, + "step": 3701 + }, + { + "epoch": 1.3959276018099547, + "grad_norm": 0.7782795963411262, + "learning_rate": 7.24273934897102e-05, + "loss": 0.4713, + "step": 3702 + }, + { + "epoch": 1.3963046757164403, + "grad_norm": 0.7644392162730114, + "learning_rate": 7.242122729689024e-05, + "loss": 0.4592, + "step": 3703 + }, + { + "epoch": 1.396681749622926, + "grad_norm": 0.6286519644055217, + "learning_rate": 7.241505885730801e-05, + "loss": 0.4555, + "step": 3704 + }, + { + "epoch": 1.3970588235294117, + "grad_norm": 0.45047771766383043, + "learning_rate": 7.240888817139094e-05, + "loss": 0.4535, + "step": 3705 + }, + { + "epoch": 1.3974358974358974, + "grad_norm": 0.4145908334512088, + "learning_rate": 7.240271523956667e-05, + "loss": 0.4675, + "step": 3706 + }, + { + "epoch": 1.397812971342383, + "grad_norm": 0.43867142152644373, + "learning_rate": 7.239654006226298e-05, + "loss": 0.4482, + "step": 3707 + }, + { + "epoch": 1.3981900452488687, + "grad_norm": 0.39739602877157776, + "learning_rate": 7.239036263990778e-05, + "loss": 0.4494, + "step": 3708 + }, + { + "epoch": 1.3985671191553544, + "grad_norm": 0.34102495840365177, + "learning_rate": 7.238418297292921e-05, + "loss": 0.4589, + "step": 3709 + }, + { + "epoch": 1.39894419306184, + "grad_norm": 0.2465496107842345, + "learning_rate": 7.237800106175546e-05, + "loss": 0.4521, + "step": 3710 + }, + { + "epoch": 1.3993212669683257, + "grad_norm": 0.20549025926233022, + "learning_rate": 7.237181690681497e-05, + "loss": 0.442, + "step": 3711 + }, + { + "epoch": 1.3996983408748114, + "grad_norm": 0.2725441152539738, + "learning_rate": 7.236563050853627e-05, + "loss": 0.4369, + "step": 3712 + }, + { + "epoch": 1.400075414781297, + "grad_norm": 0.354692277313513, + "learning_rate": 7.23594418673481e-05, + "loss": 0.4557, + "step": 3713 + }, + { + "epoch": 1.4004524886877827, + "grad_norm": 0.3193725358279911, + "learning_rate": 7.235325098367929e-05, + "loss": 0.4435, + "step": 3714 + }, + { + "epoch": 1.4008295625942684, + "grad_norm": 0.2574970604738164, + "learning_rate": 7.234705785795889e-05, + "loss": 0.4458, + "step": 3715 + }, + { + "epoch": 1.401206636500754, + "grad_norm": 0.19869135202048788, + "learning_rate": 7.234086249061609e-05, + "loss": 0.4406, + "step": 3716 + }, + { + "epoch": 1.4015837104072397, + "grad_norm": 0.21076693770421767, + "learning_rate": 7.233466488208018e-05, + "loss": 0.4514, + "step": 3717 + }, + { + "epoch": 1.4019607843137254, + "grad_norm": 0.29010724010316175, + "learning_rate": 7.232846503278069e-05, + "loss": 0.4473, + "step": 3718 + }, + { + "epoch": 1.402337858220211, + "grad_norm": 0.31980364196979866, + "learning_rate": 7.232226294314725e-05, + "loss": 0.4538, + "step": 3719 + }, + { + "epoch": 1.4027149321266967, + "grad_norm": 0.3206064710837235, + "learning_rate": 7.231605861360966e-05, + "loss": 0.444, + "step": 3720 + }, + { + "epoch": 1.4030920060331824, + "grad_norm": 0.2665259013944338, + "learning_rate": 7.230985204459786e-05, + "loss": 0.4401, + "step": 3721 + }, + { + "epoch": 1.403469079939668, + "grad_norm": 0.302547783632354, + "learning_rate": 7.2303643236542e-05, + "loss": 0.4408, + "step": 3722 + }, + { + "epoch": 1.4038461538461537, + "grad_norm": 0.31483373941341614, + "learning_rate": 7.229743218987229e-05, + "loss": 0.445, + "step": 3723 + }, + { + "epoch": 1.4042232277526394, + "grad_norm": 0.3264271679904733, + "learning_rate": 7.229121890501917e-05, + "loss": 0.4531, + "step": 3724 + }, + { + "epoch": 1.404600301659125, + "grad_norm": 0.31772867602095617, + "learning_rate": 7.228500338241324e-05, + "loss": 0.4434, + "step": 3725 + }, + { + "epoch": 1.4049773755656108, + "grad_norm": 0.24986026604080958, + "learning_rate": 7.22787856224852e-05, + "loss": 0.4645, + "step": 3726 + }, + { + "epoch": 1.4053544494720964, + "grad_norm": 0.24284935657867926, + "learning_rate": 7.227256562566595e-05, + "loss": 0.4509, + "step": 3727 + }, + { + "epoch": 1.405731523378582, + "grad_norm": 0.27465954021310074, + "learning_rate": 7.226634339238655e-05, + "loss": 0.4436, + "step": 3728 + }, + { + "epoch": 1.4061085972850678, + "grad_norm": 0.3356960337011572, + "learning_rate": 7.226011892307813e-05, + "loss": 0.4548, + "step": 3729 + }, + { + "epoch": 1.4064856711915534, + "grad_norm": 0.3660754902582047, + "learning_rate": 7.225389221817212e-05, + "loss": 0.4414, + "step": 3730 + }, + { + "epoch": 1.406862745098039, + "grad_norm": 0.43175766045960934, + "learning_rate": 7.224766327809998e-05, + "loss": 0.4419, + "step": 3731 + }, + { + "epoch": 1.4072398190045248, + "grad_norm": 0.48484961361371803, + "learning_rate": 7.224143210329337e-05, + "loss": 0.4442, + "step": 3732 + }, + { + "epoch": 1.4076168929110104, + "grad_norm": 0.5346525853207982, + "learning_rate": 7.223519869418413e-05, + "loss": 0.4555, + "step": 3733 + }, + { + "epoch": 1.4079939668174961, + "grad_norm": 0.5464881696425025, + "learning_rate": 7.22289630512042e-05, + "loss": 0.4443, + "step": 3734 + }, + { + "epoch": 1.4083710407239818, + "grad_norm": 0.5760211711615832, + "learning_rate": 7.222272517478571e-05, + "loss": 0.4378, + "step": 3735 + }, + { + "epoch": 1.4087481146304675, + "grad_norm": 0.5314873674133159, + "learning_rate": 7.221648506536097e-05, + "loss": 0.4476, + "step": 3736 + }, + { + "epoch": 1.4091251885369531, + "grad_norm": 0.5330238413037722, + "learning_rate": 7.221024272336238e-05, + "loss": 0.4496, + "step": 3737 + }, + { + "epoch": 1.4095022624434388, + "grad_norm": 0.5176151955237664, + "learning_rate": 7.220399814922254e-05, + "loss": 0.4431, + "step": 3738 + }, + { + "epoch": 1.4098793363499245, + "grad_norm": 0.47855126567968237, + "learning_rate": 7.219775134337418e-05, + "loss": 0.4453, + "step": 3739 + }, + { + "epoch": 1.4102564102564101, + "grad_norm": 0.4222990306807528, + "learning_rate": 7.219150230625022e-05, + "loss": 0.452, + "step": 3740 + }, + { + "epoch": 1.4106334841628958, + "grad_norm": 0.3656492199411692, + "learning_rate": 7.21852510382837e-05, + "loss": 0.4435, + "step": 3741 + }, + { + "epoch": 1.4110105580693815, + "grad_norm": 0.364755085538251, + "learning_rate": 7.217899753990784e-05, + "loss": 0.4621, + "step": 3742 + }, + { + "epoch": 1.4113876319758671, + "grad_norm": 0.4183760355431186, + "learning_rate": 7.217274181155599e-05, + "loss": 0.4354, + "step": 3743 + }, + { + "epoch": 1.4117647058823528, + "grad_norm": 0.45965564242504214, + "learning_rate": 7.216648385366167e-05, + "loss": 0.4472, + "step": 3744 + }, + { + "epoch": 1.4121417797888385, + "grad_norm": 0.469046706410134, + "learning_rate": 7.216022366665856e-05, + "loss": 0.4506, + "step": 3745 + }, + { + "epoch": 1.4125188536953242, + "grad_norm": 0.5108093144928646, + "learning_rate": 7.215396125098046e-05, + "loss": 0.4427, + "step": 3746 + }, + { + "epoch": 1.4128959276018098, + "grad_norm": 0.5144002874761306, + "learning_rate": 7.214769660706138e-05, + "loss": 0.4374, + "step": 3747 + }, + { + "epoch": 1.4132730015082955, + "grad_norm": 0.4879280661008998, + "learning_rate": 7.214142973533544e-05, + "loss": 0.4551, + "step": 3748 + }, + { + "epoch": 1.4136500754147812, + "grad_norm": 0.46546352190614454, + "learning_rate": 7.21351606362369e-05, + "loss": 0.4502, + "step": 3749 + }, + { + "epoch": 1.4140271493212668, + "grad_norm": 0.36808888966548153, + "learning_rate": 7.212888931020025e-05, + "loss": 0.4564, + "step": 3750 + }, + { + "epoch": 1.4144042232277527, + "grad_norm": 0.35980821878465213, + "learning_rate": 7.212261575766008e-05, + "loss": 0.4489, + "step": 3751 + }, + { + "epoch": 1.4147812971342384, + "grad_norm": 0.4473803430896227, + "learning_rate": 7.211633997905112e-05, + "loss": 0.4551, + "step": 3752 + }, + { + "epoch": 1.415158371040724, + "grad_norm": 0.4875688725926671, + "learning_rate": 7.21100619748083e-05, + "loss": 0.4444, + "step": 3753 + }, + { + "epoch": 1.4155354449472097, + "grad_norm": 0.4510664868381571, + "learning_rate": 7.210378174536665e-05, + "loss": 0.4482, + "step": 3754 + }, + { + "epoch": 1.4159125188536954, + "grad_norm": 0.4017353934211454, + "learning_rate": 7.209749929116142e-05, + "loss": 0.4406, + "step": 3755 + }, + { + "epoch": 1.416289592760181, + "grad_norm": 0.428825355333433, + "learning_rate": 7.209121461262793e-05, + "loss": 0.4416, + "step": 3756 + }, + { + "epoch": 1.4166666666666667, + "grad_norm": 0.51552915851126, + "learning_rate": 7.208492771020176e-05, + "loss": 0.4541, + "step": 3757 + }, + { + "epoch": 1.4170437405731524, + "grad_norm": 0.5641961987025289, + "learning_rate": 7.207863858431855e-05, + "loss": 0.4554, + "step": 3758 + }, + { + "epoch": 1.417420814479638, + "grad_norm": 0.48984231224298114, + "learning_rate": 7.207234723541414e-05, + "loss": 0.4511, + "step": 3759 + }, + { + "epoch": 1.4177978883861238, + "grad_norm": 0.2785150357784455, + "learning_rate": 7.206605366392451e-05, + "loss": 0.4424, + "step": 3760 + }, + { + "epoch": 1.4181749622926094, + "grad_norm": 0.4155205679365201, + "learning_rate": 7.20597578702858e-05, + "loss": 0.4369, + "step": 3761 + }, + { + "epoch": 1.418552036199095, + "grad_norm": 0.5631486835041335, + "learning_rate": 7.205345985493432e-05, + "loss": 0.4498, + "step": 3762 + }, + { + "epoch": 1.4189291101055808, + "grad_norm": 0.5328799087270558, + "learning_rate": 7.20471596183065e-05, + "loss": 0.4567, + "step": 3763 + }, + { + "epoch": 1.4193061840120664, + "grad_norm": 0.44375853583352587, + "learning_rate": 7.204085716083894e-05, + "loss": 0.4393, + "step": 3764 + }, + { + "epoch": 1.419683257918552, + "grad_norm": 0.4626711590613103, + "learning_rate": 7.20345524829684e-05, + "loss": 0.4563, + "step": 3765 + }, + { + "epoch": 1.4200603318250378, + "grad_norm": 0.5169785996423025, + "learning_rate": 7.202824558513178e-05, + "loss": 0.4452, + "step": 3766 + }, + { + "epoch": 1.4204374057315234, + "grad_norm": 0.452900675946321, + "learning_rate": 7.202193646776616e-05, + "loss": 0.4387, + "step": 3767 + }, + { + "epoch": 1.420814479638009, + "grad_norm": 0.3049962866320599, + "learning_rate": 7.201562513130875e-05, + "loss": 0.443, + "step": 3768 + }, + { + "epoch": 1.4211915535444948, + "grad_norm": 0.4086907514616655, + "learning_rate": 7.20093115761969e-05, + "loss": 0.4368, + "step": 3769 + }, + { + "epoch": 1.4215686274509804, + "grad_norm": 0.5133553899516639, + "learning_rate": 7.200299580286816e-05, + "loss": 0.4497, + "step": 3770 + }, + { + "epoch": 1.4219457013574661, + "grad_norm": 0.49282788433122027, + "learning_rate": 7.19966778117602e-05, + "loss": 0.4564, + "step": 3771 + }, + { + "epoch": 1.4223227752639518, + "grad_norm": 0.5220017479806158, + "learning_rate": 7.199035760331085e-05, + "loss": 0.4421, + "step": 3772 + }, + { + "epoch": 1.4226998491704375, + "grad_norm": 0.5326390428526943, + "learning_rate": 7.198403517795807e-05, + "loss": 0.4438, + "step": 3773 + }, + { + "epoch": 1.4230769230769231, + "grad_norm": 0.4023729291629337, + "learning_rate": 7.197771053614006e-05, + "loss": 0.4491, + "step": 3774 + }, + { + "epoch": 1.4234539969834088, + "grad_norm": 0.3438379700344434, + "learning_rate": 7.197138367829505e-05, + "loss": 0.4453, + "step": 3775 + }, + { + "epoch": 1.4238310708898945, + "grad_norm": 0.35757413106722985, + "learning_rate": 7.196505460486152e-05, + "loss": 0.444, + "step": 3776 + }, + { + "epoch": 1.4242081447963801, + "grad_norm": 0.34600231772472817, + "learning_rate": 7.195872331627804e-05, + "loss": 0.4511, + "step": 3777 + }, + { + "epoch": 1.4245852187028658, + "grad_norm": 0.33510112969218286, + "learning_rate": 7.19523898129834e-05, + "loss": 0.4364, + "step": 3778 + }, + { + "epoch": 1.4249622926093515, + "grad_norm": 0.40440469494648873, + "learning_rate": 7.194605409541648e-05, + "loss": 0.4463, + "step": 3779 + }, + { + "epoch": 1.4253393665158371, + "grad_norm": 0.39214328859334996, + "learning_rate": 7.193971616401636e-05, + "loss": 0.44, + "step": 3780 + }, + { + "epoch": 1.4257164404223228, + "grad_norm": 0.294055119020361, + "learning_rate": 7.193337601922222e-05, + "loss": 0.4561, + "step": 3781 + }, + { + "epoch": 1.4260935143288085, + "grad_norm": 0.2699444943533393, + "learning_rate": 7.192703366147346e-05, + "loss": 0.4538, + "step": 3782 + }, + { + "epoch": 1.4264705882352942, + "grad_norm": 0.2747901015586808, + "learning_rate": 7.192068909120959e-05, + "loss": 0.4337, + "step": 3783 + }, + { + "epoch": 1.4268476621417798, + "grad_norm": 0.30413842991076606, + "learning_rate": 7.191434230887027e-05, + "loss": 0.4429, + "step": 3784 + }, + { + "epoch": 1.4272247360482655, + "grad_norm": 0.3503983679390567, + "learning_rate": 7.190799331489533e-05, + "loss": 0.449, + "step": 3785 + }, + { + "epoch": 1.4276018099547512, + "grad_norm": 0.3851638668951508, + "learning_rate": 7.190164210972478e-05, + "loss": 0.444, + "step": 3786 + }, + { + "epoch": 1.4279788838612368, + "grad_norm": 0.3660290488653645, + "learning_rate": 7.189528869379871e-05, + "loss": 0.4447, + "step": 3787 + }, + { + "epoch": 1.4283559577677225, + "grad_norm": 0.28681215870186955, + "learning_rate": 7.188893306755742e-05, + "loss": 0.4401, + "step": 3788 + }, + { + "epoch": 1.4287330316742082, + "grad_norm": 0.2986256327136213, + "learning_rate": 7.188257523144136e-05, + "loss": 0.4355, + "step": 3789 + }, + { + "epoch": 1.4291101055806938, + "grad_norm": 0.2574575475932134, + "learning_rate": 7.187621518589113e-05, + "loss": 0.4505, + "step": 3790 + }, + { + "epoch": 1.4294871794871795, + "grad_norm": 0.2641349443874158, + "learning_rate": 7.186985293134744e-05, + "loss": 0.444, + "step": 3791 + }, + { + "epoch": 1.4298642533936652, + "grad_norm": 0.24277873162338698, + "learning_rate": 7.186348846825123e-05, + "loss": 0.4491, + "step": 3792 + }, + { + "epoch": 1.4302413273001509, + "grad_norm": 0.2736855072497202, + "learning_rate": 7.185712179704352e-05, + "loss": 0.4396, + "step": 3793 + }, + { + "epoch": 1.4306184012066365, + "grad_norm": 0.32308993920329604, + "learning_rate": 7.185075291816553e-05, + "loss": 0.4339, + "step": 3794 + }, + { + "epoch": 1.4309954751131222, + "grad_norm": 0.3282172398344041, + "learning_rate": 7.184438183205862e-05, + "loss": 0.437, + "step": 3795 + }, + { + "epoch": 1.4313725490196079, + "grad_norm": 0.34785807654558154, + "learning_rate": 7.183800853916427e-05, + "loss": 0.4365, + "step": 3796 + }, + { + "epoch": 1.4317496229260935, + "grad_norm": 0.3067457047795582, + "learning_rate": 7.18316330399242e-05, + "loss": 0.4365, + "step": 3797 + }, + { + "epoch": 1.4321266968325792, + "grad_norm": 0.28472567891859996, + "learning_rate": 7.182525533478019e-05, + "loss": 0.4577, + "step": 3798 + }, + { + "epoch": 1.4325037707390649, + "grad_norm": 0.3221739099392441, + "learning_rate": 7.18188754241742e-05, + "loss": 0.4409, + "step": 3799 + }, + { + "epoch": 1.4328808446455505, + "grad_norm": 0.2586815843314352, + "learning_rate": 7.181249330854837e-05, + "loss": 0.4423, + "step": 3800 + }, + { + "epoch": 1.4332579185520362, + "grad_norm": 0.30176296776444583, + "learning_rate": 7.180610898834497e-05, + "loss": 0.4576, + "step": 3801 + }, + { + "epoch": 1.4336349924585219, + "grad_norm": 0.294027524719235, + "learning_rate": 7.179972246400643e-05, + "loss": 0.4431, + "step": 3802 + }, + { + "epoch": 1.4340120663650076, + "grad_norm": 0.2347042560542869, + "learning_rate": 7.179333373597533e-05, + "loss": 0.4438, + "step": 3803 + }, + { + "epoch": 1.4343891402714932, + "grad_norm": 0.3027741056445804, + "learning_rate": 7.178694280469442e-05, + "loss": 0.4524, + "step": 3804 + }, + { + "epoch": 1.434766214177979, + "grad_norm": 0.3778380352428815, + "learning_rate": 7.178054967060655e-05, + "loss": 0.4552, + "step": 3805 + }, + { + "epoch": 1.4351432880844646, + "grad_norm": 0.38129390057449425, + "learning_rate": 7.177415433415478e-05, + "loss": 0.4402, + "step": 3806 + }, + { + "epoch": 1.4355203619909502, + "grad_norm": 0.32280810843294233, + "learning_rate": 7.17677567957823e-05, + "loss": 0.4473, + "step": 3807 + }, + { + "epoch": 1.435897435897436, + "grad_norm": 0.28967776380598365, + "learning_rate": 7.176135705593245e-05, + "loss": 0.4401, + "step": 3808 + }, + { + "epoch": 1.4362745098039216, + "grad_norm": 0.2288921682853404, + "learning_rate": 7.175495511504873e-05, + "loss": 0.4507, + "step": 3809 + }, + { + "epoch": 1.4366515837104072, + "grad_norm": 0.20547435176907156, + "learning_rate": 7.174855097357478e-05, + "loss": 0.4522, + "step": 3810 + }, + { + "epoch": 1.437028657616893, + "grad_norm": 0.25463237577030173, + "learning_rate": 7.174214463195442e-05, + "loss": 0.4444, + "step": 3811 + }, + { + "epoch": 1.4374057315233786, + "grad_norm": 0.30107952481345074, + "learning_rate": 7.17357360906316e-05, + "loss": 0.431, + "step": 3812 + }, + { + "epoch": 1.4377828054298643, + "grad_norm": 0.2382090774936678, + "learning_rate": 7.17293253500504e-05, + "loss": 0.4599, + "step": 3813 + }, + { + "epoch": 1.43815987933635, + "grad_norm": 0.23326628584286868, + "learning_rate": 7.17229124106551e-05, + "loss": 0.4416, + "step": 3814 + }, + { + "epoch": 1.4385369532428356, + "grad_norm": 0.2616708412776749, + "learning_rate": 7.171649727289012e-05, + "loss": 0.4453, + "step": 3815 + }, + { + "epoch": 1.4389140271493213, + "grad_norm": 0.262094257427101, + "learning_rate": 7.17100799372e-05, + "loss": 0.4398, + "step": 3816 + }, + { + "epoch": 1.439291101055807, + "grad_norm": 0.3013577985523395, + "learning_rate": 7.170366040402946e-05, + "loss": 0.4311, + "step": 3817 + }, + { + "epoch": 1.4396681749622926, + "grad_norm": 0.2900256531224377, + "learning_rate": 7.16972386738234e-05, + "loss": 0.444, + "step": 3818 + }, + { + "epoch": 1.4400452488687783, + "grad_norm": 0.29124404650471925, + "learning_rate": 7.169081474702679e-05, + "loss": 0.4424, + "step": 3819 + }, + { + "epoch": 1.440422322775264, + "grad_norm": 0.2341514833159437, + "learning_rate": 7.168438862408484e-05, + "loss": 0.4424, + "step": 3820 + }, + { + "epoch": 1.4407993966817496, + "grad_norm": 0.16977942458039974, + "learning_rate": 7.167796030544286e-05, + "loss": 0.4512, + "step": 3821 + }, + { + "epoch": 1.4411764705882353, + "grad_norm": 0.21519103212785795, + "learning_rate": 7.167152979154634e-05, + "loss": 0.4483, + "step": 3822 + }, + { + "epoch": 1.441553544494721, + "grad_norm": 0.27495936824545064, + "learning_rate": 7.166509708284087e-05, + "loss": 0.4522, + "step": 3823 + }, + { + "epoch": 1.4419306184012066, + "grad_norm": 0.3419907507280817, + "learning_rate": 7.165866217977229e-05, + "loss": 0.4469, + "step": 3824 + }, + { + "epoch": 1.4423076923076923, + "grad_norm": 0.38647223913733925, + "learning_rate": 7.165222508278649e-05, + "loss": 0.4432, + "step": 3825 + }, + { + "epoch": 1.442684766214178, + "grad_norm": 0.43195331504172696, + "learning_rate": 7.164578579232956e-05, + "loss": 0.4317, + "step": 3826 + }, + { + "epoch": 1.4430618401206636, + "grad_norm": 0.45854612974547154, + "learning_rate": 7.163934430884775e-05, + "loss": 0.4367, + "step": 3827 + }, + { + "epoch": 1.4434389140271493, + "grad_norm": 0.5442592976046581, + "learning_rate": 7.163290063278745e-05, + "loss": 0.452, + "step": 3828 + }, + { + "epoch": 1.443815987933635, + "grad_norm": 0.5605113136001371, + "learning_rate": 7.162645476459518e-05, + "loss": 0.447, + "step": 3829 + }, + { + "epoch": 1.4441930618401206, + "grad_norm": 0.5592617900462507, + "learning_rate": 7.162000670471765e-05, + "loss": 0.4398, + "step": 3830 + }, + { + "epoch": 1.4445701357466063, + "grad_norm": 0.5757871633778024, + "learning_rate": 7.161355645360171e-05, + "loss": 0.4489, + "step": 3831 + }, + { + "epoch": 1.444947209653092, + "grad_norm": 0.6069488362522036, + "learning_rate": 7.160710401169437e-05, + "loss": 0.4446, + "step": 3832 + }, + { + "epoch": 1.4453242835595776, + "grad_norm": 0.5867610504573431, + "learning_rate": 7.160064937944272e-05, + "loss": 0.4425, + "step": 3833 + }, + { + "epoch": 1.4457013574660633, + "grad_norm": 0.49205197038066045, + "learning_rate": 7.159419255729412e-05, + "loss": 0.447, + "step": 3834 + }, + { + "epoch": 1.446078431372549, + "grad_norm": 0.40570376470362685, + "learning_rate": 7.158773354569599e-05, + "loss": 0.449, + "step": 3835 + }, + { + "epoch": 1.4464555052790347, + "grad_norm": 0.3555986744956309, + "learning_rate": 7.158127234509594e-05, + "loss": 0.4376, + "step": 3836 + }, + { + "epoch": 1.4468325791855203, + "grad_norm": 0.3518639539599214, + "learning_rate": 7.157480895594173e-05, + "loss": 0.4316, + "step": 3837 + }, + { + "epoch": 1.447209653092006, + "grad_norm": 0.3609753859278796, + "learning_rate": 7.156834337868127e-05, + "loss": 0.4449, + "step": 3838 + }, + { + "epoch": 1.4475867269984917, + "grad_norm": 0.32723675824433834, + "learning_rate": 7.156187561376262e-05, + "loss": 0.4377, + "step": 3839 + }, + { + "epoch": 1.4479638009049773, + "grad_norm": 0.34130309392505787, + "learning_rate": 7.155540566163398e-05, + "loss": 0.4502, + "step": 3840 + }, + { + "epoch": 1.448340874811463, + "grad_norm": 0.318685071424223, + "learning_rate": 7.154893352274373e-05, + "loss": 0.4387, + "step": 3841 + }, + { + "epoch": 1.4487179487179487, + "grad_norm": 0.278328381702656, + "learning_rate": 7.154245919754034e-05, + "loss": 0.4385, + "step": 3842 + }, + { + "epoch": 1.4490950226244343, + "grad_norm": 0.31823786933676634, + "learning_rate": 7.153598268647253e-05, + "loss": 0.4472, + "step": 3843 + }, + { + "epoch": 1.44947209653092, + "grad_norm": 0.33356485271206265, + "learning_rate": 7.152950398998909e-05, + "loss": 0.4481, + "step": 3844 + }, + { + "epoch": 1.4498491704374057, + "grad_norm": 0.30602678978163805, + "learning_rate": 7.1523023108539e-05, + "loss": 0.4403, + "step": 3845 + }, + { + "epoch": 1.4502262443438914, + "grad_norm": 0.2603002442248912, + "learning_rate": 7.151654004257134e-05, + "loss": 0.4472, + "step": 3846 + }, + { + "epoch": 1.450603318250377, + "grad_norm": 0.2616581747929248, + "learning_rate": 7.151005479253544e-05, + "loss": 0.4397, + "step": 3847 + }, + { + "epoch": 1.4509803921568627, + "grad_norm": 0.23037167896971972, + "learning_rate": 7.150356735888068e-05, + "loss": 0.447, + "step": 3848 + }, + { + "epoch": 1.4513574660633484, + "grad_norm": 0.29676872123776127, + "learning_rate": 7.149707774205664e-05, + "loss": 0.4419, + "step": 3849 + }, + { + "epoch": 1.451734539969834, + "grad_norm": 0.33189797366384, + "learning_rate": 7.149058594251306e-05, + "loss": 0.4365, + "step": 3850 + }, + { + "epoch": 1.4521116138763197, + "grad_norm": 0.3315802887166259, + "learning_rate": 7.148409196069979e-05, + "loss": 0.4345, + "step": 3851 + }, + { + "epoch": 1.4524886877828054, + "grad_norm": 0.34665511048564135, + "learning_rate": 7.14775957970669e-05, + "loss": 0.4368, + "step": 3852 + }, + { + "epoch": 1.452865761689291, + "grad_norm": 0.40095916645874174, + "learning_rate": 7.147109745206452e-05, + "loss": 0.4373, + "step": 3853 + }, + { + "epoch": 1.4532428355957767, + "grad_norm": 0.4216342660190462, + "learning_rate": 7.146459692614301e-05, + "loss": 0.4526, + "step": 3854 + }, + { + "epoch": 1.4536199095022624, + "grad_norm": 0.4411629204334967, + "learning_rate": 7.145809421975284e-05, + "loss": 0.4407, + "step": 3855 + }, + { + "epoch": 1.453996983408748, + "grad_norm": 0.42415995098791515, + "learning_rate": 7.145158933334465e-05, + "loss": 0.4351, + "step": 3856 + }, + { + "epoch": 1.4543740573152337, + "grad_norm": 0.34486512057316554, + "learning_rate": 7.14450822673692e-05, + "loss": 0.4451, + "step": 3857 + }, + { + "epoch": 1.4547511312217194, + "grad_norm": 0.3398110608059866, + "learning_rate": 7.143857302227746e-05, + "loss": 0.4377, + "step": 3858 + }, + { + "epoch": 1.455128205128205, + "grad_norm": 0.30969637880019735, + "learning_rate": 7.143206159852048e-05, + "loss": 0.4428, + "step": 3859 + }, + { + "epoch": 1.4555052790346907, + "grad_norm": 0.35698291112606245, + "learning_rate": 7.142554799654952e-05, + "loss": 0.4417, + "step": 3860 + }, + { + "epoch": 1.4558823529411764, + "grad_norm": 0.4172435719726421, + "learning_rate": 7.141903221681595e-05, + "loss": 0.4344, + "step": 3861 + }, + { + "epoch": 1.456259426847662, + "grad_norm": 0.48822403231851863, + "learning_rate": 7.141251425977134e-05, + "loss": 0.4394, + "step": 3862 + }, + { + "epoch": 1.4566365007541477, + "grad_norm": 0.5130786000498097, + "learning_rate": 7.140599412586733e-05, + "loss": 0.4319, + "step": 3863 + }, + { + "epoch": 1.4570135746606334, + "grad_norm": 0.37941802360557136, + "learning_rate": 7.13994718155558e-05, + "loss": 0.4621, + "step": 3864 + }, + { + "epoch": 1.457390648567119, + "grad_norm": 0.25889995633062335, + "learning_rate": 7.13929473292887e-05, + "loss": 0.4376, + "step": 3865 + }, + { + "epoch": 1.4577677224736048, + "grad_norm": 0.3596725341224806, + "learning_rate": 7.138642066751822e-05, + "loss": 0.4515, + "step": 3866 + }, + { + "epoch": 1.4581447963800904, + "grad_norm": 0.5740208217172762, + "learning_rate": 7.137989183069662e-05, + "loss": 0.4535, + "step": 3867 + }, + { + "epoch": 1.458521870286576, + "grad_norm": 0.6562857731128946, + "learning_rate": 7.137336081927635e-05, + "loss": 0.4415, + "step": 3868 + }, + { + "epoch": 1.4588989441930618, + "grad_norm": 0.5643475320365131, + "learning_rate": 7.136682763371e-05, + "loss": 0.4471, + "step": 3869 + }, + { + "epoch": 1.4592760180995474, + "grad_norm": 0.4676575512313902, + "learning_rate": 7.136029227445032e-05, + "loss": 0.4572, + "step": 3870 + }, + { + "epoch": 1.459653092006033, + "grad_norm": 0.4130395508931317, + "learning_rate": 7.13537547419502e-05, + "loss": 0.4373, + "step": 3871 + }, + { + "epoch": 1.4600301659125188, + "grad_norm": 0.35065920413770485, + "learning_rate": 7.134721503666269e-05, + "loss": 0.4378, + "step": 3872 + }, + { + "epoch": 1.4604072398190044, + "grad_norm": 0.31725579852025626, + "learning_rate": 7.134067315904099e-05, + "loss": 0.4405, + "step": 3873 + }, + { + "epoch": 1.4607843137254901, + "grad_norm": 0.35044977462038146, + "learning_rate": 7.133412910953843e-05, + "loss": 0.4409, + "step": 3874 + }, + { + "epoch": 1.461161387631976, + "grad_norm": 0.3274453144814339, + "learning_rate": 7.132758288860852e-05, + "loss": 0.4381, + "step": 3875 + }, + { + "epoch": 1.4615384615384617, + "grad_norm": 0.36296666145470363, + "learning_rate": 7.13210344967049e-05, + "loss": 0.4511, + "step": 3876 + }, + { + "epoch": 1.4619155354449473, + "grad_norm": 0.4115883732478063, + "learning_rate": 7.131448393428136e-05, + "loss": 0.4488, + "step": 3877 + }, + { + "epoch": 1.462292609351433, + "grad_norm": 0.3467278769328098, + "learning_rate": 7.130793120179188e-05, + "loss": 0.4337, + "step": 3878 + }, + { + "epoch": 1.4626696832579187, + "grad_norm": 0.32366731681450966, + "learning_rate": 7.130137629969053e-05, + "loss": 0.4351, + "step": 3879 + }, + { + "epoch": 1.4630467571644044, + "grad_norm": 0.2722835612049922, + "learning_rate": 7.129481922843156e-05, + "loss": 0.4345, + "step": 3880 + }, + { + "epoch": 1.46342383107089, + "grad_norm": 0.239348462392053, + "learning_rate": 7.128825998846939e-05, + "loss": 0.4349, + "step": 3881 + }, + { + "epoch": 1.4638009049773757, + "grad_norm": 0.2631152088674806, + "learning_rate": 7.128169858025854e-05, + "loss": 0.4441, + "step": 3882 + }, + { + "epoch": 1.4641779788838614, + "grad_norm": 0.3578921726535083, + "learning_rate": 7.127513500425373e-05, + "loss": 0.4334, + "step": 3883 + }, + { + "epoch": 1.464555052790347, + "grad_norm": 0.39747511408023817, + "learning_rate": 7.126856926090981e-05, + "loss": 0.4378, + "step": 3884 + }, + { + "epoch": 1.4649321266968327, + "grad_norm": 0.42271793466774726, + "learning_rate": 7.126200135068177e-05, + "loss": 0.4409, + "step": 3885 + }, + { + "epoch": 1.4653092006033184, + "grad_norm": 0.39963359256057646, + "learning_rate": 7.125543127402477e-05, + "loss": 0.4478, + "step": 3886 + }, + { + "epoch": 1.465686274509804, + "grad_norm": 0.4025640885716322, + "learning_rate": 7.124885903139409e-05, + "loss": 0.4439, + "step": 3887 + }, + { + "epoch": 1.4660633484162897, + "grad_norm": 0.39514098048129676, + "learning_rate": 7.124228462324523e-05, + "loss": 0.4599, + "step": 3888 + }, + { + "epoch": 1.4664404223227754, + "grad_norm": 0.26594605124181725, + "learning_rate": 7.123570805003374e-05, + "loss": 0.4395, + "step": 3889 + }, + { + "epoch": 1.466817496229261, + "grad_norm": 0.34184956907612596, + "learning_rate": 7.122912931221539e-05, + "loss": 0.4345, + "step": 3890 + }, + { + "epoch": 1.4671945701357467, + "grad_norm": 0.49069253205506447, + "learning_rate": 7.122254841024607e-05, + "loss": 0.4431, + "step": 3891 + }, + { + "epoch": 1.4675716440422324, + "grad_norm": 0.5620066065902342, + "learning_rate": 7.121596534458185e-05, + "loss": 0.4467, + "step": 3892 + }, + { + "epoch": 1.467948717948718, + "grad_norm": 0.5468345856839049, + "learning_rate": 7.120938011567891e-05, + "loss": 0.4416, + "step": 3893 + }, + { + "epoch": 1.4683257918552037, + "grad_norm": 0.4368156941158934, + "learning_rate": 7.120279272399363e-05, + "loss": 0.437, + "step": 3894 + }, + { + "epoch": 1.4687028657616894, + "grad_norm": 0.30697564500821206, + "learning_rate": 7.119620316998247e-05, + "loss": 0.436, + "step": 3895 + }, + { + "epoch": 1.469079939668175, + "grad_norm": 0.25985498169657184, + "learning_rate": 7.118961145410211e-05, + "loss": 0.4459, + "step": 3896 + }, + { + "epoch": 1.4694570135746607, + "grad_norm": 0.28933338733250735, + "learning_rate": 7.118301757680934e-05, + "loss": 0.4601, + "step": 3897 + }, + { + "epoch": 1.4698340874811464, + "grad_norm": 0.3265316175153447, + "learning_rate": 7.117642153856111e-05, + "loss": 0.4417, + "step": 3898 + }, + { + "epoch": 1.470211161387632, + "grad_norm": 0.3819356456823847, + "learning_rate": 7.116982333981452e-05, + "loss": 0.4478, + "step": 3899 + }, + { + "epoch": 1.4705882352941178, + "grad_norm": 0.38412819551899735, + "learning_rate": 7.116322298102681e-05, + "loss": 0.4493, + "step": 3900 + }, + { + "epoch": 1.4709653092006034, + "grad_norm": 0.3269189612296741, + "learning_rate": 7.115662046265541e-05, + "loss": 0.4387, + "step": 3901 + }, + { + "epoch": 1.471342383107089, + "grad_norm": 0.26491002356058485, + "learning_rate": 7.115001578515784e-05, + "loss": 0.4414, + "step": 3902 + }, + { + "epoch": 1.4717194570135748, + "grad_norm": 0.22486237609526957, + "learning_rate": 7.114340894899181e-05, + "loss": 0.4373, + "step": 3903 + }, + { + "epoch": 1.4720965309200604, + "grad_norm": 0.2252708057063429, + "learning_rate": 7.113679995461515e-05, + "loss": 0.4395, + "step": 3904 + }, + { + "epoch": 1.472473604826546, + "grad_norm": 0.2895293844043787, + "learning_rate": 7.11301888024859e-05, + "loss": 0.4465, + "step": 3905 + }, + { + "epoch": 1.4728506787330318, + "grad_norm": 0.38923580223213305, + "learning_rate": 7.112357549306214e-05, + "loss": 0.4487, + "step": 3906 + }, + { + "epoch": 1.4732277526395174, + "grad_norm": 0.40041690775882366, + "learning_rate": 7.111696002680224e-05, + "loss": 0.453, + "step": 3907 + }, + { + "epoch": 1.473604826546003, + "grad_norm": 0.3448962627957126, + "learning_rate": 7.11103424041646e-05, + "loss": 0.4374, + "step": 3908 + }, + { + "epoch": 1.4739819004524888, + "grad_norm": 0.2591738231611585, + "learning_rate": 7.11037226256078e-05, + "loss": 0.4331, + "step": 3909 + }, + { + "epoch": 1.4743589743589745, + "grad_norm": 0.28834157255412063, + "learning_rate": 7.109710069159065e-05, + "loss": 0.459, + "step": 3910 + }, + { + "epoch": 1.4747360482654601, + "grad_norm": 0.3327465358960893, + "learning_rate": 7.109047660257198e-05, + "loss": 0.4385, + "step": 3911 + }, + { + "epoch": 1.4751131221719458, + "grad_norm": 0.3337144392384451, + "learning_rate": 7.108385035901088e-05, + "loss": 0.4525, + "step": 3912 + }, + { + "epoch": 1.4754901960784315, + "grad_norm": 0.33486784514160367, + "learning_rate": 7.107722196136649e-05, + "loss": 0.441, + "step": 3913 + }, + { + "epoch": 1.4758672699849171, + "grad_norm": 0.30886821464983694, + "learning_rate": 7.10705914100982e-05, + "loss": 0.4686, + "step": 3914 + }, + { + "epoch": 1.4762443438914028, + "grad_norm": 0.27667563581120025, + "learning_rate": 7.10639587056655e-05, + "loss": 0.4454, + "step": 3915 + }, + { + "epoch": 1.4766214177978885, + "grad_norm": 0.29129867648156854, + "learning_rate": 7.1057323848528e-05, + "loss": 0.4333, + "step": 3916 + }, + { + "epoch": 1.4769984917043741, + "grad_norm": 0.31609601618834976, + "learning_rate": 7.10506868391455e-05, + "loss": 0.4363, + "step": 3917 + }, + { + "epoch": 1.4773755656108598, + "grad_norm": 0.4161003335580922, + "learning_rate": 7.104404767797796e-05, + "loss": 0.4356, + "step": 3918 + }, + { + "epoch": 1.4777526395173455, + "grad_norm": 0.4857843719439363, + "learning_rate": 7.103740636548544e-05, + "loss": 0.4339, + "step": 3919 + }, + { + "epoch": 1.4781297134238311, + "grad_norm": 0.4505068456769768, + "learning_rate": 7.103076290212819e-05, + "loss": 0.4442, + "step": 3920 + }, + { + "epoch": 1.4785067873303168, + "grad_norm": 0.3709337148361179, + "learning_rate": 7.102411728836659e-05, + "loss": 0.4517, + "step": 3921 + }, + { + "epoch": 1.4788838612368025, + "grad_norm": 0.2891228578054361, + "learning_rate": 7.101746952466119e-05, + "loss": 0.4455, + "step": 3922 + }, + { + "epoch": 1.4792609351432882, + "grad_norm": 0.27519131168561767, + "learning_rate": 7.101081961147266e-05, + "loss": 0.4363, + "step": 3923 + }, + { + "epoch": 1.4796380090497738, + "grad_norm": 0.32540631018431515, + "learning_rate": 7.100416754926185e-05, + "loss": 0.4568, + "step": 3924 + }, + { + "epoch": 1.4800150829562595, + "grad_norm": 0.3406133257898234, + "learning_rate": 7.099751333848972e-05, + "loss": 0.4437, + "step": 3925 + }, + { + "epoch": 1.4803921568627452, + "grad_norm": 0.3451327052001363, + "learning_rate": 7.099085697961739e-05, + "loss": 0.4358, + "step": 3926 + }, + { + "epoch": 1.4807692307692308, + "grad_norm": 0.4095929657429527, + "learning_rate": 7.098419847310618e-05, + "loss": 0.4443, + "step": 3927 + }, + { + "epoch": 1.4811463046757165, + "grad_norm": 0.48447908900683956, + "learning_rate": 7.097753781941751e-05, + "loss": 0.448, + "step": 3928 + }, + { + "epoch": 1.4815233785822022, + "grad_norm": 0.47396925076381274, + "learning_rate": 7.097087501901293e-05, + "loss": 0.4407, + "step": 3929 + }, + { + "epoch": 1.4819004524886878, + "grad_norm": 0.42041469791102226, + "learning_rate": 7.09642100723542e-05, + "loss": 0.4415, + "step": 3930 + }, + { + "epoch": 1.4822775263951735, + "grad_norm": 0.3311910244183117, + "learning_rate": 7.095754297990315e-05, + "loss": 0.4414, + "step": 3931 + }, + { + "epoch": 1.4826546003016592, + "grad_norm": 0.2637490621048268, + "learning_rate": 7.095087374212184e-05, + "loss": 0.4553, + "step": 3932 + }, + { + "epoch": 1.4830316742081449, + "grad_norm": 0.2397007991704973, + "learning_rate": 7.094420235947244e-05, + "loss": 0.4452, + "step": 3933 + }, + { + "epoch": 1.4834087481146305, + "grad_norm": 0.25054581429573075, + "learning_rate": 7.093752883241726e-05, + "loss": 0.4514, + "step": 3934 + }, + { + "epoch": 1.4837858220211162, + "grad_norm": 0.28546023774244433, + "learning_rate": 7.093085316141878e-05, + "loss": 0.447, + "step": 3935 + }, + { + "epoch": 1.4841628959276019, + "grad_norm": 0.3333054657124553, + "learning_rate": 7.09241753469396e-05, + "loss": 0.4458, + "step": 3936 + }, + { + "epoch": 1.4845399698340875, + "grad_norm": 0.3376075618950727, + "learning_rate": 7.091749538944252e-05, + "loss": 0.4412, + "step": 3937 + }, + { + "epoch": 1.4849170437405732, + "grad_norm": 0.32886217083168595, + "learning_rate": 7.09108132893904e-05, + "loss": 0.4382, + "step": 3938 + }, + { + "epoch": 1.4852941176470589, + "grad_norm": 0.35195792182165225, + "learning_rate": 7.090412904724636e-05, + "loss": 0.4525, + "step": 3939 + }, + { + "epoch": 1.4856711915535445, + "grad_norm": 0.39098536649940885, + "learning_rate": 7.089744266347358e-05, + "loss": 0.4435, + "step": 3940 + }, + { + "epoch": 1.4860482654600302, + "grad_norm": 0.37283011210605305, + "learning_rate": 7.089075413853544e-05, + "loss": 0.4351, + "step": 3941 + }, + { + "epoch": 1.4864253393665159, + "grad_norm": 0.2715480464167121, + "learning_rate": 7.088406347289543e-05, + "loss": 0.4389, + "step": 3942 + }, + { + "epoch": 1.4868024132730016, + "grad_norm": 0.30119328045912586, + "learning_rate": 7.087737066701723e-05, + "loss": 0.4443, + "step": 3943 + }, + { + "epoch": 1.4871794871794872, + "grad_norm": 0.3365133494096851, + "learning_rate": 7.087067572136461e-05, + "loss": 0.4461, + "step": 3944 + }, + { + "epoch": 1.487556561085973, + "grad_norm": 0.2468091769346131, + "learning_rate": 7.086397863640156e-05, + "loss": 0.4266, + "step": 3945 + }, + { + "epoch": 1.4879336349924586, + "grad_norm": 0.25021440776655074, + "learning_rate": 7.085727941259216e-05, + "loss": 0.43, + "step": 3946 + }, + { + "epoch": 1.4883107088989442, + "grad_norm": 0.33412690069128587, + "learning_rate": 7.085057805040065e-05, + "loss": 0.4545, + "step": 3947 + }, + { + "epoch": 1.48868778280543, + "grad_norm": 0.4169221944888126, + "learning_rate": 7.084387455029146e-05, + "loss": 0.4477, + "step": 3948 + }, + { + "epoch": 1.4890648567119156, + "grad_norm": 0.39466020219101183, + "learning_rate": 7.083716891272913e-05, + "loss": 0.4477, + "step": 3949 + }, + { + "epoch": 1.4894419306184012, + "grad_norm": 0.3197759884239231, + "learning_rate": 7.083046113817834e-05, + "loss": 0.4443, + "step": 3950 + }, + { + "epoch": 1.489819004524887, + "grad_norm": 0.247575385180991, + "learning_rate": 7.082375122710394e-05, + "loss": 0.4377, + "step": 3951 + }, + { + "epoch": 1.4901960784313726, + "grad_norm": 0.303314884776533, + "learning_rate": 7.081703917997094e-05, + "loss": 0.4484, + "step": 3952 + }, + { + "epoch": 1.4905731523378583, + "grad_norm": 0.41887737856001, + "learning_rate": 7.081032499724443e-05, + "loss": 0.4461, + "step": 3953 + }, + { + "epoch": 1.490950226244344, + "grad_norm": 0.4917973585582202, + "learning_rate": 7.080360867938973e-05, + "loss": 0.434, + "step": 3954 + }, + { + "epoch": 1.4913273001508296, + "grad_norm": 0.5501593553647864, + "learning_rate": 7.079689022687229e-05, + "loss": 0.4421, + "step": 3955 + }, + { + "epoch": 1.4917043740573153, + "grad_norm": 0.5590893734681295, + "learning_rate": 7.079016964015765e-05, + "loss": 0.4357, + "step": 3956 + }, + { + "epoch": 1.492081447963801, + "grad_norm": 0.573202892956801, + "learning_rate": 7.078344691971158e-05, + "loss": 0.449, + "step": 3957 + }, + { + "epoch": 1.4924585218702866, + "grad_norm": 0.5366228705850121, + "learning_rate": 7.077672206599994e-05, + "loss": 0.4491, + "step": 3958 + }, + { + "epoch": 1.4928355957767723, + "grad_norm": 0.43480181806847046, + "learning_rate": 7.076999507948876e-05, + "loss": 0.4443, + "step": 3959 + }, + { + "epoch": 1.493212669683258, + "grad_norm": 0.2821327729002442, + "learning_rate": 7.076326596064421e-05, + "loss": 0.4421, + "step": 3960 + }, + { + "epoch": 1.4935897435897436, + "grad_norm": 0.26700308086045554, + "learning_rate": 7.075653470993263e-05, + "loss": 0.4462, + "step": 3961 + }, + { + "epoch": 1.4939668174962293, + "grad_norm": 0.302856657634353, + "learning_rate": 7.074980132782046e-05, + "loss": 0.4487, + "step": 3962 + }, + { + "epoch": 1.494343891402715, + "grad_norm": 0.2815760087894496, + "learning_rate": 7.074306581477434e-05, + "loss": 0.4526, + "step": 3963 + }, + { + "epoch": 1.4947209653092006, + "grad_norm": 0.26434664836566696, + "learning_rate": 7.073632817126102e-05, + "loss": 0.4416, + "step": 3964 + }, + { + "epoch": 1.4950980392156863, + "grad_norm": 0.3068853724470137, + "learning_rate": 7.072958839774744e-05, + "loss": 0.4394, + "step": 3965 + }, + { + "epoch": 1.495475113122172, + "grad_norm": 0.3231174532785238, + "learning_rate": 7.072284649470062e-05, + "loss": 0.4407, + "step": 3966 + }, + { + "epoch": 1.4958521870286576, + "grad_norm": 0.34868107158610656, + "learning_rate": 7.07161024625878e-05, + "loss": 0.4502, + "step": 3967 + }, + { + "epoch": 1.4962292609351433, + "grad_norm": 0.3641658975221842, + "learning_rate": 7.070935630187633e-05, + "loss": 0.4467, + "step": 3968 + }, + { + "epoch": 1.496606334841629, + "grad_norm": 0.31076747664137516, + "learning_rate": 7.070260801303371e-05, + "loss": 0.4522, + "step": 3969 + }, + { + "epoch": 1.4969834087481146, + "grad_norm": 0.302675619226733, + "learning_rate": 7.069585759652758e-05, + "loss": 0.4487, + "step": 3970 + }, + { + "epoch": 1.4973604826546003, + "grad_norm": 0.31185814290732355, + "learning_rate": 7.068910505282573e-05, + "loss": 0.4326, + "step": 3971 + }, + { + "epoch": 1.497737556561086, + "grad_norm": 0.3577820425000388, + "learning_rate": 7.068235038239614e-05, + "loss": 0.4411, + "step": 3972 + }, + { + "epoch": 1.4981146304675717, + "grad_norm": 0.44712610536159597, + "learning_rate": 7.067559358570687e-05, + "loss": 0.4455, + "step": 3973 + }, + { + "epoch": 1.4984917043740573, + "grad_norm": 0.5229882448912352, + "learning_rate": 7.06688346632262e-05, + "loss": 0.4352, + "step": 3974 + }, + { + "epoch": 1.498868778280543, + "grad_norm": 0.5427825555840445, + "learning_rate": 7.066207361542246e-05, + "loss": 0.4452, + "step": 3975 + }, + { + "epoch": 1.4992458521870287, + "grad_norm": 0.5562366070178245, + "learning_rate": 7.065531044276424e-05, + "loss": 0.443, + "step": 3976 + }, + { + "epoch": 1.4996229260935143, + "grad_norm": 0.5289066365239034, + "learning_rate": 7.064854514572018e-05, + "loss": 0.4442, + "step": 3977 + }, + { + "epoch": 1.5, + "grad_norm": 0.44409489883900544, + "learning_rate": 7.064177772475912e-05, + "loss": 0.4461, + "step": 3978 + }, + { + "epoch": 1.5003770739064857, + "grad_norm": 0.32519735331152555, + "learning_rate": 7.063500818035006e-05, + "loss": 0.4221, + "step": 3979 + }, + { + "epoch": 1.5007541478129713, + "grad_norm": 0.2548466913036733, + "learning_rate": 7.06282365129621e-05, + "loss": 0.443, + "step": 3980 + }, + { + "epoch": 1.501131221719457, + "grad_norm": 0.28569519716173347, + "learning_rate": 7.062146272306452e-05, + "loss": 0.4496, + "step": 3981 + }, + { + "epoch": 1.5015082956259427, + "grad_norm": 0.3210042999579412, + "learning_rate": 7.061468681112674e-05, + "loss": 0.4421, + "step": 3982 + }, + { + "epoch": 1.5018853695324283, + "grad_norm": 0.36056799212574525, + "learning_rate": 7.060790877761831e-05, + "loss": 0.4422, + "step": 3983 + }, + { + "epoch": 1.502262443438914, + "grad_norm": 0.33471971484875995, + "learning_rate": 7.060112862300896e-05, + "loss": 0.4491, + "step": 3984 + }, + { + "epoch": 1.5026395173453997, + "grad_norm": 0.27267775955060924, + "learning_rate": 7.059434634776853e-05, + "loss": 0.4416, + "step": 3985 + }, + { + "epoch": 1.5030165912518854, + "grad_norm": 0.23016232803888043, + "learning_rate": 7.058756195236704e-05, + "loss": 0.4498, + "step": 3986 + }, + { + "epoch": 1.503393665158371, + "grad_norm": 0.27555415453999327, + "learning_rate": 7.058077543727463e-05, + "loss": 0.4495, + "step": 3987 + }, + { + "epoch": 1.5037707390648567, + "grad_norm": 0.2957092346805793, + "learning_rate": 7.057398680296163e-05, + "loss": 0.4424, + "step": 3988 + }, + { + "epoch": 1.5041478129713424, + "grad_norm": 0.24061947961527072, + "learning_rate": 7.056719604989843e-05, + "loss": 0.4279, + "step": 3989 + }, + { + "epoch": 1.504524886877828, + "grad_norm": 0.2632831347533601, + "learning_rate": 7.056040317855567e-05, + "loss": 0.4323, + "step": 3990 + }, + { + "epoch": 1.5049019607843137, + "grad_norm": 0.3271375638922219, + "learning_rate": 7.05536081894041e-05, + "loss": 0.4413, + "step": 3991 + }, + { + "epoch": 1.5052790346907994, + "grad_norm": 0.3803328953143058, + "learning_rate": 7.054681108291455e-05, + "loss": 0.4385, + "step": 3992 + }, + { + "epoch": 1.505656108597285, + "grad_norm": 0.4360394853623666, + "learning_rate": 7.054001185955811e-05, + "loss": 0.4423, + "step": 3993 + }, + { + "epoch": 1.5060331825037707, + "grad_norm": 0.4162787259705446, + "learning_rate": 7.053321051980592e-05, + "loss": 0.4343, + "step": 3994 + }, + { + "epoch": 1.5064102564102564, + "grad_norm": 0.44960366793289935, + "learning_rate": 7.052640706412935e-05, + "loss": 0.4422, + "step": 3995 + }, + { + "epoch": 1.506787330316742, + "grad_norm": 0.4801790239353092, + "learning_rate": 7.051960149299982e-05, + "loss": 0.4467, + "step": 3996 + }, + { + "epoch": 1.5071644042232277, + "grad_norm": 0.43713828601673094, + "learning_rate": 7.051279380688899e-05, + "loss": 0.4482, + "step": 3997 + }, + { + "epoch": 1.5075414781297134, + "grad_norm": 0.39628948822263227, + "learning_rate": 7.050598400626861e-05, + "loss": 0.4338, + "step": 3998 + }, + { + "epoch": 1.507918552036199, + "grad_norm": 0.38908497336987524, + "learning_rate": 7.049917209161059e-05, + "loss": 0.4412, + "step": 3999 + }, + { + "epoch": 1.5082956259426847, + "grad_norm": 0.46712527674969667, + "learning_rate": 7.049235806338701e-05, + "loss": 0.4417, + "step": 4000 + }, + { + "epoch": 1.5086726998491704, + "grad_norm": 0.40938823756470155, + "learning_rate": 7.048554192207006e-05, + "loss": 0.4406, + "step": 4001 + }, + { + "epoch": 1.509049773755656, + "grad_norm": 0.3549653879869995, + "learning_rate": 7.047872366813209e-05, + "loss": 0.4424, + "step": 4002 + }, + { + "epoch": 1.5094268476621417, + "grad_norm": 0.2875947806507905, + "learning_rate": 7.047190330204561e-05, + "loss": 0.4467, + "step": 4003 + }, + { + "epoch": 1.5098039215686274, + "grad_norm": 0.3265292429508642, + "learning_rate": 7.046508082428326e-05, + "loss": 0.4521, + "step": 4004 + }, + { + "epoch": 1.510180995475113, + "grad_norm": 0.3881323446734123, + "learning_rate": 7.045825623531782e-05, + "loss": 0.4414, + "step": 4005 + }, + { + "epoch": 1.5105580693815988, + "grad_norm": 0.3809249199996812, + "learning_rate": 7.045142953562226e-05, + "loss": 0.4386, + "step": 4006 + }, + { + "epoch": 1.5109351432880844, + "grad_norm": 0.37843323449282124, + "learning_rate": 7.044460072566963e-05, + "loss": 0.457, + "step": 4007 + }, + { + "epoch": 1.51131221719457, + "grad_norm": 0.32704972896286655, + "learning_rate": 7.043776980593317e-05, + "loss": 0.428, + "step": 4008 + }, + { + "epoch": 1.5116892911010558, + "grad_norm": 0.25194424567655205, + "learning_rate": 7.043093677688626e-05, + "loss": 0.4343, + "step": 4009 + }, + { + "epoch": 1.5120663650075414, + "grad_norm": 0.2428608616278345, + "learning_rate": 7.042410163900243e-05, + "loss": 0.4315, + "step": 4010 + }, + { + "epoch": 1.512443438914027, + "grad_norm": 0.2709325643710776, + "learning_rate": 7.041726439275536e-05, + "loss": 0.435, + "step": 4011 + }, + { + "epoch": 1.5128205128205128, + "grad_norm": 0.3353389773626573, + "learning_rate": 7.041042503861881e-05, + "loss": 0.4587, + "step": 4012 + }, + { + "epoch": 1.5131975867269984, + "grad_norm": 0.37604842039436354, + "learning_rate": 7.040358357706681e-05, + "loss": 0.4497, + "step": 4013 + }, + { + "epoch": 1.5135746606334841, + "grad_norm": 0.3273135065418242, + "learning_rate": 7.039674000857343e-05, + "loss": 0.448, + "step": 4014 + }, + { + "epoch": 1.5139517345399698, + "grad_norm": 0.29818719865618687, + "learning_rate": 7.038989433361292e-05, + "loss": 0.4303, + "step": 4015 + }, + { + "epoch": 1.5143288084464555, + "grad_norm": 0.3310864071429825, + "learning_rate": 7.03830465526597e-05, + "loss": 0.4431, + "step": 4016 + }, + { + "epoch": 1.5147058823529411, + "grad_norm": 0.3443526716459486, + "learning_rate": 7.037619666618829e-05, + "loss": 0.442, + "step": 4017 + }, + { + "epoch": 1.5150829562594268, + "grad_norm": 0.3253436630578418, + "learning_rate": 7.036934467467341e-05, + "loss": 0.4272, + "step": 4018 + }, + { + "epoch": 1.5154600301659125, + "grad_norm": 0.33545270581573416, + "learning_rate": 7.036249057858986e-05, + "loss": 0.4429, + "step": 4019 + }, + { + "epoch": 1.5158371040723981, + "grad_norm": 0.348665160305097, + "learning_rate": 7.035563437841266e-05, + "loss": 0.4366, + "step": 4020 + }, + { + "epoch": 1.5162141779788838, + "grad_norm": 0.368611289833, + "learning_rate": 7.034877607461692e-05, + "loss": 0.4326, + "step": 4021 + }, + { + "epoch": 1.5165912518853695, + "grad_norm": 0.387176212540118, + "learning_rate": 7.034191566767792e-05, + "loss": 0.4316, + "step": 4022 + }, + { + "epoch": 1.5169683257918551, + "grad_norm": 0.3662473508189614, + "learning_rate": 7.033505315807107e-05, + "loss": 0.437, + "step": 4023 + }, + { + "epoch": 1.5173453996983408, + "grad_norm": 0.3345140001290856, + "learning_rate": 7.032818854627195e-05, + "loss": 0.4526, + "step": 4024 + }, + { + "epoch": 1.5177224736048265, + "grad_norm": 0.2745546990168941, + "learning_rate": 7.032132183275627e-05, + "loss": 0.4495, + "step": 4025 + }, + { + "epoch": 1.5180995475113122, + "grad_norm": 0.23987750885479903, + "learning_rate": 7.031445301799988e-05, + "loss": 0.4359, + "step": 4026 + }, + { + "epoch": 1.5184766214177978, + "grad_norm": 0.2267394144994783, + "learning_rate": 7.030758210247877e-05, + "loss": 0.4386, + "step": 4027 + }, + { + "epoch": 1.5188536953242835, + "grad_norm": 0.27723017853512316, + "learning_rate": 7.030070908666912e-05, + "loss": 0.4487, + "step": 4028 + }, + { + "epoch": 1.5192307692307692, + "grad_norm": 0.3319862123495655, + "learning_rate": 7.029383397104721e-05, + "loss": 0.4309, + "step": 4029 + }, + { + "epoch": 1.5196078431372548, + "grad_norm": 0.29376941066276996, + "learning_rate": 7.028695675608946e-05, + "loss": 0.4476, + "step": 4030 + }, + { + "epoch": 1.5199849170437405, + "grad_norm": 0.4124740135334057, + "learning_rate": 7.028007744227247e-05, + "loss": 0.4445, + "step": 4031 + }, + { + "epoch": 1.5203619909502262, + "grad_norm": 0.5419984469982921, + "learning_rate": 7.027319603007299e-05, + "loss": 0.4504, + "step": 4032 + }, + { + "epoch": 1.5207390648567118, + "grad_norm": 0.5653405676016576, + "learning_rate": 7.026631251996787e-05, + "loss": 0.4436, + "step": 4033 + }, + { + "epoch": 1.5211161387631975, + "grad_norm": 0.4792597035005107, + "learning_rate": 7.025942691243412e-05, + "loss": 0.4407, + "step": 4034 + }, + { + "epoch": 1.5214932126696832, + "grad_norm": 0.42647642466258684, + "learning_rate": 7.025253920794894e-05, + "loss": 0.4478, + "step": 4035 + }, + { + "epoch": 1.5218702865761689, + "grad_norm": 0.4119713186386823, + "learning_rate": 7.024564940698963e-05, + "loss": 0.4472, + "step": 4036 + }, + { + "epoch": 1.5222473604826545, + "grad_norm": 0.37370387570001673, + "learning_rate": 7.023875751003365e-05, + "loss": 0.4418, + "step": 4037 + }, + { + "epoch": 1.5226244343891402, + "grad_norm": 0.3515051088415819, + "learning_rate": 7.023186351755858e-05, + "loss": 0.4492, + "step": 4038 + }, + { + "epoch": 1.5230015082956259, + "grad_norm": 0.34461709950709735, + "learning_rate": 7.022496743004219e-05, + "loss": 0.4348, + "step": 4039 + }, + { + "epoch": 1.5233785822021115, + "grad_norm": 0.375210674355887, + "learning_rate": 7.021806924796237e-05, + "loss": 0.4496, + "step": 4040 + }, + { + "epoch": 1.5237556561085972, + "grad_norm": 0.38374912959944457, + "learning_rate": 7.021116897179715e-05, + "loss": 0.4422, + "step": 4041 + }, + { + "epoch": 1.5241327300150829, + "grad_norm": 0.30222035249726603, + "learning_rate": 7.020426660202473e-05, + "loss": 0.4494, + "step": 4042 + }, + { + "epoch": 1.5245098039215685, + "grad_norm": 0.2866619422733543, + "learning_rate": 7.019736213912341e-05, + "loss": 0.4443, + "step": 4043 + }, + { + "epoch": 1.5248868778280542, + "grad_norm": 0.349949359409903, + "learning_rate": 7.019045558357169e-05, + "loss": 0.4441, + "step": 4044 + }, + { + "epoch": 1.5252639517345399, + "grad_norm": 0.4134362297749217, + "learning_rate": 7.018354693584818e-05, + "loss": 0.4432, + "step": 4045 + }, + { + "epoch": 1.5256410256410255, + "grad_norm": 0.4250806987253934, + "learning_rate": 7.017663619643161e-05, + "loss": 0.4506, + "step": 4046 + }, + { + "epoch": 1.5260180995475112, + "grad_norm": 0.37802354611304106, + "learning_rate": 7.016972336580095e-05, + "loss": 0.4376, + "step": 4047 + }, + { + "epoch": 1.526395173453997, + "grad_norm": 0.39729240785963926, + "learning_rate": 7.016280844443521e-05, + "loss": 0.4398, + "step": 4048 + }, + { + "epoch": 1.5267722473604826, + "grad_norm": 0.43907826902655867, + "learning_rate": 7.015589143281359e-05, + "loss": 0.4361, + "step": 4049 + }, + { + "epoch": 1.5271493212669682, + "grad_norm": 0.40242870011163334, + "learning_rate": 7.014897233141546e-05, + "loss": 0.4412, + "step": 4050 + }, + { + "epoch": 1.527526395173454, + "grad_norm": 0.4137800263449328, + "learning_rate": 7.014205114072027e-05, + "loss": 0.4404, + "step": 4051 + }, + { + "epoch": 1.5279034690799396, + "grad_norm": 0.43426450070806927, + "learning_rate": 7.01351278612077e-05, + "loss": 0.4401, + "step": 4052 + }, + { + "epoch": 1.5282805429864252, + "grad_norm": 0.39312746021579953, + "learning_rate": 7.012820249335748e-05, + "loss": 0.4467, + "step": 4053 + }, + { + "epoch": 1.528657616892911, + "grad_norm": 0.37500052481287727, + "learning_rate": 7.012127503764956e-05, + "loss": 0.4485, + "step": 4054 + }, + { + "epoch": 1.5290346907993966, + "grad_norm": 0.4335169592462598, + "learning_rate": 7.011434549456399e-05, + "loss": 0.4405, + "step": 4055 + }, + { + "epoch": 1.5294117647058822, + "grad_norm": 0.4177905296226285, + "learning_rate": 7.010741386458099e-05, + "loss": 0.4399, + "step": 4056 + }, + { + "epoch": 1.529788838612368, + "grad_norm": 0.31943769351015505, + "learning_rate": 7.010048014818091e-05, + "loss": 0.4397, + "step": 4057 + }, + { + "epoch": 1.5301659125188536, + "grad_norm": 0.27480123785420746, + "learning_rate": 7.009354434584426e-05, + "loss": 0.444, + "step": 4058 + }, + { + "epoch": 1.5305429864253393, + "grad_norm": 0.23065431889796742, + "learning_rate": 7.008660645805167e-05, + "loss": 0.4312, + "step": 4059 + }, + { + "epoch": 1.530920060331825, + "grad_norm": 0.2601681765630667, + "learning_rate": 7.007966648528396e-05, + "loss": 0.4361, + "step": 4060 + }, + { + "epoch": 1.5312971342383106, + "grad_norm": 0.2987894312077151, + "learning_rate": 7.007272442802203e-05, + "loss": 0.4406, + "step": 4061 + }, + { + "epoch": 1.5316742081447963, + "grad_norm": 0.387975815183997, + "learning_rate": 7.006578028674696e-05, + "loss": 0.4549, + "step": 4062 + }, + { + "epoch": 1.532051282051282, + "grad_norm": 0.4664919132825584, + "learning_rate": 7.005883406194001e-05, + "loss": 0.4413, + "step": 4063 + }, + { + "epoch": 1.5324283559577676, + "grad_norm": 0.4497094504276014, + "learning_rate": 7.00518857540825e-05, + "loss": 0.438, + "step": 4064 + }, + { + "epoch": 1.5328054298642533, + "grad_norm": 0.3676551013561914, + "learning_rate": 7.004493536365598e-05, + "loss": 0.4335, + "step": 4065 + }, + { + "epoch": 1.533182503770739, + "grad_norm": 0.28174947140807044, + "learning_rate": 7.003798289114207e-05, + "loss": 0.4337, + "step": 4066 + }, + { + "epoch": 1.5335595776772246, + "grad_norm": 0.23346359643433956, + "learning_rate": 7.003102833702261e-05, + "loss": 0.4403, + "step": 4067 + }, + { + "epoch": 1.5339366515837103, + "grad_norm": 0.24413264186975173, + "learning_rate": 7.002407170177952e-05, + "loss": 0.4264, + "step": 4068 + }, + { + "epoch": 1.534313725490196, + "grad_norm": 0.2618132511039234, + "learning_rate": 7.001711298589488e-05, + "loss": 0.4404, + "step": 4069 + }, + { + "epoch": 1.5346907993966816, + "grad_norm": 0.2720058062436509, + "learning_rate": 7.001015218985095e-05, + "loss": 0.4329, + "step": 4070 + }, + { + "epoch": 1.5350678733031673, + "grad_norm": 0.29869188348655684, + "learning_rate": 7.000318931413007e-05, + "loss": 0.4251, + "step": 4071 + }, + { + "epoch": 1.535444947209653, + "grad_norm": 0.33067532513886994, + "learning_rate": 6.99962243592148e-05, + "loss": 0.4482, + "step": 4072 + }, + { + "epoch": 1.5358220211161386, + "grad_norm": 0.3626394633624191, + "learning_rate": 6.998925732558778e-05, + "loss": 0.4438, + "step": 4073 + }, + { + "epoch": 1.5361990950226243, + "grad_norm": 0.3437204798844522, + "learning_rate": 6.998228821373185e-05, + "loss": 0.4378, + "step": 4074 + }, + { + "epoch": 1.53657616892911, + "grad_norm": 0.38632923150533316, + "learning_rate": 6.997531702412992e-05, + "loss": 0.4413, + "step": 4075 + }, + { + "epoch": 1.5369532428355956, + "grad_norm": 0.3585183642668211, + "learning_rate": 6.996834375726512e-05, + "loss": 0.4372, + "step": 4076 + }, + { + "epoch": 1.5373303167420813, + "grad_norm": 0.33966308048548194, + "learning_rate": 6.996136841362067e-05, + "loss": 0.4445, + "step": 4077 + }, + { + "epoch": 1.537707390648567, + "grad_norm": 0.3465160639649871, + "learning_rate": 6.995439099367997e-05, + "loss": 0.4429, + "step": 4078 + }, + { + "epoch": 1.5380844645550527, + "grad_norm": 0.32492260801283845, + "learning_rate": 6.994741149792653e-05, + "loss": 0.4507, + "step": 4079 + }, + { + "epoch": 1.5384615384615383, + "grad_norm": 0.27982428827644285, + "learning_rate": 6.994042992684406e-05, + "loss": 0.4431, + "step": 4080 + }, + { + "epoch": 1.538838612368024, + "grad_norm": 0.25224654869893426, + "learning_rate": 6.993344628091632e-05, + "loss": 0.4414, + "step": 4081 + }, + { + "epoch": 1.5392156862745097, + "grad_norm": 0.23184295228263735, + "learning_rate": 6.992646056062735e-05, + "loss": 0.4426, + "step": 4082 + }, + { + "epoch": 1.5395927601809953, + "grad_norm": 0.25049724587245314, + "learning_rate": 6.991947276646116e-05, + "loss": 0.4312, + "step": 4083 + }, + { + "epoch": 1.539969834087481, + "grad_norm": 0.26246777793082776, + "learning_rate": 6.991248289890208e-05, + "loss": 0.4468, + "step": 4084 + }, + { + "epoch": 1.5403469079939667, + "grad_norm": 0.3258785645015844, + "learning_rate": 6.990549095843446e-05, + "loss": 0.4385, + "step": 4085 + }, + { + "epoch": 1.5407239819004523, + "grad_norm": 0.3954892653192255, + "learning_rate": 6.989849694554282e-05, + "loss": 0.4311, + "step": 4086 + }, + { + "epoch": 1.541101055806938, + "grad_norm": 0.3715304664127426, + "learning_rate": 6.989150086071189e-05, + "loss": 0.4512, + "step": 4087 + }, + { + "epoch": 1.5414781297134237, + "grad_norm": 0.3594421939149424, + "learning_rate": 6.988450270442645e-05, + "loss": 0.4338, + "step": 4088 + }, + { + "epoch": 1.5418552036199094, + "grad_norm": 0.3843129685928035, + "learning_rate": 6.987750247717149e-05, + "loss": 0.4433, + "step": 4089 + }, + { + "epoch": 1.542232277526395, + "grad_norm": 0.37174434377408233, + "learning_rate": 6.98705001794321e-05, + "loss": 0.4384, + "step": 4090 + }, + { + "epoch": 1.5426093514328807, + "grad_norm": 0.27518283411200445, + "learning_rate": 6.986349581169355e-05, + "loss": 0.423, + "step": 4091 + }, + { + "epoch": 1.5429864253393664, + "grad_norm": 0.20866506060511672, + "learning_rate": 6.985648937444123e-05, + "loss": 0.4457, + "step": 4092 + }, + { + "epoch": 1.543363499245852, + "grad_norm": 0.2111448514569059, + "learning_rate": 6.984948086816067e-05, + "loss": 0.4444, + "step": 4093 + }, + { + "epoch": 1.5437405731523377, + "grad_norm": 0.1973209797080346, + "learning_rate": 6.984247029333757e-05, + "loss": 0.4299, + "step": 4094 + }, + { + "epoch": 1.5441176470588234, + "grad_norm": 0.19188514111893437, + "learning_rate": 6.983545765045774e-05, + "loss": 0.4429, + "step": 4095 + }, + { + "epoch": 1.544494720965309, + "grad_norm": 0.24146995543661645, + "learning_rate": 6.982844294000716e-05, + "loss": 0.439, + "step": 4096 + }, + { + "epoch": 1.5448717948717947, + "grad_norm": 0.2702914677319482, + "learning_rate": 6.982142616247195e-05, + "loss": 0.457, + "step": 4097 + }, + { + "epoch": 1.5452488687782804, + "grad_norm": 0.2996848026550202, + "learning_rate": 6.981440731833837e-05, + "loss": 0.441, + "step": 4098 + }, + { + "epoch": 1.545625942684766, + "grad_norm": 0.3717385578824724, + "learning_rate": 6.980738640809278e-05, + "loss": 0.4439, + "step": 4099 + }, + { + "epoch": 1.5460030165912517, + "grad_norm": 0.38713280402446915, + "learning_rate": 6.980036343222177e-05, + "loss": 0.4522, + "step": 4100 + }, + { + "epoch": 1.5463800904977374, + "grad_norm": 0.43175168312980033, + "learning_rate": 6.9793338391212e-05, + "loss": 0.4453, + "step": 4101 + }, + { + "epoch": 1.546757164404223, + "grad_norm": 0.4797747023221676, + "learning_rate": 6.97863112855503e-05, + "loss": 0.4444, + "step": 4102 + }, + { + "epoch": 1.547134238310709, + "grad_norm": 0.5483533306745701, + "learning_rate": 6.977928211572366e-05, + "loss": 0.431, + "step": 4103 + }, + { + "epoch": 1.5475113122171946, + "grad_norm": 0.5288279299394928, + "learning_rate": 6.977225088221918e-05, + "loss": 0.4234, + "step": 4104 + }, + { + "epoch": 1.5478883861236803, + "grad_norm": 0.4729234164379677, + "learning_rate": 6.976521758552412e-05, + "loss": 0.4307, + "step": 4105 + }, + { + "epoch": 1.548265460030166, + "grad_norm": 0.3632743237908047, + "learning_rate": 6.975818222612589e-05, + "loss": 0.4384, + "step": 4106 + }, + { + "epoch": 1.5486425339366516, + "grad_norm": 0.23636263451092893, + "learning_rate": 6.975114480451201e-05, + "loss": 0.4481, + "step": 4107 + }, + { + "epoch": 1.5490196078431373, + "grad_norm": 0.2322379984789217, + "learning_rate": 6.97441053211702e-05, + "loss": 0.4607, + "step": 4108 + }, + { + "epoch": 1.549396681749623, + "grad_norm": 0.26490557853941243, + "learning_rate": 6.973706377658827e-05, + "loss": 0.4317, + "step": 4109 + }, + { + "epoch": 1.5497737556561086, + "grad_norm": 0.3274326034546734, + "learning_rate": 6.97300201712542e-05, + "loss": 0.4442, + "step": 4110 + }, + { + "epoch": 1.5501508295625943, + "grad_norm": 0.3991209917432144, + "learning_rate": 6.97229745056561e-05, + "loss": 0.4472, + "step": 4111 + }, + { + "epoch": 1.55052790346908, + "grad_norm": 0.48311140874929853, + "learning_rate": 6.971592678028223e-05, + "loss": 0.4422, + "step": 4112 + }, + { + "epoch": 1.5509049773755657, + "grad_norm": 0.5591971738358802, + "learning_rate": 6.970887699562099e-05, + "loss": 0.4493, + "step": 4113 + }, + { + "epoch": 1.5512820512820513, + "grad_norm": 0.6391100739348847, + "learning_rate": 6.970182515216094e-05, + "loss": 0.4461, + "step": 4114 + }, + { + "epoch": 1.551659125188537, + "grad_norm": 0.6972001616546802, + "learning_rate": 6.969477125039073e-05, + "loss": 0.4391, + "step": 4115 + }, + { + "epoch": 1.5520361990950227, + "grad_norm": 0.6682769228420533, + "learning_rate": 6.968771529079922e-05, + "loss": 0.4341, + "step": 4116 + }, + { + "epoch": 1.5524132730015083, + "grad_norm": 0.6037077926475116, + "learning_rate": 6.968065727387538e-05, + "loss": 0.4387, + "step": 4117 + }, + { + "epoch": 1.552790346907994, + "grad_norm": 0.5238018535103407, + "learning_rate": 6.967359720010832e-05, + "loss": 0.4403, + "step": 4118 + }, + { + "epoch": 1.5531674208144797, + "grad_norm": 0.48227328599341307, + "learning_rate": 6.966653506998728e-05, + "loss": 0.4383, + "step": 4119 + }, + { + "epoch": 1.5535444947209653, + "grad_norm": 0.38695916441128575, + "learning_rate": 6.965947088400169e-05, + "loss": 0.446, + "step": 4120 + }, + { + "epoch": 1.553921568627451, + "grad_norm": 0.37921174745705416, + "learning_rate": 6.965240464264107e-05, + "loss": 0.4294, + "step": 4121 + }, + { + "epoch": 1.5542986425339367, + "grad_norm": 0.4342708122715024, + "learning_rate": 6.96453363463951e-05, + "loss": 0.4456, + "step": 4122 + }, + { + "epoch": 1.5546757164404224, + "grad_norm": 0.39168668989802796, + "learning_rate": 6.963826599575362e-05, + "loss": 0.435, + "step": 4123 + }, + { + "epoch": 1.555052790346908, + "grad_norm": 0.3124183702828326, + "learning_rate": 6.96311935912066e-05, + "loss": 0.4428, + "step": 4124 + }, + { + "epoch": 1.5554298642533937, + "grad_norm": 0.3095495080809144, + "learning_rate": 6.962411913324415e-05, + "loss": 0.4339, + "step": 4125 + }, + { + "epoch": 1.5558069381598794, + "grad_norm": 0.35166049851154846, + "learning_rate": 6.961704262235651e-05, + "loss": 0.4459, + "step": 4126 + }, + { + "epoch": 1.556184012066365, + "grad_norm": 0.34390239857013544, + "learning_rate": 6.96099640590341e-05, + "loss": 0.4382, + "step": 4127 + }, + { + "epoch": 1.5565610859728507, + "grad_norm": 0.2392759337105144, + "learning_rate": 6.960288344376744e-05, + "loss": 0.4364, + "step": 4128 + }, + { + "epoch": 1.5569381598793364, + "grad_norm": 0.29541461227996846, + "learning_rate": 6.959580077704721e-05, + "loss": 0.4318, + "step": 4129 + }, + { + "epoch": 1.557315233785822, + "grad_norm": 0.33763891269303786, + "learning_rate": 6.958871605936424e-05, + "loss": 0.4376, + "step": 4130 + }, + { + "epoch": 1.5576923076923077, + "grad_norm": 0.3236501824875437, + "learning_rate": 6.95816292912095e-05, + "loss": 0.4449, + "step": 4131 + }, + { + "epoch": 1.5580693815987934, + "grad_norm": 0.29170241395521856, + "learning_rate": 6.957454047307409e-05, + "loss": 0.4407, + "step": 4132 + }, + { + "epoch": 1.558446455505279, + "grad_norm": 0.2763172938228937, + "learning_rate": 6.956744960544925e-05, + "loss": 0.4372, + "step": 4133 + }, + { + "epoch": 1.5588235294117647, + "grad_norm": 0.3719576801177354, + "learning_rate": 6.956035668882637e-05, + "loss": 0.4351, + "step": 4134 + }, + { + "epoch": 1.5592006033182504, + "grad_norm": 0.3939708792824328, + "learning_rate": 6.955326172369701e-05, + "loss": 0.4422, + "step": 4135 + }, + { + "epoch": 1.559577677224736, + "grad_norm": 0.3482887882407022, + "learning_rate": 6.95461647105528e-05, + "loss": 0.4354, + "step": 4136 + }, + { + "epoch": 1.5599547511312217, + "grad_norm": 0.2981890349042097, + "learning_rate": 6.953906564988562e-05, + "loss": 0.4252, + "step": 4137 + }, + { + "epoch": 1.5603318250377074, + "grad_norm": 0.2658146200601253, + "learning_rate": 6.953196454218737e-05, + "loss": 0.4379, + "step": 4138 + }, + { + "epoch": 1.560708898944193, + "grad_norm": 0.26756416470337935, + "learning_rate": 6.952486138795015e-05, + "loss": 0.4397, + "step": 4139 + }, + { + "epoch": 1.5610859728506787, + "grad_norm": 0.23933591531367998, + "learning_rate": 6.951775618766625e-05, + "loss": 0.4409, + "step": 4140 + }, + { + "epoch": 1.5614630467571644, + "grad_norm": 0.24672818121378096, + "learning_rate": 6.951064894182804e-05, + "loss": 0.4385, + "step": 4141 + }, + { + "epoch": 1.56184012066365, + "grad_norm": 0.254272715374114, + "learning_rate": 6.950353965092802e-05, + "loss": 0.4386, + "step": 4142 + }, + { + "epoch": 1.5622171945701357, + "grad_norm": 0.25883427785811114, + "learning_rate": 6.949642831545887e-05, + "loss": 0.4331, + "step": 4143 + }, + { + "epoch": 1.5625942684766214, + "grad_norm": 0.22596287505580753, + "learning_rate": 6.948931493591339e-05, + "loss": 0.4367, + "step": 4144 + }, + { + "epoch": 1.562971342383107, + "grad_norm": 0.275124249365915, + "learning_rate": 6.948219951278456e-05, + "loss": 0.4342, + "step": 4145 + }, + { + "epoch": 1.5633484162895928, + "grad_norm": 0.3164696266063443, + "learning_rate": 6.947508204656546e-05, + "loss": 0.4308, + "step": 4146 + }, + { + "epoch": 1.5637254901960784, + "grad_norm": 0.3462349792892341, + "learning_rate": 6.946796253774931e-05, + "loss": 0.445, + "step": 4147 + }, + { + "epoch": 1.564102564102564, + "grad_norm": 0.36946922564238643, + "learning_rate": 6.946084098682948e-05, + "loss": 0.4439, + "step": 4148 + }, + { + "epoch": 1.5644796380090498, + "grad_norm": 0.370340054573371, + "learning_rate": 6.945371739429952e-05, + "loss": 0.4467, + "step": 4149 + }, + { + "epoch": 1.5648567119155354, + "grad_norm": 0.31042899883154634, + "learning_rate": 6.944659176065306e-05, + "loss": 0.4389, + "step": 4150 + }, + { + "epoch": 1.565233785822021, + "grad_norm": 0.30282662907699354, + "learning_rate": 6.943946408638392e-05, + "loss": 0.4347, + "step": 4151 + }, + { + "epoch": 1.5656108597285068, + "grad_norm": 0.35593832485065247, + "learning_rate": 6.943233437198603e-05, + "loss": 0.4558, + "step": 4152 + }, + { + "epoch": 1.5659879336349924, + "grad_norm": 0.3614150849833712, + "learning_rate": 6.942520261795347e-05, + "loss": 0.4302, + "step": 4153 + }, + { + "epoch": 1.5663650075414781, + "grad_norm": 0.344001445371523, + "learning_rate": 6.941806882478046e-05, + "loss": 0.4391, + "step": 4154 + }, + { + "epoch": 1.5667420814479638, + "grad_norm": 0.31410210795023724, + "learning_rate": 6.941093299296139e-05, + "loss": 0.4373, + "step": 4155 + }, + { + "epoch": 1.5671191553544495, + "grad_norm": 0.3124024692709055, + "learning_rate": 6.940379512299074e-05, + "loss": 0.4505, + "step": 4156 + }, + { + "epoch": 1.5674962292609351, + "grad_norm": 0.2543054317713581, + "learning_rate": 6.939665521536316e-05, + "loss": 0.4405, + "step": 4157 + }, + { + "epoch": 1.5678733031674208, + "grad_norm": 0.2524851220206609, + "learning_rate": 6.938951327057347e-05, + "loss": 0.4339, + "step": 4158 + }, + { + "epoch": 1.5682503770739065, + "grad_norm": 0.2861141041577763, + "learning_rate": 6.938236928911657e-05, + "loss": 0.4365, + "step": 4159 + }, + { + "epoch": 1.5686274509803921, + "grad_norm": 0.3214731255684187, + "learning_rate": 6.937522327148754e-05, + "loss": 0.4308, + "step": 4160 + }, + { + "epoch": 1.5690045248868778, + "grad_norm": 0.3520289034753588, + "learning_rate": 6.93680752181816e-05, + "loss": 0.4418, + "step": 4161 + }, + { + "epoch": 1.5693815987933635, + "grad_norm": 0.3802657985145378, + "learning_rate": 6.936092512969408e-05, + "loss": 0.4434, + "step": 4162 + }, + { + "epoch": 1.5697586726998491, + "grad_norm": 0.4079782951265708, + "learning_rate": 6.93537730065205e-05, + "loss": 0.4452, + "step": 4163 + }, + { + "epoch": 1.5701357466063348, + "grad_norm": 0.3879534772672815, + "learning_rate": 6.93466188491565e-05, + "loss": 0.4366, + "step": 4164 + }, + { + "epoch": 1.5705128205128205, + "grad_norm": 0.34346409878375217, + "learning_rate": 6.933946265809784e-05, + "loss": 0.4376, + "step": 4165 + }, + { + "epoch": 1.5708898944193062, + "grad_norm": 0.3207452855319987, + "learning_rate": 6.933230443384043e-05, + "loss": 0.4263, + "step": 4166 + }, + { + "epoch": 1.5712669683257918, + "grad_norm": 0.3099461020189314, + "learning_rate": 6.932514417688035e-05, + "loss": 0.4486, + "step": 4167 + }, + { + "epoch": 1.5716440422322775, + "grad_norm": 0.3697799134507283, + "learning_rate": 6.931798188771378e-05, + "loss": 0.4441, + "step": 4168 + }, + { + "epoch": 1.5720211161387632, + "grad_norm": 0.35728153967419574, + "learning_rate": 6.931081756683708e-05, + "loss": 0.4372, + "step": 4169 + }, + { + "epoch": 1.5723981900452488, + "grad_norm": 0.2900819712861887, + "learning_rate": 6.930365121474672e-05, + "loss": 0.4541, + "step": 4170 + }, + { + "epoch": 1.5727752639517345, + "grad_norm": 0.27895321976253395, + "learning_rate": 6.929648283193932e-05, + "loss": 0.4475, + "step": 4171 + }, + { + "epoch": 1.5731523378582202, + "grad_norm": 0.1763616085421161, + "learning_rate": 6.928931241891164e-05, + "loss": 0.441, + "step": 4172 + }, + { + "epoch": 1.5735294117647058, + "grad_norm": 0.26858806079757136, + "learning_rate": 6.928213997616059e-05, + "loss": 0.4397, + "step": 4173 + }, + { + "epoch": 1.5739064856711915, + "grad_norm": 0.33146384563641956, + "learning_rate": 6.927496550418322e-05, + "loss": 0.4517, + "step": 4174 + }, + { + "epoch": 1.5742835595776772, + "grad_norm": 0.3863592055615072, + "learning_rate": 6.926778900347672e-05, + "loss": 0.4362, + "step": 4175 + }, + { + "epoch": 1.5746606334841629, + "grad_norm": 0.4790733184692391, + "learning_rate": 6.926061047453838e-05, + "loss": 0.4366, + "step": 4176 + }, + { + "epoch": 1.5750377073906485, + "grad_norm": 0.5334041339876409, + "learning_rate": 6.925342991786568e-05, + "loss": 0.434, + "step": 4177 + }, + { + "epoch": 1.5754147812971342, + "grad_norm": 0.5758902288259653, + "learning_rate": 6.924624733395626e-05, + "loss": 0.4482, + "step": 4178 + }, + { + "epoch": 1.5757918552036199, + "grad_norm": 0.5736243087201014, + "learning_rate": 6.923906272330784e-05, + "loss": 0.4377, + "step": 4179 + }, + { + "epoch": 1.5761689291101055, + "grad_norm": 0.5190579860464932, + "learning_rate": 6.92318760864183e-05, + "loss": 0.4347, + "step": 4180 + }, + { + "epoch": 1.5765460030165912, + "grad_norm": 0.4107624480410451, + "learning_rate": 6.922468742378567e-05, + "loss": 0.4419, + "step": 4181 + }, + { + "epoch": 1.5769230769230769, + "grad_norm": 0.30944093791514027, + "learning_rate": 6.921749673590813e-05, + "loss": 0.444, + "step": 4182 + }, + { + "epoch": 1.5773001508295625, + "grad_norm": 0.25015327073308846, + "learning_rate": 6.921030402328399e-05, + "loss": 0.4301, + "step": 4183 + }, + { + "epoch": 1.5776772247360482, + "grad_norm": 0.2462074313009292, + "learning_rate": 6.920310928641167e-05, + "loss": 0.4346, + "step": 4184 + }, + { + "epoch": 1.5780542986425339, + "grad_norm": 0.2992759288106259, + "learning_rate": 6.91959125257898e-05, + "loss": 0.4452, + "step": 4185 + }, + { + "epoch": 1.5784313725490198, + "grad_norm": 0.3469927119352085, + "learning_rate": 6.918871374191708e-05, + "loss": 0.4324, + "step": 4186 + }, + { + "epoch": 1.5788084464555054, + "grad_norm": 0.35898525975320444, + "learning_rate": 6.918151293529239e-05, + "loss": 0.4345, + "step": 4187 + }, + { + "epoch": 1.5791855203619911, + "grad_norm": 0.33736463008272477, + "learning_rate": 6.917431010641473e-05, + "loss": 0.4399, + "step": 4188 + }, + { + "epoch": 1.5795625942684768, + "grad_norm": 0.3418869797684304, + "learning_rate": 6.916710525578327e-05, + "loss": 0.4386, + "step": 4189 + }, + { + "epoch": 1.5799396681749625, + "grad_norm": 0.31180162468756145, + "learning_rate": 6.915989838389728e-05, + "loss": 0.4584, + "step": 4190 + }, + { + "epoch": 1.5803167420814481, + "grad_norm": 0.25252429044041236, + "learning_rate": 6.91526894912562e-05, + "loss": 0.4388, + "step": 4191 + }, + { + "epoch": 1.5806938159879338, + "grad_norm": 0.25486367044038355, + "learning_rate": 6.914547857835959e-05, + "loss": 0.4411, + "step": 4192 + }, + { + "epoch": 1.5810708898944195, + "grad_norm": 0.2703174078469553, + "learning_rate": 6.913826564570716e-05, + "loss": 0.4484, + "step": 4193 + }, + { + "epoch": 1.5814479638009051, + "grad_norm": 0.2579810122830218, + "learning_rate": 6.913105069379878e-05, + "loss": 0.4423, + "step": 4194 + }, + { + "epoch": 1.5818250377073908, + "grad_norm": 0.25377958628782854, + "learning_rate": 6.912383372313443e-05, + "loss": 0.4398, + "step": 4195 + }, + { + "epoch": 1.5822021116138765, + "grad_norm": 0.2698624824797701, + "learning_rate": 6.911661473421422e-05, + "loss": 0.4423, + "step": 4196 + }, + { + "epoch": 1.5825791855203621, + "grad_norm": 0.22565002734925385, + "learning_rate": 6.910939372753845e-05, + "loss": 0.4339, + "step": 4197 + }, + { + "epoch": 1.5829562594268478, + "grad_norm": 0.2415683168850231, + "learning_rate": 6.91021707036075e-05, + "loss": 0.4318, + "step": 4198 + }, + { + "epoch": 1.5833333333333335, + "grad_norm": 0.2859620730720779, + "learning_rate": 6.909494566292195e-05, + "loss": 0.4473, + "step": 4199 + }, + { + "epoch": 1.5837104072398192, + "grad_norm": 0.3248449625527654, + "learning_rate": 6.908771860598247e-05, + "loss": 0.4339, + "step": 4200 + }, + { + "epoch": 1.5840874811463048, + "grad_norm": 0.35645045450336604, + "learning_rate": 6.908048953328989e-05, + "loss": 0.4508, + "step": 4201 + }, + { + "epoch": 1.5844645550527905, + "grad_norm": 0.331712208464722, + "learning_rate": 6.907325844534518e-05, + "loss": 0.4272, + "step": 4202 + }, + { + "epoch": 1.5848416289592762, + "grad_norm": 0.30771791922761926, + "learning_rate": 6.906602534264944e-05, + "loss": 0.4427, + "step": 4203 + }, + { + "epoch": 1.5852187028657618, + "grad_norm": 0.37685651148988253, + "learning_rate": 6.905879022570393e-05, + "loss": 0.4392, + "step": 4204 + }, + { + "epoch": 1.5855957767722475, + "grad_norm": 0.47877274682599236, + "learning_rate": 6.905155309501002e-05, + "loss": 0.4515, + "step": 4205 + }, + { + "epoch": 1.5859728506787332, + "grad_norm": 0.5005763193712642, + "learning_rate": 6.904431395106927e-05, + "loss": 0.4408, + "step": 4206 + }, + { + "epoch": 1.5863499245852188, + "grad_norm": 0.4603801286195965, + "learning_rate": 6.903707279438331e-05, + "loss": 0.4305, + "step": 4207 + }, + { + "epoch": 1.5867269984917045, + "grad_norm": 0.37320623530018737, + "learning_rate": 6.902982962545398e-05, + "loss": 0.4273, + "step": 4208 + }, + { + "epoch": 1.5871040723981902, + "grad_norm": 0.3217639727599722, + "learning_rate": 6.90225844447832e-05, + "loss": 0.4341, + "step": 4209 + }, + { + "epoch": 1.5874811463046758, + "grad_norm": 0.26666528687648416, + "learning_rate": 6.901533725287304e-05, + "loss": 0.4273, + "step": 4210 + }, + { + "epoch": 1.5878582202111615, + "grad_norm": 0.26842913524847767, + "learning_rate": 6.900808805022576e-05, + "loss": 0.437, + "step": 4211 + }, + { + "epoch": 1.5882352941176472, + "grad_norm": 0.2480627500196315, + "learning_rate": 6.900083683734372e-05, + "loss": 0.456, + "step": 4212 + }, + { + "epoch": 1.5886123680241329, + "grad_norm": 0.29715319532223017, + "learning_rate": 6.899358361472939e-05, + "loss": 0.4445, + "step": 4213 + }, + { + "epoch": 1.5889894419306185, + "grad_norm": 0.4301824523887715, + "learning_rate": 6.898632838288544e-05, + "loss": 0.4246, + "step": 4214 + }, + { + "epoch": 1.5893665158371042, + "grad_norm": 0.47804479823897505, + "learning_rate": 6.897907114231465e-05, + "loss": 0.4368, + "step": 4215 + }, + { + "epoch": 1.5897435897435899, + "grad_norm": 0.3597860613274733, + "learning_rate": 6.897181189351993e-05, + "loss": 0.4434, + "step": 4216 + }, + { + "epoch": 1.5901206636500755, + "grad_norm": 0.36306852519034033, + "learning_rate": 6.896455063700435e-05, + "loss": 0.4359, + "step": 4217 + }, + { + "epoch": 1.5904977375565612, + "grad_norm": 0.34165833097974396, + "learning_rate": 6.895728737327108e-05, + "loss": 0.4402, + "step": 4218 + }, + { + "epoch": 1.5908748114630469, + "grad_norm": 0.3121157046577594, + "learning_rate": 6.89500221028235e-05, + "loss": 0.4396, + "step": 4219 + }, + { + "epoch": 1.5912518853695325, + "grad_norm": 0.3294073900147279, + "learning_rate": 6.894275482616506e-05, + "loss": 0.4513, + "step": 4220 + }, + { + "epoch": 1.5916289592760182, + "grad_norm": 0.3936451592383952, + "learning_rate": 6.89354855437994e-05, + "loss": 0.4327, + "step": 4221 + }, + { + "epoch": 1.5920060331825039, + "grad_norm": 0.5173143740425545, + "learning_rate": 6.892821425623024e-05, + "loss": 0.4369, + "step": 4222 + }, + { + "epoch": 1.5923831070889896, + "grad_norm": 0.5699308920538618, + "learning_rate": 6.892094096396149e-05, + "loss": 0.4529, + "step": 4223 + }, + { + "epoch": 1.5927601809954752, + "grad_norm": 0.520198369030742, + "learning_rate": 6.89136656674972e-05, + "loss": 0.4488, + "step": 4224 + }, + { + "epoch": 1.593137254901961, + "grad_norm": 0.4952531514544274, + "learning_rate": 6.890638836734151e-05, + "loss": 0.4329, + "step": 4225 + }, + { + "epoch": 1.5935143288084466, + "grad_norm": 0.49652328722653016, + "learning_rate": 6.889910906399875e-05, + "loss": 0.4374, + "step": 4226 + }, + { + "epoch": 1.5938914027149322, + "grad_norm": 0.4859652981615277, + "learning_rate": 6.889182775797336e-05, + "loss": 0.4398, + "step": 4227 + }, + { + "epoch": 1.594268476621418, + "grad_norm": 0.49340273995333517, + "learning_rate": 6.888454444976995e-05, + "loss": 0.4541, + "step": 4228 + }, + { + "epoch": 1.5946455505279036, + "grad_norm": 0.5199378860615049, + "learning_rate": 6.887725913989322e-05, + "loss": 0.4462, + "step": 4229 + }, + { + "epoch": 1.5950226244343892, + "grad_norm": 0.4796672880602571, + "learning_rate": 6.886997182884803e-05, + "loss": 0.4363, + "step": 4230 + }, + { + "epoch": 1.595399698340875, + "grad_norm": 0.4101734022269485, + "learning_rate": 6.886268251713942e-05, + "loss": 0.4455, + "step": 4231 + }, + { + "epoch": 1.5957767722473606, + "grad_norm": 0.3411888080615738, + "learning_rate": 6.885539120527251e-05, + "loss": 0.4308, + "step": 4232 + }, + { + "epoch": 1.5961538461538463, + "grad_norm": 0.3087747536927945, + "learning_rate": 6.884809789375258e-05, + "loss": 0.4426, + "step": 4233 + }, + { + "epoch": 1.596530920060332, + "grad_norm": 0.3165486095060931, + "learning_rate": 6.884080258308507e-05, + "loss": 0.4442, + "step": 4234 + }, + { + "epoch": 1.5969079939668176, + "grad_norm": 0.3055711705352523, + "learning_rate": 6.883350527377551e-05, + "loss": 0.4407, + "step": 4235 + }, + { + "epoch": 1.5972850678733033, + "grad_norm": 0.2617207196157129, + "learning_rate": 6.88262059663296e-05, + "loss": 0.4335, + "step": 4236 + }, + { + "epoch": 1.597662141779789, + "grad_norm": 0.2757584956559199, + "learning_rate": 6.88189046612532e-05, + "loss": 0.4303, + "step": 4237 + }, + { + "epoch": 1.5980392156862746, + "grad_norm": 0.29392899156333363, + "learning_rate": 6.881160135905227e-05, + "loss": 0.4365, + "step": 4238 + }, + { + "epoch": 1.5984162895927603, + "grad_norm": 0.3068211767203431, + "learning_rate": 6.88042960602329e-05, + "loss": 0.4623, + "step": 4239 + }, + { + "epoch": 1.598793363499246, + "grad_norm": 0.31518909817859075, + "learning_rate": 6.879698876530139e-05, + "loss": 0.4294, + "step": 4240 + }, + { + "epoch": 1.5991704374057316, + "grad_norm": 0.3317662796260494, + "learning_rate": 6.878967947476409e-05, + "loss": 0.4411, + "step": 4241 + }, + { + "epoch": 1.5995475113122173, + "grad_norm": 0.3554215524298698, + "learning_rate": 6.878236818912752e-05, + "loss": 0.4329, + "step": 4242 + }, + { + "epoch": 1.599924585218703, + "grad_norm": 0.3619656593877066, + "learning_rate": 6.877505490889838e-05, + "loss": 0.4314, + "step": 4243 + }, + { + "epoch": 1.6003016591251886, + "grad_norm": 0.3251864085607395, + "learning_rate": 6.876773963458348e-05, + "loss": 0.4304, + "step": 4244 + }, + { + "epoch": 1.6006787330316743, + "grad_norm": 0.3025932505586567, + "learning_rate": 6.876042236668972e-05, + "loss": 0.4556, + "step": 4245 + }, + { + "epoch": 1.60105580693816, + "grad_norm": 0.2925769494920588, + "learning_rate": 6.875310310572419e-05, + "loss": 0.4477, + "step": 4246 + }, + { + "epoch": 1.6014328808446456, + "grad_norm": 0.24242365458426687, + "learning_rate": 6.874578185219413e-05, + "loss": 0.4344, + "step": 4247 + }, + { + "epoch": 1.6018099547511313, + "grad_norm": 0.27438133800466946, + "learning_rate": 6.873845860660689e-05, + "loss": 0.4367, + "step": 4248 + }, + { + "epoch": 1.602187028657617, + "grad_norm": 0.2701159974908222, + "learning_rate": 6.873113336946996e-05, + "loss": 0.4361, + "step": 4249 + }, + { + "epoch": 1.6025641025641026, + "grad_norm": 0.222518630685992, + "learning_rate": 6.872380614129097e-05, + "loss": 0.4389, + "step": 4250 + }, + { + "epoch": 1.6029411764705883, + "grad_norm": 0.22625845745780987, + "learning_rate": 6.871647692257768e-05, + "loss": 0.4437, + "step": 4251 + }, + { + "epoch": 1.603318250377074, + "grad_norm": 0.27363918883700966, + "learning_rate": 6.870914571383802e-05, + "loss": 0.4461, + "step": 4252 + }, + { + "epoch": 1.6036953242835597, + "grad_norm": 0.25508696934975095, + "learning_rate": 6.870181251558004e-05, + "loss": 0.4337, + "step": 4253 + }, + { + "epoch": 1.6040723981900453, + "grad_norm": 0.21748359717432397, + "learning_rate": 6.869447732831189e-05, + "loss": 0.4454, + "step": 4254 + }, + { + "epoch": 1.604449472096531, + "grad_norm": 0.2640047551283804, + "learning_rate": 6.868714015254192e-05, + "loss": 0.4483, + "step": 4255 + }, + { + "epoch": 1.6048265460030167, + "grad_norm": 0.2937162466309208, + "learning_rate": 6.867980098877858e-05, + "loss": 0.4334, + "step": 4256 + }, + { + "epoch": 1.6052036199095023, + "grad_norm": 0.35033899403004065, + "learning_rate": 6.867245983753048e-05, + "loss": 0.4432, + "step": 4257 + }, + { + "epoch": 1.605580693815988, + "grad_norm": 0.4459534149529148, + "learning_rate": 6.866511669930633e-05, + "loss": 0.438, + "step": 4258 + }, + { + "epoch": 1.6059577677224737, + "grad_norm": 0.5871537046869686, + "learning_rate": 6.865777157461503e-05, + "loss": 0.446, + "step": 4259 + }, + { + "epoch": 1.6063348416289593, + "grad_norm": 0.5888837183688879, + "learning_rate": 6.865042446396558e-05, + "loss": 0.4363, + "step": 4260 + }, + { + "epoch": 1.606711915535445, + "grad_norm": 0.5513019293673238, + "learning_rate": 6.864307536786714e-05, + "loss": 0.4376, + "step": 4261 + }, + { + "epoch": 1.6070889894419307, + "grad_norm": 0.4825520822845758, + "learning_rate": 6.863572428682897e-05, + "loss": 0.4412, + "step": 4262 + }, + { + "epoch": 1.6074660633484164, + "grad_norm": 0.41947135524148527, + "learning_rate": 6.86283712213605e-05, + "loss": 0.4541, + "step": 4263 + }, + { + "epoch": 1.607843137254902, + "grad_norm": 0.40761383227119596, + "learning_rate": 6.86210161719713e-05, + "loss": 0.4474, + "step": 4264 + }, + { + "epoch": 1.6082202111613877, + "grad_norm": 0.42457296070488326, + "learning_rate": 6.861365913917108e-05, + "loss": 0.4459, + "step": 4265 + }, + { + "epoch": 1.6085972850678734, + "grad_norm": 0.32634243297869936, + "learning_rate": 6.860630012346964e-05, + "loss": 0.4271, + "step": 4266 + }, + { + "epoch": 1.608974358974359, + "grad_norm": 0.2914855158382274, + "learning_rate": 6.859893912537699e-05, + "loss": 0.4411, + "step": 4267 + }, + { + "epoch": 1.6093514328808447, + "grad_norm": 0.3080142700441029, + "learning_rate": 6.859157614540322e-05, + "loss": 0.443, + "step": 4268 + }, + { + "epoch": 1.6097285067873304, + "grad_norm": 0.2707537154288312, + "learning_rate": 6.858421118405858e-05, + "loss": 0.4375, + "step": 4269 + }, + { + "epoch": 1.610105580693816, + "grad_norm": 0.2875222473071264, + "learning_rate": 6.857684424185347e-05, + "loss": 0.4412, + "step": 4270 + }, + { + "epoch": 1.6104826546003017, + "grad_norm": 0.3173729706327943, + "learning_rate": 6.856947531929841e-05, + "loss": 0.4347, + "step": 4271 + }, + { + "epoch": 1.6108597285067874, + "grad_norm": 0.36445671438756705, + "learning_rate": 6.856210441690404e-05, + "loss": 0.4427, + "step": 4272 + }, + { + "epoch": 1.611236802413273, + "grad_norm": 0.4057241529330195, + "learning_rate": 6.855473153518117e-05, + "loss": 0.4587, + "step": 4273 + }, + { + "epoch": 1.6116138763197587, + "grad_norm": 0.3648742205560921, + "learning_rate": 6.854735667464072e-05, + "loss": 0.4414, + "step": 4274 + }, + { + "epoch": 1.6119909502262444, + "grad_norm": 0.2825398331101147, + "learning_rate": 6.853997983579378e-05, + "loss": 0.4426, + "step": 4275 + }, + { + "epoch": 1.61236802413273, + "grad_norm": 0.23624317241036946, + "learning_rate": 6.853260101915156e-05, + "loss": 0.4332, + "step": 4276 + }, + { + "epoch": 1.6127450980392157, + "grad_norm": 0.26876298044049374, + "learning_rate": 6.85252202252254e-05, + "loss": 0.4395, + "step": 4277 + }, + { + "epoch": 1.6131221719457014, + "grad_norm": 0.316902878591125, + "learning_rate": 6.851783745452677e-05, + "loss": 0.4406, + "step": 4278 + }, + { + "epoch": 1.613499245852187, + "grad_norm": 0.34621779236705535, + "learning_rate": 6.85104527075673e-05, + "loss": 0.455, + "step": 4279 + }, + { + "epoch": 1.6138763197586727, + "grad_norm": 0.345121991725609, + "learning_rate": 6.850306598485875e-05, + "loss": 0.4407, + "step": 4280 + }, + { + "epoch": 1.6142533936651584, + "grad_norm": 0.42728888807897275, + "learning_rate": 6.8495677286913e-05, + "loss": 0.4452, + "step": 4281 + }, + { + "epoch": 1.614630467571644, + "grad_norm": 0.5481936157553662, + "learning_rate": 6.848828661424212e-05, + "loss": 0.4432, + "step": 4282 + }, + { + "epoch": 1.6150075414781297, + "grad_norm": 0.6200827575829961, + "learning_rate": 6.848089396735821e-05, + "loss": 0.4364, + "step": 4283 + }, + { + "epoch": 1.6153846153846154, + "grad_norm": 0.5812900984128622, + "learning_rate": 6.847349934677363e-05, + "loss": 0.445, + "step": 4284 + }, + { + "epoch": 1.615761689291101, + "grad_norm": 0.46333377031227174, + "learning_rate": 6.84661027530008e-05, + "loss": 0.4524, + "step": 4285 + }, + { + "epoch": 1.6161387631975868, + "grad_norm": 0.3051768866063295, + "learning_rate": 6.84587041865523e-05, + "loss": 0.4444, + "step": 4286 + }, + { + "epoch": 1.6165158371040724, + "grad_norm": 0.3038752154539943, + "learning_rate": 6.845130364794083e-05, + "loss": 0.4402, + "step": 4287 + }, + { + "epoch": 1.616892911010558, + "grad_norm": 0.3243383711034976, + "learning_rate": 6.844390113767926e-05, + "loss": 0.4484, + "step": 4288 + }, + { + "epoch": 1.6172699849170438, + "grad_norm": 0.3828058609544681, + "learning_rate": 6.843649665628058e-05, + "loss": 0.442, + "step": 4289 + }, + { + "epoch": 1.6176470588235294, + "grad_norm": 0.4428742528359022, + "learning_rate": 6.842909020425789e-05, + "loss": 0.4382, + "step": 4290 + }, + { + "epoch": 1.618024132730015, + "grad_norm": 0.3829026110959642, + "learning_rate": 6.842168178212448e-05, + "loss": 0.44, + "step": 4291 + }, + { + "epoch": 1.6184012066365008, + "grad_norm": 0.34553748900476006, + "learning_rate": 6.841427139039372e-05, + "loss": 0.4424, + "step": 4292 + }, + { + "epoch": 1.6187782805429864, + "grad_norm": 0.37704640833173164, + "learning_rate": 6.840685902957917e-05, + "loss": 0.4528, + "step": 4293 + }, + { + "epoch": 1.6191553544494721, + "grad_norm": 0.4434384444161037, + "learning_rate": 6.839944470019447e-05, + "loss": 0.4356, + "step": 4294 + }, + { + "epoch": 1.6195324283559578, + "grad_norm": 0.480676165791573, + "learning_rate": 6.839202840275345e-05, + "loss": 0.4417, + "step": 4295 + }, + { + "epoch": 1.6199095022624435, + "grad_norm": 0.4401111468190734, + "learning_rate": 6.838461013777006e-05, + "loss": 0.4388, + "step": 4296 + }, + { + "epoch": 1.6202865761689291, + "grad_norm": 0.460756217707888, + "learning_rate": 6.837718990575834e-05, + "loss": 0.4361, + "step": 4297 + }, + { + "epoch": 1.6206636500754148, + "grad_norm": 0.5226737388693795, + "learning_rate": 6.836976770723255e-05, + "loss": 0.4504, + "step": 4298 + }, + { + "epoch": 1.6210407239819005, + "grad_norm": 0.585870717616534, + "learning_rate": 6.836234354270701e-05, + "loss": 0.448, + "step": 4299 + }, + { + "epoch": 1.6214177978883861, + "grad_norm": 0.5742385340689194, + "learning_rate": 6.835491741269622e-05, + "loss": 0.4363, + "step": 4300 + }, + { + "epoch": 1.6217948717948718, + "grad_norm": 0.48181286684879787, + "learning_rate": 6.834748931771481e-05, + "loss": 0.4383, + "step": 4301 + }, + { + "epoch": 1.6221719457013575, + "grad_norm": 0.3319005042435525, + "learning_rate": 6.834005925827752e-05, + "loss": 0.4287, + "step": 4302 + }, + { + "epoch": 1.6225490196078431, + "grad_norm": 0.33805117526483924, + "learning_rate": 6.833262723489927e-05, + "loss": 0.4372, + "step": 4303 + }, + { + "epoch": 1.6229260935143288, + "grad_norm": 0.36640508123700677, + "learning_rate": 6.832519324809509e-05, + "loss": 0.4412, + "step": 4304 + }, + { + "epoch": 1.6233031674208145, + "grad_norm": 0.29489120345912984, + "learning_rate": 6.831775729838015e-05, + "loss": 0.4345, + "step": 4305 + }, + { + "epoch": 1.6236802413273002, + "grad_norm": 0.3241043300164632, + "learning_rate": 6.831031938626974e-05, + "loss": 0.4384, + "step": 4306 + }, + { + "epoch": 1.6240573152337858, + "grad_norm": 0.3842727042279726, + "learning_rate": 6.830287951227929e-05, + "loss": 0.4315, + "step": 4307 + }, + { + "epoch": 1.6244343891402715, + "grad_norm": 0.36729881392706754, + "learning_rate": 6.829543767692442e-05, + "loss": 0.4397, + "step": 4308 + }, + { + "epoch": 1.6248114630467572, + "grad_norm": 0.27986150648308333, + "learning_rate": 6.82879938807208e-05, + "loss": 0.4441, + "step": 4309 + }, + { + "epoch": 1.6251885369532428, + "grad_norm": 0.2916986671462384, + "learning_rate": 6.82805481241843e-05, + "loss": 0.4514, + "step": 4310 + }, + { + "epoch": 1.6255656108597285, + "grad_norm": 0.2859681493928355, + "learning_rate": 6.82731004078309e-05, + "loss": 0.442, + "step": 4311 + }, + { + "epoch": 1.6259426847662142, + "grad_norm": 0.30386504432674927, + "learning_rate": 6.82656507321767e-05, + "loss": 0.4446, + "step": 4312 + }, + { + "epoch": 1.6263197586726998, + "grad_norm": 0.33321263963886566, + "learning_rate": 6.825819909773801e-05, + "loss": 0.439, + "step": 4313 + }, + { + "epoch": 1.6266968325791855, + "grad_norm": 0.34312652780628494, + "learning_rate": 6.825074550503117e-05, + "loss": 0.4344, + "step": 4314 + }, + { + "epoch": 1.6270739064856712, + "grad_norm": 0.3730477129128219, + "learning_rate": 6.824328995457273e-05, + "loss": 0.4337, + "step": 4315 + }, + { + "epoch": 1.6274509803921569, + "grad_norm": 0.3565267147726921, + "learning_rate": 6.823583244687935e-05, + "loss": 0.4317, + "step": 4316 + }, + { + "epoch": 1.6278280542986425, + "grad_norm": 0.3055115906607039, + "learning_rate": 6.822837298246781e-05, + "loss": 0.4271, + "step": 4317 + }, + { + "epoch": 1.6282051282051282, + "grad_norm": 0.2535291600198815, + "learning_rate": 6.822091156185509e-05, + "loss": 0.4378, + "step": 4318 + }, + { + "epoch": 1.6285822021116139, + "grad_norm": 0.2079512437575077, + "learning_rate": 6.821344818555821e-05, + "loss": 0.446, + "step": 4319 + }, + { + "epoch": 1.6289592760180995, + "grad_norm": 0.2480025130650315, + "learning_rate": 6.820598285409439e-05, + "loss": 0.4397, + "step": 4320 + }, + { + "epoch": 1.6293363499245852, + "grad_norm": 0.2879223982533125, + "learning_rate": 6.819851556798098e-05, + "loss": 0.4486, + "step": 4321 + }, + { + "epoch": 1.6297134238310709, + "grad_norm": 0.24762515349259018, + "learning_rate": 6.819104632773547e-05, + "loss": 0.4379, + "step": 4322 + }, + { + "epoch": 1.6300904977375565, + "grad_norm": 0.241133441515131, + "learning_rate": 6.818357513387544e-05, + "loss": 0.4263, + "step": 4323 + }, + { + "epoch": 1.6304675716440422, + "grad_norm": 0.300493733744354, + "learning_rate": 6.817610198691866e-05, + "loss": 0.4391, + "step": 4324 + }, + { + "epoch": 1.6308446455505279, + "grad_norm": 0.3221308803003292, + "learning_rate": 6.816862688738302e-05, + "loss": 0.4333, + "step": 4325 + }, + { + "epoch": 1.6312217194570136, + "grad_norm": 0.32189419306439476, + "learning_rate": 6.816114983578649e-05, + "loss": 0.4355, + "step": 4326 + }, + { + "epoch": 1.6315987933634992, + "grad_norm": 0.3105658152513302, + "learning_rate": 6.815367083264727e-05, + "loss": 0.4458, + "step": 4327 + }, + { + "epoch": 1.631975867269985, + "grad_norm": 0.37903511633174425, + "learning_rate": 6.814618987848363e-05, + "loss": 0.4344, + "step": 4328 + }, + { + "epoch": 1.6323529411764706, + "grad_norm": 0.4278913973504527, + "learning_rate": 6.8138706973814e-05, + "loss": 0.4304, + "step": 4329 + }, + { + "epoch": 1.6327300150829562, + "grad_norm": 0.46953229453096373, + "learning_rate": 6.813122211915693e-05, + "loss": 0.4218, + "step": 4330 + }, + { + "epoch": 1.633107088989442, + "grad_norm": 0.5409735470213104, + "learning_rate": 6.812373531503114e-05, + "loss": 0.452, + "step": 4331 + }, + { + "epoch": 1.6334841628959276, + "grad_norm": 0.5695653362419237, + "learning_rate": 6.811624656195542e-05, + "loss": 0.4402, + "step": 4332 + }, + { + "epoch": 1.6338612368024132, + "grad_norm": 0.5872483160191732, + "learning_rate": 6.810875586044875e-05, + "loss": 0.4431, + "step": 4333 + }, + { + "epoch": 1.634238310708899, + "grad_norm": 0.5453046060938702, + "learning_rate": 6.810126321103023e-05, + "loss": 0.4317, + "step": 4334 + }, + { + "epoch": 1.6346153846153846, + "grad_norm": 0.4894685261860172, + "learning_rate": 6.80937686142191e-05, + "loss": 0.4316, + "step": 4335 + }, + { + "epoch": 1.6349924585218703, + "grad_norm": 0.41656104155826157, + "learning_rate": 6.808627207053473e-05, + "loss": 0.4287, + "step": 4336 + }, + { + "epoch": 1.635369532428356, + "grad_norm": 0.3375553827236319, + "learning_rate": 6.807877358049661e-05, + "loss": 0.4459, + "step": 4337 + }, + { + "epoch": 1.6357466063348416, + "grad_norm": 0.3260790143569269, + "learning_rate": 6.80712731446244e-05, + "loss": 0.4313, + "step": 4338 + }, + { + "epoch": 1.6361236802413273, + "grad_norm": 0.267597609307703, + "learning_rate": 6.806377076343784e-05, + "loss": 0.4248, + "step": 4339 + }, + { + "epoch": 1.636500754147813, + "grad_norm": 0.21836167687790692, + "learning_rate": 6.805626643745688e-05, + "loss": 0.441, + "step": 4340 + }, + { + "epoch": 1.6368778280542986, + "grad_norm": 0.2546952668396906, + "learning_rate": 6.804876016720153e-05, + "loss": 0.4404, + "step": 4341 + }, + { + "epoch": 1.6372549019607843, + "grad_norm": 0.2786071134338802, + "learning_rate": 6.804125195319197e-05, + "loss": 0.4544, + "step": 4342 + }, + { + "epoch": 1.63763197586727, + "grad_norm": 0.3069814141752978, + "learning_rate": 6.803374179594854e-05, + "loss": 0.437, + "step": 4343 + }, + { + "epoch": 1.6380090497737556, + "grad_norm": 0.2983832270453821, + "learning_rate": 6.802622969599167e-05, + "loss": 0.4368, + "step": 4344 + }, + { + "epoch": 1.6383861236802413, + "grad_norm": 0.29241839421202126, + "learning_rate": 6.801871565384193e-05, + "loss": 0.441, + "step": 4345 + }, + { + "epoch": 1.638763197586727, + "grad_norm": 0.27405505598820795, + "learning_rate": 6.801119967002006e-05, + "loss": 0.4287, + "step": 4346 + }, + { + "epoch": 1.6391402714932126, + "grad_norm": 0.3159255345827021, + "learning_rate": 6.800368174504688e-05, + "loss": 0.4307, + "step": 4347 + }, + { + "epoch": 1.6395173453996983, + "grad_norm": 0.3419758150398991, + "learning_rate": 6.799616187944341e-05, + "loss": 0.4541, + "step": 4348 + }, + { + "epoch": 1.639894419306184, + "grad_norm": 0.2877420178393324, + "learning_rate": 6.798864007373075e-05, + "loss": 0.4474, + "step": 4349 + }, + { + "epoch": 1.6402714932126696, + "grad_norm": 0.24075604373565962, + "learning_rate": 6.798111632843015e-05, + "loss": 0.4463, + "step": 4350 + }, + { + "epoch": 1.6406485671191553, + "grad_norm": 0.25924928179446116, + "learning_rate": 6.797359064406302e-05, + "loss": 0.4338, + "step": 4351 + }, + { + "epoch": 1.641025641025641, + "grad_norm": 0.33781025214432486, + "learning_rate": 6.796606302115086e-05, + "loss": 0.446, + "step": 4352 + }, + { + "epoch": 1.6414027149321266, + "grad_norm": 0.3830074435768156, + "learning_rate": 6.795853346021535e-05, + "loss": 0.4337, + "step": 4353 + }, + { + "epoch": 1.6417797888386123, + "grad_norm": 0.42338359136611853, + "learning_rate": 6.795100196177826e-05, + "loss": 0.4248, + "step": 4354 + }, + { + "epoch": 1.642156862745098, + "grad_norm": 0.41190589319619664, + "learning_rate": 6.794346852636153e-05, + "loss": 0.4398, + "step": 4355 + }, + { + "epoch": 1.6425339366515836, + "grad_norm": 0.36361983387880875, + "learning_rate": 6.793593315448723e-05, + "loss": 0.4467, + "step": 4356 + }, + { + "epoch": 1.6429110105580693, + "grad_norm": 0.2710652554488024, + "learning_rate": 6.792839584667753e-05, + "loss": 0.4386, + "step": 4357 + }, + { + "epoch": 1.643288084464555, + "grad_norm": 0.30342414873837037, + "learning_rate": 6.792085660345477e-05, + "loss": 0.4399, + "step": 4358 + }, + { + "epoch": 1.6436651583710407, + "grad_norm": 0.36157050386039297, + "learning_rate": 6.791331542534142e-05, + "loss": 0.4434, + "step": 4359 + }, + { + "epoch": 1.6440422322775263, + "grad_norm": 0.3029388508011339, + "learning_rate": 6.790577231286006e-05, + "loss": 0.4347, + "step": 4360 + }, + { + "epoch": 1.644419306184012, + "grad_norm": 0.24032673664899853, + "learning_rate": 6.789822726653344e-05, + "loss": 0.4409, + "step": 4361 + }, + { + "epoch": 1.6447963800904977, + "grad_norm": 0.27137559765128344, + "learning_rate": 6.78906802868844e-05, + "loss": 0.4406, + "step": 4362 + }, + { + "epoch": 1.6451734539969833, + "grad_norm": 0.3454681341323082, + "learning_rate": 6.788313137443597e-05, + "loss": 0.4359, + "step": 4363 + }, + { + "epoch": 1.645550527903469, + "grad_norm": 0.3803457769478639, + "learning_rate": 6.787558052971127e-05, + "loss": 0.4509, + "step": 4364 + }, + { + "epoch": 1.6459276018099547, + "grad_norm": 0.36265722642790016, + "learning_rate": 6.786802775323356e-05, + "loss": 0.4516, + "step": 4365 + }, + { + "epoch": 1.6463046757164403, + "grad_norm": 0.33851180163215466, + "learning_rate": 6.786047304552623e-05, + "loss": 0.4466, + "step": 4366 + }, + { + "epoch": 1.646681749622926, + "grad_norm": 0.31018110630615664, + "learning_rate": 6.785291640711286e-05, + "loss": 0.4441, + "step": 4367 + }, + { + "epoch": 1.6470588235294117, + "grad_norm": 0.3153340951257154, + "learning_rate": 6.784535783851708e-05, + "loss": 0.4517, + "step": 4368 + }, + { + "epoch": 1.6474358974358974, + "grad_norm": 0.37304612430043954, + "learning_rate": 6.783779734026268e-05, + "loss": 0.4482, + "step": 4369 + }, + { + "epoch": 1.647812971342383, + "grad_norm": 0.4025743677130827, + "learning_rate": 6.783023491287364e-05, + "loss": 0.4323, + "step": 4370 + }, + { + "epoch": 1.6481900452488687, + "grad_norm": 0.44874016257295807, + "learning_rate": 6.782267055687399e-05, + "loss": 0.4412, + "step": 4371 + }, + { + "epoch": 1.6485671191553544, + "grad_norm": 0.52898958886781, + "learning_rate": 6.781510427278795e-05, + "loss": 0.4374, + "step": 4372 + }, + { + "epoch": 1.64894419306184, + "grad_norm": 0.5681208853114399, + "learning_rate": 6.780753606113986e-05, + "loss": 0.4358, + "step": 4373 + }, + { + "epoch": 1.6493212669683257, + "grad_norm": 0.535966629051567, + "learning_rate": 6.77999659224542e-05, + "loss": 0.4445, + "step": 4374 + }, + { + "epoch": 1.6496983408748114, + "grad_norm": 0.5186039066787096, + "learning_rate": 6.779239385725554e-05, + "loss": 0.4385, + "step": 4375 + }, + { + "epoch": 1.650075414781297, + "grad_norm": 0.4775513231900373, + "learning_rate": 6.778481986606866e-05, + "loss": 0.4272, + "step": 4376 + }, + { + "epoch": 1.6504524886877827, + "grad_norm": 0.34505364797935567, + "learning_rate": 6.777724394941839e-05, + "loss": 0.431, + "step": 4377 + }, + { + "epoch": 1.6508295625942684, + "grad_norm": 0.3070300439972083, + "learning_rate": 6.776966610782975e-05, + "loss": 0.4292, + "step": 4378 + }, + { + "epoch": 1.651206636500754, + "grad_norm": 0.3107506518555868, + "learning_rate": 6.776208634182789e-05, + "loss": 0.423, + "step": 4379 + }, + { + "epoch": 1.6515837104072397, + "grad_norm": 0.3470784849517065, + "learning_rate": 6.775450465193807e-05, + "loss": 0.4483, + "step": 4380 + }, + { + "epoch": 1.6519607843137254, + "grad_norm": 0.3653102285151275, + "learning_rate": 6.774692103868569e-05, + "loss": 0.4382, + "step": 4381 + }, + { + "epoch": 1.652337858220211, + "grad_norm": 0.4078009260890774, + "learning_rate": 6.77393355025963e-05, + "loss": 0.4405, + "step": 4382 + }, + { + "epoch": 1.6527149321266967, + "grad_norm": 0.43861725582610156, + "learning_rate": 6.773174804419556e-05, + "loss": 0.4363, + "step": 4383 + }, + { + "epoch": 1.6530920060331824, + "grad_norm": 0.40891881177801664, + "learning_rate": 6.772415866400928e-05, + "loss": 0.4466, + "step": 4384 + }, + { + "epoch": 1.653469079939668, + "grad_norm": 0.3649629653019532, + "learning_rate": 6.77165673625634e-05, + "loss": 0.438, + "step": 4385 + }, + { + "epoch": 1.6538461538461537, + "grad_norm": 0.27341327596108406, + "learning_rate": 6.770897414038398e-05, + "loss": 0.4411, + "step": 4386 + }, + { + "epoch": 1.6542232277526394, + "grad_norm": 0.28197254255122356, + "learning_rate": 6.770137899799723e-05, + "loss": 0.4317, + "step": 4387 + }, + { + "epoch": 1.654600301659125, + "grad_norm": 0.39396670381607357, + "learning_rate": 6.76937819359295e-05, + "loss": 0.4331, + "step": 4388 + }, + { + "epoch": 1.6549773755656108, + "grad_norm": 0.43905989518400473, + "learning_rate": 6.768618295470723e-05, + "loss": 0.4396, + "step": 4389 + }, + { + "epoch": 1.6553544494720964, + "grad_norm": 0.3899245834226773, + "learning_rate": 6.767858205485704e-05, + "loss": 0.4325, + "step": 4390 + }, + { + "epoch": 1.655731523378582, + "grad_norm": 0.33333286978804855, + "learning_rate": 6.767097923690568e-05, + "loss": 0.439, + "step": 4391 + }, + { + "epoch": 1.6561085972850678, + "grad_norm": 0.3028414376186835, + "learning_rate": 6.766337450137999e-05, + "loss": 0.4403, + "step": 4392 + }, + { + "epoch": 1.6564856711915534, + "grad_norm": 0.3401951493418656, + "learning_rate": 6.765576784880699e-05, + "loss": 0.4442, + "step": 4393 + }, + { + "epoch": 1.656862745098039, + "grad_norm": 0.36449216118082417, + "learning_rate": 6.76481592797138e-05, + "loss": 0.4388, + "step": 4394 + }, + { + "epoch": 1.6572398190045248, + "grad_norm": 0.3857304231456996, + "learning_rate": 6.76405487946277e-05, + "loss": 0.4265, + "step": 4395 + }, + { + "epoch": 1.6576168929110104, + "grad_norm": 0.4161914259530549, + "learning_rate": 6.76329363940761e-05, + "loss": 0.4287, + "step": 4396 + }, + { + "epoch": 1.6579939668174961, + "grad_norm": 0.4273700117835423, + "learning_rate": 6.762532207858649e-05, + "loss": 0.4388, + "step": 4397 + }, + { + "epoch": 1.6583710407239818, + "grad_norm": 0.4387570306304613, + "learning_rate": 6.761770584868659e-05, + "loss": 0.4392, + "step": 4398 + }, + { + "epoch": 1.6587481146304675, + "grad_norm": 0.42428402712359203, + "learning_rate": 6.761008770490416e-05, + "loss": 0.4562, + "step": 4399 + }, + { + "epoch": 1.6591251885369531, + "grad_norm": 0.4366479806259187, + "learning_rate": 6.760246764776713e-05, + "loss": 0.4447, + "step": 4400 + }, + { + "epoch": 1.6595022624434388, + "grad_norm": 0.45758418648865445, + "learning_rate": 6.759484567780359e-05, + "loss": 0.4351, + "step": 4401 + }, + { + "epoch": 1.6598793363499245, + "grad_norm": 0.46691808684479263, + "learning_rate": 6.758722179554172e-05, + "loss": 0.4364, + "step": 4402 + }, + { + "epoch": 1.6602564102564101, + "grad_norm": 0.4306851132075273, + "learning_rate": 6.757959600150985e-05, + "loss": 0.439, + "step": 4403 + }, + { + "epoch": 1.6606334841628958, + "grad_norm": 0.341999007113645, + "learning_rate": 6.757196829623644e-05, + "loss": 0.4394, + "step": 4404 + }, + { + "epoch": 1.6610105580693815, + "grad_norm": 0.3447987153877928, + "learning_rate": 6.756433868025007e-05, + "loss": 0.4339, + "step": 4405 + }, + { + "epoch": 1.6613876319758671, + "grad_norm": 0.3425770583542036, + "learning_rate": 6.75567071540795e-05, + "loss": 0.4505, + "step": 4406 + }, + { + "epoch": 1.6617647058823528, + "grad_norm": 0.33047378496649304, + "learning_rate": 6.754907371825355e-05, + "loss": 0.4273, + "step": 4407 + }, + { + "epoch": 1.6621417797888385, + "grad_norm": 0.389021157123877, + "learning_rate": 6.754143837330124e-05, + "loss": 0.421, + "step": 4408 + }, + { + "epoch": 1.6625188536953242, + "grad_norm": 0.40082297870173844, + "learning_rate": 6.753380111975168e-05, + "loss": 0.4423, + "step": 4409 + }, + { + "epoch": 1.6628959276018098, + "grad_norm": 0.3790174248185573, + "learning_rate": 6.752616195813411e-05, + "loss": 0.434, + "step": 4410 + }, + { + "epoch": 1.6632730015082955, + "grad_norm": 0.34954363788851583, + "learning_rate": 6.751852088897795e-05, + "loss": 0.4271, + "step": 4411 + }, + { + "epoch": 1.6636500754147812, + "grad_norm": 0.3763244760502502, + "learning_rate": 6.751087791281269e-05, + "loss": 0.4447, + "step": 4412 + }, + { + "epoch": 1.6640271493212668, + "grad_norm": 0.4337089600397443, + "learning_rate": 6.750323303016799e-05, + "loss": 0.4394, + "step": 4413 + }, + { + "epoch": 1.6644042232277525, + "grad_norm": 0.37944814082034756, + "learning_rate": 6.749558624157365e-05, + "loss": 0.4254, + "step": 4414 + }, + { + "epoch": 1.6647812971342382, + "grad_norm": 0.3842933783223546, + "learning_rate": 6.748793754755957e-05, + "loss": 0.4404, + "step": 4415 + }, + { + "epoch": 1.6651583710407238, + "grad_norm": 0.4399432095275675, + "learning_rate": 6.748028694865579e-05, + "loss": 0.4427, + "step": 4416 + }, + { + "epoch": 1.6655354449472095, + "grad_norm": 0.5014540542908136, + "learning_rate": 6.74726344453925e-05, + "loss": 0.4393, + "step": 4417 + }, + { + "epoch": 1.6659125188536952, + "grad_norm": 0.5049131406571536, + "learning_rate": 6.74649800383e-05, + "loss": 0.4412, + "step": 4418 + }, + { + "epoch": 1.6662895927601808, + "grad_norm": 0.3791138424286965, + "learning_rate": 6.745732372790877e-05, + "loss": 0.427, + "step": 4419 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.28852162277916316, + "learning_rate": 6.744966551474936e-05, + "loss": 0.4317, + "step": 4420 + }, + { + "epoch": 1.6670437405731522, + "grad_norm": 0.3458565537261191, + "learning_rate": 6.744200539935246e-05, + "loss": 0.4537, + "step": 4421 + }, + { + "epoch": 1.6674208144796379, + "grad_norm": 0.40700034399459323, + "learning_rate": 6.743434338224892e-05, + "loss": 0.444, + "step": 4422 + }, + { + "epoch": 1.6677978883861235, + "grad_norm": 0.38028705342148583, + "learning_rate": 6.742667946396975e-05, + "loss": 0.4459, + "step": 4423 + }, + { + "epoch": 1.6681749622926092, + "grad_norm": 0.32401523587700604, + "learning_rate": 6.7419013645046e-05, + "loss": 0.446, + "step": 4424 + }, + { + "epoch": 1.6685520361990949, + "grad_norm": 0.3368292296290719, + "learning_rate": 6.741134592600892e-05, + "loss": 0.4257, + "step": 4425 + }, + { + "epoch": 1.6689291101055805, + "grad_norm": 0.32797714765904357, + "learning_rate": 6.74036763073899e-05, + "loss": 0.436, + "step": 4426 + }, + { + "epoch": 1.6693061840120662, + "grad_norm": 0.2674519175132533, + "learning_rate": 6.73960047897204e-05, + "loss": 0.4379, + "step": 4427 + }, + { + "epoch": 1.6696832579185519, + "grad_norm": 0.24569043850094072, + "learning_rate": 6.738833137353207e-05, + "loss": 0.4321, + "step": 4428 + }, + { + "epoch": 1.6700603318250375, + "grad_norm": 0.35818015121990227, + "learning_rate": 6.738065605935668e-05, + "loss": 0.4369, + "step": 4429 + }, + { + "epoch": 1.6704374057315232, + "grad_norm": 0.36775982086543985, + "learning_rate": 6.73729788477261e-05, + "loss": 0.45, + "step": 4430 + }, + { + "epoch": 1.6708144796380089, + "grad_norm": 0.256018858259803, + "learning_rate": 6.736529973917238e-05, + "loss": 0.4339, + "step": 4431 + }, + { + "epoch": 1.6711915535444946, + "grad_norm": 0.18625262053006517, + "learning_rate": 6.735761873422766e-05, + "loss": 0.4264, + "step": 4432 + }, + { + "epoch": 1.6715686274509802, + "grad_norm": 0.2663275654921983, + "learning_rate": 6.734993583342422e-05, + "loss": 0.4311, + "step": 4433 + }, + { + "epoch": 1.6719457013574661, + "grad_norm": 0.3349221508906374, + "learning_rate": 6.734225103729449e-05, + "loss": 0.4429, + "step": 4434 + }, + { + "epoch": 1.6723227752639518, + "grad_norm": 0.3462803851273169, + "learning_rate": 6.7334564346371e-05, + "loss": 0.4461, + "step": 4435 + }, + { + "epoch": 1.6726998491704375, + "grad_norm": 0.32540732439816256, + "learning_rate": 6.732687576118647e-05, + "loss": 0.4475, + "step": 4436 + }, + { + "epoch": 1.6730769230769231, + "grad_norm": 0.33440372426971426, + "learning_rate": 6.731918528227367e-05, + "loss": 0.4285, + "step": 4437 + }, + { + "epoch": 1.6734539969834088, + "grad_norm": 0.3858665009716081, + "learning_rate": 6.731149291016557e-05, + "loss": 0.4322, + "step": 4438 + }, + { + "epoch": 1.6738310708898945, + "grad_norm": 0.42715397919960496, + "learning_rate": 6.730379864539525e-05, + "loss": 0.4363, + "step": 4439 + }, + { + "epoch": 1.6742081447963801, + "grad_norm": 0.41583421867291753, + "learning_rate": 6.729610248849588e-05, + "loss": 0.4442, + "step": 4440 + }, + { + "epoch": 1.6745852187028658, + "grad_norm": 0.3986046059110098, + "learning_rate": 6.728840444000083e-05, + "loss": 0.4285, + "step": 4441 + }, + { + "epoch": 1.6749622926093515, + "grad_norm": 0.4389676958350751, + "learning_rate": 6.728070450044354e-05, + "loss": 0.4489, + "step": 4442 + }, + { + "epoch": 1.6753393665158371, + "grad_norm": 0.41036940869266636, + "learning_rate": 6.727300267035765e-05, + "loss": 0.4296, + "step": 4443 + }, + { + "epoch": 1.6757164404223228, + "grad_norm": 0.3214082828881047, + "learning_rate": 6.726529895027686e-05, + "loss": 0.4321, + "step": 4444 + }, + { + "epoch": 1.6760935143288085, + "grad_norm": 0.24488163079132194, + "learning_rate": 6.725759334073503e-05, + "loss": 0.4311, + "step": 4445 + }, + { + "epoch": 1.6764705882352942, + "grad_norm": 0.2955978006701843, + "learning_rate": 6.724988584226616e-05, + "loss": 0.4227, + "step": 4446 + }, + { + "epoch": 1.6768476621417798, + "grad_norm": 0.3493278810096719, + "learning_rate": 6.724217645540438e-05, + "loss": 0.4359, + "step": 4447 + }, + { + "epoch": 1.6772247360482655, + "grad_norm": 0.3923379964849147, + "learning_rate": 6.723446518068393e-05, + "loss": 0.4367, + "step": 4448 + }, + { + "epoch": 1.6776018099547512, + "grad_norm": 0.4698133901970702, + "learning_rate": 6.722675201863921e-05, + "loss": 0.439, + "step": 4449 + }, + { + "epoch": 1.6779788838612368, + "grad_norm": 0.5014545540935188, + "learning_rate": 6.721903696980471e-05, + "loss": 0.4481, + "step": 4450 + }, + { + "epoch": 1.6783559577677225, + "grad_norm": 0.4770444721418162, + "learning_rate": 6.721132003471511e-05, + "loss": 0.4466, + "step": 4451 + }, + { + "epoch": 1.6787330316742082, + "grad_norm": 0.43371620080013157, + "learning_rate": 6.720360121390514e-05, + "loss": 0.4356, + "step": 4452 + }, + { + "epoch": 1.6791101055806938, + "grad_norm": 0.4076533528520783, + "learning_rate": 6.719588050790976e-05, + "loss": 0.4381, + "step": 4453 + }, + { + "epoch": 1.6794871794871795, + "grad_norm": 0.37271615220738497, + "learning_rate": 6.718815791726398e-05, + "loss": 0.4363, + "step": 4454 + }, + { + "epoch": 1.6798642533936652, + "grad_norm": 0.33100103747852744, + "learning_rate": 6.718043344250296e-05, + "loss": 0.4389, + "step": 4455 + }, + { + "epoch": 1.6802413273001509, + "grad_norm": 0.3575895894753704, + "learning_rate": 6.717270708416201e-05, + "loss": 0.4324, + "step": 4456 + }, + { + "epoch": 1.6806184012066365, + "grad_norm": 0.37846165378709473, + "learning_rate": 6.716497884277658e-05, + "loss": 0.4467, + "step": 4457 + }, + { + "epoch": 1.6809954751131222, + "grad_norm": 0.40909414626757895, + "learning_rate": 6.71572487188822e-05, + "loss": 0.4388, + "step": 4458 + }, + { + "epoch": 1.6813725490196079, + "grad_norm": 0.4405909843609411, + "learning_rate": 6.714951671301457e-05, + "loss": 0.433, + "step": 4459 + }, + { + "epoch": 1.6817496229260935, + "grad_norm": 0.4421873298150646, + "learning_rate": 6.714178282570952e-05, + "loss": 0.4301, + "step": 4460 + }, + { + "epoch": 1.6821266968325792, + "grad_norm": 0.4319865675387642, + "learning_rate": 6.713404705750298e-05, + "loss": 0.428, + "step": 4461 + }, + { + "epoch": 1.6825037707390649, + "grad_norm": 0.4021816601558157, + "learning_rate": 6.712630940893107e-05, + "loss": 0.4379, + "step": 4462 + }, + { + "epoch": 1.6828808446455505, + "grad_norm": 0.39836411938717714, + "learning_rate": 6.711856988052995e-05, + "loss": 0.4485, + "step": 4463 + }, + { + "epoch": 1.6832579185520362, + "grad_norm": 0.36527396135763485, + "learning_rate": 6.711082847283602e-05, + "loss": 0.4422, + "step": 4464 + }, + { + "epoch": 1.6836349924585219, + "grad_norm": 0.37879266153890856, + "learning_rate": 6.710308518638571e-05, + "loss": 0.4417, + "step": 4465 + }, + { + "epoch": 1.6840120663650076, + "grad_norm": 0.4404692434282577, + "learning_rate": 6.709534002171564e-05, + "loss": 0.4392, + "step": 4466 + }, + { + "epoch": 1.6843891402714932, + "grad_norm": 0.5257381709768694, + "learning_rate": 6.708759297936254e-05, + "loss": 0.4374, + "step": 4467 + }, + { + "epoch": 1.684766214177979, + "grad_norm": 0.4738288307589306, + "learning_rate": 6.707984405986328e-05, + "loss": 0.4475, + "step": 4468 + }, + { + "epoch": 1.6851432880844646, + "grad_norm": 0.3616668969328696, + "learning_rate": 6.707209326375483e-05, + "loss": 0.4362, + "step": 4469 + }, + { + "epoch": 1.6855203619909502, + "grad_norm": 0.3503816120067374, + "learning_rate": 6.706434059157433e-05, + "loss": 0.4263, + "step": 4470 + }, + { + "epoch": 1.685897435897436, + "grad_norm": 0.43003133044034497, + "learning_rate": 6.705658604385905e-05, + "loss": 0.4378, + "step": 4471 + }, + { + "epoch": 1.6862745098039216, + "grad_norm": 0.4065752364234508, + "learning_rate": 6.704882962114636e-05, + "loss": 0.4395, + "step": 4472 + }, + { + "epoch": 1.6866515837104072, + "grad_norm": 0.31514345617795997, + "learning_rate": 6.704107132397376e-05, + "loss": 0.4499, + "step": 4473 + }, + { + "epoch": 1.687028657616893, + "grad_norm": 0.3962099950588375, + "learning_rate": 6.703331115287889e-05, + "loss": 0.4314, + "step": 4474 + }, + { + "epoch": 1.6874057315233786, + "grad_norm": 0.4255066317736969, + "learning_rate": 6.702554910839954e-05, + "loss": 0.4376, + "step": 4475 + }, + { + "epoch": 1.6877828054298643, + "grad_norm": 0.3681768083986044, + "learning_rate": 6.701778519107362e-05, + "loss": 0.4355, + "step": 4476 + }, + { + "epoch": 1.68815987933635, + "grad_norm": 0.3301163930864235, + "learning_rate": 6.701001940143914e-05, + "loss": 0.4366, + "step": 4477 + }, + { + "epoch": 1.6885369532428356, + "grad_norm": 0.3621617303349387, + "learning_rate": 6.700225174003429e-05, + "loss": 0.4412, + "step": 4478 + }, + { + "epoch": 1.6889140271493213, + "grad_norm": 0.5331038758700437, + "learning_rate": 6.699448220739733e-05, + "loss": 0.4287, + "step": 4479 + }, + { + "epoch": 1.689291101055807, + "grad_norm": 0.63814519081138, + "learning_rate": 6.69867108040667e-05, + "loss": 0.4403, + "step": 4480 + }, + { + "epoch": 1.6896681749622926, + "grad_norm": 2.7746610592942034, + "learning_rate": 6.697893753058095e-05, + "loss": 0.4413, + "step": 4481 + }, + { + "epoch": 1.6900452488687783, + "grad_norm": 0.8985309264384407, + "learning_rate": 6.697116238747876e-05, + "loss": 0.4437, + "step": 4482 + }, + { + "epoch": 1.690422322775264, + "grad_norm": 1.6237049375937576, + "learning_rate": 6.696338537529894e-05, + "loss": 0.4393, + "step": 4483 + }, + { + "epoch": 1.6907993966817496, + "grad_norm": 0.7507120697992066, + "learning_rate": 6.695560649458043e-05, + "loss": 0.4548, + "step": 4484 + }, + { + "epoch": 1.6911764705882353, + "grad_norm": 2.0503416800608223, + "learning_rate": 6.69478257458623e-05, + "loss": 0.4577, + "step": 4485 + }, + { + "epoch": 1.691553544494721, + "grad_norm": 1.29909146345261, + "learning_rate": 6.694004312968374e-05, + "loss": 0.45, + "step": 4486 + }, + { + "epoch": 1.6919306184012066, + "grad_norm": 1.4883262645559654, + "learning_rate": 6.693225864658409e-05, + "loss": 0.4625, + "step": 4487 + }, + { + "epoch": 1.6923076923076923, + "grad_norm": 1.2382146893317139, + "learning_rate": 6.69244722971028e-05, + "loss": 0.4588, + "step": 4488 + }, + { + "epoch": 1.692684766214178, + "grad_norm": 0.7387846907895208, + "learning_rate": 6.691668408177948e-05, + "loss": 0.4407, + "step": 4489 + }, + { + "epoch": 1.6930618401206636, + "grad_norm": 0.8844283880128734, + "learning_rate": 6.690889400115383e-05, + "loss": 0.4413, + "step": 4490 + }, + { + "epoch": 1.6934389140271493, + "grad_norm": 0.5719895914552798, + "learning_rate": 6.690110205576568e-05, + "loss": 0.4424, + "step": 4491 + }, + { + "epoch": 1.693815987933635, + "grad_norm": 0.7297294148282588, + "learning_rate": 6.689330824615501e-05, + "loss": 0.4446, + "step": 4492 + }, + { + "epoch": 1.6941930618401206, + "grad_norm": 0.5911242692441914, + "learning_rate": 6.688551257286196e-05, + "loss": 0.4439, + "step": 4493 + }, + { + "epoch": 1.6945701357466063, + "grad_norm": 0.7934221849388416, + "learning_rate": 6.68777150364267e-05, + "loss": 0.4583, + "step": 4494 + }, + { + "epoch": 1.694947209653092, + "grad_norm": 0.8484001574968267, + "learning_rate": 6.686991563738967e-05, + "loss": 0.4385, + "step": 4495 + }, + { + "epoch": 1.6953242835595776, + "grad_norm": 0.6336652500171331, + "learning_rate": 6.686211437629129e-05, + "loss": 0.44, + "step": 4496 + }, + { + "epoch": 1.6957013574660633, + "grad_norm": 0.6857333711675672, + "learning_rate": 6.685431125367223e-05, + "loss": 0.4434, + "step": 4497 + }, + { + "epoch": 1.696078431372549, + "grad_norm": 0.5026810403724723, + "learning_rate": 6.68465062700732e-05, + "loss": 0.447, + "step": 4498 + }, + { + "epoch": 1.6964555052790347, + "grad_norm": 0.4974691399746339, + "learning_rate": 6.68386994260351e-05, + "loss": 0.4438, + "step": 4499 + }, + { + "epoch": 1.6968325791855203, + "grad_norm": 0.4297279110041207, + "learning_rate": 6.683089072209894e-05, + "loss": 0.4375, + "step": 4500 + }, + { + "epoch": 1.697209653092006, + "grad_norm": 0.45189475065001244, + "learning_rate": 6.682308015880586e-05, + "loss": 0.4358, + "step": 4501 + }, + { + "epoch": 1.6975867269984917, + "grad_norm": 0.3934531278066316, + "learning_rate": 6.681526773669711e-05, + "loss": 0.4514, + "step": 4502 + }, + { + "epoch": 1.6979638009049773, + "grad_norm": 0.4118593423556584, + "learning_rate": 6.68074534563141e-05, + "loss": 0.438, + "step": 4503 + }, + { + "epoch": 1.698340874811463, + "grad_norm": 0.4022968413621368, + "learning_rate": 6.679963731819833e-05, + "loss": 0.4265, + "step": 4504 + }, + { + "epoch": 1.6987179487179487, + "grad_norm": 0.46100802928990986, + "learning_rate": 6.679181932289148e-05, + "loss": 0.4463, + "step": 4505 + }, + { + "epoch": 1.6990950226244343, + "grad_norm": 0.44639871780625306, + "learning_rate": 6.67839994709353e-05, + "loss": 0.4404, + "step": 4506 + }, + { + "epoch": 1.69947209653092, + "grad_norm": 0.47712903356419367, + "learning_rate": 6.677617776287172e-05, + "loss": 0.4312, + "step": 4507 + }, + { + "epoch": 1.6998491704374057, + "grad_norm": 0.2740299925467893, + "learning_rate": 6.676835419924278e-05, + "loss": 0.4351, + "step": 4508 + }, + { + "epoch": 1.7002262443438914, + "grad_norm": 0.43869894083046906, + "learning_rate": 6.676052878059062e-05, + "loss": 0.4311, + "step": 4509 + }, + { + "epoch": 1.700603318250377, + "grad_norm": 0.34994398010552175, + "learning_rate": 6.675270150745757e-05, + "loss": 0.441, + "step": 4510 + }, + { + "epoch": 1.7009803921568627, + "grad_norm": 0.3756724656453144, + "learning_rate": 6.674487238038603e-05, + "loss": 0.4419, + "step": 4511 + }, + { + "epoch": 1.7013574660633484, + "grad_norm": 0.3866312482894926, + "learning_rate": 6.673704139991855e-05, + "loss": 0.4309, + "step": 4512 + }, + { + "epoch": 1.701734539969834, + "grad_norm": 0.23083927464966267, + "learning_rate": 6.672920856659781e-05, + "loss": 0.4433, + "step": 4513 + }, + { + "epoch": 1.7021116138763197, + "grad_norm": 0.3246489932760953, + "learning_rate": 6.672137388096665e-05, + "loss": 0.442, + "step": 4514 + }, + { + "epoch": 1.7024886877828054, + "grad_norm": 0.33959566430212673, + "learning_rate": 6.671353734356797e-05, + "loss": 0.4308, + "step": 4515 + }, + { + "epoch": 1.702865761689291, + "grad_norm": 0.26850914142883214, + "learning_rate": 6.670569895494483e-05, + "loss": 0.4303, + "step": 4516 + }, + { + "epoch": 1.703242835595777, + "grad_norm": 0.2780259201170997, + "learning_rate": 6.669785871564047e-05, + "loss": 0.4396, + "step": 4517 + }, + { + "epoch": 1.7036199095022626, + "grad_norm": 0.25342235448532413, + "learning_rate": 6.669001662619815e-05, + "loss": 0.441, + "step": 4518 + }, + { + "epoch": 1.7039969834087483, + "grad_norm": 0.25075515299313694, + "learning_rate": 6.668217268716139e-05, + "loss": 0.4314, + "step": 4519 + }, + { + "epoch": 1.704374057315234, + "grad_norm": 0.30039261958588487, + "learning_rate": 6.66743268990737e-05, + "loss": 0.4491, + "step": 4520 + }, + { + "epoch": 1.7047511312217196, + "grad_norm": 0.3284879667079072, + "learning_rate": 6.666647926247881e-05, + "loss": 0.4467, + "step": 4521 + }, + { + "epoch": 1.7051282051282053, + "grad_norm": 0.36213558856062183, + "learning_rate": 6.665862977792058e-05, + "loss": 0.4379, + "step": 4522 + }, + { + "epoch": 1.705505279034691, + "grad_norm": 0.3487849809298496, + "learning_rate": 6.665077844594295e-05, + "loss": 0.4306, + "step": 4523 + }, + { + "epoch": 1.7058823529411766, + "grad_norm": 0.264431601818324, + "learning_rate": 6.664292526709001e-05, + "loss": 0.4356, + "step": 4524 + }, + { + "epoch": 1.7062594268476623, + "grad_norm": 0.3133095867907822, + "learning_rate": 6.663507024190598e-05, + "loss": 0.4425, + "step": 4525 + }, + { + "epoch": 1.706636500754148, + "grad_norm": 0.3309422424411422, + "learning_rate": 6.662721337093522e-05, + "loss": 0.4388, + "step": 4526 + }, + { + "epoch": 1.7070135746606336, + "grad_norm": 0.3208622785739004, + "learning_rate": 6.661935465472217e-05, + "loss": 0.4288, + "step": 4527 + }, + { + "epoch": 1.7073906485671193, + "grad_norm": 0.329201110551855, + "learning_rate": 6.661149409381147e-05, + "loss": 0.4363, + "step": 4528 + }, + { + "epoch": 1.707767722473605, + "grad_norm": 0.3206326642632692, + "learning_rate": 6.660363168874783e-05, + "loss": 0.4415, + "step": 4529 + }, + { + "epoch": 1.7081447963800906, + "grad_norm": 0.31734011799850786, + "learning_rate": 6.659576744007611e-05, + "loss": 0.4436, + "step": 4530 + }, + { + "epoch": 1.7085218702865763, + "grad_norm": 0.2614999604004437, + "learning_rate": 6.658790134834129e-05, + "loss": 0.4433, + "step": 4531 + }, + { + "epoch": 1.708898944193062, + "grad_norm": 0.23273563355036755, + "learning_rate": 6.658003341408849e-05, + "loss": 0.4329, + "step": 4532 + }, + { + "epoch": 1.7092760180995477, + "grad_norm": 0.29639735142709056, + "learning_rate": 6.657216363786297e-05, + "loss": 0.4416, + "step": 4533 + }, + { + "epoch": 1.7096530920060333, + "grad_norm": 0.3169954139527416, + "learning_rate": 6.656429202021006e-05, + "loss": 0.4237, + "step": 4534 + }, + { + "epoch": 1.710030165912519, + "grad_norm": 0.2853425501581903, + "learning_rate": 6.655641856167527e-05, + "loss": 0.4455, + "step": 4535 + }, + { + "epoch": 1.7104072398190047, + "grad_norm": 0.26478466010228335, + "learning_rate": 6.654854326280426e-05, + "loss": 0.4412, + "step": 4536 + }, + { + "epoch": 1.7107843137254903, + "grad_norm": 0.3449800223656547, + "learning_rate": 6.654066612414273e-05, + "loss": 0.4318, + "step": 4537 + }, + { + "epoch": 1.711161387631976, + "grad_norm": 0.2957349689295796, + "learning_rate": 6.653278714623657e-05, + "loss": 0.4343, + "step": 4538 + }, + { + "epoch": 1.7115384615384617, + "grad_norm": 0.25363792071826663, + "learning_rate": 6.652490632963182e-05, + "loss": 0.4359, + "step": 4539 + }, + { + "epoch": 1.7119155354449473, + "grad_norm": 0.2926215659892614, + "learning_rate": 6.651702367487456e-05, + "loss": 0.4417, + "step": 4540 + }, + { + "epoch": 1.712292609351433, + "grad_norm": 0.3286454699697494, + "learning_rate": 6.650913918251111e-05, + "loss": 0.4424, + "step": 4541 + }, + { + "epoch": 1.7126696832579187, + "grad_norm": 0.29621538810833564, + "learning_rate": 6.650125285308781e-05, + "loss": 0.4334, + "step": 4542 + }, + { + "epoch": 1.7130467571644044, + "grad_norm": 0.25348952828269933, + "learning_rate": 6.64933646871512e-05, + "loss": 0.4356, + "step": 4543 + }, + { + "epoch": 1.71342383107089, + "grad_norm": 0.2709947885337897, + "learning_rate": 6.64854746852479e-05, + "loss": 0.4341, + "step": 4544 + }, + { + "epoch": 1.7138009049773757, + "grad_norm": 0.30652716737745017, + "learning_rate": 6.647758284792473e-05, + "loss": 0.4491, + "step": 4545 + }, + { + "epoch": 1.7141779788838614, + "grad_norm": 0.3237427154094533, + "learning_rate": 6.646968917572852e-05, + "loss": 0.4429, + "step": 4546 + }, + { + "epoch": 1.714555052790347, + "grad_norm": 0.34551481989331945, + "learning_rate": 6.646179366920635e-05, + "loss": 0.429, + "step": 4547 + }, + { + "epoch": 1.7149321266968327, + "grad_norm": 0.3689591067082058, + "learning_rate": 6.645389632890535e-05, + "loss": 0.446, + "step": 4548 + }, + { + "epoch": 1.7153092006033184, + "grad_norm": 0.3231208853035886, + "learning_rate": 6.644599715537278e-05, + "loss": 0.4387, + "step": 4549 + }, + { + "epoch": 1.715686274509804, + "grad_norm": 0.22196347918060366, + "learning_rate": 6.643809614915608e-05, + "loss": 0.4394, + "step": 4550 + }, + { + "epoch": 1.7160633484162897, + "grad_norm": 0.2220945019965009, + "learning_rate": 6.643019331080278e-05, + "loss": 0.4445, + "step": 4551 + }, + { + "epoch": 1.7164404223227754, + "grad_norm": 0.24620005390169428, + "learning_rate": 6.64222886408605e-05, + "loss": 0.4457, + "step": 4552 + }, + { + "epoch": 1.716817496229261, + "grad_norm": 0.27606961406810154, + "learning_rate": 6.641438213987707e-05, + "loss": 0.4365, + "step": 4553 + }, + { + "epoch": 1.7171945701357467, + "grad_norm": 0.22658766837206323, + "learning_rate": 6.640647380840039e-05, + "loss": 0.438, + "step": 4554 + }, + { + "epoch": 1.7175716440422324, + "grad_norm": 0.268464147413527, + "learning_rate": 6.639856364697848e-05, + "loss": 0.4417, + "step": 4555 + }, + { + "epoch": 1.717948717948718, + "grad_norm": 0.35995822344490436, + "learning_rate": 6.639065165615954e-05, + "loss": 0.4333, + "step": 4556 + }, + { + "epoch": 1.7183257918552037, + "grad_norm": 0.30119410315236445, + "learning_rate": 6.638273783649184e-05, + "loss": 0.4226, + "step": 4557 + }, + { + "epoch": 1.7187028657616894, + "grad_norm": 0.3121553921087156, + "learning_rate": 6.637482218852382e-05, + "loss": 0.4445, + "step": 4558 + }, + { + "epoch": 1.719079939668175, + "grad_norm": 0.3807784358803629, + "learning_rate": 6.6366904712804e-05, + "loss": 0.4416, + "step": 4559 + }, + { + "epoch": 1.7194570135746607, + "grad_norm": 0.3315405242788428, + "learning_rate": 6.63589854098811e-05, + "loss": 0.442, + "step": 4560 + }, + { + "epoch": 1.7198340874811464, + "grad_norm": 0.29535916803603807, + "learning_rate": 6.635106428030386e-05, + "loss": 0.4277, + "step": 4561 + }, + { + "epoch": 1.720211161387632, + "grad_norm": 0.2755942483881739, + "learning_rate": 6.634314132462125e-05, + "loss": 0.4271, + "step": 4562 + }, + { + "epoch": 1.7205882352941178, + "grad_norm": 0.2818667616276914, + "learning_rate": 6.633521654338233e-05, + "loss": 0.441, + "step": 4563 + }, + { + "epoch": 1.7209653092006034, + "grad_norm": 0.27166726947678893, + "learning_rate": 6.632728993713623e-05, + "loss": 0.4402, + "step": 4564 + }, + { + "epoch": 1.721342383107089, + "grad_norm": 0.23763306252934432, + "learning_rate": 6.63193615064323e-05, + "loss": 0.4398, + "step": 4565 + }, + { + "epoch": 1.7217194570135748, + "grad_norm": 0.2472293414246907, + "learning_rate": 6.631143125181998e-05, + "loss": 0.4381, + "step": 4566 + }, + { + "epoch": 1.7220965309200604, + "grad_norm": 0.23284072296232486, + "learning_rate": 6.63034991738488e-05, + "loss": 0.4291, + "step": 4567 + }, + { + "epoch": 1.722473604826546, + "grad_norm": 0.24900336823057126, + "learning_rate": 6.629556527306845e-05, + "loss": 0.4246, + "step": 4568 + }, + { + "epoch": 1.7228506787330318, + "grad_norm": 0.26027176265698204, + "learning_rate": 6.628762955002878e-05, + "loss": 0.4393, + "step": 4569 + }, + { + "epoch": 1.7232277526395174, + "grad_norm": 0.3286285286760001, + "learning_rate": 6.627969200527966e-05, + "loss": 0.4258, + "step": 4570 + }, + { + "epoch": 1.723604826546003, + "grad_norm": 0.3851589837096971, + "learning_rate": 6.627175263937123e-05, + "loss": 0.443, + "step": 4571 + }, + { + "epoch": 1.7239819004524888, + "grad_norm": 0.4137410539623589, + "learning_rate": 6.626381145285363e-05, + "loss": 0.4261, + "step": 4572 + }, + { + "epoch": 1.7243589743589745, + "grad_norm": 0.35868113837405935, + "learning_rate": 6.62558684462772e-05, + "loss": 0.4469, + "step": 4573 + }, + { + "epoch": 1.7247360482654601, + "grad_norm": 0.2911751528140088, + "learning_rate": 6.624792362019235e-05, + "loss": 0.4279, + "step": 4574 + }, + { + "epoch": 1.7251131221719458, + "grad_norm": 0.27380843267014887, + "learning_rate": 6.62399769751497e-05, + "loss": 0.4408, + "step": 4575 + }, + { + "epoch": 1.7254901960784315, + "grad_norm": 0.29054237232132374, + "learning_rate": 6.623202851169992e-05, + "loss": 0.4427, + "step": 4576 + }, + { + "epoch": 1.7258672699849171, + "grad_norm": 0.3619620650538472, + "learning_rate": 6.622407823039381e-05, + "loss": 0.4273, + "step": 4577 + }, + { + "epoch": 1.7262443438914028, + "grad_norm": 0.40151372940566504, + "learning_rate": 6.621612613178235e-05, + "loss": 0.4418, + "step": 4578 + }, + { + "epoch": 1.7266214177978885, + "grad_norm": 0.41832967103284324, + "learning_rate": 6.620817221641661e-05, + "loss": 0.437, + "step": 4579 + }, + { + "epoch": 1.7269984917043741, + "grad_norm": 0.3677390272075487, + "learning_rate": 6.620021648484779e-05, + "loss": 0.4353, + "step": 4580 + }, + { + "epoch": 1.7273755656108598, + "grad_norm": 0.2688410269760475, + "learning_rate": 6.619225893762719e-05, + "loss": 0.4295, + "step": 4581 + }, + { + "epoch": 1.7277526395173455, + "grad_norm": 0.31090025981925135, + "learning_rate": 6.618429957530628e-05, + "loss": 0.438, + "step": 4582 + }, + { + "epoch": 1.7281297134238311, + "grad_norm": 0.34975756601715763, + "learning_rate": 6.617633839843664e-05, + "loss": 0.424, + "step": 4583 + }, + { + "epoch": 1.7285067873303168, + "grad_norm": 0.3711154141017657, + "learning_rate": 6.616837540756997e-05, + "loss": 0.4331, + "step": 4584 + }, + { + "epoch": 1.7288838612368025, + "grad_norm": 0.3794113218448038, + "learning_rate": 6.616041060325808e-05, + "loss": 0.4451, + "step": 4585 + }, + { + "epoch": 1.7292609351432882, + "grad_norm": 0.32881909548777005, + "learning_rate": 6.615244398605296e-05, + "loss": 0.4368, + "step": 4586 + }, + { + "epoch": 1.7296380090497738, + "grad_norm": 0.26843062617371444, + "learning_rate": 6.614447555650666e-05, + "loss": 0.4309, + "step": 4587 + }, + { + "epoch": 1.7300150829562595, + "grad_norm": 0.2347324955602365, + "learning_rate": 6.613650531517138e-05, + "loss": 0.4311, + "step": 4588 + }, + { + "epoch": 1.7303921568627452, + "grad_norm": 0.26351923366974883, + "learning_rate": 6.612853326259948e-05, + "loss": 0.4326, + "step": 4589 + }, + { + "epoch": 1.7307692307692308, + "grad_norm": 0.2759792247400359, + "learning_rate": 6.61205593993434e-05, + "loss": 0.4552, + "step": 4590 + }, + { + "epoch": 1.7311463046757165, + "grad_norm": 0.22984710788642204, + "learning_rate": 6.611258372595573e-05, + "loss": 0.4404, + "step": 4591 + }, + { + "epoch": 1.7315233785822022, + "grad_norm": 0.22722404733151214, + "learning_rate": 6.610460624298918e-05, + "loss": 0.4315, + "step": 4592 + }, + { + "epoch": 1.7319004524886878, + "grad_norm": 0.22515334232962483, + "learning_rate": 6.609662695099653e-05, + "loss": 0.4442, + "step": 4593 + }, + { + "epoch": 1.7322775263951735, + "grad_norm": 0.16914867863882602, + "learning_rate": 6.608864585053082e-05, + "loss": 0.4365, + "step": 4594 + }, + { + "epoch": 1.7326546003016592, + "grad_norm": 0.19341004127843997, + "learning_rate": 6.608066294214507e-05, + "loss": 0.4312, + "step": 4595 + }, + { + "epoch": 1.7330316742081449, + "grad_norm": 0.1893089603227325, + "learning_rate": 6.607267822639251e-05, + "loss": 0.4407, + "step": 4596 + }, + { + "epoch": 1.7334087481146305, + "grad_norm": 0.22227055422599723, + "learning_rate": 6.60646917038265e-05, + "loss": 0.4206, + "step": 4597 + }, + { + "epoch": 1.7337858220211162, + "grad_norm": 0.27676865625004804, + "learning_rate": 6.605670337500045e-05, + "loss": 0.4355, + "step": 4598 + }, + { + "epoch": 1.7341628959276019, + "grad_norm": 0.25969826269707574, + "learning_rate": 6.604871324046798e-05, + "loss": 0.4291, + "step": 4599 + }, + { + "epoch": 1.7345399698340875, + "grad_norm": 0.2804795922255373, + "learning_rate": 6.604072130078279e-05, + "loss": 0.439, + "step": 4600 + }, + { + "epoch": 1.7349170437405732, + "grad_norm": 0.28084666319290347, + "learning_rate": 6.603272755649871e-05, + "loss": 0.4396, + "step": 4601 + }, + { + "epoch": 1.7352941176470589, + "grad_norm": 0.33944812491187454, + "learning_rate": 6.602473200816969e-05, + "loss": 0.4347, + "step": 4602 + }, + { + "epoch": 1.7356711915535445, + "grad_norm": 0.32915400597767697, + "learning_rate": 6.601673465634983e-05, + "loss": 0.4256, + "step": 4603 + }, + { + "epoch": 1.7360482654600302, + "grad_norm": 0.30275600042197326, + "learning_rate": 6.600873550159334e-05, + "loss": 0.4431, + "step": 4604 + }, + { + "epoch": 1.7364253393665159, + "grad_norm": 0.36202875450250954, + "learning_rate": 6.600073454445455e-05, + "loss": 0.4402, + "step": 4605 + }, + { + "epoch": 1.7368024132730016, + "grad_norm": 0.4344467856123818, + "learning_rate": 6.599273178548792e-05, + "loss": 0.4341, + "step": 4606 + }, + { + "epoch": 1.7371794871794872, + "grad_norm": 0.46924482434832276, + "learning_rate": 6.598472722524802e-05, + "loss": 0.453, + "step": 4607 + }, + { + "epoch": 1.737556561085973, + "grad_norm": 0.44332281284104397, + "learning_rate": 6.597672086428959e-05, + "loss": 0.4383, + "step": 4608 + }, + { + "epoch": 1.7379336349924586, + "grad_norm": 0.46920175202280023, + "learning_rate": 6.596871270316744e-05, + "loss": 0.4495, + "step": 4609 + }, + { + "epoch": 1.7383107088989442, + "grad_norm": 0.5432367239115966, + "learning_rate": 6.596070274243653e-05, + "loss": 0.4394, + "step": 4610 + }, + { + "epoch": 1.73868778280543, + "grad_norm": 0.5102327660768813, + "learning_rate": 6.595269098265194e-05, + "loss": 0.4281, + "step": 4611 + }, + { + "epoch": 1.7390648567119156, + "grad_norm": 0.410942434141237, + "learning_rate": 6.59446774243689e-05, + "loss": 0.4381, + "step": 4612 + }, + { + "epoch": 1.7394419306184012, + "grad_norm": 0.2918456969341894, + "learning_rate": 6.593666206814271e-05, + "loss": 0.4245, + "step": 4613 + }, + { + "epoch": 1.739819004524887, + "grad_norm": 0.24397587593609074, + "learning_rate": 6.592864491452887e-05, + "loss": 0.4296, + "step": 4614 + }, + { + "epoch": 1.7401960784313726, + "grad_norm": 0.24377020477594616, + "learning_rate": 6.592062596408291e-05, + "loss": 0.4316, + "step": 4615 + }, + { + "epoch": 1.7405731523378583, + "grad_norm": 0.3266289617617688, + "learning_rate": 6.591260521736058e-05, + "loss": 0.4341, + "step": 4616 + }, + { + "epoch": 1.740950226244344, + "grad_norm": 0.36941115310676365, + "learning_rate": 6.590458267491769e-05, + "loss": 0.4398, + "step": 4617 + }, + { + "epoch": 1.7413273001508296, + "grad_norm": 0.35986905287723625, + "learning_rate": 6.589655833731019e-05, + "loss": 0.4436, + "step": 4618 + }, + { + "epoch": 1.7417043740573153, + "grad_norm": 0.2901915714404463, + "learning_rate": 6.588853220509417e-05, + "loss": 0.4375, + "step": 4619 + }, + { + "epoch": 1.742081447963801, + "grad_norm": 0.2527715178931932, + "learning_rate": 6.588050427882582e-05, + "loss": 0.4405, + "step": 4620 + }, + { + "epoch": 1.7424585218702866, + "grad_norm": 0.20375566536351605, + "learning_rate": 6.58724745590615e-05, + "loss": 0.4405, + "step": 4621 + }, + { + "epoch": 1.7428355957767723, + "grad_norm": 0.21388586747571006, + "learning_rate": 6.586444304635764e-05, + "loss": 0.4446, + "step": 4622 + }, + { + "epoch": 1.743212669683258, + "grad_norm": 0.25763099057648514, + "learning_rate": 6.585640974127081e-05, + "loss": 0.4352, + "step": 4623 + }, + { + "epoch": 1.7435897435897436, + "grad_norm": 0.2896531771107546, + "learning_rate": 6.584837464435773e-05, + "loss": 0.4385, + "step": 4624 + }, + { + "epoch": 1.7439668174962293, + "grad_norm": 0.29960650637278907, + "learning_rate": 6.58403377561752e-05, + "loss": 0.4375, + "step": 4625 + }, + { + "epoch": 1.744343891402715, + "grad_norm": 0.27655082761160815, + "learning_rate": 6.583229907728018e-05, + "loss": 0.4318, + "step": 4626 + }, + { + "epoch": 1.7447209653092006, + "grad_norm": 0.29692650515383434, + "learning_rate": 6.582425860822977e-05, + "loss": 0.4534, + "step": 4627 + }, + { + "epoch": 1.7450980392156863, + "grad_norm": 0.3315877624348048, + "learning_rate": 6.581621634958113e-05, + "loss": 0.4427, + "step": 4628 + }, + { + "epoch": 1.745475113122172, + "grad_norm": 0.3046831414761734, + "learning_rate": 6.580817230189159e-05, + "loss": 0.4452, + "step": 4629 + }, + { + "epoch": 1.7458521870286576, + "grad_norm": 0.2686050959923558, + "learning_rate": 6.580012646571862e-05, + "loss": 0.4436, + "step": 4630 + }, + { + "epoch": 1.7462292609351433, + "grad_norm": 0.26467420919765544, + "learning_rate": 6.579207884161975e-05, + "loss": 0.4498, + "step": 4631 + }, + { + "epoch": 1.746606334841629, + "grad_norm": 0.24385716149966638, + "learning_rate": 6.578402943015271e-05, + "loss": 0.4363, + "step": 4632 + }, + { + "epoch": 1.7469834087481146, + "grad_norm": 0.2978084499650328, + "learning_rate": 6.57759782318753e-05, + "loss": 0.4339, + "step": 4633 + }, + { + "epoch": 1.7473604826546003, + "grad_norm": 0.41967294020278517, + "learning_rate": 6.576792524734545e-05, + "loss": 0.436, + "step": 4634 + }, + { + "epoch": 1.747737556561086, + "grad_norm": 0.5077085769749327, + "learning_rate": 6.575987047712124e-05, + "loss": 0.4396, + "step": 4635 + }, + { + "epoch": 1.7481146304675717, + "grad_norm": 0.509975145717908, + "learning_rate": 6.575181392176087e-05, + "loss": 0.4313, + "step": 4636 + }, + { + "epoch": 1.7484917043740573, + "grad_norm": 0.4372531246377716, + "learning_rate": 6.574375558182261e-05, + "loss": 0.4344, + "step": 4637 + }, + { + "epoch": 1.748868778280543, + "grad_norm": 0.37460785398422414, + "learning_rate": 6.573569545786494e-05, + "loss": 0.4356, + "step": 4638 + }, + { + "epoch": 1.7492458521870287, + "grad_norm": 0.318664255921202, + "learning_rate": 6.572763355044641e-05, + "loss": 0.4392, + "step": 4639 + }, + { + "epoch": 1.7496229260935143, + "grad_norm": 0.268999418720337, + "learning_rate": 6.571956986012568e-05, + "loss": 0.4544, + "step": 4640 + }, + { + "epoch": 1.75, + "grad_norm": 0.30008405547457373, + "learning_rate": 6.571150438746157e-05, + "loss": 0.4338, + "step": 4641 + }, + { + "epoch": 1.7503770739064857, + "grad_norm": 0.33455847433847474, + "learning_rate": 6.570343713301302e-05, + "loss": 0.4439, + "step": 4642 + }, + { + "epoch": 1.7507541478129713, + "grad_norm": 0.34310584389021626, + "learning_rate": 6.569536809733908e-05, + "loss": 0.4409, + "step": 4643 + }, + { + "epoch": 1.751131221719457, + "grad_norm": 0.3226920084884255, + "learning_rate": 6.568729728099893e-05, + "loss": 0.437, + "step": 4644 + }, + { + "epoch": 1.7515082956259427, + "grad_norm": 0.3021544304936396, + "learning_rate": 6.567922468455184e-05, + "loss": 0.4427, + "step": 4645 + }, + { + "epoch": 1.7518853695324283, + "grad_norm": 0.2593908745932738, + "learning_rate": 6.567115030855727e-05, + "loss": 0.4457, + "step": 4646 + }, + { + "epoch": 1.752262443438914, + "grad_norm": 0.3154787536812744, + "learning_rate": 6.566307415357476e-05, + "loss": 0.4256, + "step": 4647 + }, + { + "epoch": 1.7526395173453997, + "grad_norm": 0.36796533708748946, + "learning_rate": 6.565499622016396e-05, + "loss": 0.4337, + "step": 4648 + }, + { + "epoch": 1.7530165912518854, + "grad_norm": 0.34688988752541294, + "learning_rate": 6.564691650888468e-05, + "loss": 0.4288, + "step": 4649 + }, + { + "epoch": 1.753393665158371, + "grad_norm": 0.2937302787338568, + "learning_rate": 6.563883502029684e-05, + "loss": 0.4384, + "step": 4650 + }, + { + "epoch": 1.7537707390648567, + "grad_norm": 0.22830312855136664, + "learning_rate": 6.563075175496048e-05, + "loss": 0.449, + "step": 4651 + }, + { + "epoch": 1.7541478129713424, + "grad_norm": 0.23433989197080382, + "learning_rate": 6.562266671343574e-05, + "loss": 0.4322, + "step": 4652 + }, + { + "epoch": 1.754524886877828, + "grad_norm": 0.2291590802310671, + "learning_rate": 6.561457989628293e-05, + "loss": 0.4528, + "step": 4653 + }, + { + "epoch": 1.7549019607843137, + "grad_norm": 0.2589458967438959, + "learning_rate": 6.560649130406246e-05, + "loss": 0.4353, + "step": 4654 + }, + { + "epoch": 1.7552790346907994, + "grad_norm": 0.2994067660888456, + "learning_rate": 6.559840093733486e-05, + "loss": 0.4368, + "step": 4655 + }, + { + "epoch": 1.755656108597285, + "grad_norm": 0.3421205734015615, + "learning_rate": 6.559030879666077e-05, + "loss": 0.4388, + "step": 4656 + }, + { + "epoch": 1.7560331825037707, + "grad_norm": 0.3851334605217363, + "learning_rate": 6.5582214882601e-05, + "loss": 0.4291, + "step": 4657 + }, + { + "epoch": 1.7564102564102564, + "grad_norm": 0.4305552162377593, + "learning_rate": 6.557411919571641e-05, + "loss": 0.4315, + "step": 4658 + }, + { + "epoch": 1.756787330316742, + "grad_norm": 0.4326869628296667, + "learning_rate": 6.556602173656805e-05, + "loss": 0.4506, + "step": 4659 + }, + { + "epoch": 1.7571644042232277, + "grad_norm": 0.41039906101469753, + "learning_rate": 6.555792250571705e-05, + "loss": 0.4397, + "step": 4660 + }, + { + "epoch": 1.7575414781297134, + "grad_norm": 0.3577946590633223, + "learning_rate": 6.55498215037247e-05, + "loss": 0.4335, + "step": 4661 + }, + { + "epoch": 1.757918552036199, + "grad_norm": 0.34967827195815865, + "learning_rate": 6.554171873115238e-05, + "loss": 0.433, + "step": 4662 + }, + { + "epoch": 1.7582956259426847, + "grad_norm": 0.292365128061322, + "learning_rate": 6.55336141885616e-05, + "loss": 0.4357, + "step": 4663 + }, + { + "epoch": 1.7586726998491704, + "grad_norm": 0.23918879490521122, + "learning_rate": 6.552550787651403e-05, + "loss": 0.4304, + "step": 4664 + }, + { + "epoch": 1.759049773755656, + "grad_norm": 0.2417658258955777, + "learning_rate": 6.551739979557138e-05, + "loss": 0.4328, + "step": 4665 + }, + { + "epoch": 1.7594268476621417, + "grad_norm": 0.3161773788288283, + "learning_rate": 6.550928994629557e-05, + "loss": 0.4547, + "step": 4666 + }, + { + "epoch": 1.7598039215686274, + "grad_norm": 0.3649169026464407, + "learning_rate": 6.55011783292486e-05, + "loss": 0.4324, + "step": 4667 + }, + { + "epoch": 1.760180995475113, + "grad_norm": 0.39525524274284746, + "learning_rate": 6.549306494499256e-05, + "loss": 0.4389, + "step": 4668 + }, + { + "epoch": 1.7605580693815988, + "grad_norm": 0.42103851313914575, + "learning_rate": 6.548494979408976e-05, + "loss": 0.4377, + "step": 4669 + }, + { + "epoch": 1.7609351432880844, + "grad_norm": 0.45832424203342553, + "learning_rate": 6.547683287710254e-05, + "loss": 0.4426, + "step": 4670 + }, + { + "epoch": 1.76131221719457, + "grad_norm": 0.4486564354014962, + "learning_rate": 6.546871419459339e-05, + "loss": 0.4429, + "step": 4671 + }, + { + "epoch": 1.7616892911010558, + "grad_norm": 0.4110859456199352, + "learning_rate": 6.546059374712494e-05, + "loss": 0.4374, + "step": 4672 + }, + { + "epoch": 1.7620663650075414, + "grad_norm": 0.3838723776185939, + "learning_rate": 6.545247153525994e-05, + "loss": 0.438, + "step": 4673 + }, + { + "epoch": 1.762443438914027, + "grad_norm": 0.4088552148748501, + "learning_rate": 6.544434755956123e-05, + "loss": 0.4384, + "step": 4674 + }, + { + "epoch": 1.7628205128205128, + "grad_norm": 0.36504065097403926, + "learning_rate": 6.54362218205918e-05, + "loss": 0.4272, + "step": 4675 + }, + { + "epoch": 1.7631975867269984, + "grad_norm": 0.3025839359284935, + "learning_rate": 6.542809431891476e-05, + "loss": 0.4307, + "step": 4676 + }, + { + "epoch": 1.7635746606334841, + "grad_norm": 0.36501129214237354, + "learning_rate": 6.541996505509335e-05, + "loss": 0.4392, + "step": 4677 + }, + { + "epoch": 1.7639517345399698, + "grad_norm": 0.4593926102224298, + "learning_rate": 6.54118340296909e-05, + "loss": 0.4332, + "step": 4678 + }, + { + "epoch": 1.7643288084464555, + "grad_norm": 0.46423009336437204, + "learning_rate": 6.540370124327089e-05, + "loss": 0.4315, + "step": 4679 + }, + { + "epoch": 1.7647058823529411, + "grad_norm": 0.35584053513345915, + "learning_rate": 6.539556669639691e-05, + "loss": 0.437, + "step": 4680 + }, + { + "epoch": 1.7650829562594268, + "grad_norm": 0.24467824089024998, + "learning_rate": 6.53874303896327e-05, + "loss": 0.4462, + "step": 4681 + }, + { + "epoch": 1.7654600301659125, + "grad_norm": 0.31975922459865874, + "learning_rate": 6.537929232354208e-05, + "loss": 0.4422, + "step": 4682 + }, + { + "epoch": 1.7658371040723981, + "grad_norm": 0.3739481645267712, + "learning_rate": 6.537115249868902e-05, + "loss": 0.4358, + "step": 4683 + }, + { + "epoch": 1.7662141779788838, + "grad_norm": 0.3717467385334578, + "learning_rate": 6.536301091563757e-05, + "loss": 0.4231, + "step": 4684 + }, + { + "epoch": 1.7665912518853695, + "grad_norm": 0.33087312624095244, + "learning_rate": 6.535486757495198e-05, + "loss": 0.4456, + "step": 4685 + }, + { + "epoch": 1.7669683257918551, + "grad_norm": 0.2935985494580059, + "learning_rate": 6.534672247719655e-05, + "loss": 0.4259, + "step": 4686 + }, + { + "epoch": 1.7673453996983408, + "grad_norm": 0.29223935665027717, + "learning_rate": 6.533857562293573e-05, + "loss": 0.4399, + "step": 4687 + }, + { + "epoch": 1.7677224736048265, + "grad_norm": 0.3624851501662224, + "learning_rate": 6.533042701273412e-05, + "loss": 0.4351, + "step": 4688 + }, + { + "epoch": 1.7680995475113122, + "grad_norm": 0.3522475511204694, + "learning_rate": 6.532227664715636e-05, + "loss": 0.4483, + "step": 4689 + }, + { + "epoch": 1.7684766214177978, + "grad_norm": 0.2576854124709336, + "learning_rate": 6.53141245267673e-05, + "loss": 0.4325, + "step": 4690 + }, + { + "epoch": 1.7688536953242835, + "grad_norm": 0.24392115351800892, + "learning_rate": 6.530597065213186e-05, + "loss": 0.4228, + "step": 4691 + }, + { + "epoch": 1.7692307692307692, + "grad_norm": 0.2560295958460675, + "learning_rate": 6.529781502381509e-05, + "loss": 0.4514, + "step": 4692 + }, + { + "epoch": 1.7696078431372548, + "grad_norm": 0.2610772467099813, + "learning_rate": 6.528965764238219e-05, + "loss": 0.4376, + "step": 4693 + }, + { + "epoch": 1.7699849170437405, + "grad_norm": 0.29235431949644924, + "learning_rate": 6.528149850839845e-05, + "loss": 0.4382, + "step": 4694 + }, + { + "epoch": 1.7703619909502262, + "grad_norm": 0.33532695568896337, + "learning_rate": 6.527333762242928e-05, + "loss": 0.4347, + "step": 4695 + }, + { + "epoch": 1.7707390648567118, + "grad_norm": 0.32539332138408905, + "learning_rate": 6.526517498504024e-05, + "loss": 0.4289, + "step": 4696 + }, + { + "epoch": 1.7711161387631975, + "grad_norm": 0.2939172032726036, + "learning_rate": 6.525701059679697e-05, + "loss": 0.4371, + "step": 4697 + }, + { + "epoch": 1.7714932126696832, + "grad_norm": 0.32057222598944024, + "learning_rate": 6.524884445826529e-05, + "loss": 0.4378, + "step": 4698 + }, + { + "epoch": 1.7718702865761689, + "grad_norm": 0.3247342586854191, + "learning_rate": 6.524067657001107e-05, + "loss": 0.4301, + "step": 4699 + }, + { + "epoch": 1.7722473604826545, + "grad_norm": 0.25939466689904633, + "learning_rate": 6.523250693260035e-05, + "loss": 0.4406, + "step": 4700 + }, + { + "epoch": 1.7726244343891402, + "grad_norm": 0.2727435864269429, + "learning_rate": 6.522433554659929e-05, + "loss": 0.4466, + "step": 4701 + }, + { + "epoch": 1.7730015082956259, + "grad_norm": 0.29979885225172626, + "learning_rate": 6.521616241257415e-05, + "loss": 0.4433, + "step": 4702 + }, + { + "epoch": 1.7733785822021115, + "grad_norm": 0.2716849292687178, + "learning_rate": 6.520798753109131e-05, + "loss": 0.4345, + "step": 4703 + }, + { + "epoch": 1.7737556561085972, + "grad_norm": 0.32491709801073887, + "learning_rate": 6.51998109027173e-05, + "loss": 0.4329, + "step": 4704 + }, + { + "epoch": 1.7741327300150829, + "grad_norm": 0.36589967038215715, + "learning_rate": 6.519163252801875e-05, + "loss": 0.4394, + "step": 4705 + }, + { + "epoch": 1.7745098039215685, + "grad_norm": 0.39369052350878814, + "learning_rate": 6.518345240756241e-05, + "loss": 0.4404, + "step": 4706 + }, + { + "epoch": 1.7748868778280542, + "grad_norm": 0.40817844974101575, + "learning_rate": 6.517527054191515e-05, + "loss": 0.4323, + "step": 4707 + }, + { + "epoch": 1.7752639517345399, + "grad_norm": 0.3790662079407362, + "learning_rate": 6.516708693164397e-05, + "loss": 0.4312, + "step": 4708 + }, + { + "epoch": 1.7756410256410255, + "grad_norm": 0.24777849261399493, + "learning_rate": 6.515890157731599e-05, + "loss": 0.433, + "step": 4709 + }, + { + "epoch": 1.7760180995475112, + "grad_norm": 0.2088611420789491, + "learning_rate": 6.515071447949846e-05, + "loss": 0.4372, + "step": 4710 + }, + { + "epoch": 1.776395173453997, + "grad_norm": 0.2645151572968029, + "learning_rate": 6.514252563875872e-05, + "loss": 0.4309, + "step": 4711 + }, + { + "epoch": 1.7767722473604826, + "grad_norm": 0.282069484310066, + "learning_rate": 6.513433505566425e-05, + "loss": 0.4428, + "step": 4712 + }, + { + "epoch": 1.7771493212669682, + "grad_norm": 0.3016623820129838, + "learning_rate": 6.512614273078265e-05, + "loss": 0.429, + "step": 4713 + }, + { + "epoch": 1.777526395173454, + "grad_norm": 0.32727792028917374, + "learning_rate": 6.511794866468165e-05, + "loss": 0.4445, + "step": 4714 + }, + { + "epoch": 1.7779034690799396, + "grad_norm": 0.35283449511224846, + "learning_rate": 6.510975285792908e-05, + "loss": 0.4388, + "step": 4715 + }, + { + "epoch": 1.7782805429864252, + "grad_norm": 0.33613164719105554, + "learning_rate": 6.510155531109294e-05, + "loss": 0.4419, + "step": 4716 + }, + { + "epoch": 1.778657616892911, + "grad_norm": 0.33184926559986433, + "learning_rate": 6.509335602474125e-05, + "loss": 0.4376, + "step": 4717 + }, + { + "epoch": 1.7790346907993966, + "grad_norm": 0.3249296187533255, + "learning_rate": 6.508515499944224e-05, + "loss": 0.4373, + "step": 4718 + }, + { + "epoch": 1.7794117647058822, + "grad_norm": 0.29280072573061694, + "learning_rate": 6.507695223576428e-05, + "loss": 0.4542, + "step": 4719 + }, + { + "epoch": 1.779788838612368, + "grad_norm": 0.26561915383164775, + "learning_rate": 6.506874773427574e-05, + "loss": 0.4262, + "step": 4720 + }, + { + "epoch": 1.7801659125188536, + "grad_norm": 0.22769923394803845, + "learning_rate": 6.506054149554521e-05, + "loss": 0.4382, + "step": 4721 + }, + { + "epoch": 1.7805429864253393, + "grad_norm": 0.22679094654281964, + "learning_rate": 6.505233352014141e-05, + "loss": 0.4285, + "step": 4722 + }, + { + "epoch": 1.780920060331825, + "grad_norm": 0.22517296545155363, + "learning_rate": 6.50441238086331e-05, + "loss": 0.4466, + "step": 4723 + }, + { + "epoch": 1.7812971342383106, + "grad_norm": 0.2850167435023046, + "learning_rate": 6.503591236158924e-05, + "loss": 0.4267, + "step": 4724 + }, + { + "epoch": 1.7816742081447963, + "grad_norm": 0.3767522347474853, + "learning_rate": 6.502769917957884e-05, + "loss": 0.4285, + "step": 4725 + }, + { + "epoch": 1.782051282051282, + "grad_norm": 0.3728615979042167, + "learning_rate": 6.50194842631711e-05, + "loss": 0.4312, + "step": 4726 + }, + { + "epoch": 1.7824283559577676, + "grad_norm": 0.36521395009090424, + "learning_rate": 6.501126761293529e-05, + "loss": 0.4383, + "step": 4727 + }, + { + "epoch": 1.7828054298642533, + "grad_norm": 0.42919843962129484, + "learning_rate": 6.500304922944082e-05, + "loss": 0.444, + "step": 4728 + }, + { + "epoch": 1.783182503770739, + "grad_norm": 0.46831256203611554, + "learning_rate": 6.499482911325721e-05, + "loss": 0.4354, + "step": 4729 + }, + { + "epoch": 1.7835595776772246, + "grad_norm": 0.4751171964175993, + "learning_rate": 6.498660726495413e-05, + "loss": 0.4325, + "step": 4730 + }, + { + "epoch": 1.7839366515837103, + "grad_norm": 0.44722707044814447, + "learning_rate": 6.497838368510131e-05, + "loss": 0.4343, + "step": 4731 + }, + { + "epoch": 1.784313725490196, + "grad_norm": 0.42652088810098127, + "learning_rate": 6.497015837426868e-05, + "loss": 0.4456, + "step": 4732 + }, + { + "epoch": 1.7846907993966816, + "grad_norm": 0.3598498526930446, + "learning_rate": 6.496193133302619e-05, + "loss": 0.4484, + "step": 4733 + }, + { + "epoch": 1.7850678733031673, + "grad_norm": 0.2774937926426895, + "learning_rate": 6.495370256194402e-05, + "loss": 0.4407, + "step": 4734 + }, + { + "epoch": 1.785444947209653, + "grad_norm": 0.3528566219894574, + "learning_rate": 6.49454720615924e-05, + "loss": 0.4422, + "step": 4735 + }, + { + "epoch": 1.7858220211161386, + "grad_norm": 0.3991314198185257, + "learning_rate": 6.493723983254168e-05, + "loss": 0.4301, + "step": 4736 + }, + { + "epoch": 1.7861990950226243, + "grad_norm": 0.39959372099663676, + "learning_rate": 6.492900587536237e-05, + "loss": 0.4307, + "step": 4737 + }, + { + "epoch": 1.78657616892911, + "grad_norm": 0.3745070736930385, + "learning_rate": 6.492077019062506e-05, + "loss": 0.4358, + "step": 4738 + }, + { + "epoch": 1.7869532428355956, + "grad_norm": 0.3626535490913078, + "learning_rate": 6.491253277890048e-05, + "loss": 0.4395, + "step": 4739 + }, + { + "epoch": 1.7873303167420813, + "grad_norm": 0.3550349082378268, + "learning_rate": 6.490429364075948e-05, + "loss": 0.4365, + "step": 4740 + }, + { + "epoch": 1.787707390648567, + "grad_norm": 0.3149290400216713, + "learning_rate": 6.4896052776773e-05, + "loss": 0.4336, + "step": 4741 + }, + { + "epoch": 1.7880844645550527, + "grad_norm": 0.2776146378655148, + "learning_rate": 6.488781018751218e-05, + "loss": 0.4406, + "step": 4742 + }, + { + "epoch": 1.7884615384615383, + "grad_norm": 0.27503940489279316, + "learning_rate": 6.487956587354816e-05, + "loss": 0.4309, + "step": 4743 + }, + { + "epoch": 1.788838612368024, + "grad_norm": 0.3223165177938816, + "learning_rate": 6.487131983545231e-05, + "loss": 0.433, + "step": 4744 + }, + { + "epoch": 1.7892156862745097, + "grad_norm": 0.3488631358664068, + "learning_rate": 6.486307207379606e-05, + "loss": 0.4321, + "step": 4745 + }, + { + "epoch": 1.7895927601809953, + "grad_norm": 0.29271061922369584, + "learning_rate": 6.485482258915096e-05, + "loss": 0.4314, + "step": 4746 + }, + { + "epoch": 1.789969834087481, + "grad_norm": 0.2690798365345002, + "learning_rate": 6.484657138208869e-05, + "loss": 0.4422, + "step": 4747 + }, + { + "epoch": 1.7903469079939667, + "grad_norm": 0.27892919243460923, + "learning_rate": 6.483831845318108e-05, + "loss": 0.4495, + "step": 4748 + }, + { + "epoch": 1.7907239819004523, + "grad_norm": 0.2923513100684061, + "learning_rate": 6.483006380300003e-05, + "loss": 0.4511, + "step": 4749 + }, + { + "epoch": 1.791101055806938, + "grad_norm": 0.30762191788996496, + "learning_rate": 6.482180743211757e-05, + "loss": 0.436, + "step": 4750 + }, + { + "epoch": 1.7914781297134237, + "grad_norm": 0.30922913569804916, + "learning_rate": 6.481354934110588e-05, + "loss": 0.4399, + "step": 4751 + }, + { + "epoch": 1.7918552036199094, + "grad_norm": 0.28061548352501847, + "learning_rate": 6.480528953053723e-05, + "loss": 0.44, + "step": 4752 + }, + { + "epoch": 1.792232277526395, + "grad_norm": 0.28157863369992525, + "learning_rate": 6.479702800098401e-05, + "loss": 0.4318, + "step": 4753 + }, + { + "epoch": 1.7926093514328807, + "grad_norm": 0.29112725710564347, + "learning_rate": 6.478876475301877e-05, + "loss": 0.4448, + "step": 4754 + }, + { + "epoch": 1.7929864253393664, + "grad_norm": 0.3503480784999252, + "learning_rate": 6.478049978721411e-05, + "loss": 0.4308, + "step": 4755 + }, + { + "epoch": 1.793363499245852, + "grad_norm": 0.41802493676413, + "learning_rate": 6.477223310414278e-05, + "loss": 0.4409, + "step": 4756 + }, + { + "epoch": 1.7937405731523377, + "grad_norm": 0.34825325793002726, + "learning_rate": 6.476396470437766e-05, + "loss": 0.423, + "step": 4757 + }, + { + "epoch": 1.7941176470588234, + "grad_norm": 0.24299063759993877, + "learning_rate": 6.475569458849178e-05, + "loss": 0.4403, + "step": 4758 + }, + { + "epoch": 1.794494720965309, + "grad_norm": 0.25246195065382326, + "learning_rate": 6.474742275705822e-05, + "loss": 0.4397, + "step": 4759 + }, + { + "epoch": 1.7948717948717947, + "grad_norm": 0.34259533402750353, + "learning_rate": 6.47391492106502e-05, + "loss": 0.4338, + "step": 4760 + }, + { + "epoch": 1.7952488687782804, + "grad_norm": 0.33990298777984285, + "learning_rate": 6.47308739498411e-05, + "loss": 0.4339, + "step": 4761 + }, + { + "epoch": 1.795625942684766, + "grad_norm": 0.29621105609980497, + "learning_rate": 6.472259697520436e-05, + "loss": 0.4406, + "step": 4762 + }, + { + "epoch": 1.7960030165912517, + "grad_norm": 0.4351238790757566, + "learning_rate": 6.471431828731357e-05, + "loss": 0.433, + "step": 4763 + }, + { + "epoch": 1.7963800904977374, + "grad_norm": 0.6026055773041136, + "learning_rate": 6.470603788674246e-05, + "loss": 0.4459, + "step": 4764 + }, + { + "epoch": 1.796757164404223, + "grad_norm": 0.5942628125077624, + "learning_rate": 6.469775577406483e-05, + "loss": 0.4332, + "step": 4765 + }, + { + "epoch": 1.797134238310709, + "grad_norm": 0.4824102936985714, + "learning_rate": 6.468947194985463e-05, + "loss": 0.4373, + "step": 4766 + }, + { + "epoch": 1.7975113122171946, + "grad_norm": 0.39758896895961476, + "learning_rate": 6.468118641468593e-05, + "loss": 0.4265, + "step": 4767 + }, + { + "epoch": 1.7978883861236803, + "grad_norm": 0.35545851472522827, + "learning_rate": 6.46728991691329e-05, + "loss": 0.4372, + "step": 4768 + }, + { + "epoch": 1.798265460030166, + "grad_norm": 0.3682802307243223, + "learning_rate": 6.466461021376984e-05, + "loss": 0.432, + "step": 4769 + }, + { + "epoch": 1.7986425339366516, + "grad_norm": 0.4201226440402914, + "learning_rate": 6.465631954917118e-05, + "loss": 0.4411, + "step": 4770 + }, + { + "epoch": 1.7990196078431373, + "grad_norm": 0.37876766902900705, + "learning_rate": 6.464802717591143e-05, + "loss": 0.4322, + "step": 4771 + }, + { + "epoch": 1.799396681749623, + "grad_norm": 0.3688569112375617, + "learning_rate": 6.463973309456528e-05, + "loss": 0.4409, + "step": 4772 + }, + { + "epoch": 1.7997737556561086, + "grad_norm": 0.36627808150062086, + "learning_rate": 6.463143730570746e-05, + "loss": 0.444, + "step": 4773 + }, + { + "epoch": 1.8001508295625943, + "grad_norm": 0.37531440163372165, + "learning_rate": 6.462313980991291e-05, + "loss": 0.4215, + "step": 4774 + }, + { + "epoch": 1.80052790346908, + "grad_norm": 0.32098333803104734, + "learning_rate": 6.461484060775658e-05, + "loss": 0.4414, + "step": 4775 + }, + { + "epoch": 1.8009049773755657, + "grad_norm": 0.2577622871797308, + "learning_rate": 6.460653969981365e-05, + "loss": 0.4531, + "step": 4776 + }, + { + "epoch": 1.8012820512820513, + "grad_norm": 0.386121410905289, + "learning_rate": 6.459823708665934e-05, + "loss": 0.4306, + "step": 4777 + }, + { + "epoch": 1.801659125188537, + "grad_norm": 0.4945790196731151, + "learning_rate": 6.458993276886901e-05, + "loss": 0.4277, + "step": 4778 + }, + { + "epoch": 1.8020361990950227, + "grad_norm": 0.4691096730850483, + "learning_rate": 6.458162674701816e-05, + "loss": 0.4378, + "step": 4779 + }, + { + "epoch": 1.8024132730015083, + "grad_norm": 0.3168265089622196, + "learning_rate": 6.457331902168237e-05, + "loss": 0.4332, + "step": 4780 + }, + { + "epoch": 1.802790346907994, + "grad_norm": 0.26271736230176407, + "learning_rate": 6.456500959343739e-05, + "loss": 0.4401, + "step": 4781 + }, + { + "epoch": 1.8031674208144797, + "grad_norm": 0.3408693947121041, + "learning_rate": 6.455669846285901e-05, + "loss": 0.4392, + "step": 4782 + }, + { + "epoch": 1.8035444947209653, + "grad_norm": 0.3494956199133135, + "learning_rate": 6.454838563052322e-05, + "loss": 0.4386, + "step": 4783 + }, + { + "epoch": 1.803921568627451, + "grad_norm": 0.2672897098457332, + "learning_rate": 6.454007109700608e-05, + "loss": 0.4426, + "step": 4784 + }, + { + "epoch": 1.8042986425339367, + "grad_norm": 0.3093816215475023, + "learning_rate": 6.453175486288379e-05, + "loss": 0.4295, + "step": 4785 + }, + { + "epoch": 1.8046757164404224, + "grad_norm": 0.375381466368877, + "learning_rate": 6.452343692873262e-05, + "loss": 0.4357, + "step": 4786 + }, + { + "epoch": 1.805052790346908, + "grad_norm": 0.32116868125189785, + "learning_rate": 6.451511729512904e-05, + "loss": 0.4325, + "step": 4787 + }, + { + "epoch": 1.8054298642533937, + "grad_norm": 0.29874268082547584, + "learning_rate": 6.450679596264957e-05, + "loss": 0.4274, + "step": 4788 + }, + { + "epoch": 1.8058069381598794, + "grad_norm": 0.33196508230415794, + "learning_rate": 6.449847293187089e-05, + "loss": 0.4264, + "step": 4789 + }, + { + "epoch": 1.806184012066365, + "grad_norm": 0.28917343810656343, + "learning_rate": 6.449014820336977e-05, + "loss": 0.4343, + "step": 4790 + }, + { + "epoch": 1.8065610859728507, + "grad_norm": 0.2528030511892693, + "learning_rate": 6.448182177772308e-05, + "loss": 0.4404, + "step": 4791 + }, + { + "epoch": 1.8069381598793364, + "grad_norm": 0.22829341899725786, + "learning_rate": 6.447349365550787e-05, + "loss": 0.4442, + "step": 4792 + }, + { + "epoch": 1.807315233785822, + "grad_norm": 0.2583498702505945, + "learning_rate": 6.446516383730127e-05, + "loss": 0.4382, + "step": 4793 + }, + { + "epoch": 1.8076923076923077, + "grad_norm": 0.29530964694057904, + "learning_rate": 6.44568323236805e-05, + "loss": 0.4458, + "step": 4794 + }, + { + "epoch": 1.8080693815987934, + "grad_norm": 0.2889267555520883, + "learning_rate": 6.444849911522294e-05, + "loss": 0.4313, + "step": 4795 + }, + { + "epoch": 1.808446455505279, + "grad_norm": 0.3010179300325856, + "learning_rate": 6.444016421250607e-05, + "loss": 0.4282, + "step": 4796 + }, + { + "epoch": 1.8088235294117647, + "grad_norm": 0.3262884530629475, + "learning_rate": 6.443182761610752e-05, + "loss": 0.4344, + "step": 4797 + }, + { + "epoch": 1.8092006033182504, + "grad_norm": 0.2897847904967042, + "learning_rate": 6.442348932660498e-05, + "loss": 0.4317, + "step": 4798 + }, + { + "epoch": 1.809577677224736, + "grad_norm": 0.23808213279396964, + "learning_rate": 6.441514934457628e-05, + "loss": 0.4287, + "step": 4799 + }, + { + "epoch": 1.8099547511312217, + "grad_norm": 0.22843235517673033, + "learning_rate": 6.440680767059941e-05, + "loss": 0.4259, + "step": 4800 + }, + { + "epoch": 1.8103318250377074, + "grad_norm": 0.28615980712963945, + "learning_rate": 6.43984643052524e-05, + "loss": 0.4334, + "step": 4801 + }, + { + "epoch": 1.810708898944193, + "grad_norm": 0.28319687435404506, + "learning_rate": 6.439011924911346e-05, + "loss": 0.4334, + "step": 4802 + }, + { + "epoch": 1.8110859728506787, + "grad_norm": 0.2922057466080407, + "learning_rate": 6.438177250276089e-05, + "loss": 0.435, + "step": 4803 + }, + { + "epoch": 1.8114630467571644, + "grad_norm": 0.32120927346630196, + "learning_rate": 6.437342406677311e-05, + "loss": 0.4292, + "step": 4804 + }, + { + "epoch": 1.81184012066365, + "grad_norm": 0.3457162302728207, + "learning_rate": 6.436507394172866e-05, + "loss": 0.4245, + "step": 4805 + }, + { + "epoch": 1.8122171945701357, + "grad_norm": 0.3526067016256077, + "learning_rate": 6.435672212820621e-05, + "loss": 0.4353, + "step": 4806 + }, + { + "epoch": 1.8125942684766214, + "grad_norm": 0.2972842876942046, + "learning_rate": 6.434836862678451e-05, + "loss": 0.4392, + "step": 4807 + }, + { + "epoch": 1.812971342383107, + "grad_norm": 0.2525854729497138, + "learning_rate": 6.434001343804246e-05, + "loss": 0.4442, + "step": 4808 + }, + { + "epoch": 1.8133484162895928, + "grad_norm": 0.2333786509029339, + "learning_rate": 6.433165656255907e-05, + "loss": 0.4365, + "step": 4809 + }, + { + "epoch": 1.8137254901960784, + "grad_norm": 0.19762907139464916, + "learning_rate": 6.432329800091347e-05, + "loss": 0.44, + "step": 4810 + }, + { + "epoch": 1.814102564102564, + "grad_norm": 0.20179716215206944, + "learning_rate": 6.431493775368488e-05, + "loss": 0.4287, + "step": 4811 + }, + { + "epoch": 1.8144796380090498, + "grad_norm": 0.2582278656532671, + "learning_rate": 6.430657582145268e-05, + "loss": 0.4293, + "step": 4812 + }, + { + "epoch": 1.8148567119155354, + "grad_norm": 0.29706791452422016, + "learning_rate": 6.429821220479634e-05, + "loss": 0.4412, + "step": 4813 + }, + { + "epoch": 1.815233785822021, + "grad_norm": 0.30352218320924484, + "learning_rate": 6.428984690429543e-05, + "loss": 0.4378, + "step": 4814 + }, + { + "epoch": 1.8156108597285068, + "grad_norm": 0.2948006098681184, + "learning_rate": 6.42814799205297e-05, + "loss": 0.4421, + "step": 4815 + }, + { + "epoch": 1.8159879336349924, + "grad_norm": 0.4002701242871594, + "learning_rate": 6.427311125407893e-05, + "loss": 0.4372, + "step": 4816 + }, + { + "epoch": 1.8163650075414781, + "grad_norm": 0.4660180394741998, + "learning_rate": 6.426474090552308e-05, + "loss": 0.4418, + "step": 4817 + }, + { + "epoch": 1.8167420814479638, + "grad_norm": 0.44735427995803634, + "learning_rate": 6.425636887544221e-05, + "loss": 0.4258, + "step": 4818 + }, + { + "epoch": 1.8171191553544495, + "grad_norm": 0.4042130624870835, + "learning_rate": 6.42479951644165e-05, + "loss": 0.4354, + "step": 4819 + }, + { + "epoch": 1.8174962292609351, + "grad_norm": 0.39524757623819096, + "learning_rate": 6.423961977302623e-05, + "loss": 0.4294, + "step": 4820 + }, + { + "epoch": 1.8178733031674208, + "grad_norm": 0.43329001387806204, + "learning_rate": 6.42312427018518e-05, + "loss": 0.4463, + "step": 4821 + }, + { + "epoch": 1.8182503770739065, + "grad_norm": 0.4615227508288967, + "learning_rate": 6.422286395147377e-05, + "loss": 0.428, + "step": 4822 + }, + { + "epoch": 1.8186274509803921, + "grad_norm": 0.4905719754134773, + "learning_rate": 6.421448352247272e-05, + "loss": 0.4374, + "step": 4823 + }, + { + "epoch": 1.8190045248868778, + "grad_norm": 0.47156821829321316, + "learning_rate": 6.420610141542945e-05, + "loss": 0.4289, + "step": 4824 + }, + { + "epoch": 1.8193815987933635, + "grad_norm": 0.3955744754239562, + "learning_rate": 6.419771763092484e-05, + "loss": 0.4334, + "step": 4825 + }, + { + "epoch": 1.8197586726998491, + "grad_norm": 0.3048644735797756, + "learning_rate": 6.418933216953984e-05, + "loss": 0.4436, + "step": 4826 + }, + { + "epoch": 1.8201357466063348, + "grad_norm": 0.2614849800731218, + "learning_rate": 6.41809450318556e-05, + "loss": 0.4305, + "step": 4827 + }, + { + "epoch": 1.8205128205128205, + "grad_norm": 0.18981924389648444, + "learning_rate": 6.41725562184533e-05, + "loss": 0.4413, + "step": 4828 + }, + { + "epoch": 1.8208898944193062, + "grad_norm": 0.3340921274176523, + "learning_rate": 6.41641657299143e-05, + "loss": 0.4411, + "step": 4829 + }, + { + "epoch": 1.8212669683257918, + "grad_norm": 0.40165837457749526, + "learning_rate": 6.415577356682004e-05, + "loss": 0.4264, + "step": 4830 + }, + { + "epoch": 1.8216440422322775, + "grad_norm": 0.34726249910022056, + "learning_rate": 6.414737972975211e-05, + "loss": 0.4312, + "step": 4831 + }, + { + "epoch": 1.8220211161387632, + "grad_norm": 0.2698276999729396, + "learning_rate": 6.41389842192922e-05, + "loss": 0.4262, + "step": 4832 + }, + { + "epoch": 1.8223981900452488, + "grad_norm": 0.24456576569091965, + "learning_rate": 6.413058703602206e-05, + "loss": 0.4246, + "step": 4833 + }, + { + "epoch": 1.8227752639517345, + "grad_norm": 0.24702779916529918, + "learning_rate": 6.412218818052365e-05, + "loss": 0.43, + "step": 4834 + }, + { + "epoch": 1.8231523378582202, + "grad_norm": 0.3202055592463292, + "learning_rate": 6.411378765337902e-05, + "loss": 0.4201, + "step": 4835 + }, + { + "epoch": 1.8235294117647058, + "grad_norm": 0.3194282318407174, + "learning_rate": 6.410538545517026e-05, + "loss": 0.4316, + "step": 4836 + }, + { + "epoch": 1.8239064856711915, + "grad_norm": 0.2732558689657931, + "learning_rate": 6.409698158647969e-05, + "loss": 0.4296, + "step": 4837 + }, + { + "epoch": 1.8242835595776772, + "grad_norm": 0.24418107180217438, + "learning_rate": 6.408857604788966e-05, + "loss": 0.4295, + "step": 4838 + }, + { + "epoch": 1.8246606334841629, + "grad_norm": 0.2788428602591698, + "learning_rate": 6.40801688399827e-05, + "loss": 0.436, + "step": 4839 + }, + { + "epoch": 1.8250377073906485, + "grad_norm": 0.2955233669615414, + "learning_rate": 6.407175996334137e-05, + "loss": 0.4328, + "step": 4840 + }, + { + "epoch": 1.8254147812971342, + "grad_norm": 0.2859915076747467, + "learning_rate": 6.406334941854844e-05, + "loss": 0.4304, + "step": 4841 + }, + { + "epoch": 1.8257918552036199, + "grad_norm": 0.26279825094567394, + "learning_rate": 6.405493720618673e-05, + "loss": 0.4278, + "step": 4842 + }, + { + "epoch": 1.8261689291101055, + "grad_norm": 0.3229035657248369, + "learning_rate": 6.404652332683921e-05, + "loss": 0.4374, + "step": 4843 + }, + { + "epoch": 1.8265460030165912, + "grad_norm": 0.3527394182809686, + "learning_rate": 6.403810778108897e-05, + "loss": 0.4358, + "step": 4844 + }, + { + "epoch": 1.8269230769230769, + "grad_norm": 0.338509986004157, + "learning_rate": 6.402969056951916e-05, + "loss": 0.4385, + "step": 4845 + }, + { + "epoch": 1.8273001508295625, + "grad_norm": 0.3119039416137734, + "learning_rate": 6.402127169271312e-05, + "loss": 0.4311, + "step": 4846 + }, + { + "epoch": 1.8276772247360482, + "grad_norm": 0.27592345947898794, + "learning_rate": 6.401285115125426e-05, + "loss": 0.448, + "step": 4847 + }, + { + "epoch": 1.8280542986425339, + "grad_norm": 0.26892136545570067, + "learning_rate": 6.40044289457261e-05, + "loss": 0.4333, + "step": 4848 + }, + { + "epoch": 1.8284313725490198, + "grad_norm": 0.30168499735644955, + "learning_rate": 6.399600507671233e-05, + "loss": 0.4404, + "step": 4849 + }, + { + "epoch": 1.8288084464555054, + "grad_norm": 0.2654047810863477, + "learning_rate": 6.398757954479667e-05, + "loss": 0.4413, + "step": 4850 + }, + { + "epoch": 1.8291855203619911, + "grad_norm": 0.23767389211651846, + "learning_rate": 6.397915235056304e-05, + "loss": 0.4411, + "step": 4851 + }, + { + "epoch": 1.8295625942684768, + "grad_norm": 0.2556904492129591, + "learning_rate": 6.397072349459541e-05, + "loss": 0.4445, + "step": 4852 + }, + { + "epoch": 1.8299396681749625, + "grad_norm": 0.3065090515551391, + "learning_rate": 6.396229297747791e-05, + "loss": 0.4442, + "step": 4853 + }, + { + "epoch": 1.8303167420814481, + "grad_norm": 0.31656797070427456, + "learning_rate": 6.395386079979477e-05, + "loss": 0.4342, + "step": 4854 + }, + { + "epoch": 1.8306938159879338, + "grad_norm": 0.3020168844590613, + "learning_rate": 6.394542696213031e-05, + "loss": 0.4299, + "step": 4855 + }, + { + "epoch": 1.8310708898944195, + "grad_norm": 0.3917411608406224, + "learning_rate": 6.393699146506901e-05, + "loss": 0.4406, + "step": 4856 + }, + { + "epoch": 1.8314479638009051, + "grad_norm": 0.41363413654800296, + "learning_rate": 6.392855430919543e-05, + "loss": 0.4271, + "step": 4857 + }, + { + "epoch": 1.8318250377073908, + "grad_norm": 0.3079289847722077, + "learning_rate": 6.392011549509425e-05, + "loss": 0.4314, + "step": 4858 + }, + { + "epoch": 1.8322021116138765, + "grad_norm": 0.2497630406643751, + "learning_rate": 6.39116750233503e-05, + "loss": 0.4356, + "step": 4859 + }, + { + "epoch": 1.8325791855203621, + "grad_norm": 0.3405791027917721, + "learning_rate": 6.390323289454847e-05, + "loss": 0.4399, + "step": 4860 + }, + { + "epoch": 1.8329562594268478, + "grad_norm": 0.3676151534472108, + "learning_rate": 6.389478910927381e-05, + "loss": 0.4456, + "step": 4861 + }, + { + "epoch": 1.8333333333333335, + "grad_norm": 0.3417560448465933, + "learning_rate": 6.388634366811146e-05, + "loss": 0.4471, + "step": 4862 + }, + { + "epoch": 1.8337104072398192, + "grad_norm": 0.3155746994765893, + "learning_rate": 6.387789657164668e-05, + "loss": 0.4298, + "step": 4863 + }, + { + "epoch": 1.8340874811463048, + "grad_norm": 0.32975371792892444, + "learning_rate": 6.386944782046483e-05, + "loss": 0.4564, + "step": 4864 + }, + { + "epoch": 1.8344645550527905, + "grad_norm": 0.37676605058429274, + "learning_rate": 6.386099741515144e-05, + "loss": 0.4419, + "step": 4865 + }, + { + "epoch": 1.8348416289592762, + "grad_norm": 0.3583244295489848, + "learning_rate": 6.385254535629209e-05, + "loss": 0.4382, + "step": 4866 + }, + { + "epoch": 1.8352187028657618, + "grad_norm": 0.3055962203757337, + "learning_rate": 6.38440916444725e-05, + "loss": 0.4365, + "step": 4867 + }, + { + "epoch": 1.8355957767722475, + "grad_norm": 0.2616581417417523, + "learning_rate": 6.383563628027851e-05, + "loss": 0.4248, + "step": 4868 + }, + { + "epoch": 1.8359728506787332, + "grad_norm": 0.28287785982394875, + "learning_rate": 6.382717926429606e-05, + "loss": 0.4446, + "step": 4869 + }, + { + "epoch": 1.8363499245852188, + "grad_norm": 0.3147258322782245, + "learning_rate": 6.381872059711123e-05, + "loss": 0.4468, + "step": 4870 + }, + { + "epoch": 1.8367269984917045, + "grad_norm": 0.29645865369557123, + "learning_rate": 6.381026027931018e-05, + "loss": 0.4485, + "step": 4871 + }, + { + "epoch": 1.8371040723981902, + "grad_norm": 0.2373456191514591, + "learning_rate": 6.380179831147923e-05, + "loss": 0.4374, + "step": 4872 + }, + { + "epoch": 1.8374811463046758, + "grad_norm": 0.20551606429166147, + "learning_rate": 6.379333469420476e-05, + "loss": 0.4446, + "step": 4873 + }, + { + "epoch": 1.8378582202111615, + "grad_norm": 0.2651812437025001, + "learning_rate": 6.378486942807329e-05, + "loss": 0.4446, + "step": 4874 + }, + { + "epoch": 1.8382352941176472, + "grad_norm": 0.33509479574389206, + "learning_rate": 6.377640251367148e-05, + "loss": 0.4452, + "step": 4875 + }, + { + "epoch": 1.8386123680241329, + "grad_norm": 0.2867718467134974, + "learning_rate": 6.376793395158606e-05, + "loss": 0.4316, + "step": 4876 + }, + { + "epoch": 1.8389894419306185, + "grad_norm": 0.23307077665334808, + "learning_rate": 6.37594637424039e-05, + "loss": 0.4453, + "step": 4877 + }, + { + "epoch": 1.8393665158371042, + "grad_norm": 0.25340716673411545, + "learning_rate": 6.375099188671198e-05, + "loss": 0.4458, + "step": 4878 + }, + { + "epoch": 1.8397435897435899, + "grad_norm": 0.33337773298027573, + "learning_rate": 6.374251838509738e-05, + "loss": 0.4387, + "step": 4879 + }, + { + "epoch": 1.8401206636500755, + "grad_norm": 0.3003094721788114, + "learning_rate": 6.373404323814733e-05, + "loss": 0.4271, + "step": 4880 + }, + { + "epoch": 1.8404977375565612, + "grad_norm": 0.23628002658677405, + "learning_rate": 6.372556644644913e-05, + "loss": 0.4442, + "step": 4881 + }, + { + "epoch": 1.8408748114630469, + "grad_norm": 0.3051258504446109, + "learning_rate": 6.371708801059022e-05, + "loss": 0.4264, + "step": 4882 + }, + { + "epoch": 1.8412518853695325, + "grad_norm": 0.34955702897050156, + "learning_rate": 6.370860793115814e-05, + "loss": 0.4393, + "step": 4883 + }, + { + "epoch": 1.8416289592760182, + "grad_norm": 0.3377011415715323, + "learning_rate": 6.370012620874056e-05, + "loss": 0.4417, + "step": 4884 + }, + { + "epoch": 1.8420060331825039, + "grad_norm": 0.29242327034747895, + "learning_rate": 6.369164284392528e-05, + "loss": 0.4324, + "step": 4885 + }, + { + "epoch": 1.8423831070889896, + "grad_norm": 0.2910649974635927, + "learning_rate": 6.368315783730013e-05, + "loss": 0.4416, + "step": 4886 + }, + { + "epoch": 1.8427601809954752, + "grad_norm": 0.3409112415726723, + "learning_rate": 6.367467118945318e-05, + "loss": 0.4326, + "step": 4887 + }, + { + "epoch": 1.843137254901961, + "grad_norm": 0.40148334674122843, + "learning_rate": 6.36661829009725e-05, + "loss": 0.4532, + "step": 4888 + }, + { + "epoch": 1.8435143288084466, + "grad_norm": 0.3851234624921488, + "learning_rate": 6.365769297244633e-05, + "loss": 0.4323, + "step": 4889 + }, + { + "epoch": 1.8438914027149322, + "grad_norm": 0.412388014470688, + "learning_rate": 6.364920140446304e-05, + "loss": 0.4437, + "step": 4890 + }, + { + "epoch": 1.844268476621418, + "grad_norm": 0.4332556738665668, + "learning_rate": 6.364070819761105e-05, + "loss": 0.4299, + "step": 4891 + }, + { + "epoch": 1.8446455505279036, + "grad_norm": 0.40537792183985144, + "learning_rate": 6.363221335247897e-05, + "loss": 0.4339, + "step": 4892 + }, + { + "epoch": 1.8450226244343892, + "grad_norm": 0.3816723565666037, + "learning_rate": 6.362371686965544e-05, + "loss": 0.4359, + "step": 4893 + }, + { + "epoch": 1.845399698340875, + "grad_norm": 0.3741235697473625, + "learning_rate": 6.36152187497293e-05, + "loss": 0.443, + "step": 4894 + }, + { + "epoch": 1.8457767722473606, + "grad_norm": 0.35077916147057986, + "learning_rate": 6.360671899328947e-05, + "loss": 0.4352, + "step": 4895 + }, + { + "epoch": 1.8461538461538463, + "grad_norm": 0.2362554707731588, + "learning_rate": 6.359821760092493e-05, + "loss": 0.4414, + "step": 4896 + }, + { + "epoch": 1.846530920060332, + "grad_norm": 0.19598567662341648, + "learning_rate": 6.358971457322483e-05, + "loss": 0.44, + "step": 4897 + }, + { + "epoch": 1.8469079939668176, + "grad_norm": 0.2903108335582799, + "learning_rate": 6.358120991077846e-05, + "loss": 0.4333, + "step": 4898 + }, + { + "epoch": 1.8472850678733033, + "grad_norm": 0.2990772337731361, + "learning_rate": 6.357270361417515e-05, + "loss": 0.4451, + "step": 4899 + }, + { + "epoch": 1.847662141779789, + "grad_norm": 0.3239782899821691, + "learning_rate": 6.356419568400438e-05, + "loss": 0.4251, + "step": 4900 + }, + { + "epoch": 1.8480392156862746, + "grad_norm": 0.3551535185961457, + "learning_rate": 6.355568612085575e-05, + "loss": 0.4267, + "step": 4901 + }, + { + "epoch": 1.8484162895927603, + "grad_norm": 0.37929579243157047, + "learning_rate": 6.354717492531897e-05, + "loss": 0.4378, + "step": 4902 + }, + { + "epoch": 1.848793363499246, + "grad_norm": 0.3470670395285525, + "learning_rate": 6.353866209798386e-05, + "loss": 0.4534, + "step": 4903 + }, + { + "epoch": 1.8491704374057316, + "grad_norm": 0.30693211770023054, + "learning_rate": 6.353014763944034e-05, + "loss": 0.4254, + "step": 4904 + }, + { + "epoch": 1.8495475113122173, + "grad_norm": 0.2887939090571726, + "learning_rate": 6.352163155027847e-05, + "loss": 0.4354, + "step": 4905 + }, + { + "epoch": 1.849924585218703, + "grad_norm": 0.27753288396335063, + "learning_rate": 6.351311383108838e-05, + "loss": 0.4403, + "step": 4906 + }, + { + "epoch": 1.8503016591251886, + "grad_norm": 0.2862482041926623, + "learning_rate": 6.350459448246038e-05, + "loss": 0.4421, + "step": 4907 + }, + { + "epoch": 1.8506787330316743, + "grad_norm": 0.3280334496169336, + "learning_rate": 6.349607350498482e-05, + "loss": 0.4375, + "step": 4908 + }, + { + "epoch": 1.85105580693816, + "grad_norm": 0.3955947550041615, + "learning_rate": 6.34875508992522e-05, + "loss": 0.4418, + "step": 4909 + }, + { + "epoch": 1.8514328808446456, + "grad_norm": 0.39786160271969995, + "learning_rate": 6.347902666585313e-05, + "loss": 0.4361, + "step": 4910 + }, + { + "epoch": 1.8518099547511313, + "grad_norm": 0.42544754310936417, + "learning_rate": 6.347050080537836e-05, + "loss": 0.4339, + "step": 4911 + }, + { + "epoch": 1.852187028657617, + "grad_norm": 0.4561924860979165, + "learning_rate": 6.346197331841868e-05, + "loss": 0.4241, + "step": 4912 + }, + { + "epoch": 1.8525641025641026, + "grad_norm": 0.43294401217688616, + "learning_rate": 6.345344420556508e-05, + "loss": 0.4407, + "step": 4913 + }, + { + "epoch": 1.8529411764705883, + "grad_norm": 0.3490464147551089, + "learning_rate": 6.344491346740859e-05, + "loss": 0.4337, + "step": 4914 + }, + { + "epoch": 1.853318250377074, + "grad_norm": 0.31449181892165196, + "learning_rate": 6.34363811045404e-05, + "loss": 0.4154, + "step": 4915 + }, + { + "epoch": 1.8536953242835597, + "grad_norm": 0.32247380211327525, + "learning_rate": 6.342784711755178e-05, + "loss": 0.4439, + "step": 4916 + }, + { + "epoch": 1.8540723981900453, + "grad_norm": 0.3344647905975875, + "learning_rate": 6.341931150703414e-05, + "loss": 0.4349, + "step": 4917 + }, + { + "epoch": 1.854449472096531, + "grad_norm": 0.3279393452632193, + "learning_rate": 6.341077427357897e-05, + "loss": 0.4278, + "step": 4918 + }, + { + "epoch": 1.8548265460030167, + "grad_norm": 0.2672447067746726, + "learning_rate": 6.340223541777792e-05, + "loss": 0.4357, + "step": 4919 + }, + { + "epoch": 1.8552036199095023, + "grad_norm": 0.22582024479278887, + "learning_rate": 6.339369494022273e-05, + "loss": 0.4349, + "step": 4920 + }, + { + "epoch": 1.855580693815988, + "grad_norm": 0.25658749066286995, + "learning_rate": 6.33851528415052e-05, + "loss": 0.4412, + "step": 4921 + }, + { + "epoch": 1.8559577677224737, + "grad_norm": 0.24868970211038854, + "learning_rate": 6.337660912221733e-05, + "loss": 0.4405, + "step": 4922 + }, + { + "epoch": 1.8563348416289593, + "grad_norm": 0.25200948987230737, + "learning_rate": 6.336806378295119e-05, + "loss": 0.4342, + "step": 4923 + }, + { + "epoch": 1.856711915535445, + "grad_norm": 0.23920101628261828, + "learning_rate": 6.335951682429896e-05, + "loss": 0.431, + "step": 4924 + }, + { + "epoch": 1.8570889894419307, + "grad_norm": 0.23178683177047898, + "learning_rate": 6.335096824685292e-05, + "loss": 0.4473, + "step": 4925 + }, + { + "epoch": 1.8574660633484164, + "grad_norm": 0.30237153035675873, + "learning_rate": 6.33424180512055e-05, + "loss": 0.4394, + "step": 4926 + }, + { + "epoch": 1.857843137254902, + "grad_norm": 0.3666872804326659, + "learning_rate": 6.333386623794922e-05, + "loss": 0.4366, + "step": 4927 + }, + { + "epoch": 1.8582202111613877, + "grad_norm": 0.40111920927677325, + "learning_rate": 6.33253128076767e-05, + "loss": 0.427, + "step": 4928 + }, + { + "epoch": 1.8585972850678734, + "grad_norm": 0.39849057925768155, + "learning_rate": 6.33167577609807e-05, + "loss": 0.419, + "step": 4929 + }, + { + "epoch": 1.858974358974359, + "grad_norm": 0.325942887236116, + "learning_rate": 6.330820109845405e-05, + "loss": 0.4238, + "step": 4930 + }, + { + "epoch": 1.8593514328808447, + "grad_norm": 0.38539855192367933, + "learning_rate": 6.329964282068975e-05, + "loss": 0.4434, + "step": 4931 + }, + { + "epoch": 1.8597285067873304, + "grad_norm": 0.4545248243631932, + "learning_rate": 6.329108292828089e-05, + "loss": 0.4285, + "step": 4932 + }, + { + "epoch": 1.860105580693816, + "grad_norm": 0.39309796879359155, + "learning_rate": 6.328252142182062e-05, + "loss": 0.4464, + "step": 4933 + }, + { + "epoch": 1.8604826546003017, + "grad_norm": 0.28829498921145524, + "learning_rate": 6.327395830190226e-05, + "loss": 0.4427, + "step": 4934 + }, + { + "epoch": 1.8608597285067874, + "grad_norm": 0.2440426641979394, + "learning_rate": 6.326539356911926e-05, + "loss": 0.4343, + "step": 4935 + }, + { + "epoch": 1.861236802413273, + "grad_norm": 0.35039421008864446, + "learning_rate": 6.325682722406509e-05, + "loss": 0.4259, + "step": 4936 + }, + { + "epoch": 1.8616138763197587, + "grad_norm": 0.3773999733387703, + "learning_rate": 6.324825926733343e-05, + "loss": 0.4348, + "step": 4937 + }, + { + "epoch": 1.8619909502262444, + "grad_norm": 0.32613835053728746, + "learning_rate": 6.323968969951805e-05, + "loss": 0.4305, + "step": 4938 + }, + { + "epoch": 1.86236802413273, + "grad_norm": 0.346684138296319, + "learning_rate": 6.323111852121276e-05, + "loss": 0.433, + "step": 4939 + }, + { + "epoch": 1.8627450980392157, + "grad_norm": 0.3704577765714623, + "learning_rate": 6.322254573301157e-05, + "loss": 0.4364, + "step": 4940 + }, + { + "epoch": 1.8631221719457014, + "grad_norm": 0.39943683131109775, + "learning_rate": 6.321397133550857e-05, + "loss": 0.4323, + "step": 4941 + }, + { + "epoch": 1.863499245852187, + "grad_norm": 0.3699102989857939, + "learning_rate": 6.320539532929793e-05, + "loss": 0.4237, + "step": 4942 + }, + { + "epoch": 1.8638763197586727, + "grad_norm": 0.37131168332351344, + "learning_rate": 6.319681771497399e-05, + "loss": 0.4334, + "step": 4943 + }, + { + "epoch": 1.8642533936651584, + "grad_norm": 0.3436838673396283, + "learning_rate": 6.318823849313115e-05, + "loss": 0.4359, + "step": 4944 + }, + { + "epoch": 1.864630467571644, + "grad_norm": 0.22577950145860598, + "learning_rate": 6.317965766436395e-05, + "loss": 0.4424, + "step": 4945 + }, + { + "epoch": 1.8650075414781297, + "grad_norm": 0.29488722354959884, + "learning_rate": 6.317107522926704e-05, + "loss": 0.4469, + "step": 4946 + }, + { + "epoch": 1.8653846153846154, + "grad_norm": 0.3270977129521684, + "learning_rate": 6.316249118843516e-05, + "loss": 0.4329, + "step": 4947 + }, + { + "epoch": 1.865761689291101, + "grad_norm": 0.31944483868363016, + "learning_rate": 6.31539055424632e-05, + "loss": 0.4332, + "step": 4948 + }, + { + "epoch": 1.8661387631975868, + "grad_norm": 0.33294139868599165, + "learning_rate": 6.314531829194611e-05, + "loss": 0.4319, + "step": 4949 + }, + { + "epoch": 1.8665158371040724, + "grad_norm": 0.35295408611353063, + "learning_rate": 6.3136729437479e-05, + "loss": 0.436, + "step": 4950 + }, + { + "epoch": 1.866892911010558, + "grad_norm": 0.35695709464973213, + "learning_rate": 6.312813897965706e-05, + "loss": 0.4384, + "step": 4951 + }, + { + "epoch": 1.8672699849170438, + "grad_norm": 0.37839281740873953, + "learning_rate": 6.31195469190756e-05, + "loss": 0.4394, + "step": 4952 + }, + { + "epoch": 1.8676470588235294, + "grad_norm": 0.41129456243134027, + "learning_rate": 6.311095325633006e-05, + "loss": 0.4356, + "step": 4953 + }, + { + "epoch": 1.868024132730015, + "grad_norm": 0.4211739729400994, + "learning_rate": 6.310235799201593e-05, + "loss": 0.4333, + "step": 4954 + }, + { + "epoch": 1.8684012066365008, + "grad_norm": 0.4785379249379429, + "learning_rate": 6.30937611267289e-05, + "loss": 0.4392, + "step": 4955 + }, + { + "epoch": 1.8687782805429864, + "grad_norm": 0.5705911056817825, + "learning_rate": 6.308516266106472e-05, + "loss": 0.4365, + "step": 4956 + }, + { + "epoch": 1.8691553544494721, + "grad_norm": 0.6364709131592935, + "learning_rate": 6.307656259561922e-05, + "loss": 0.4329, + "step": 4957 + }, + { + "epoch": 1.8695324283559578, + "grad_norm": 0.5899208567954177, + "learning_rate": 6.306796093098842e-05, + "loss": 0.4339, + "step": 4958 + }, + { + "epoch": 1.8699095022624435, + "grad_norm": 0.5231315865099788, + "learning_rate": 6.305935766776839e-05, + "loss": 0.4358, + "step": 4959 + }, + { + "epoch": 1.8702865761689291, + "grad_norm": 0.43430759118691314, + "learning_rate": 6.305075280655532e-05, + "loss": 0.43, + "step": 4960 + }, + { + "epoch": 1.8706636500754148, + "grad_norm": 0.30174027089528055, + "learning_rate": 6.304214634794552e-05, + "loss": 0.4354, + "step": 4961 + }, + { + "epoch": 1.8710407239819005, + "grad_norm": 0.2396315007434575, + "learning_rate": 6.303353829253543e-05, + "loss": 0.4298, + "step": 4962 + }, + { + "epoch": 1.8714177978883861, + "grad_norm": 0.35207155172594556, + "learning_rate": 6.302492864092155e-05, + "loss": 0.4394, + "step": 4963 + }, + { + "epoch": 1.8717948717948718, + "grad_norm": 0.45181716231380925, + "learning_rate": 6.301631739370055e-05, + "loss": 0.4378, + "step": 4964 + }, + { + "epoch": 1.8721719457013575, + "grad_norm": 0.5033424346502408, + "learning_rate": 6.300770455146916e-05, + "loss": 0.4413, + "step": 4965 + }, + { + "epoch": 1.8725490196078431, + "grad_norm": 0.5006198842131395, + "learning_rate": 6.299909011482428e-05, + "loss": 0.429, + "step": 4966 + }, + { + "epoch": 1.8729260935143288, + "grad_norm": 0.439575975278642, + "learning_rate": 6.299047408436282e-05, + "loss": 0.4256, + "step": 4967 + }, + { + "epoch": 1.8733031674208145, + "grad_norm": 0.3340052163938844, + "learning_rate": 6.298185646068191e-05, + "loss": 0.4234, + "step": 4968 + }, + { + "epoch": 1.8736802413273002, + "grad_norm": 0.23633709972869654, + "learning_rate": 6.297323724437875e-05, + "loss": 0.4364, + "step": 4969 + }, + { + "epoch": 1.8740573152337858, + "grad_norm": 0.23251640760569683, + "learning_rate": 6.29646164360506e-05, + "loss": 0.4337, + "step": 4970 + }, + { + "epoch": 1.8744343891402715, + "grad_norm": 0.27290049913321407, + "learning_rate": 6.295599403629492e-05, + "loss": 0.4464, + "step": 4971 + }, + { + "epoch": 1.8748114630467572, + "grad_norm": 0.3174845985756316, + "learning_rate": 6.294737004570921e-05, + "loss": 0.426, + "step": 4972 + }, + { + "epoch": 1.8751885369532428, + "grad_norm": 0.3871807740724507, + "learning_rate": 6.29387444648911e-05, + "loss": 0.4272, + "step": 4973 + }, + { + "epoch": 1.8755656108597285, + "grad_norm": 0.40080404806401265, + "learning_rate": 6.293011729443836e-05, + "loss": 0.4334, + "step": 4974 + }, + { + "epoch": 1.8759426847662142, + "grad_norm": 0.3922087678393934, + "learning_rate": 6.292148853494883e-05, + "loss": 0.4323, + "step": 4975 + }, + { + "epoch": 1.8763197586726998, + "grad_norm": 0.46733263301670946, + "learning_rate": 6.291285818702047e-05, + "loss": 0.4322, + "step": 4976 + }, + { + "epoch": 1.8766968325791855, + "grad_norm": 0.45859223162000157, + "learning_rate": 6.290422625125136e-05, + "loss": 0.4259, + "step": 4977 + }, + { + "epoch": 1.8770739064856712, + "grad_norm": 0.4050924953067819, + "learning_rate": 6.289559272823969e-05, + "loss": 0.4283, + "step": 4978 + }, + { + "epoch": 1.8774509803921569, + "grad_norm": 0.3610729078765266, + "learning_rate": 6.288695761858375e-05, + "loss": 0.4358, + "step": 4979 + }, + { + "epoch": 1.8778280542986425, + "grad_norm": 0.33388456083803664, + "learning_rate": 6.287832092288194e-05, + "loss": 0.4404, + "step": 4980 + }, + { + "epoch": 1.8782051282051282, + "grad_norm": 0.27344869231771934, + "learning_rate": 6.28696826417328e-05, + "loss": 0.4294, + "step": 4981 + }, + { + "epoch": 1.8785822021116139, + "grad_norm": 0.21966457120205027, + "learning_rate": 6.286104277573492e-05, + "loss": 0.4283, + "step": 4982 + }, + { + "epoch": 1.8789592760180995, + "grad_norm": 0.29163867412958694, + "learning_rate": 6.285240132548707e-05, + "loss": 0.4267, + "step": 4983 + }, + { + "epoch": 1.8793363499245852, + "grad_norm": 0.39058088466470253, + "learning_rate": 6.284375829158807e-05, + "loss": 0.4502, + "step": 4984 + }, + { + "epoch": 1.8797134238310709, + "grad_norm": 0.38829115347094795, + "learning_rate": 6.283511367463687e-05, + "loss": 0.4251, + "step": 4985 + }, + { + "epoch": 1.8800904977375565, + "grad_norm": 0.35561029820771106, + "learning_rate": 6.282646747523257e-05, + "loss": 0.4242, + "step": 4986 + }, + { + "epoch": 1.8804675716440422, + "grad_norm": 0.3211156591396657, + "learning_rate": 6.281781969397431e-05, + "loss": 0.4464, + "step": 4987 + }, + { + "epoch": 1.8808446455505279, + "grad_norm": 0.3651493468126658, + "learning_rate": 6.28091703314614e-05, + "loss": 0.4386, + "step": 4988 + }, + { + "epoch": 1.8812217194570136, + "grad_norm": 0.3533876747973531, + "learning_rate": 6.280051938829319e-05, + "loss": 0.4358, + "step": 4989 + }, + { + "epoch": 1.8815987933634992, + "grad_norm": 0.3705071934253224, + "learning_rate": 6.279186686506923e-05, + "loss": 0.4442, + "step": 4990 + }, + { + "epoch": 1.881975867269985, + "grad_norm": 0.3903286777412402, + "learning_rate": 6.27832127623891e-05, + "loss": 0.4314, + "step": 4991 + }, + { + "epoch": 1.8823529411764706, + "grad_norm": 0.3847218369357505, + "learning_rate": 6.277455708085255e-05, + "loss": 0.4412, + "step": 4992 + }, + { + "epoch": 1.8827300150829562, + "grad_norm": 0.3520255644287503, + "learning_rate": 6.276589982105939e-05, + "loss": 0.4418, + "step": 4993 + }, + { + "epoch": 1.883107088989442, + "grad_norm": 0.3136778770743599, + "learning_rate": 6.275724098360955e-05, + "loss": 0.4257, + "step": 4994 + }, + { + "epoch": 1.8834841628959276, + "grad_norm": 0.28383640745162375, + "learning_rate": 6.274858056910312e-05, + "loss": 0.4304, + "step": 4995 + }, + { + "epoch": 1.8838612368024132, + "grad_norm": 0.22751139341663237, + "learning_rate": 6.273991857814021e-05, + "loss": 0.4391, + "step": 4996 + }, + { + "epoch": 1.884238310708899, + "grad_norm": 0.2449086469972493, + "learning_rate": 6.273125501132113e-05, + "loss": 0.4362, + "step": 4997 + }, + { + "epoch": 1.8846153846153846, + "grad_norm": 0.26194974218957595, + "learning_rate": 6.272258986924624e-05, + "loss": 0.4491, + "step": 4998 + }, + { + "epoch": 1.8849924585218703, + "grad_norm": 0.3615236668690587, + "learning_rate": 6.271392315251601e-05, + "loss": 0.4546, + "step": 4999 + }, + { + "epoch": 1.885369532428356, + "grad_norm": 0.40251960796973674, + "learning_rate": 6.270525486173107e-05, + "loss": 0.4411, + "step": 5000 + }, + { + "epoch": 1.8857466063348416, + "grad_norm": 0.3843908573143503, + "learning_rate": 6.269658499749211e-05, + "loss": 0.4337, + "step": 5001 + }, + { + "epoch": 1.8861236802413273, + "grad_norm": 0.3521222832202886, + "learning_rate": 6.268791356039994e-05, + "loss": 0.4348, + "step": 5002 + }, + { + "epoch": 1.886500754147813, + "grad_norm": 0.34576341713300995, + "learning_rate": 6.267924055105548e-05, + "loss": 0.4418, + "step": 5003 + }, + { + "epoch": 1.8868778280542986, + "grad_norm": 0.3205894806944214, + "learning_rate": 6.267056597005977e-05, + "loss": 0.444, + "step": 5004 + }, + { + "epoch": 1.8872549019607843, + "grad_norm": 0.24140244824487475, + "learning_rate": 6.266188981801394e-05, + "loss": 0.4256, + "step": 5005 + }, + { + "epoch": 1.88763197586727, + "grad_norm": 0.28802822089440966, + "learning_rate": 6.265321209551925e-05, + "loss": 0.4375, + "step": 5006 + }, + { + "epoch": 1.8880090497737556, + "grad_norm": 0.35316191455617724, + "learning_rate": 6.264453280317704e-05, + "loss": 0.4333, + "step": 5007 + }, + { + "epoch": 1.8883861236802413, + "grad_norm": 0.39610614591050824, + "learning_rate": 6.263585194158881e-05, + "loss": 0.4232, + "step": 5008 + }, + { + "epoch": 1.888763197586727, + "grad_norm": 0.3744128253112282, + "learning_rate": 6.262716951135612e-05, + "loss": 0.4359, + "step": 5009 + }, + { + "epoch": 1.8891402714932126, + "grad_norm": 0.31955743343725995, + "learning_rate": 6.261848551308064e-05, + "loss": 0.434, + "step": 5010 + }, + { + "epoch": 1.8895173453996983, + "grad_norm": 0.2794866543045146, + "learning_rate": 6.260979994736419e-05, + "loss": 0.4429, + "step": 5011 + }, + { + "epoch": 1.889894419306184, + "grad_norm": 0.24465320266693882, + "learning_rate": 6.260111281480865e-05, + "loss": 0.4221, + "step": 5012 + }, + { + "epoch": 1.8902714932126696, + "grad_norm": 0.22976152270101574, + "learning_rate": 6.259242411601604e-05, + "loss": 0.4315, + "step": 5013 + }, + { + "epoch": 1.8906485671191553, + "grad_norm": 0.263096848261728, + "learning_rate": 6.258373385158849e-05, + "loss": 0.4354, + "step": 5014 + }, + { + "epoch": 1.891025641025641, + "grad_norm": 0.2885682061772431, + "learning_rate": 6.25750420221282e-05, + "loss": 0.4311, + "step": 5015 + }, + { + "epoch": 1.8914027149321266, + "grad_norm": 0.2929715671917859, + "learning_rate": 6.256634862823752e-05, + "loss": 0.4353, + "step": 5016 + }, + { + "epoch": 1.8917797888386123, + "grad_norm": 0.3091325208953015, + "learning_rate": 6.255765367051892e-05, + "loss": 0.4471, + "step": 5017 + }, + { + "epoch": 1.892156862745098, + "grad_norm": 0.306309637352922, + "learning_rate": 6.25489571495749e-05, + "loss": 0.4314, + "step": 5018 + }, + { + "epoch": 1.8925339366515836, + "grad_norm": 0.32671610616063407, + "learning_rate": 6.254025906600817e-05, + "loss": 0.4366, + "step": 5019 + }, + { + "epoch": 1.8929110105580693, + "grad_norm": 0.33152005518395045, + "learning_rate": 6.253155942042147e-05, + "loss": 0.4299, + "step": 5020 + }, + { + "epoch": 1.893288084464555, + "grad_norm": 0.2916045279552143, + "learning_rate": 6.25228582134177e-05, + "loss": 0.4413, + "step": 5021 + }, + { + "epoch": 1.8936651583710407, + "grad_norm": 0.27666031627814197, + "learning_rate": 6.251415544559982e-05, + "loss": 0.4436, + "step": 5022 + }, + { + "epoch": 1.8940422322775263, + "grad_norm": 0.2663171934616717, + "learning_rate": 6.250545111757095e-05, + "loss": 0.4399, + "step": 5023 + }, + { + "epoch": 1.894419306184012, + "grad_norm": 0.22312336929174326, + "learning_rate": 6.249674522993427e-05, + "loss": 0.4404, + "step": 5024 + }, + { + "epoch": 1.8947963800904977, + "grad_norm": 0.2373655932266391, + "learning_rate": 6.24880377832931e-05, + "loss": 0.4429, + "step": 5025 + }, + { + "epoch": 1.8951734539969833, + "grad_norm": 0.2747960154540673, + "learning_rate": 6.247932877825086e-05, + "loss": 0.4288, + "step": 5026 + }, + { + "epoch": 1.895550527903469, + "grad_norm": 0.26920834702434343, + "learning_rate": 6.247061821541108e-05, + "loss": 0.4336, + "step": 5027 + }, + { + "epoch": 1.8959276018099547, + "grad_norm": 0.3106704272030589, + "learning_rate": 6.246190609537738e-05, + "loss": 0.4295, + "step": 5028 + }, + { + "epoch": 1.8963046757164403, + "grad_norm": 0.3621851301602543, + "learning_rate": 6.245319241875353e-05, + "loss": 0.4298, + "step": 5029 + }, + { + "epoch": 1.896681749622926, + "grad_norm": 0.38616775541822196, + "learning_rate": 6.244447718614334e-05, + "loss": 0.4406, + "step": 5030 + }, + { + "epoch": 1.8970588235294117, + "grad_norm": 0.35245142965966253, + "learning_rate": 6.24357603981508e-05, + "loss": 0.4462, + "step": 5031 + }, + { + "epoch": 1.8974358974358974, + "grad_norm": 0.32454224243679647, + "learning_rate": 6.242704205537996e-05, + "loss": 0.4416, + "step": 5032 + }, + { + "epoch": 1.897812971342383, + "grad_norm": 0.3699309847439731, + "learning_rate": 6.2418322158435e-05, + "loss": 0.4352, + "step": 5033 + }, + { + "epoch": 1.8981900452488687, + "grad_norm": 0.435700295118881, + "learning_rate": 6.24096007079202e-05, + "loss": 0.4304, + "step": 5034 + }, + { + "epoch": 1.8985671191553544, + "grad_norm": 0.3776430431548297, + "learning_rate": 6.240087770443994e-05, + "loss": 0.4303, + "step": 5035 + }, + { + "epoch": 1.89894419306184, + "grad_norm": 0.27420214997694264, + "learning_rate": 6.239215314859874e-05, + "loss": 0.4438, + "step": 5036 + }, + { + "epoch": 1.8993212669683257, + "grad_norm": 0.23175868709249323, + "learning_rate": 6.238342704100118e-05, + "loss": 0.4268, + "step": 5037 + }, + { + "epoch": 1.8996983408748114, + "grad_norm": 0.2373396567025238, + "learning_rate": 6.237469938225198e-05, + "loss": 0.4221, + "step": 5038 + }, + { + "epoch": 1.900075414781297, + "grad_norm": 0.26437752532451586, + "learning_rate": 6.236597017295596e-05, + "loss": 0.4413, + "step": 5039 + }, + { + "epoch": 1.9004524886877827, + "grad_norm": 0.2730225842901401, + "learning_rate": 6.235723941371804e-05, + "loss": 0.4298, + "step": 5040 + }, + { + "epoch": 1.9008295625942684, + "grad_norm": 0.2778595202472365, + "learning_rate": 6.234850710514327e-05, + "loss": 0.4414, + "step": 5041 + }, + { + "epoch": 1.901206636500754, + "grad_norm": 0.2756521423575634, + "learning_rate": 6.233977324783679e-05, + "loss": 0.4329, + "step": 5042 + }, + { + "epoch": 1.9015837104072397, + "grad_norm": 0.23276525332355535, + "learning_rate": 6.233103784240381e-05, + "loss": 0.4355, + "step": 5043 + }, + { + "epoch": 1.9019607843137254, + "grad_norm": 0.23627692566942293, + "learning_rate": 6.232230088944974e-05, + "loss": 0.4216, + "step": 5044 + }, + { + "epoch": 1.902337858220211, + "grad_norm": 0.262698878788836, + "learning_rate": 6.231356238958e-05, + "loss": 0.4295, + "step": 5045 + }, + { + "epoch": 1.9027149321266967, + "grad_norm": 0.25628735089342203, + "learning_rate": 6.230482234340021e-05, + "loss": 0.4345, + "step": 5046 + }, + { + "epoch": 1.9030920060331824, + "grad_norm": 0.24548467209247193, + "learning_rate": 6.229608075151598e-05, + "loss": 0.4376, + "step": 5047 + }, + { + "epoch": 1.903469079939668, + "grad_norm": 0.23038914671678212, + "learning_rate": 6.228733761453315e-05, + "loss": 0.4282, + "step": 5048 + }, + { + "epoch": 1.9038461538461537, + "grad_norm": 0.2847988232238122, + "learning_rate": 6.227859293305758e-05, + "loss": 0.4318, + "step": 5049 + }, + { + "epoch": 1.9042232277526394, + "grad_norm": 0.25871508364926765, + "learning_rate": 6.226984670769529e-05, + "loss": 0.4388, + "step": 5050 + }, + { + "epoch": 1.904600301659125, + "grad_norm": 0.23923200825616744, + "learning_rate": 6.226109893905238e-05, + "loss": 0.4251, + "step": 5051 + }, + { + "epoch": 1.9049773755656108, + "grad_norm": 0.26344703480789033, + "learning_rate": 6.225234962773505e-05, + "loss": 0.4415, + "step": 5052 + }, + { + "epoch": 1.9053544494720964, + "grad_norm": 0.263519909662909, + "learning_rate": 6.224359877434962e-05, + "loss": 0.4303, + "step": 5053 + }, + { + "epoch": 1.905731523378582, + "grad_norm": 0.26223412828676224, + "learning_rate": 6.223484637950254e-05, + "loss": 0.4309, + "step": 5054 + }, + { + "epoch": 1.9061085972850678, + "grad_norm": 0.3522886786698316, + "learning_rate": 6.222609244380032e-05, + "loss": 0.4356, + "step": 5055 + }, + { + "epoch": 1.9064856711915534, + "grad_norm": 0.37527797039544664, + "learning_rate": 6.221733696784961e-05, + "loss": 0.4247, + "step": 5056 + }, + { + "epoch": 1.906862745098039, + "grad_norm": 0.3623585131625892, + "learning_rate": 6.220857995225717e-05, + "loss": 0.436, + "step": 5057 + }, + { + "epoch": 1.9072398190045248, + "grad_norm": 0.34019855674383215, + "learning_rate": 6.219982139762982e-05, + "loss": 0.4404, + "step": 5058 + }, + { + "epoch": 1.9076168929110104, + "grad_norm": 0.28715926861216684, + "learning_rate": 6.219106130457454e-05, + "loss": 0.4418, + "step": 5059 + }, + { + "epoch": 1.9079939668174961, + "grad_norm": 0.2569897250148576, + "learning_rate": 6.218229967369841e-05, + "loss": 0.4373, + "step": 5060 + }, + { + "epoch": 1.9083710407239818, + "grad_norm": 0.22982022252717857, + "learning_rate": 6.217353650560858e-05, + "loss": 0.4322, + "step": 5061 + }, + { + "epoch": 1.9087481146304675, + "grad_norm": 0.2339194960781127, + "learning_rate": 6.216477180091232e-05, + "loss": 0.4296, + "step": 5062 + }, + { + "epoch": 1.9091251885369531, + "grad_norm": 0.27490649741473816, + "learning_rate": 6.215600556021706e-05, + "loss": 0.4356, + "step": 5063 + }, + { + "epoch": 1.9095022624434388, + "grad_norm": 0.28234650680681594, + "learning_rate": 6.214723778413025e-05, + "loss": 0.4264, + "step": 5064 + }, + { + "epoch": 1.9098793363499245, + "grad_norm": 0.27173910495116343, + "learning_rate": 6.213846847325953e-05, + "loss": 0.4393, + "step": 5065 + }, + { + "epoch": 1.9102564102564101, + "grad_norm": 0.26695724657148673, + "learning_rate": 6.212969762821256e-05, + "loss": 0.435, + "step": 5066 + }, + { + "epoch": 1.9106334841628958, + "grad_norm": 0.2793597132279423, + "learning_rate": 6.212092524959718e-05, + "loss": 0.4426, + "step": 5067 + }, + { + "epoch": 1.9110105580693815, + "grad_norm": 0.29200347979949876, + "learning_rate": 6.211215133802129e-05, + "loss": 0.429, + "step": 5068 + }, + { + "epoch": 1.9113876319758671, + "grad_norm": 0.2579712924911403, + "learning_rate": 6.210337589409295e-05, + "loss": 0.4314, + "step": 5069 + }, + { + "epoch": 1.9117647058823528, + "grad_norm": 0.2562073653998086, + "learning_rate": 6.209459891842023e-05, + "loss": 0.4412, + "step": 5070 + }, + { + "epoch": 1.9121417797888385, + "grad_norm": 0.3040609748293148, + "learning_rate": 6.208582041161143e-05, + "loss": 0.4294, + "step": 5071 + }, + { + "epoch": 1.9125188536953242, + "grad_norm": 0.34814760731508543, + "learning_rate": 6.207704037427487e-05, + "loss": 0.4372, + "step": 5072 + }, + { + "epoch": 1.9128959276018098, + "grad_norm": 0.30275551043691856, + "learning_rate": 6.206825880701898e-05, + "loss": 0.442, + "step": 5073 + }, + { + "epoch": 1.9132730015082955, + "grad_norm": 0.28696121933489743, + "learning_rate": 6.205947571045234e-05, + "loss": 0.4412, + "step": 5074 + }, + { + "epoch": 1.9136500754147812, + "grad_norm": 0.309505340884794, + "learning_rate": 6.205069108518359e-05, + "loss": 0.4345, + "step": 5075 + }, + { + "epoch": 1.9140271493212668, + "grad_norm": 0.3147634685312897, + "learning_rate": 6.20419049318215e-05, + "loss": 0.4409, + "step": 5076 + }, + { + "epoch": 1.9144042232277525, + "grad_norm": 0.34215596831423095, + "learning_rate": 6.203311725097497e-05, + "loss": 0.4361, + "step": 5077 + }, + { + "epoch": 1.9147812971342382, + "grad_norm": 0.3900651794525659, + "learning_rate": 6.202432804325295e-05, + "loss": 0.4243, + "step": 5078 + }, + { + "epoch": 1.9151583710407238, + "grad_norm": 0.41435223854442665, + "learning_rate": 6.201553730926453e-05, + "loss": 0.4323, + "step": 5079 + }, + { + "epoch": 1.9155354449472095, + "grad_norm": 0.4514355369436633, + "learning_rate": 6.200674504961889e-05, + "loss": 0.432, + "step": 5080 + }, + { + "epoch": 1.9159125188536952, + "grad_norm": 0.4928967237911428, + "learning_rate": 6.199795126492536e-05, + "loss": 0.4498, + "step": 5081 + }, + { + "epoch": 1.9162895927601808, + "grad_norm": 0.522657237471616, + "learning_rate": 6.19891559557933e-05, + "loss": 0.4231, + "step": 5082 + }, + { + "epoch": 1.9166666666666665, + "grad_norm": 0.45640317508796474, + "learning_rate": 6.198035912283225e-05, + "loss": 0.4413, + "step": 5083 + }, + { + "epoch": 1.9170437405731522, + "grad_norm": 0.39350210096988214, + "learning_rate": 6.197156076665182e-05, + "loss": 0.4339, + "step": 5084 + }, + { + "epoch": 1.9174208144796379, + "grad_norm": 0.33120339077840594, + "learning_rate": 6.19627608878617e-05, + "loss": 0.4332, + "step": 5085 + }, + { + "epoch": 1.9177978883861235, + "grad_norm": 0.23016128408769088, + "learning_rate": 6.195395948707171e-05, + "loss": 0.422, + "step": 5086 + }, + { + "epoch": 1.9181749622926092, + "grad_norm": 0.27631275795367727, + "learning_rate": 6.194515656489182e-05, + "loss": 0.4342, + "step": 5087 + }, + { + "epoch": 1.9185520361990949, + "grad_norm": 0.3588862854019354, + "learning_rate": 6.193635212193206e-05, + "loss": 0.4352, + "step": 5088 + }, + { + "epoch": 1.9189291101055805, + "grad_norm": 0.39468321178093463, + "learning_rate": 6.192754615880254e-05, + "loss": 0.4375, + "step": 5089 + }, + { + "epoch": 1.9193061840120662, + "grad_norm": 0.39240118964781595, + "learning_rate": 6.191873867611353e-05, + "loss": 0.4356, + "step": 5090 + }, + { + "epoch": 1.9196832579185519, + "grad_norm": 0.3994155646928074, + "learning_rate": 6.190992967447538e-05, + "loss": 0.4335, + "step": 5091 + }, + { + "epoch": 1.9200603318250375, + "grad_norm": 0.3773226652594256, + "learning_rate": 6.190111915449853e-05, + "loss": 0.4369, + "step": 5092 + }, + { + "epoch": 1.9204374057315232, + "grad_norm": 0.28396382787369034, + "learning_rate": 6.189230711679354e-05, + "loss": 0.4223, + "step": 5093 + }, + { + "epoch": 1.9208144796380089, + "grad_norm": 0.24392152456770524, + "learning_rate": 6.188349356197108e-05, + "loss": 0.4413, + "step": 5094 + }, + { + "epoch": 1.9211915535444946, + "grad_norm": 0.2761108451647001, + "learning_rate": 6.187467849064195e-05, + "loss": 0.4271, + "step": 5095 + }, + { + "epoch": 1.9215686274509802, + "grad_norm": 0.3835438184154282, + "learning_rate": 6.186586190341698e-05, + "loss": 0.427, + "step": 5096 + }, + { + "epoch": 1.9219457013574661, + "grad_norm": 0.38244354966643956, + "learning_rate": 6.18570438009072e-05, + "loss": 0.4324, + "step": 5097 + }, + { + "epoch": 1.9223227752639518, + "grad_norm": 0.3560542671062101, + "learning_rate": 6.184822418372365e-05, + "loss": 0.4339, + "step": 5098 + }, + { + "epoch": 1.9226998491704375, + "grad_norm": 0.3393440027020713, + "learning_rate": 6.183940305247754e-05, + "loss": 0.4228, + "step": 5099 + }, + { + "epoch": 1.9230769230769231, + "grad_norm": 0.3202720214578123, + "learning_rate": 6.183058040778018e-05, + "loss": 0.4374, + "step": 5100 + }, + { + "epoch": 1.9234539969834088, + "grad_norm": 0.30381235284913777, + "learning_rate": 6.182175625024297e-05, + "loss": 0.4375, + "step": 5101 + }, + { + "epoch": 1.9238310708898945, + "grad_norm": 0.2808345291574228, + "learning_rate": 6.181293058047739e-05, + "loss": 0.4436, + "step": 5102 + }, + { + "epoch": 1.9242081447963801, + "grad_norm": 0.31647731610027746, + "learning_rate": 6.180410339909509e-05, + "loss": 0.4322, + "step": 5103 + }, + { + "epoch": 1.9245852187028658, + "grad_norm": 0.3541281744488772, + "learning_rate": 6.179527470670775e-05, + "loss": 0.4449, + "step": 5104 + }, + { + "epoch": 1.9249622926093515, + "grad_norm": 0.36678128304378455, + "learning_rate": 6.178644450392722e-05, + "loss": 0.435, + "step": 5105 + }, + { + "epoch": 1.9253393665158371, + "grad_norm": 0.34592552787567166, + "learning_rate": 6.177761279136538e-05, + "loss": 0.44, + "step": 5106 + }, + { + "epoch": 1.9257164404223228, + "grad_norm": 0.31790576709240376, + "learning_rate": 6.176877956963433e-05, + "loss": 0.4288, + "step": 5107 + }, + { + "epoch": 1.9260935143288085, + "grad_norm": 0.31390496756640096, + "learning_rate": 6.175994483934613e-05, + "loss": 0.4255, + "step": 5108 + }, + { + "epoch": 1.9264705882352942, + "grad_norm": 0.3420481130552604, + "learning_rate": 6.175110860111307e-05, + "loss": 0.4321, + "step": 5109 + }, + { + "epoch": 1.9268476621417798, + "grad_norm": 0.39042273606118116, + "learning_rate": 6.174227085554748e-05, + "loss": 0.4355, + "step": 5110 + }, + { + "epoch": 1.9272247360482655, + "grad_norm": 0.4327319816113501, + "learning_rate": 6.17334316032618e-05, + "loss": 0.4436, + "step": 5111 + }, + { + "epoch": 1.9276018099547512, + "grad_norm": 0.4741924524182849, + "learning_rate": 6.172459084486859e-05, + "loss": 0.4355, + "step": 5112 + }, + { + "epoch": 1.9279788838612368, + "grad_norm": 0.4767261622964244, + "learning_rate": 6.17157485809805e-05, + "loss": 0.4319, + "step": 5113 + }, + { + "epoch": 1.9283559577677225, + "grad_norm": 0.4983186298102908, + "learning_rate": 6.170690481221029e-05, + "loss": 0.4302, + "step": 5114 + }, + { + "epoch": 1.9287330316742082, + "grad_norm": 0.4174599304193042, + "learning_rate": 6.169805953917084e-05, + "loss": 0.4382, + "step": 5115 + }, + { + "epoch": 1.9291101055806938, + "grad_norm": 0.29101363063960245, + "learning_rate": 6.16892127624751e-05, + "loss": 0.4341, + "step": 5116 + }, + { + "epoch": 1.9294871794871795, + "grad_norm": 0.26207731718121036, + "learning_rate": 6.168036448273617e-05, + "loss": 0.4384, + "step": 5117 + }, + { + "epoch": 1.9298642533936652, + "grad_norm": 0.29581996839642266, + "learning_rate": 6.16715147005672e-05, + "loss": 0.4317, + "step": 5118 + }, + { + "epoch": 1.9302413273001509, + "grad_norm": 0.3384443272920711, + "learning_rate": 6.166266341658148e-05, + "loss": 0.4365, + "step": 5119 + }, + { + "epoch": 1.9306184012066365, + "grad_norm": 0.4147909031592253, + "learning_rate": 6.165381063139242e-05, + "loss": 0.4317, + "step": 5120 + }, + { + "epoch": 1.9309954751131222, + "grad_norm": 0.4832251408757751, + "learning_rate": 6.164495634561347e-05, + "loss": 0.424, + "step": 5121 + }, + { + "epoch": 1.9313725490196079, + "grad_norm": 0.4325949649371779, + "learning_rate": 6.163610055985824e-05, + "loss": 0.4384, + "step": 5122 + }, + { + "epoch": 1.9317496229260935, + "grad_norm": 0.3855725253531961, + "learning_rate": 6.162724327474045e-05, + "loss": 0.4407, + "step": 5123 + }, + { + "epoch": 1.9321266968325792, + "grad_norm": 0.43021652060109306, + "learning_rate": 6.161838449087389e-05, + "loss": 0.4343, + "step": 5124 + }, + { + "epoch": 1.9325037707390649, + "grad_norm": 0.45779123454174936, + "learning_rate": 6.160952420887244e-05, + "loss": 0.4345, + "step": 5125 + }, + { + "epoch": 1.9328808446455505, + "grad_norm": 0.3755909981633619, + "learning_rate": 6.160066242935014e-05, + "loss": 0.4247, + "step": 5126 + }, + { + "epoch": 1.9332579185520362, + "grad_norm": 0.33792513263731855, + "learning_rate": 6.15917991529211e-05, + "loss": 0.4347, + "step": 5127 + }, + { + "epoch": 1.9336349924585219, + "grad_norm": 0.3116183496142885, + "learning_rate": 6.158293438019952e-05, + "loss": 0.4399, + "step": 5128 + }, + { + "epoch": 1.9340120663650076, + "grad_norm": 0.25681139922820206, + "learning_rate": 6.157406811179972e-05, + "loss": 0.435, + "step": 5129 + }, + { + "epoch": 1.9343891402714932, + "grad_norm": 0.2658925171539, + "learning_rate": 6.156520034833616e-05, + "loss": 0.4352, + "step": 5130 + }, + { + "epoch": 1.934766214177979, + "grad_norm": 0.26394800665677565, + "learning_rate": 6.155633109042333e-05, + "loss": 0.4263, + "step": 5131 + }, + { + "epoch": 1.9351432880844646, + "grad_norm": 0.2623668637258554, + "learning_rate": 6.154746033867588e-05, + "loss": 0.4471, + "step": 5132 + }, + { + "epoch": 1.9355203619909502, + "grad_norm": 0.2951312912794262, + "learning_rate": 6.153858809370854e-05, + "loss": 0.4328, + "step": 5133 + }, + { + "epoch": 1.935897435897436, + "grad_norm": 0.29676404448407295, + "learning_rate": 6.152971435613616e-05, + "loss": 0.4446, + "step": 5134 + }, + { + "epoch": 1.9362745098039216, + "grad_norm": 0.2741968730234263, + "learning_rate": 6.152083912657364e-05, + "loss": 0.4339, + "step": 5135 + }, + { + "epoch": 1.9366515837104072, + "grad_norm": 0.25650483291896314, + "learning_rate": 6.151196240563609e-05, + "loss": 0.4303, + "step": 5136 + }, + { + "epoch": 1.937028657616893, + "grad_norm": 0.21840623780554827, + "learning_rate": 6.150308419393863e-05, + "loss": 0.4344, + "step": 5137 + }, + { + "epoch": 1.9374057315233786, + "grad_norm": 0.253284512694045, + "learning_rate": 6.149420449209649e-05, + "loss": 0.4402, + "step": 5138 + }, + { + "epoch": 1.9377828054298643, + "grad_norm": 0.29967584848479967, + "learning_rate": 6.148532330072504e-05, + "loss": 0.4224, + "step": 5139 + }, + { + "epoch": 1.93815987933635, + "grad_norm": 0.31141731603378364, + "learning_rate": 6.147644062043976e-05, + "loss": 0.4436, + "step": 5140 + }, + { + "epoch": 1.9385369532428356, + "grad_norm": 0.34009520901268925, + "learning_rate": 6.146755645185618e-05, + "loss": 0.4383, + "step": 5141 + }, + { + "epoch": 1.9389140271493213, + "grad_norm": 0.31419408915629915, + "learning_rate": 6.145867079558999e-05, + "loss": 0.4199, + "step": 5142 + }, + { + "epoch": 1.939291101055807, + "grad_norm": 0.3088652044875387, + "learning_rate": 6.144978365225695e-05, + "loss": 0.4258, + "step": 5143 + }, + { + "epoch": 1.9396681749622926, + "grad_norm": 0.32451214302923653, + "learning_rate": 6.144089502247292e-05, + "loss": 0.43, + "step": 5144 + }, + { + "epoch": 1.9400452488687783, + "grad_norm": 0.30249263296983403, + "learning_rate": 6.143200490685389e-05, + "loss": 0.4317, + "step": 5145 + }, + { + "epoch": 1.940422322775264, + "grad_norm": 0.36740246954784633, + "learning_rate": 6.142311330601592e-05, + "loss": 0.4349, + "step": 5146 + }, + { + "epoch": 1.9407993966817496, + "grad_norm": 0.41531203818828827, + "learning_rate": 6.14142202205752e-05, + "loss": 0.4337, + "step": 5147 + }, + { + "epoch": 1.9411764705882353, + "grad_norm": 0.37292915950086886, + "learning_rate": 6.140532565114801e-05, + "loss": 0.4305, + "step": 5148 + }, + { + "epoch": 1.941553544494721, + "grad_norm": 0.2858715305555173, + "learning_rate": 6.139642959835076e-05, + "loss": 0.4176, + "step": 5149 + }, + { + "epoch": 1.9419306184012066, + "grad_norm": 0.23084934557221204, + "learning_rate": 6.13875320627999e-05, + "loss": 0.4435, + "step": 5150 + }, + { + "epoch": 1.9423076923076923, + "grad_norm": 0.2845221237237062, + "learning_rate": 6.137863304511205e-05, + "loss": 0.4277, + "step": 5151 + }, + { + "epoch": 1.942684766214178, + "grad_norm": 0.3270369983769582, + "learning_rate": 6.136973254590389e-05, + "loss": 0.4221, + "step": 5152 + }, + { + "epoch": 1.9430618401206636, + "grad_norm": 0.2974672875326515, + "learning_rate": 6.136083056579222e-05, + "loss": 0.4384, + "step": 5153 + }, + { + "epoch": 1.9434389140271493, + "grad_norm": 0.23302630326246343, + "learning_rate": 6.135192710539394e-05, + "loss": 0.4311, + "step": 5154 + }, + { + "epoch": 1.943815987933635, + "grad_norm": 0.21666923990423315, + "learning_rate": 6.134302216532604e-05, + "loss": 0.4268, + "step": 5155 + }, + { + "epoch": 1.9441930618401206, + "grad_norm": 0.2577218306979524, + "learning_rate": 6.133411574620564e-05, + "loss": 0.4358, + "step": 5156 + }, + { + "epoch": 1.9445701357466063, + "grad_norm": 0.27616172651987086, + "learning_rate": 6.132520784864994e-05, + "loss": 0.4325, + "step": 5157 + }, + { + "epoch": 1.944947209653092, + "grad_norm": 0.3423581580572692, + "learning_rate": 6.131629847327625e-05, + "loss": 0.4299, + "step": 5158 + }, + { + "epoch": 1.9453242835595776, + "grad_norm": 0.377233403819589, + "learning_rate": 6.130738762070199e-05, + "loss": 0.4365, + "step": 5159 + }, + { + "epoch": 1.9457013574660633, + "grad_norm": 0.34825413662929977, + "learning_rate": 6.129847529154467e-05, + "loss": 0.4443, + "step": 5160 + }, + { + "epoch": 1.946078431372549, + "grad_norm": 0.3390637656519629, + "learning_rate": 6.12895614864219e-05, + "loss": 0.431, + "step": 5161 + }, + { + "epoch": 1.9464555052790347, + "grad_norm": 0.2849894529882499, + "learning_rate": 6.12806462059514e-05, + "loss": 0.4231, + "step": 5162 + }, + { + "epoch": 1.9468325791855203, + "grad_norm": 0.2701831458293802, + "learning_rate": 6.1271729450751e-05, + "loss": 0.4327, + "step": 5163 + }, + { + "epoch": 1.947209653092006, + "grad_norm": 0.2899526279093692, + "learning_rate": 6.12628112214386e-05, + "loss": 0.4352, + "step": 5164 + }, + { + "epoch": 1.9475867269984917, + "grad_norm": 0.2775975572305508, + "learning_rate": 6.125389151863226e-05, + "loss": 0.4333, + "step": 5165 + }, + { + "epoch": 1.9479638009049773, + "grad_norm": 0.25732804682381827, + "learning_rate": 6.124497034295009e-05, + "loss": 0.425, + "step": 5166 + }, + { + "epoch": 1.948340874811463, + "grad_norm": 0.2405872178157052, + "learning_rate": 6.123604769501032e-05, + "loss": 0.4354, + "step": 5167 + }, + { + "epoch": 1.9487179487179487, + "grad_norm": 0.2921553099119894, + "learning_rate": 6.122712357543128e-05, + "loss": 0.4363, + "step": 5168 + }, + { + "epoch": 1.9490950226244343, + "grad_norm": 0.2865211741823676, + "learning_rate": 6.121819798483139e-05, + "loss": 0.4295, + "step": 5169 + }, + { + "epoch": 1.94947209653092, + "grad_norm": 0.24945509369809762, + "learning_rate": 6.120927092382922e-05, + "loss": 0.4195, + "step": 5170 + }, + { + "epoch": 1.9498491704374057, + "grad_norm": 0.38509983208695225, + "learning_rate": 6.120034239304337e-05, + "loss": 0.44, + "step": 5171 + }, + { + "epoch": 1.9502262443438914, + "grad_norm": 0.4808564233716553, + "learning_rate": 6.119141239309259e-05, + "loss": 0.4306, + "step": 5172 + }, + { + "epoch": 1.950603318250377, + "grad_norm": 0.45752688577078887, + "learning_rate": 6.118248092459576e-05, + "loss": 0.4278, + "step": 5173 + }, + { + "epoch": 1.9509803921568627, + "grad_norm": 0.4032579498405219, + "learning_rate": 6.117354798817177e-05, + "loss": 0.4292, + "step": 5174 + }, + { + "epoch": 1.9513574660633484, + "grad_norm": 0.40489417897388774, + "learning_rate": 6.11646135844397e-05, + "loss": 0.43, + "step": 5175 + }, + { + "epoch": 1.951734539969834, + "grad_norm": 0.42741185807235854, + "learning_rate": 6.115567771401867e-05, + "loss": 0.429, + "step": 5176 + }, + { + "epoch": 1.9521116138763197, + "grad_norm": 0.3980856354019884, + "learning_rate": 6.114674037752794e-05, + "loss": 0.4294, + "step": 5177 + }, + { + "epoch": 1.9524886877828054, + "grad_norm": 0.3300600171975561, + "learning_rate": 6.113780157558685e-05, + "loss": 0.4256, + "step": 5178 + }, + { + "epoch": 1.952865761689291, + "grad_norm": 0.2581220489140906, + "learning_rate": 6.112886130881488e-05, + "loss": 0.4374, + "step": 5179 + }, + { + "epoch": 1.953242835595777, + "grad_norm": 0.2810200281817808, + "learning_rate": 6.111991957783155e-05, + "loss": 0.4435, + "step": 5180 + }, + { + "epoch": 1.9536199095022626, + "grad_norm": 0.235030034922998, + "learning_rate": 6.111097638325653e-05, + "loss": 0.4455, + "step": 5181 + }, + { + "epoch": 1.9539969834087483, + "grad_norm": 0.25854916617233914, + "learning_rate": 6.110203172570957e-05, + "loss": 0.4341, + "step": 5182 + }, + { + "epoch": 1.954374057315234, + "grad_norm": 0.2743483350651568, + "learning_rate": 6.109308560581053e-05, + "loss": 0.4322, + "step": 5183 + }, + { + "epoch": 1.9547511312217196, + "grad_norm": 0.3465787710791588, + "learning_rate": 6.108413802417935e-05, + "loss": 0.439, + "step": 5184 + }, + { + "epoch": 1.9551282051282053, + "grad_norm": 0.44132830523925654, + "learning_rate": 6.107518898143614e-05, + "loss": 0.4335, + "step": 5185 + }, + { + "epoch": 1.955505279034691, + "grad_norm": 0.4991487947521824, + "learning_rate": 6.106623847820101e-05, + "loss": 0.4312, + "step": 5186 + }, + { + "epoch": 1.9558823529411766, + "grad_norm": 0.5450973639850172, + "learning_rate": 6.105728651509424e-05, + "loss": 0.4376, + "step": 5187 + }, + { + "epoch": 1.9562594268476623, + "grad_norm": 0.5191347518211165, + "learning_rate": 6.10483330927362e-05, + "loss": 0.4213, + "step": 5188 + }, + { + "epoch": 1.956636500754148, + "grad_norm": 0.44344819880860714, + "learning_rate": 6.103937821174733e-05, + "loss": 0.4343, + "step": 5189 + }, + { + "epoch": 1.9570135746606336, + "grad_norm": 0.32009047865691714, + "learning_rate": 6.1030421872748226e-05, + "loss": 0.4302, + "step": 5190 + }, + { + "epoch": 1.9573906485671193, + "grad_norm": 0.245429902379328, + "learning_rate": 6.102146407635953e-05, + "loss": 0.4273, + "step": 5191 + }, + { + "epoch": 1.957767722473605, + "grad_norm": 0.27018695079735844, + "learning_rate": 6.1012504823202024e-05, + "loss": 0.4367, + "step": 5192 + }, + { + "epoch": 1.9581447963800906, + "grad_norm": 0.35663915779632777, + "learning_rate": 6.100354411389656e-05, + "loss": 0.4339, + "step": 5193 + }, + { + "epoch": 1.9585218702865763, + "grad_norm": 0.4297029927924005, + "learning_rate": 6.0994581949064146e-05, + "loss": 0.4369, + "step": 5194 + }, + { + "epoch": 1.958898944193062, + "grad_norm": 0.3941990869926259, + "learning_rate": 6.0985618329325804e-05, + "loss": 0.4343, + "step": 5195 + }, + { + "epoch": 1.9592760180995477, + "grad_norm": 0.33469858717514095, + "learning_rate": 6.097665325530274e-05, + "loss": 0.4424, + "step": 5196 + }, + { + "epoch": 1.9596530920060333, + "grad_norm": 0.3364702650767368, + "learning_rate": 6.09676867276162e-05, + "loss": 0.4329, + "step": 5197 + }, + { + "epoch": 1.960030165912519, + "grad_norm": 0.34001163951059077, + "learning_rate": 6.095871874688757e-05, + "loss": 0.4183, + "step": 5198 + }, + { + "epoch": 1.9604072398190047, + "grad_norm": 0.3044168804684423, + "learning_rate": 6.0949749313738325e-05, + "loss": 0.4262, + "step": 5199 + }, + { + "epoch": 1.9607843137254903, + "grad_norm": 0.22155034432979895, + "learning_rate": 6.094077842879003e-05, + "loss": 0.4184, + "step": 5200 + }, + { + "epoch": 1.961161387631976, + "grad_norm": 0.274020437656579, + "learning_rate": 6.093180609266437e-05, + "loss": 0.4372, + "step": 5201 + }, + { + "epoch": 1.9615384615384617, + "grad_norm": 0.35270386898057526, + "learning_rate": 6.092283230598311e-05, + "loss": 0.4402, + "step": 5202 + }, + { + "epoch": 1.9619155354449473, + "grad_norm": 0.4274337387778576, + "learning_rate": 6.0913857069368145e-05, + "loss": 0.4233, + "step": 5203 + }, + { + "epoch": 1.962292609351433, + "grad_norm": 0.4379504253561969, + "learning_rate": 6.090488038344141e-05, + "loss": 0.4285, + "step": 5204 + }, + { + "epoch": 1.9626696832579187, + "grad_norm": 0.3906688113948547, + "learning_rate": 6.089590224882502e-05, + "loss": 0.4269, + "step": 5205 + }, + { + "epoch": 1.9630467571644044, + "grad_norm": 0.369797307352933, + "learning_rate": 6.088692266614114e-05, + "loss": 0.434, + "step": 5206 + }, + { + "epoch": 1.96342383107089, + "grad_norm": 0.4053884318899659, + "learning_rate": 6.087794163601205e-05, + "loss": 0.4277, + "step": 5207 + }, + { + "epoch": 1.9638009049773757, + "grad_norm": 0.36239798942832463, + "learning_rate": 6.086895915906011e-05, + "loss": 0.4368, + "step": 5208 + }, + { + "epoch": 1.9641779788838614, + "grad_norm": 0.21648974231725077, + "learning_rate": 6.0859975235907825e-05, + "loss": 0.4313, + "step": 5209 + }, + { + "epoch": 1.964555052790347, + "grad_norm": 0.34171498792439403, + "learning_rate": 6.085098986717775e-05, + "loss": 0.4467, + "step": 5210 + }, + { + "epoch": 1.9649321266968327, + "grad_norm": 0.44967996701420765, + "learning_rate": 6.084200305349257e-05, + "loss": 0.4395, + "step": 5211 + }, + { + "epoch": 1.9653092006033184, + "grad_norm": 0.40513090856886513, + "learning_rate": 6.0833014795475065e-05, + "loss": 0.4335, + "step": 5212 + }, + { + "epoch": 1.965686274509804, + "grad_norm": 0.28704319999000033, + "learning_rate": 6.0824025093748123e-05, + "loss": 0.4299, + "step": 5213 + }, + { + "epoch": 1.9660633484162897, + "grad_norm": 0.3033713315144454, + "learning_rate": 6.081503394893471e-05, + "loss": 0.4231, + "step": 5214 + }, + { + "epoch": 1.9664404223227754, + "grad_norm": 0.36227227051635286, + "learning_rate": 6.0806041361657894e-05, + "loss": 0.4281, + "step": 5215 + }, + { + "epoch": 1.966817496229261, + "grad_norm": 0.3457053851749164, + "learning_rate": 6.079704733254089e-05, + "loss": 0.4367, + "step": 5216 + }, + { + "epoch": 1.9671945701357467, + "grad_norm": 0.3144644608259517, + "learning_rate": 6.0788051862206934e-05, + "loss": 0.4427, + "step": 5217 + }, + { + "epoch": 1.9675716440422324, + "grad_norm": 0.3418556512227856, + "learning_rate": 6.077905495127942e-05, + "loss": 0.4336, + "step": 5218 + }, + { + "epoch": 1.967948717948718, + "grad_norm": 0.3938755826264275, + "learning_rate": 6.077005660038184e-05, + "loss": 0.4287, + "step": 5219 + }, + { + "epoch": 1.9683257918552037, + "grad_norm": 0.33669837140681, + "learning_rate": 6.076105681013777e-05, + "loss": 0.4374, + "step": 5220 + }, + { + "epoch": 1.9687028657616894, + "grad_norm": 0.34540364596844086, + "learning_rate": 6.075205558117086e-05, + "loss": 0.4231, + "step": 5221 + }, + { + "epoch": 1.969079939668175, + "grad_norm": 0.4538900508046896, + "learning_rate": 6.0743052914104906e-05, + "loss": 0.4365, + "step": 5222 + }, + { + "epoch": 1.9694570135746607, + "grad_norm": 0.4085009375423182, + "learning_rate": 6.0734048809563804e-05, + "loss": 0.4182, + "step": 5223 + }, + { + "epoch": 1.9698340874811464, + "grad_norm": 0.3259670330109517, + "learning_rate": 6.072504326817148e-05, + "loss": 0.4474, + "step": 5224 + }, + { + "epoch": 1.970211161387632, + "grad_norm": 0.3206178451867778, + "learning_rate": 6.0716036290552065e-05, + "loss": 0.4339, + "step": 5225 + }, + { + "epoch": 1.9705882352941178, + "grad_norm": 0.30428179635631913, + "learning_rate": 6.070702787732971e-05, + "loss": 0.4191, + "step": 5226 + }, + { + "epoch": 1.9709653092006034, + "grad_norm": 0.33594057499542535, + "learning_rate": 6.0698018029128673e-05, + "loss": 0.4374, + "step": 5227 + }, + { + "epoch": 1.971342383107089, + "grad_norm": 0.2691607585156506, + "learning_rate": 6.068900674657336e-05, + "loss": 0.4264, + "step": 5228 + }, + { + "epoch": 1.9717194570135748, + "grad_norm": 0.2086656417700571, + "learning_rate": 6.067999403028823e-05, + "loss": 0.441, + "step": 5229 + }, + { + "epoch": 1.9720965309200604, + "grad_norm": 0.21431837960597844, + "learning_rate": 6.067097988089785e-05, + "loss": 0.4319, + "step": 5230 + }, + { + "epoch": 1.972473604826546, + "grad_norm": 0.23872825910144094, + "learning_rate": 6.0661964299026894e-05, + "loss": 0.4311, + "step": 5231 + }, + { + "epoch": 1.9728506787330318, + "grad_norm": 0.2552691432188204, + "learning_rate": 6.065294728530015e-05, + "loss": 0.4301, + "step": 5232 + }, + { + "epoch": 1.9732277526395174, + "grad_norm": 0.23925697613229308, + "learning_rate": 6.0643928840342485e-05, + "loss": 0.4197, + "step": 5233 + }, + { + "epoch": 1.973604826546003, + "grad_norm": 0.3289151877752341, + "learning_rate": 6.063490896477884e-05, + "loss": 0.4406, + "step": 5234 + }, + { + "epoch": 1.9739819004524888, + "grad_norm": 0.36031749769330923, + "learning_rate": 6.062588765923433e-05, + "loss": 0.4312, + "step": 5235 + }, + { + "epoch": 1.9743589743589745, + "grad_norm": 0.23832218389726004, + "learning_rate": 6.061686492433409e-05, + "loss": 0.4263, + "step": 5236 + }, + { + "epoch": 1.9747360482654601, + "grad_norm": 0.23213034156543932, + "learning_rate": 6.06078407607034e-05, + "loss": 0.4441, + "step": 5237 + }, + { + "epoch": 1.9751131221719458, + "grad_norm": 0.2503517116301244, + "learning_rate": 6.059881516896761e-05, + "loss": 0.4336, + "step": 5238 + }, + { + "epoch": 1.9754901960784315, + "grad_norm": 0.2552730429571951, + "learning_rate": 6.058978814975222e-05, + "loss": 0.4449, + "step": 5239 + }, + { + "epoch": 1.9758672699849171, + "grad_norm": 0.31733589788662164, + "learning_rate": 6.0580759703682774e-05, + "loss": 0.4323, + "step": 5240 + }, + { + "epoch": 1.9762443438914028, + "grad_norm": 0.3317275203561303, + "learning_rate": 6.057172983138492e-05, + "loss": 0.4277, + "step": 5241 + }, + { + "epoch": 1.9766214177978885, + "grad_norm": 0.33960071619745674, + "learning_rate": 6.056269853348444e-05, + "loss": 0.4258, + "step": 5242 + }, + { + "epoch": 1.9769984917043741, + "grad_norm": 0.294282769820963, + "learning_rate": 6.055366581060721e-05, + "loss": 0.4282, + "step": 5243 + }, + { + "epoch": 1.9773755656108598, + "grad_norm": 0.26627319764759333, + "learning_rate": 6.054463166337915e-05, + "loss": 0.4287, + "step": 5244 + }, + { + "epoch": 1.9777526395173455, + "grad_norm": 0.2491614407988511, + "learning_rate": 6.053559609242634e-05, + "loss": 0.4301, + "step": 5245 + }, + { + "epoch": 1.9781297134238311, + "grad_norm": 0.26766538049930166, + "learning_rate": 6.052655909837495e-05, + "loss": 0.4349, + "step": 5246 + }, + { + "epoch": 1.9785067873303168, + "grad_norm": 0.28131083663388823, + "learning_rate": 6.051752068185121e-05, + "loss": 0.4386, + "step": 5247 + }, + { + "epoch": 1.9788838612368025, + "grad_norm": 0.2869843211730475, + "learning_rate": 6.050848084348148e-05, + "loss": 0.415, + "step": 5248 + }, + { + "epoch": 1.9792609351432882, + "grad_norm": 0.2937224689841907, + "learning_rate": 6.0499439583892234e-05, + "loss": 0.4251, + "step": 5249 + }, + { + "epoch": 1.9796380090497738, + "grad_norm": 0.33481186558282444, + "learning_rate": 6.049039690371e-05, + "loss": 0.4472, + "step": 5250 + }, + { + "epoch": 1.9800150829562595, + "grad_norm": 0.36297966937144244, + "learning_rate": 6.048135280356142e-05, + "loss": 0.431, + "step": 5251 + }, + { + "epoch": 1.9803921568627452, + "grad_norm": 0.337687372542343, + "learning_rate": 6.047230728407329e-05, + "loss": 0.4457, + "step": 5252 + }, + { + "epoch": 1.9807692307692308, + "grad_norm": 0.33694902442793684, + "learning_rate": 6.046326034587241e-05, + "loss": 0.4535, + "step": 5253 + }, + { + "epoch": 1.9811463046757165, + "grad_norm": 0.36532205348241475, + "learning_rate": 6.045421198958572e-05, + "loss": 0.4293, + "step": 5254 + }, + { + "epoch": 1.9815233785822022, + "grad_norm": 0.37043385962352215, + "learning_rate": 6.044516221584032e-05, + "loss": 0.4371, + "step": 5255 + }, + { + "epoch": 1.9819004524886878, + "grad_norm": 0.304860041547654, + "learning_rate": 6.043611102526328e-05, + "loss": 0.4335, + "step": 5256 + }, + { + "epoch": 1.9822775263951735, + "grad_norm": 0.25234796720449304, + "learning_rate": 6.042705841848188e-05, + "loss": 0.4427, + "step": 5257 + }, + { + "epoch": 1.9826546003016592, + "grad_norm": 0.2576232212016, + "learning_rate": 6.0418004396123456e-05, + "loss": 0.4385, + "step": 5258 + }, + { + "epoch": 1.9830316742081449, + "grad_norm": 0.2757144599436326, + "learning_rate": 6.040894895881543e-05, + "loss": 0.4421, + "step": 5259 + }, + { + "epoch": 1.9834087481146305, + "grad_norm": 0.28721270711159763, + "learning_rate": 6.039989210718536e-05, + "loss": 0.4409, + "step": 5260 + }, + { + "epoch": 1.9837858220211162, + "grad_norm": 0.28718346778434467, + "learning_rate": 6.039083384186086e-05, + "loss": 0.4309, + "step": 5261 + }, + { + "epoch": 1.9841628959276019, + "grad_norm": 0.2968432450566863, + "learning_rate": 6.038177416346965e-05, + "loss": 0.4377, + "step": 5262 + }, + { + "epoch": 1.9845399698340875, + "grad_norm": 0.27663278999706853, + "learning_rate": 6.037271307263957e-05, + "loss": 0.4335, + "step": 5263 + }, + { + "epoch": 1.9849170437405732, + "grad_norm": 0.2792440711647132, + "learning_rate": 6.036365056999854e-05, + "loss": 0.4316, + "step": 5264 + }, + { + "epoch": 1.9852941176470589, + "grad_norm": 0.27784803547563613, + "learning_rate": 6.0354586656174606e-05, + "loss": 0.4328, + "step": 5265 + }, + { + "epoch": 1.9856711915535445, + "grad_norm": 0.2681620356994337, + "learning_rate": 6.034552133179585e-05, + "loss": 0.4293, + "step": 5266 + }, + { + "epoch": 1.9860482654600302, + "grad_norm": 0.3288374639342613, + "learning_rate": 6.033645459749054e-05, + "loss": 0.432, + "step": 5267 + }, + { + "epoch": 1.9864253393665159, + "grad_norm": 0.390930108413126, + "learning_rate": 6.0327386453886944e-05, + "loss": 0.4518, + "step": 5268 + }, + { + "epoch": 1.9868024132730016, + "grad_norm": 0.400639471817605, + "learning_rate": 6.031831690161349e-05, + "loss": 0.4361, + "step": 5269 + }, + { + "epoch": 1.9871794871794872, + "grad_norm": 0.44516170981528, + "learning_rate": 6.030924594129872e-05, + "loss": 0.4397, + "step": 5270 + }, + { + "epoch": 1.987556561085973, + "grad_norm": 0.4698772194473896, + "learning_rate": 6.0300173573571195e-05, + "loss": 0.427, + "step": 5271 + }, + { + "epoch": 1.9879336349924586, + "grad_norm": 0.42471234710543193, + "learning_rate": 6.0291099799059664e-05, + "loss": 0.4354, + "step": 5272 + }, + { + "epoch": 1.9883107088989442, + "grad_norm": 0.3689162014081315, + "learning_rate": 6.028202461839291e-05, + "loss": 0.4351, + "step": 5273 + }, + { + "epoch": 1.98868778280543, + "grad_norm": 0.33631984414010646, + "learning_rate": 6.027294803219985e-05, + "loss": 0.428, + "step": 5274 + }, + { + "epoch": 1.9890648567119156, + "grad_norm": 0.33419180589755976, + "learning_rate": 6.0263870041109465e-05, + "loss": 0.4315, + "step": 5275 + }, + { + "epoch": 1.9894419306184012, + "grad_norm": 0.3679526405413062, + "learning_rate": 6.025479064575085e-05, + "loss": 0.4279, + "step": 5276 + }, + { + "epoch": 1.989819004524887, + "grad_norm": 0.39271896942358375, + "learning_rate": 6.0245709846753226e-05, + "loss": 0.4375, + "step": 5277 + }, + { + "epoch": 1.9901960784313726, + "grad_norm": 0.37110997663531664, + "learning_rate": 6.023662764474586e-05, + "loss": 0.4415, + "step": 5278 + }, + { + "epoch": 1.9905731523378583, + "grad_norm": 0.3283093866402883, + "learning_rate": 6.0227544040358144e-05, + "loss": 0.4233, + "step": 5279 + }, + { + "epoch": 1.990950226244344, + "grad_norm": 0.318910605344668, + "learning_rate": 6.021845903421958e-05, + "loss": 0.4243, + "step": 5280 + }, + { + "epoch": 1.9913273001508296, + "grad_norm": 0.28610877029081455, + "learning_rate": 6.020937262695973e-05, + "loss": 0.4356, + "step": 5281 + }, + { + "epoch": 1.9917043740573153, + "grad_norm": 0.2368541519806284, + "learning_rate": 6.020028481920829e-05, + "loss": 0.4177, + "step": 5282 + }, + { + "epoch": 1.992081447963801, + "grad_norm": 0.21751475143908008, + "learning_rate": 6.0191195611595024e-05, + "loss": 0.4233, + "step": 5283 + }, + { + "epoch": 1.9924585218702866, + "grad_norm": 0.25532105059048404, + "learning_rate": 6.018210500474981e-05, + "loss": 0.426, + "step": 5284 + }, + { + "epoch": 1.9928355957767723, + "grad_norm": 0.26919545688759994, + "learning_rate": 6.0173012999302624e-05, + "loss": 0.4355, + "step": 5285 + }, + { + "epoch": 1.993212669683258, + "grad_norm": 0.2752113787354001, + "learning_rate": 6.016391959588353e-05, + "loss": 0.4314, + "step": 5286 + }, + { + "epoch": 1.9935897435897436, + "grad_norm": 0.32130559310356904, + "learning_rate": 6.015482479512269e-05, + "loss": 0.4206, + "step": 5287 + }, + { + "epoch": 1.9939668174962293, + "grad_norm": 0.3063754303914869, + "learning_rate": 6.0145728597650367e-05, + "loss": 0.4225, + "step": 5288 + }, + { + "epoch": 1.994343891402715, + "grad_norm": 0.2539690634427835, + "learning_rate": 6.013663100409693e-05, + "loss": 0.43, + "step": 5289 + }, + { + "epoch": 1.9947209653092006, + "grad_norm": 0.24975933841280917, + "learning_rate": 6.0127532015092823e-05, + "loss": 0.4373, + "step": 5290 + }, + { + "epoch": 1.9950980392156863, + "grad_norm": 0.23309175929042708, + "learning_rate": 6.01184316312686e-05, + "loss": 0.438, + "step": 5291 + }, + { + "epoch": 1.995475113122172, + "grad_norm": 0.30569922203177147, + "learning_rate": 6.0109329853254916e-05, + "loss": 0.4326, + "step": 5292 + }, + { + "epoch": 1.9958521870286576, + "grad_norm": 0.39918124035019287, + "learning_rate": 6.0100226681682496e-05, + "loss": 0.4332, + "step": 5293 + }, + { + "epoch": 1.9962292609351433, + "grad_norm": 0.3692641329055498, + "learning_rate": 6.00911221171822e-05, + "loss": 0.437, + "step": 5294 + }, + { + "epoch": 1.996606334841629, + "grad_norm": 0.26788177812665637, + "learning_rate": 6.0082016160384966e-05, + "loss": 0.4171, + "step": 5295 + }, + { + "epoch": 1.9969834087481146, + "grad_norm": 0.23561853482707298, + "learning_rate": 6.007290881192182e-05, + "loss": 0.4254, + "step": 5296 + }, + { + "epoch": 1.9973604826546003, + "grad_norm": 0.3130166177605249, + "learning_rate": 6.0063800072423905e-05, + "loss": 0.42, + "step": 5297 + }, + { + "epoch": 1.997737556561086, + "grad_norm": 0.3687001761637817, + "learning_rate": 6.0054689942522444e-05, + "loss": 0.4344, + "step": 5298 + }, + { + "epoch": 1.9981146304675717, + "grad_norm": 0.38417468064503835, + "learning_rate": 6.0045578422848754e-05, + "loss": 0.431, + "step": 5299 + }, + { + "epoch": 1.9984917043740573, + "grad_norm": 0.34768782800199766, + "learning_rate": 6.0036465514034254e-05, + "loss": 0.434, + "step": 5300 + }, + { + "epoch": 1.998868778280543, + "grad_norm": 0.37057403898582814, + "learning_rate": 6.002735121671047e-05, + "loss": 0.4316, + "step": 5301 + }, + { + "epoch": 1.9992458521870287, + "grad_norm": 0.44752961036230593, + "learning_rate": 6.001823553150902e-05, + "loss": 0.4513, + "step": 5302 + }, + { + "epoch": 1.9996229260935143, + "grad_norm": 0.4623159235422908, + "learning_rate": 6.0009118459061585e-05, + "loss": 0.4432, + "step": 5303 + }, + { + "epoch": 2.0, + "grad_norm": 0.3977088496930432, + "learning_rate": 6.000000000000001e-05, + "loss": 0.4352, + "step": 5304 + }, + { + "epoch": 2.0003770739064857, + "grad_norm": 0.3026910210329173, + "learning_rate": 5.999088015495616e-05, + "loss": 0.4121, + "step": 5305 + }, + { + "epoch": 2.0007541478129713, + "grad_norm": 0.23402981793227415, + "learning_rate": 5.998175892456205e-05, + "loss": 0.4124, + "step": 5306 + }, + { + "epoch": 2.001131221719457, + "grad_norm": 0.2739766678769215, + "learning_rate": 5.997263630944978e-05, + "loss": 0.4089, + "step": 5307 + }, + { + "epoch": 2.0015082956259427, + "grad_norm": 0.30734851416407644, + "learning_rate": 5.996351231025152e-05, + "loss": 0.4218, + "step": 5308 + }, + { + "epoch": 2.0018853695324283, + "grad_norm": 0.3263654389903649, + "learning_rate": 5.9954386927599555e-05, + "loss": 0.4136, + "step": 5309 + }, + { + "epoch": 2.002262443438914, + "grad_norm": 0.3091215881385145, + "learning_rate": 5.9945260162126295e-05, + "loss": 0.4181, + "step": 5310 + }, + { + "epoch": 2.0026395173453997, + "grad_norm": 0.3049572636261895, + "learning_rate": 5.993613201446419e-05, + "loss": 0.4199, + "step": 5311 + }, + { + "epoch": 2.0030165912518854, + "grad_norm": 0.30922747900196573, + "learning_rate": 5.992700248524581e-05, + "loss": 0.427, + "step": 5312 + }, + { + "epoch": 2.003393665158371, + "grad_norm": 0.2497687469523008, + "learning_rate": 5.991787157510385e-05, + "loss": 0.4119, + "step": 5313 + }, + { + "epoch": 2.0037707390648567, + "grad_norm": 0.20966370239389473, + "learning_rate": 5.9908739284671054e-05, + "loss": 0.4042, + "step": 5314 + }, + { + "epoch": 2.0041478129713424, + "grad_norm": 0.24073591898738655, + "learning_rate": 5.989960561458027e-05, + "loss": 0.4147, + "step": 5315 + }, + { + "epoch": 2.004524886877828, + "grad_norm": 0.2829656056189382, + "learning_rate": 5.989047056546448e-05, + "loss": 0.4189, + "step": 5316 + }, + { + "epoch": 2.0049019607843137, + "grad_norm": 0.3691886985104506, + "learning_rate": 5.9881334137956715e-05, + "loss": 0.4196, + "step": 5317 + }, + { + "epoch": 2.0052790346907994, + "grad_norm": 0.3420599544507286, + "learning_rate": 5.9872196332690135e-05, + "loss": 0.4182, + "step": 5318 + }, + { + "epoch": 2.005656108597285, + "grad_norm": 0.3216531345906275, + "learning_rate": 5.9863057150297975e-05, + "loss": 0.4211, + "step": 5319 + }, + { + "epoch": 2.0060331825037707, + "grad_norm": 0.35251846147469096, + "learning_rate": 5.985391659141357e-05, + "loss": 0.4261, + "step": 5320 + }, + { + "epoch": 2.0064102564102564, + "grad_norm": 0.3457132594888152, + "learning_rate": 5.984477465667036e-05, + "loss": 0.4183, + "step": 5321 + }, + { + "epoch": 2.006787330316742, + "grad_norm": 0.34100577814889454, + "learning_rate": 5.9835631346701866e-05, + "loss": 0.4093, + "step": 5322 + }, + { + "epoch": 2.0071644042232277, + "grad_norm": 0.28981597140251236, + "learning_rate": 5.982648666214172e-05, + "loss": 0.4149, + "step": 5323 + }, + { + "epoch": 2.0075414781297134, + "grad_norm": 0.24982968716748322, + "learning_rate": 5.981734060362362e-05, + "loss": 0.4202, + "step": 5324 + }, + { + "epoch": 2.007918552036199, + "grad_norm": 0.2274848886382616, + "learning_rate": 5.980819317178141e-05, + "loss": 0.4177, + "step": 5325 + }, + { + "epoch": 2.0082956259426847, + "grad_norm": 0.25131373221720005, + "learning_rate": 5.979904436724897e-05, + "loss": 0.4144, + "step": 5326 + }, + { + "epoch": 2.0086726998491704, + "grad_norm": 0.24487963701622975, + "learning_rate": 5.9789894190660325e-05, + "loss": 0.4163, + "step": 5327 + }, + { + "epoch": 2.009049773755656, + "grad_norm": 0.29683169741386983, + "learning_rate": 5.9780742642649565e-05, + "loss": 0.425, + "step": 5328 + }, + { + "epoch": 2.0094268476621417, + "grad_norm": 0.3217771903538917, + "learning_rate": 5.977158972385089e-05, + "loss": 0.4122, + "step": 5329 + }, + { + "epoch": 2.0098039215686274, + "grad_norm": 0.2896492443813514, + "learning_rate": 5.976243543489856e-05, + "loss": 0.4102, + "step": 5330 + }, + { + "epoch": 2.010180995475113, + "grad_norm": 0.2971265197748785, + "learning_rate": 5.975327977642702e-05, + "loss": 0.416, + "step": 5331 + }, + { + "epoch": 2.0105580693815988, + "grad_norm": 0.26293153908274364, + "learning_rate": 5.974412274907069e-05, + "loss": 0.4173, + "step": 5332 + }, + { + "epoch": 2.0109351432880844, + "grad_norm": 0.23078556006818024, + "learning_rate": 5.973496435346417e-05, + "loss": 0.4132, + "step": 5333 + }, + { + "epoch": 2.01131221719457, + "grad_norm": 0.24244832369185934, + "learning_rate": 5.972580459024214e-05, + "loss": 0.4125, + "step": 5334 + }, + { + "epoch": 2.0116892911010558, + "grad_norm": 0.25885330568837794, + "learning_rate": 5.971664346003932e-05, + "loss": 0.4085, + "step": 5335 + }, + { + "epoch": 2.0120663650075414, + "grad_norm": 0.29736007353721705, + "learning_rate": 5.970748096349061e-05, + "loss": 0.4238, + "step": 5336 + }, + { + "epoch": 2.012443438914027, + "grad_norm": 0.3316770142107208, + "learning_rate": 5.969831710123095e-05, + "loss": 0.4141, + "step": 5337 + }, + { + "epoch": 2.0128205128205128, + "grad_norm": 0.3138490875961291, + "learning_rate": 5.96891518738954e-05, + "loss": 0.4173, + "step": 5338 + }, + { + "epoch": 2.0131975867269984, + "grad_norm": 0.2481384247127631, + "learning_rate": 5.967998528211907e-05, + "loss": 0.4042, + "step": 5339 + }, + { + "epoch": 2.013574660633484, + "grad_norm": 0.23843802818091836, + "learning_rate": 5.967081732653722e-05, + "loss": 0.4079, + "step": 5340 + }, + { + "epoch": 2.01395173453997, + "grad_norm": 0.265881718248369, + "learning_rate": 5.96616480077852e-05, + "loss": 0.4174, + "step": 5341 + }, + { + "epoch": 2.0143288084464555, + "grad_norm": 0.2623551692683074, + "learning_rate": 5.9652477326498386e-05, + "loss": 0.4165, + "step": 5342 + }, + { + "epoch": 2.014705882352941, + "grad_norm": 0.23892743422517146, + "learning_rate": 5.964330528331234e-05, + "loss": 0.425, + "step": 5343 + }, + { + "epoch": 2.015082956259427, + "grad_norm": 0.27012846164429266, + "learning_rate": 5.963413187886265e-05, + "loss": 0.4171, + "step": 5344 + }, + { + "epoch": 2.0154600301659125, + "grad_norm": 0.27501192200066243, + "learning_rate": 5.962495711378504e-05, + "loss": 0.4226, + "step": 5345 + }, + { + "epoch": 2.015837104072398, + "grad_norm": 0.26748950674962235, + "learning_rate": 5.961578098871531e-05, + "loss": 0.4177, + "step": 5346 + }, + { + "epoch": 2.016214177978884, + "grad_norm": 0.32568660949274103, + "learning_rate": 5.960660350428937e-05, + "loss": 0.4172, + "step": 5347 + }, + { + "epoch": 2.0165912518853695, + "grad_norm": 0.3217772465256499, + "learning_rate": 5.9597424661143177e-05, + "loss": 0.4103, + "step": 5348 + }, + { + "epoch": 2.016968325791855, + "grad_norm": 0.27211421920128553, + "learning_rate": 5.958824445991284e-05, + "loss": 0.4131, + "step": 5349 + }, + { + "epoch": 2.017345399698341, + "grad_norm": 0.17842137607283118, + "learning_rate": 5.957906290123456e-05, + "loss": 0.4129, + "step": 5350 + }, + { + "epoch": 2.0177224736048265, + "grad_norm": 0.26469878136866176, + "learning_rate": 5.956987998574456e-05, + "loss": 0.4201, + "step": 5351 + }, + { + "epoch": 2.018099547511312, + "grad_norm": 0.3424893996235769, + "learning_rate": 5.956069571407924e-05, + "loss": 0.4278, + "step": 5352 + }, + { + "epoch": 2.018476621417798, + "grad_norm": 0.3115721079114088, + "learning_rate": 5.955151008687508e-05, + "loss": 0.417, + "step": 5353 + }, + { + "epoch": 2.0188536953242835, + "grad_norm": 0.2711050063759208, + "learning_rate": 5.9542323104768594e-05, + "loss": 0.4257, + "step": 5354 + }, + { + "epoch": 2.019230769230769, + "grad_norm": 0.21983934177101522, + "learning_rate": 5.9533134768396455e-05, + "loss": 0.4208, + "step": 5355 + }, + { + "epoch": 2.019607843137255, + "grad_norm": 0.236839532276414, + "learning_rate": 5.95239450783954e-05, + "loss": 0.4253, + "step": 5356 + }, + { + "epoch": 2.0199849170437405, + "grad_norm": 0.29153320789952686, + "learning_rate": 5.9514754035402266e-05, + "loss": 0.4245, + "step": 5357 + }, + { + "epoch": 2.020361990950226, + "grad_norm": 0.3091693901514053, + "learning_rate": 5.950556164005399e-05, + "loss": 0.431, + "step": 5358 + }, + { + "epoch": 2.020739064856712, + "grad_norm": 0.29948351327523604, + "learning_rate": 5.94963678929876e-05, + "loss": 0.419, + "step": 5359 + }, + { + "epoch": 2.0211161387631975, + "grad_norm": 0.32739658858391046, + "learning_rate": 5.94871727948402e-05, + "loss": 0.4361, + "step": 5360 + }, + { + "epoch": 2.021493212669683, + "grad_norm": 0.3857460356083034, + "learning_rate": 5.947797634624901e-05, + "loss": 0.4274, + "step": 5361 + }, + { + "epoch": 2.021870286576169, + "grad_norm": 0.3747378684201688, + "learning_rate": 5.9468778547851344e-05, + "loss": 0.4141, + "step": 5362 + }, + { + "epoch": 2.0222473604826545, + "grad_norm": 0.3147697560768918, + "learning_rate": 5.945957940028459e-05, + "loss": 0.4223, + "step": 5363 + }, + { + "epoch": 2.02262443438914, + "grad_norm": 0.267217488438826, + "learning_rate": 5.945037890418624e-05, + "loss": 0.4135, + "step": 5364 + }, + { + "epoch": 2.023001508295626, + "grad_norm": 0.22038287903147494, + "learning_rate": 5.9441177060193895e-05, + "loss": 0.4117, + "step": 5365 + }, + { + "epoch": 2.0233785822021115, + "grad_norm": 0.2252448852142345, + "learning_rate": 5.943197386894521e-05, + "loss": 0.4237, + "step": 5366 + }, + { + "epoch": 2.023755656108597, + "grad_norm": 0.18765722944760752, + "learning_rate": 5.942276933107798e-05, + "loss": 0.4261, + "step": 5367 + }, + { + "epoch": 2.024132730015083, + "grad_norm": 0.17720870707978462, + "learning_rate": 5.941356344723007e-05, + "loss": 0.4147, + "step": 5368 + }, + { + "epoch": 2.0245098039215685, + "grad_norm": 0.21297477304822574, + "learning_rate": 5.940435621803941e-05, + "loss": 0.4297, + "step": 5369 + }, + { + "epoch": 2.024886877828054, + "grad_norm": 0.24475765165269575, + "learning_rate": 5.939514764414409e-05, + "loss": 0.4284, + "step": 5370 + }, + { + "epoch": 2.02526395173454, + "grad_norm": 0.2427615760831688, + "learning_rate": 5.938593772618225e-05, + "loss": 0.4166, + "step": 5371 + }, + { + "epoch": 2.0256410256410255, + "grad_norm": 0.20436336490788862, + "learning_rate": 5.937672646479211e-05, + "loss": 0.4166, + "step": 5372 + }, + { + "epoch": 2.026018099547511, + "grad_norm": 0.2108788102904402, + "learning_rate": 5.936751386061202e-05, + "loss": 0.4193, + "step": 5373 + }, + { + "epoch": 2.026395173453997, + "grad_norm": 0.23904369994295743, + "learning_rate": 5.935829991428039e-05, + "loss": 0.4208, + "step": 5374 + }, + { + "epoch": 2.0267722473604826, + "grad_norm": 0.24292303489817324, + "learning_rate": 5.934908462643575e-05, + "loss": 0.4192, + "step": 5375 + }, + { + "epoch": 2.0271493212669682, + "grad_norm": 0.22236554336071715, + "learning_rate": 5.9339867997716714e-05, + "loss": 0.4144, + "step": 5376 + }, + { + "epoch": 2.027526395173454, + "grad_norm": 0.21643783852350462, + "learning_rate": 5.9330650028761974e-05, + "loss": 0.4174, + "step": 5377 + }, + { + "epoch": 2.0279034690799396, + "grad_norm": 0.24587284719310698, + "learning_rate": 5.932143072021034e-05, + "loss": 0.4249, + "step": 5378 + }, + { + "epoch": 2.0282805429864252, + "grad_norm": 0.3045723655502357, + "learning_rate": 5.931221007270069e-05, + "loss": 0.4117, + "step": 5379 + }, + { + "epoch": 2.028657616892911, + "grad_norm": 0.3438689744890974, + "learning_rate": 5.930298808687201e-05, + "loss": 0.4169, + "step": 5380 + }, + { + "epoch": 2.0290346907993966, + "grad_norm": 0.32788178180444116, + "learning_rate": 5.929376476336338e-05, + "loss": 0.4288, + "step": 5381 + }, + { + "epoch": 2.0294117647058822, + "grad_norm": 0.3111740786328955, + "learning_rate": 5.9284540102813964e-05, + "loss": 0.4134, + "step": 5382 + }, + { + "epoch": 2.029788838612368, + "grad_norm": 0.32872395878832716, + "learning_rate": 5.927531410586302e-05, + "loss": 0.4133, + "step": 5383 + }, + { + "epoch": 2.0301659125188536, + "grad_norm": 0.3413451906982112, + "learning_rate": 5.926608677314991e-05, + "loss": 0.4277, + "step": 5384 + }, + { + "epoch": 2.0305429864253393, + "grad_norm": 0.34727832087937477, + "learning_rate": 5.925685810531408e-05, + "loss": 0.4112, + "step": 5385 + }, + { + "epoch": 2.030920060331825, + "grad_norm": 0.41718347765799413, + "learning_rate": 5.9247628102995054e-05, + "loss": 0.4259, + "step": 5386 + }, + { + "epoch": 2.0312971342383106, + "grad_norm": 0.4424627036976475, + "learning_rate": 5.923839676683248e-05, + "loss": 0.4118, + "step": 5387 + }, + { + "epoch": 2.0316742081447963, + "grad_norm": 0.4162440665163693, + "learning_rate": 5.9229164097466076e-05, + "loss": 0.4233, + "step": 5388 + }, + { + "epoch": 2.032051282051282, + "grad_norm": 0.35756098638828043, + "learning_rate": 5.921993009553566e-05, + "loss": 0.4144, + "step": 5389 + }, + { + "epoch": 2.0324283559577676, + "grad_norm": 0.2714373295439461, + "learning_rate": 5.9210694761681126e-05, + "loss": 0.4263, + "step": 5390 + }, + { + "epoch": 2.0328054298642533, + "grad_norm": 0.22937398435407155, + "learning_rate": 5.920145809654249e-05, + "loss": 0.4113, + "step": 5391 + }, + { + "epoch": 2.033182503770739, + "grad_norm": 0.25675505857124126, + "learning_rate": 5.919222010075985e-05, + "loss": 0.4282, + "step": 5392 + }, + { + "epoch": 2.0335595776772246, + "grad_norm": 0.2719003001878944, + "learning_rate": 5.9182980774973365e-05, + "loss": 0.4192, + "step": 5393 + }, + { + "epoch": 2.0339366515837103, + "grad_norm": 0.28257053998711434, + "learning_rate": 5.917374011982333e-05, + "loss": 0.4199, + "step": 5394 + }, + { + "epoch": 2.034313725490196, + "grad_norm": 0.3387219652094811, + "learning_rate": 5.9164498135950116e-05, + "loss": 0.4166, + "step": 5395 + }, + { + "epoch": 2.0346907993966816, + "grad_norm": 0.32927176660391316, + "learning_rate": 5.915525482399417e-05, + "loss": 0.4202, + "step": 5396 + }, + { + "epoch": 2.0350678733031673, + "grad_norm": 0.25420102798836797, + "learning_rate": 5.914601018459606e-05, + "loss": 0.4131, + "step": 5397 + }, + { + "epoch": 2.035444947209653, + "grad_norm": 0.2736706681902737, + "learning_rate": 5.9136764218396435e-05, + "loss": 0.4113, + "step": 5398 + }, + { + "epoch": 2.0358220211161386, + "grad_norm": 0.31608962858198186, + "learning_rate": 5.912751692603602e-05, + "loss": 0.4177, + "step": 5399 + }, + { + "epoch": 2.0361990950226243, + "grad_norm": 0.28405024674565443, + "learning_rate": 5.911826830815565e-05, + "loss": 0.4026, + "step": 5400 + }, + { + "epoch": 2.03657616892911, + "grad_norm": 0.18659612600520067, + "learning_rate": 5.910901836539624e-05, + "loss": 0.4185, + "step": 5401 + }, + { + "epoch": 2.0369532428355956, + "grad_norm": 0.22983947353557307, + "learning_rate": 5.909976709839881e-05, + "loss": 0.4042, + "step": 5402 + }, + { + "epoch": 2.0373303167420813, + "grad_norm": 0.3177288658273607, + "learning_rate": 5.9090514507804456e-05, + "loss": 0.4188, + "step": 5403 + }, + { + "epoch": 2.037707390648567, + "grad_norm": 0.3286937456099659, + "learning_rate": 5.9081260594254386e-05, + "loss": 0.4196, + "step": 5404 + }, + { + "epoch": 2.0380844645550527, + "grad_norm": 0.32993506928959876, + "learning_rate": 5.9072005358389886e-05, + "loss": 0.4222, + "step": 5405 + }, + { + "epoch": 2.0384615384615383, + "grad_norm": 0.3550095436218124, + "learning_rate": 5.9062748800852315e-05, + "loss": 0.4296, + "step": 5406 + }, + { + "epoch": 2.038838612368024, + "grad_norm": 0.31298122828579056, + "learning_rate": 5.905349092228317e-05, + "loss": 0.4097, + "step": 5407 + }, + { + "epoch": 2.0392156862745097, + "grad_norm": 0.2697374703312884, + "learning_rate": 5.904423172332401e-05, + "loss": 0.408, + "step": 5408 + }, + { + "epoch": 2.0395927601809953, + "grad_norm": 0.34781717314832256, + "learning_rate": 5.903497120461648e-05, + "loss": 0.4114, + "step": 5409 + }, + { + "epoch": 2.039969834087481, + "grad_norm": 0.3466924214113255, + "learning_rate": 5.902570936680232e-05, + "loss": 0.4265, + "step": 5410 + }, + { + "epoch": 2.0403469079939667, + "grad_norm": 0.2896066706199965, + "learning_rate": 5.90164462105234e-05, + "loss": 0.4252, + "step": 5411 + }, + { + "epoch": 2.0407239819004523, + "grad_norm": 0.25130280257540893, + "learning_rate": 5.90071817364216e-05, + "loss": 0.4114, + "step": 5412 + }, + { + "epoch": 2.041101055806938, + "grad_norm": 0.3198464657261538, + "learning_rate": 5.8997915945138974e-05, + "loss": 0.4162, + "step": 5413 + }, + { + "epoch": 2.0414781297134237, + "grad_norm": 0.4242042309919969, + "learning_rate": 5.898864883731763e-05, + "loss": 0.4097, + "step": 5414 + }, + { + "epoch": 2.0418552036199094, + "grad_norm": 0.4016406329837886, + "learning_rate": 5.897938041359976e-05, + "loss": 0.4133, + "step": 5415 + }, + { + "epoch": 2.042232277526395, + "grad_norm": 0.3216124361767717, + "learning_rate": 5.897011067462766e-05, + "loss": 0.4191, + "step": 5416 + }, + { + "epoch": 2.0426093514328807, + "grad_norm": 0.32791510834476345, + "learning_rate": 5.896083962104372e-05, + "loss": 0.4317, + "step": 5417 + }, + { + "epoch": 2.0429864253393664, + "grad_norm": 0.3334809840860963, + "learning_rate": 5.895156725349042e-05, + "loss": 0.4248, + "step": 5418 + }, + { + "epoch": 2.043363499245852, + "grad_norm": 0.37663528023457543, + "learning_rate": 5.8942293572610306e-05, + "loss": 0.4117, + "step": 5419 + }, + { + "epoch": 2.0437405731523377, + "grad_norm": 0.39580938258244414, + "learning_rate": 5.893301857904606e-05, + "loss": 0.4216, + "step": 5420 + }, + { + "epoch": 2.0441176470588234, + "grad_norm": 0.303241714082815, + "learning_rate": 5.892374227344041e-05, + "loss": 0.4145, + "step": 5421 + }, + { + "epoch": 2.044494720965309, + "grad_norm": 0.22557686807008553, + "learning_rate": 5.8914464656436204e-05, + "loss": 0.4242, + "step": 5422 + }, + { + "epoch": 2.0448717948717947, + "grad_norm": 0.2444144567497157, + "learning_rate": 5.890518572867638e-05, + "loss": 0.4178, + "step": 5423 + }, + { + "epoch": 2.0452488687782804, + "grad_norm": 0.2989429046386716, + "learning_rate": 5.889590549080396e-05, + "loss": 0.4318, + "step": 5424 + }, + { + "epoch": 2.045625942684766, + "grad_norm": 0.31514965053726024, + "learning_rate": 5.888662394346204e-05, + "loss": 0.4206, + "step": 5425 + }, + { + "epoch": 2.0460030165912517, + "grad_norm": 0.28509402058163225, + "learning_rate": 5.8877341087293845e-05, + "loss": 0.4204, + "step": 5426 + }, + { + "epoch": 2.0463800904977374, + "grad_norm": 0.27525675623628004, + "learning_rate": 5.8868056922942643e-05, + "loss": 0.4231, + "step": 5427 + }, + { + "epoch": 2.046757164404223, + "grad_norm": 0.23460280133926936, + "learning_rate": 5.885877145105184e-05, + "loss": 0.4228, + "step": 5428 + }, + { + "epoch": 2.0471342383107087, + "grad_norm": 0.2158542603687244, + "learning_rate": 5.8849484672264906e-05, + "loss": 0.4106, + "step": 5429 + }, + { + "epoch": 2.0475113122171944, + "grad_norm": 0.27256172049704036, + "learning_rate": 5.884019658722539e-05, + "loss": 0.4255, + "step": 5430 + }, + { + "epoch": 2.04788838612368, + "grad_norm": 0.2828522864860686, + "learning_rate": 5.883090719657698e-05, + "loss": 0.3997, + "step": 5431 + }, + { + "epoch": 2.0482654600301657, + "grad_norm": 0.234913554516001, + "learning_rate": 5.8821616500963396e-05, + "loss": 0.4146, + "step": 5432 + }, + { + "epoch": 2.0486425339366514, + "grad_norm": 0.2554542317000975, + "learning_rate": 5.881232450102848e-05, + "loss": 0.4152, + "step": 5433 + }, + { + "epoch": 2.049019607843137, + "grad_norm": 0.2887647043044569, + "learning_rate": 5.880303119741616e-05, + "loss": 0.4258, + "step": 5434 + }, + { + "epoch": 2.0493966817496228, + "grad_norm": 0.2997149279815265, + "learning_rate": 5.879373659077047e-05, + "loss": 0.4097, + "step": 5435 + }, + { + "epoch": 2.0497737556561084, + "grad_norm": 0.2788285729052227, + "learning_rate": 5.878444068173549e-05, + "loss": 0.4138, + "step": 5436 + }, + { + "epoch": 2.050150829562594, + "grad_norm": 0.25177931135407894, + "learning_rate": 5.877514347095544e-05, + "loss": 0.424, + "step": 5437 + }, + { + "epoch": 2.0505279034690798, + "grad_norm": 0.24794273296402572, + "learning_rate": 5.8765844959074596e-05, + "loss": 0.4317, + "step": 5438 + }, + { + "epoch": 2.0509049773755654, + "grad_norm": 0.19730989567510598, + "learning_rate": 5.875654514673734e-05, + "loss": 0.4143, + "step": 5439 + }, + { + "epoch": 2.051282051282051, + "grad_norm": 0.19254455399746873, + "learning_rate": 5.874724403458815e-05, + "loss": 0.4191, + "step": 5440 + }, + { + "epoch": 2.0516591251885368, + "grad_norm": 0.24022733392708256, + "learning_rate": 5.873794162327159e-05, + "loss": 0.4049, + "step": 5441 + }, + { + "epoch": 2.0520361990950224, + "grad_norm": 0.2744238776779503, + "learning_rate": 5.8728637913432267e-05, + "loss": 0.4241, + "step": 5442 + }, + { + "epoch": 2.052413273001508, + "grad_norm": 0.28932097712948757, + "learning_rate": 5.871933290571496e-05, + "loss": 0.414, + "step": 5443 + }, + { + "epoch": 2.052790346907994, + "grad_norm": 0.27934932091003284, + "learning_rate": 5.87100266007645e-05, + "loss": 0.4244, + "step": 5444 + }, + { + "epoch": 2.0531674208144794, + "grad_norm": 0.25717094320344464, + "learning_rate": 5.8700718999225775e-05, + "loss": 0.4026, + "step": 5445 + }, + { + "epoch": 2.053544494720965, + "grad_norm": 0.22504041333689168, + "learning_rate": 5.8691410101743815e-05, + "loss": 0.4287, + "step": 5446 + }, + { + "epoch": 2.053921568627451, + "grad_norm": 0.24924608536460194, + "learning_rate": 5.8682099908963716e-05, + "loss": 0.4173, + "step": 5447 + }, + { + "epoch": 2.0542986425339365, + "grad_norm": 0.267706798629298, + "learning_rate": 5.867278842153067e-05, + "loss": 0.4087, + "step": 5448 + }, + { + "epoch": 2.054675716440422, + "grad_norm": 0.25441755319111464, + "learning_rate": 5.866347564008994e-05, + "loss": 0.4192, + "step": 5449 + }, + { + "epoch": 2.055052790346908, + "grad_norm": 0.2499388843944412, + "learning_rate": 5.8654161565286904e-05, + "loss": 0.4197, + "step": 5450 + }, + { + "epoch": 2.0554298642533935, + "grad_norm": 0.34207753676159225, + "learning_rate": 5.8644846197767025e-05, + "loss": 0.4321, + "step": 5451 + }, + { + "epoch": 2.055806938159879, + "grad_norm": 0.4057798537985801, + "learning_rate": 5.863552953817584e-05, + "loss": 0.4285, + "step": 5452 + }, + { + "epoch": 2.056184012066365, + "grad_norm": 0.32738388480781694, + "learning_rate": 5.8626211587158985e-05, + "loss": 0.4145, + "step": 5453 + }, + { + "epoch": 2.0565610859728505, + "grad_norm": 0.22809007608924453, + "learning_rate": 5.861689234536219e-05, + "loss": 0.4111, + "step": 5454 + }, + { + "epoch": 2.056938159879336, + "grad_norm": 0.23005723982501208, + "learning_rate": 5.8607571813431264e-05, + "loss": 0.4232, + "step": 5455 + }, + { + "epoch": 2.057315233785822, + "grad_norm": 0.3140045840838473, + "learning_rate": 5.859824999201213e-05, + "loss": 0.4241, + "step": 5456 + }, + { + "epoch": 2.0576923076923075, + "grad_norm": 0.34031304728327777, + "learning_rate": 5.858892688175075e-05, + "loss": 0.4162, + "step": 5457 + }, + { + "epoch": 2.058069381598793, + "grad_norm": 0.2898725287757458, + "learning_rate": 5.857960248329323e-05, + "loss": 0.4134, + "step": 5458 + }, + { + "epoch": 2.058446455505279, + "grad_norm": 0.22828345668677097, + "learning_rate": 5.8570276797285745e-05, + "loss": 0.4203, + "step": 5459 + }, + { + "epoch": 2.0588235294117645, + "grad_norm": 0.23816161957511112, + "learning_rate": 5.856094982437454e-05, + "loss": 0.4141, + "step": 5460 + }, + { + "epoch": 2.05920060331825, + "grad_norm": 0.2935523155404572, + "learning_rate": 5.8551621565205984e-05, + "loss": 0.4221, + "step": 5461 + }, + { + "epoch": 2.059577677224736, + "grad_norm": 0.3266524212581728, + "learning_rate": 5.8542292020426506e-05, + "loss": 0.439, + "step": 5462 + }, + { + "epoch": 2.0599547511312215, + "grad_norm": 0.31131356104803487, + "learning_rate": 5.8532961190682644e-05, + "loss": 0.4209, + "step": 5463 + }, + { + "epoch": 2.060331825037707, + "grad_norm": 0.316851145444407, + "learning_rate": 5.8523629076621e-05, + "loss": 0.4245, + "step": 5464 + }, + { + "epoch": 2.060708898944193, + "grad_norm": 0.32739365219621414, + "learning_rate": 5.851429567888829e-05, + "loss": 0.434, + "step": 5465 + }, + { + "epoch": 2.0610859728506785, + "grad_norm": 0.2708724747424875, + "learning_rate": 5.850496099813133e-05, + "loss": 0.4135, + "step": 5466 + }, + { + "epoch": 2.061463046757164, + "grad_norm": 0.22508497406772685, + "learning_rate": 5.849562503499697e-05, + "loss": 0.4186, + "step": 5467 + }, + { + "epoch": 2.06184012066365, + "grad_norm": 0.21510531194983137, + "learning_rate": 5.84862877901322e-05, + "loss": 0.4157, + "step": 5468 + }, + { + "epoch": 2.0622171945701355, + "grad_norm": 0.21202927447914338, + "learning_rate": 5.84769492641841e-05, + "loss": 0.4319, + "step": 5469 + }, + { + "epoch": 2.062594268476621, + "grad_norm": 0.2092378072097199, + "learning_rate": 5.846760945779979e-05, + "loss": 0.4185, + "step": 5470 + }, + { + "epoch": 2.062971342383107, + "grad_norm": 0.191808866546014, + "learning_rate": 5.8458268371626525e-05, + "loss": 0.4168, + "step": 5471 + }, + { + "epoch": 2.0633484162895925, + "grad_norm": 0.21641164827745665, + "learning_rate": 5.844892600631165e-05, + "loss": 0.4196, + "step": 5472 + }, + { + "epoch": 2.063725490196078, + "grad_norm": 0.26281250154833136, + "learning_rate": 5.843958236250254e-05, + "loss": 0.4162, + "step": 5473 + }, + { + "epoch": 2.064102564102564, + "grad_norm": 0.3007569656736331, + "learning_rate": 5.843023744084675e-05, + "loss": 0.4178, + "step": 5474 + }, + { + "epoch": 2.06447963800905, + "grad_norm": 0.2733935827567231, + "learning_rate": 5.842089124199184e-05, + "loss": 0.4231, + "step": 5475 + }, + { + "epoch": 2.0648567119155357, + "grad_norm": 0.2851306891620789, + "learning_rate": 5.841154376658552e-05, + "loss": 0.4164, + "step": 5476 + }, + { + "epoch": 2.0652337858220213, + "grad_norm": 0.27620903878328196, + "learning_rate": 5.8402195015275535e-05, + "loss": 0.4216, + "step": 5477 + }, + { + "epoch": 2.065610859728507, + "grad_norm": 0.31095717136029694, + "learning_rate": 5.839284498870977e-05, + "loss": 0.4161, + "step": 5478 + }, + { + "epoch": 2.0659879336349927, + "grad_norm": 0.28318201581012387, + "learning_rate": 5.838349368753616e-05, + "loss": 0.417, + "step": 5479 + }, + { + "epoch": 2.0663650075414783, + "grad_norm": 0.28595390832332646, + "learning_rate": 5.8374141112402735e-05, + "loss": 0.4203, + "step": 5480 + }, + { + "epoch": 2.066742081447964, + "grad_norm": 0.2904690311447995, + "learning_rate": 5.836478726395763e-05, + "loss": 0.4174, + "step": 5481 + }, + { + "epoch": 2.0671191553544497, + "grad_norm": 0.28216036866854544, + "learning_rate": 5.8355432142849065e-05, + "loss": 0.4146, + "step": 5482 + }, + { + "epoch": 2.0674962292609353, + "grad_norm": 0.26237020739050504, + "learning_rate": 5.8346075749725326e-05, + "loss": 0.4271, + "step": 5483 + }, + { + "epoch": 2.067873303167421, + "grad_norm": 0.26601985367709474, + "learning_rate": 5.8336718085234814e-05, + "loss": 0.4232, + "step": 5484 + }, + { + "epoch": 2.0682503770739067, + "grad_norm": 0.27462790973637846, + "learning_rate": 5.8327359150026e-05, + "loss": 0.4087, + "step": 5485 + }, + { + "epoch": 2.0686274509803924, + "grad_norm": 0.2340754217367879, + "learning_rate": 5.8317998944747455e-05, + "loss": 0.4255, + "step": 5486 + }, + { + "epoch": 2.069004524886878, + "grad_norm": 0.24429732865662268, + "learning_rate": 5.8308637470047836e-05, + "loss": 0.4203, + "step": 5487 + }, + { + "epoch": 2.0693815987933637, + "grad_norm": 0.28053816559931966, + "learning_rate": 5.829927472657588e-05, + "loss": 0.4209, + "step": 5488 + }, + { + "epoch": 2.0697586726998494, + "grad_norm": 0.32216536783805416, + "learning_rate": 5.82899107149804e-05, + "loss": 0.4191, + "step": 5489 + }, + { + "epoch": 2.070135746606335, + "grad_norm": 0.3107611475800745, + "learning_rate": 5.8280545435910356e-05, + "loss": 0.4215, + "step": 5490 + }, + { + "epoch": 2.0705128205128207, + "grad_norm": 0.29937743682188483, + "learning_rate": 5.827117889001472e-05, + "loss": 0.4267, + "step": 5491 + }, + { + "epoch": 2.0708898944193064, + "grad_norm": 0.2866905238365436, + "learning_rate": 5.826181107794258e-05, + "loss": 0.4082, + "step": 5492 + }, + { + "epoch": 2.071266968325792, + "grad_norm": 0.3551196547557424, + "learning_rate": 5.825244200034314e-05, + "loss": 0.4154, + "step": 5493 + }, + { + "epoch": 2.0716440422322777, + "grad_norm": 0.3799394913115776, + "learning_rate": 5.824307165786567e-05, + "loss": 0.4234, + "step": 5494 + }, + { + "epoch": 2.0720211161387634, + "grad_norm": 0.33401452570370077, + "learning_rate": 5.8233700051159494e-05, + "loss": 0.4122, + "step": 5495 + }, + { + "epoch": 2.072398190045249, + "grad_norm": 0.25573660100158435, + "learning_rate": 5.82243271808741e-05, + "loss": 0.424, + "step": 5496 + }, + { + "epoch": 2.0727752639517347, + "grad_norm": 0.34484012258915386, + "learning_rate": 5.821495304765897e-05, + "loss": 0.406, + "step": 5497 + }, + { + "epoch": 2.0731523378582204, + "grad_norm": 0.30090864677511564, + "learning_rate": 5.820557765216377e-05, + "loss": 0.4145, + "step": 5498 + }, + { + "epoch": 2.073529411764706, + "grad_norm": 0.2640148000228213, + "learning_rate": 5.819620099503818e-05, + "loss": 0.4121, + "step": 5499 + }, + { + "epoch": 2.0739064856711917, + "grad_norm": 0.26525553321969003, + "learning_rate": 5.818682307693201e-05, + "loss": 0.4207, + "step": 5500 + }, + { + "epoch": 2.0742835595776774, + "grad_norm": 0.28246130997983043, + "learning_rate": 5.8177443898495126e-05, + "loss": 0.4171, + "step": 5501 + }, + { + "epoch": 2.074660633484163, + "grad_norm": 0.2681267749485559, + "learning_rate": 5.8168063460377506e-05, + "loss": 0.4207, + "step": 5502 + }, + { + "epoch": 2.0750377073906487, + "grad_norm": 0.20896747638130542, + "learning_rate": 5.815868176322919e-05, + "loss": 0.4186, + "step": 5503 + }, + { + "epoch": 2.0754147812971344, + "grad_norm": 0.29034406652942685, + "learning_rate": 5.814929880770035e-05, + "loss": 0.4304, + "step": 5504 + }, + { + "epoch": 2.07579185520362, + "grad_norm": 0.2696933536782547, + "learning_rate": 5.8139914594441204e-05, + "loss": 0.4097, + "step": 5505 + }, + { + "epoch": 2.0761689291101058, + "grad_norm": 0.2418870114776767, + "learning_rate": 5.813052912410206e-05, + "loss": 0.419, + "step": 5506 + }, + { + "epoch": 2.0765460030165914, + "grad_norm": 0.28910719951731384, + "learning_rate": 5.8121142397333326e-05, + "loss": 0.4162, + "step": 5507 + }, + { + "epoch": 2.076923076923077, + "grad_norm": 0.3213417470291067, + "learning_rate": 5.8111754414785504e-05, + "loss": 0.4265, + "step": 5508 + }, + { + "epoch": 2.0773001508295628, + "grad_norm": 0.3465815726760489, + "learning_rate": 5.810236517710916e-05, + "loss": 0.4116, + "step": 5509 + }, + { + "epoch": 2.0776772247360484, + "grad_norm": 0.38823605314159537, + "learning_rate": 5.8092974684954965e-05, + "loss": 0.4228, + "step": 5510 + }, + { + "epoch": 2.078054298642534, + "grad_norm": 0.37582569306111147, + "learning_rate": 5.8083582938973676e-05, + "loss": 0.4126, + "step": 5511 + }, + { + "epoch": 2.0784313725490198, + "grad_norm": 0.2967014668753079, + "learning_rate": 5.8074189939816134e-05, + "loss": 0.4164, + "step": 5512 + }, + { + "epoch": 2.0788084464555054, + "grad_norm": 0.2378741291361747, + "learning_rate": 5.806479568813325e-05, + "loss": 0.4175, + "step": 5513 + }, + { + "epoch": 2.079185520361991, + "grad_norm": 0.23555553687567374, + "learning_rate": 5.805540018457606e-05, + "loss": 0.4122, + "step": 5514 + }, + { + "epoch": 2.079562594268477, + "grad_norm": 0.279316368174325, + "learning_rate": 5.804600342979564e-05, + "loss": 0.4131, + "step": 5515 + }, + { + "epoch": 2.0799396681749625, + "grad_norm": 0.3058123886075919, + "learning_rate": 5.80366054244432e-05, + "loss": 0.4109, + "step": 5516 + }, + { + "epoch": 2.080316742081448, + "grad_norm": 0.31281694804199356, + "learning_rate": 5.8027206169170005e-05, + "loss": 0.4149, + "step": 5517 + }, + { + "epoch": 2.080693815987934, + "grad_norm": 0.30578574544403725, + "learning_rate": 5.80178056646274e-05, + "loss": 0.4213, + "step": 5518 + }, + { + "epoch": 2.0810708898944195, + "grad_norm": 0.31032679773017874, + "learning_rate": 5.800840391146685e-05, + "loss": 0.4106, + "step": 5519 + }, + { + "epoch": 2.081447963800905, + "grad_norm": 0.3059451720855536, + "learning_rate": 5.799900091033988e-05, + "loss": 0.4226, + "step": 5520 + }, + { + "epoch": 2.081825037707391, + "grad_norm": 0.3098903570095159, + "learning_rate": 5.798959666189812e-05, + "loss": 0.4203, + "step": 5521 + }, + { + "epoch": 2.0822021116138765, + "grad_norm": 0.3169252145188887, + "learning_rate": 5.798019116679326e-05, + "loss": 0.4165, + "step": 5522 + }, + { + "epoch": 2.082579185520362, + "grad_norm": 0.31343982079158983, + "learning_rate": 5.79707844256771e-05, + "loss": 0.421, + "step": 5523 + }, + { + "epoch": 2.082956259426848, + "grad_norm": 0.24690150222325094, + "learning_rate": 5.796137643920153e-05, + "loss": 0.4205, + "step": 5524 + }, + { + "epoch": 2.0833333333333335, + "grad_norm": 0.23386650831185196, + "learning_rate": 5.79519672080185e-05, + "loss": 0.4121, + "step": 5525 + }, + { + "epoch": 2.083710407239819, + "grad_norm": 0.2948401060023413, + "learning_rate": 5.794255673278006e-05, + "loss": 0.4139, + "step": 5526 + }, + { + "epoch": 2.084087481146305, + "grad_norm": 0.35711398586105453, + "learning_rate": 5.793314501413836e-05, + "loss": 0.4244, + "step": 5527 + }, + { + "epoch": 2.0844645550527905, + "grad_norm": 0.3275456315663045, + "learning_rate": 5.792373205274561e-05, + "loss": 0.4157, + "step": 5528 + }, + { + "epoch": 2.084841628959276, + "grad_norm": 0.3084580796876282, + "learning_rate": 5.7914317849254134e-05, + "loss": 0.4233, + "step": 5529 + }, + { + "epoch": 2.085218702865762, + "grad_norm": 0.31918171381328886, + "learning_rate": 5.790490240431631e-05, + "loss": 0.426, + "step": 5530 + }, + { + "epoch": 2.0855957767722475, + "grad_norm": 0.33518180911347434, + "learning_rate": 5.789548571858464e-05, + "loss": 0.4293, + "step": 5531 + }, + { + "epoch": 2.085972850678733, + "grad_norm": 0.3438382587236895, + "learning_rate": 5.7886067792711685e-05, + "loss": 0.4077, + "step": 5532 + }, + { + "epoch": 2.086349924585219, + "grad_norm": 0.27508713696784987, + "learning_rate": 5.7876648627350086e-05, + "loss": 0.4177, + "step": 5533 + }, + { + "epoch": 2.0867269984917045, + "grad_norm": 0.21647641285568126, + "learning_rate": 5.786722822315261e-05, + "loss": 0.4206, + "step": 5534 + }, + { + "epoch": 2.08710407239819, + "grad_norm": 0.21829082476677836, + "learning_rate": 5.7857806580772054e-05, + "loss": 0.4239, + "step": 5535 + }, + { + "epoch": 2.087481146304676, + "grad_norm": 0.2386233545631551, + "learning_rate": 5.7848383700861346e-05, + "loss": 0.4181, + "step": 5536 + }, + { + "epoch": 2.0878582202111615, + "grad_norm": 0.22935233006101574, + "learning_rate": 5.783895958407348e-05, + "loss": 0.4169, + "step": 5537 + }, + { + "epoch": 2.088235294117647, + "grad_norm": 0.26099144615542147, + "learning_rate": 5.782953423106154e-05, + "loss": 0.4218, + "step": 5538 + }, + { + "epoch": 2.088612368024133, + "grad_norm": 0.27900345372549096, + "learning_rate": 5.782010764247869e-05, + "loss": 0.4185, + "step": 5539 + }, + { + "epoch": 2.0889894419306185, + "grad_norm": 0.30944412555168843, + "learning_rate": 5.781067981897818e-05, + "loss": 0.4194, + "step": 5540 + }, + { + "epoch": 2.089366515837104, + "grad_norm": 0.31186120182978316, + "learning_rate": 5.7801250761213366e-05, + "loss": 0.4158, + "step": 5541 + }, + { + "epoch": 2.08974358974359, + "grad_norm": 0.32028537127116596, + "learning_rate": 5.7791820469837655e-05, + "loss": 0.4083, + "step": 5542 + }, + { + "epoch": 2.0901206636500755, + "grad_norm": 0.32222809592208795, + "learning_rate": 5.778238894550457e-05, + "loss": 0.4201, + "step": 5543 + }, + { + "epoch": 2.090497737556561, + "grad_norm": 0.328178798832803, + "learning_rate": 5.777295618886771e-05, + "loss": 0.4187, + "step": 5544 + }, + { + "epoch": 2.090874811463047, + "grad_norm": 0.3453863276485141, + "learning_rate": 5.7763522200580746e-05, + "loss": 0.4218, + "step": 5545 + }, + { + "epoch": 2.0912518853695325, + "grad_norm": 0.3373155967279912, + "learning_rate": 5.775408698129745e-05, + "loss": 0.4319, + "step": 5546 + }, + { + "epoch": 2.091628959276018, + "grad_norm": 0.3545640253399273, + "learning_rate": 5.774465053167167e-05, + "loss": 0.4291, + "step": 5547 + }, + { + "epoch": 2.092006033182504, + "grad_norm": 0.3389323281757615, + "learning_rate": 5.7735212852357364e-05, + "loss": 0.4202, + "step": 5548 + }, + { + "epoch": 2.0923831070889896, + "grad_norm": 0.3213205954363844, + "learning_rate": 5.772577394400852e-05, + "loss": 0.4266, + "step": 5549 + }, + { + "epoch": 2.0927601809954752, + "grad_norm": 0.3500626781113008, + "learning_rate": 5.771633380727928e-05, + "loss": 0.4308, + "step": 5550 + }, + { + "epoch": 2.093137254901961, + "grad_norm": 0.383367101903558, + "learning_rate": 5.77068924428238e-05, + "loss": 0.4133, + "step": 5551 + }, + { + "epoch": 2.0935143288084466, + "grad_norm": 0.36014958081000303, + "learning_rate": 5.76974498512964e-05, + "loss": 0.4251, + "step": 5552 + }, + { + "epoch": 2.0938914027149322, + "grad_norm": 0.2552509282043228, + "learning_rate": 5.768800603335141e-05, + "loss": 0.4168, + "step": 5553 + }, + { + "epoch": 2.094268476621418, + "grad_norm": 0.20008477025024832, + "learning_rate": 5.767856098964331e-05, + "loss": 0.421, + "step": 5554 + }, + { + "epoch": 2.0946455505279036, + "grad_norm": 0.2011844639582152, + "learning_rate": 5.766911472082659e-05, + "loss": 0.4201, + "step": 5555 + }, + { + "epoch": 2.0950226244343892, + "grad_norm": 0.2566131663633258, + "learning_rate": 5.7659667227555905e-05, + "loss": 0.4104, + "step": 5556 + }, + { + "epoch": 2.095399698340875, + "grad_norm": 0.2801598052429333, + "learning_rate": 5.765021851048595e-05, + "loss": 0.4215, + "step": 5557 + }, + { + "epoch": 2.0957767722473606, + "grad_norm": 0.21678030782640015, + "learning_rate": 5.7640768570271496e-05, + "loss": 0.4229, + "step": 5558 + }, + { + "epoch": 2.0961538461538463, + "grad_norm": 0.20766179122296646, + "learning_rate": 5.763131740756743e-05, + "loss": 0.417, + "step": 5559 + }, + { + "epoch": 2.096530920060332, + "grad_norm": 0.2637381658769069, + "learning_rate": 5.7621865023028714e-05, + "loss": 0.421, + "step": 5560 + }, + { + "epoch": 2.0969079939668176, + "grad_norm": 0.2874014392005131, + "learning_rate": 5.761241141731038e-05, + "loss": 0.4103, + "step": 5561 + }, + { + "epoch": 2.0972850678733033, + "grad_norm": 0.26190023699471926, + "learning_rate": 5.760295659106755e-05, + "loss": 0.4175, + "step": 5562 + }, + { + "epoch": 2.097662141779789, + "grad_norm": 0.26070411739835525, + "learning_rate": 5.7593500544955455e-05, + "loss": 0.4146, + "step": 5563 + }, + { + "epoch": 2.0980392156862746, + "grad_norm": 0.2395617928129469, + "learning_rate": 5.758404327962936e-05, + "loss": 0.4279, + "step": 5564 + }, + { + "epoch": 2.0984162895927603, + "grad_norm": 0.18446172921941828, + "learning_rate": 5.7574584795744674e-05, + "loss": 0.394, + "step": 5565 + }, + { + "epoch": 2.098793363499246, + "grad_norm": 0.17210582034139704, + "learning_rate": 5.756512509395686e-05, + "loss": 0.4179, + "step": 5566 + }, + { + "epoch": 2.0991704374057316, + "grad_norm": 0.21210016776582172, + "learning_rate": 5.755566417492144e-05, + "loss": 0.4179, + "step": 5567 + }, + { + "epoch": 2.0995475113122173, + "grad_norm": 0.21307396206215598, + "learning_rate": 5.754620203929406e-05, + "loss": 0.4091, + "step": 5568 + }, + { + "epoch": 2.099924585218703, + "grad_norm": 0.21045833116881585, + "learning_rate": 5.753673868773046e-05, + "loss": 0.4267, + "step": 5569 + }, + { + "epoch": 2.1003016591251886, + "grad_norm": 0.21456415589823313, + "learning_rate": 5.752727412088642e-05, + "loss": 0.4292, + "step": 5570 + }, + { + "epoch": 2.1006787330316743, + "grad_norm": 0.19229489617955178, + "learning_rate": 5.751780833941782e-05, + "loss": 0.4147, + "step": 5571 + }, + { + "epoch": 2.10105580693816, + "grad_norm": 0.24067248749408476, + "learning_rate": 5.750834134398066e-05, + "loss": 0.42, + "step": 5572 + }, + { + "epoch": 2.1014328808446456, + "grad_norm": 0.26823236599052075, + "learning_rate": 5.749887313523095e-05, + "loss": 0.4015, + "step": 5573 + }, + { + "epoch": 2.1018099547511313, + "grad_norm": 0.26450047540379656, + "learning_rate": 5.7489403713824874e-05, + "loss": 0.4218, + "step": 5574 + }, + { + "epoch": 2.102187028657617, + "grad_norm": 0.30485083123058787, + "learning_rate": 5.747993308041862e-05, + "loss": 0.415, + "step": 5575 + }, + { + "epoch": 2.1025641025641026, + "grad_norm": 0.27678140087501724, + "learning_rate": 5.747046123566852e-05, + "loss": 0.4304, + "step": 5576 + }, + { + "epoch": 2.1029411764705883, + "grad_norm": 0.257707518828476, + "learning_rate": 5.746098818023093e-05, + "loss": 0.4152, + "step": 5577 + }, + { + "epoch": 2.103318250377074, + "grad_norm": 0.24250019004352638, + "learning_rate": 5.745151391476238e-05, + "loss": 0.4121, + "step": 5578 + }, + { + "epoch": 2.1036953242835597, + "grad_norm": 0.26183775362762496, + "learning_rate": 5.744203843991937e-05, + "loss": 0.4253, + "step": 5579 + }, + { + "epoch": 2.1040723981900453, + "grad_norm": 0.3036711146544233, + "learning_rate": 5.7432561756358575e-05, + "loss": 0.4085, + "step": 5580 + }, + { + "epoch": 2.104449472096531, + "grad_norm": 0.3290594742726747, + "learning_rate": 5.742308386473672e-05, + "loss": 0.4145, + "step": 5581 + }, + { + "epoch": 2.1048265460030167, + "grad_norm": 0.3779760961916899, + "learning_rate": 5.74136047657106e-05, + "loss": 0.4207, + "step": 5582 + }, + { + "epoch": 2.1052036199095023, + "grad_norm": 0.3772937831237461, + "learning_rate": 5.740412445993711e-05, + "loss": 0.4129, + "step": 5583 + }, + { + "epoch": 2.105580693815988, + "grad_norm": 0.34399807860847975, + "learning_rate": 5.739464294807324e-05, + "loss": 0.4182, + "step": 5584 + }, + { + "epoch": 2.1059577677224737, + "grad_norm": 0.2724085466875675, + "learning_rate": 5.738516023077604e-05, + "loss": 0.413, + "step": 5585 + }, + { + "epoch": 2.1063348416289593, + "grad_norm": 0.27582634734993633, + "learning_rate": 5.737567630870265e-05, + "loss": 0.4092, + "step": 5586 + }, + { + "epoch": 2.106711915535445, + "grad_norm": 0.34153538283483686, + "learning_rate": 5.736619118251032e-05, + "loss": 0.4257, + "step": 5587 + }, + { + "epoch": 2.1070889894419307, + "grad_norm": 0.32193882208039526, + "learning_rate": 5.7356704852856336e-05, + "loss": 0.404, + "step": 5588 + }, + { + "epoch": 2.1074660633484164, + "grad_norm": 0.2586780602598645, + "learning_rate": 5.73472173203981e-05, + "loss": 0.4301, + "step": 5589 + }, + { + "epoch": 2.107843137254902, + "grad_norm": 0.24450387684975314, + "learning_rate": 5.733772858579309e-05, + "loss": 0.4166, + "step": 5590 + }, + { + "epoch": 2.1082202111613877, + "grad_norm": 0.23763736504352292, + "learning_rate": 5.732823864969888e-05, + "loss": 0.4176, + "step": 5591 + }, + { + "epoch": 2.1085972850678734, + "grad_norm": 0.2664265222892081, + "learning_rate": 5.7318747512773085e-05, + "loss": 0.4117, + "step": 5592 + }, + { + "epoch": 2.108974358974359, + "grad_norm": 0.26694112666055764, + "learning_rate": 5.730925517567345e-05, + "loss": 0.421, + "step": 5593 + }, + { + "epoch": 2.1093514328808447, + "grad_norm": 0.2656978885858578, + "learning_rate": 5.72997616390578e-05, + "loss": 0.4182, + "step": 5594 + }, + { + "epoch": 2.1097285067873304, + "grad_norm": 0.25781684327714177, + "learning_rate": 5.7290266903583994e-05, + "loss": 0.4139, + "step": 5595 + }, + { + "epoch": 2.110105580693816, + "grad_norm": 0.22995428100270485, + "learning_rate": 5.728077096991004e-05, + "loss": 0.4155, + "step": 5596 + }, + { + "epoch": 2.1104826546003017, + "grad_norm": 0.1940205182659553, + "learning_rate": 5.7271273838693984e-05, + "loss": 0.4248, + "step": 5597 + }, + { + "epoch": 2.1108597285067874, + "grad_norm": 0.17064800851290987, + "learning_rate": 5.7261775510593974e-05, + "loss": 0.4122, + "step": 5598 + }, + { + "epoch": 2.111236802413273, + "grad_norm": 0.21354562181924902, + "learning_rate": 5.7252275986268215e-05, + "loss": 0.4288, + "step": 5599 + }, + { + "epoch": 2.1116138763197587, + "grad_norm": 0.27372542164843744, + "learning_rate": 5.724277526637506e-05, + "loss": 0.4193, + "step": 5600 + }, + { + "epoch": 2.1119909502262444, + "grad_norm": 0.25966367636440973, + "learning_rate": 5.7233273351572855e-05, + "loss": 0.4257, + "step": 5601 + }, + { + "epoch": 2.11236802413273, + "grad_norm": 0.27833894468801923, + "learning_rate": 5.72237702425201e-05, + "loss": 0.43, + "step": 5602 + }, + { + "epoch": 2.1127450980392157, + "grad_norm": 0.34304749232094145, + "learning_rate": 5.7214265939875344e-05, + "loss": 0.4215, + "step": 5603 + }, + { + "epoch": 2.1131221719457014, + "grad_norm": 0.35356143529145406, + "learning_rate": 5.720476044429722e-05, + "loss": 0.4136, + "step": 5604 + }, + { + "epoch": 2.113499245852187, + "grad_norm": 0.35389060319256027, + "learning_rate": 5.719525375644446e-05, + "loss": 0.4258, + "step": 5605 + }, + { + "epoch": 2.1138763197586727, + "grad_norm": 0.3127844025548493, + "learning_rate": 5.7185745876975874e-05, + "loss": 0.4127, + "step": 5606 + }, + { + "epoch": 2.1142533936651584, + "grad_norm": 0.25536672992652076, + "learning_rate": 5.717623680655034e-05, + "loss": 0.429, + "step": 5607 + }, + { + "epoch": 2.114630467571644, + "grad_norm": 0.24418997160625283, + "learning_rate": 5.7166726545826826e-05, + "loss": 0.4178, + "step": 5608 + }, + { + "epoch": 2.1150075414781297, + "grad_norm": 0.21040126455614527, + "learning_rate": 5.71572150954644e-05, + "loss": 0.4181, + "step": 5609 + }, + { + "epoch": 2.1153846153846154, + "grad_norm": 0.2685586942434249, + "learning_rate": 5.714770245612217e-05, + "loss": 0.4307, + "step": 5610 + }, + { + "epoch": 2.115761689291101, + "grad_norm": 0.2845935507391867, + "learning_rate": 5.7138188628459383e-05, + "loss": 0.4251, + "step": 5611 + }, + { + "epoch": 2.1161387631975868, + "grad_norm": 0.28409650974972034, + "learning_rate": 5.7128673613135316e-05, + "loss": 0.4333, + "step": 5612 + }, + { + "epoch": 2.1165158371040724, + "grad_norm": 0.38433512603456493, + "learning_rate": 5.711915741080937e-05, + "loss": 0.4261, + "step": 5613 + }, + { + "epoch": 2.116892911010558, + "grad_norm": 0.4667743809439844, + "learning_rate": 5.7109640022140996e-05, + "loss": 0.4209, + "step": 5614 + }, + { + "epoch": 2.1172699849170438, + "grad_norm": 0.43484088421312844, + "learning_rate": 5.710012144778975e-05, + "loss": 0.4273, + "step": 5615 + }, + { + "epoch": 2.1176470588235294, + "grad_norm": 0.39086212156683176, + "learning_rate": 5.709060168841524e-05, + "loss": 0.4259, + "step": 5616 + }, + { + "epoch": 2.118024132730015, + "grad_norm": 0.3722608332120948, + "learning_rate": 5.7081080744677205e-05, + "loss": 0.4206, + "step": 5617 + }, + { + "epoch": 2.1184012066365008, + "grad_norm": 0.3464785790333316, + "learning_rate": 5.707155861723543e-05, + "loss": 0.4221, + "step": 5618 + }, + { + "epoch": 2.1187782805429864, + "grad_norm": 0.25832287135782817, + "learning_rate": 5.706203530674977e-05, + "loss": 0.4176, + "step": 5619 + }, + { + "epoch": 2.119155354449472, + "grad_norm": 0.2607370003804401, + "learning_rate": 5.705251081388021e-05, + "loss": 0.4241, + "step": 5620 + }, + { + "epoch": 2.119532428355958, + "grad_norm": 0.3958821855709329, + "learning_rate": 5.704298513928678e-05, + "loss": 0.4171, + "step": 5621 + }, + { + "epoch": 2.1199095022624435, + "grad_norm": 0.4004286107837725, + "learning_rate": 5.703345828362958e-05, + "loss": 0.4234, + "step": 5622 + }, + { + "epoch": 2.120286576168929, + "grad_norm": 0.31652423221055587, + "learning_rate": 5.702393024756883e-05, + "loss": 0.4135, + "step": 5623 + }, + { + "epoch": 2.120663650075415, + "grad_norm": 0.2554947400724556, + "learning_rate": 5.701440103176483e-05, + "loss": 0.4153, + "step": 5624 + }, + { + "epoch": 2.1210407239819005, + "grad_norm": 0.24474063178091224, + "learning_rate": 5.7004870636877925e-05, + "loss": 0.4095, + "step": 5625 + }, + { + "epoch": 2.121417797888386, + "grad_norm": 0.2836345483853234, + "learning_rate": 5.6995339063568555e-05, + "loss": 0.4191, + "step": 5626 + }, + { + "epoch": 2.121794871794872, + "grad_norm": 0.357267276803534, + "learning_rate": 5.6985806312497284e-05, + "loss": 0.4308, + "step": 5627 + }, + { + "epoch": 2.1221719457013575, + "grad_norm": 0.3650651608007004, + "learning_rate": 5.697627238432469e-05, + "loss": 0.4165, + "step": 5628 + }, + { + "epoch": 2.122549019607843, + "grad_norm": 0.3481808383485934, + "learning_rate": 5.6966737279711474e-05, + "loss": 0.4299, + "step": 5629 + }, + { + "epoch": 2.122926093514329, + "grad_norm": 0.30754140847330036, + "learning_rate": 5.695720099931842e-05, + "loss": 0.4149, + "step": 5630 + }, + { + "epoch": 2.1233031674208145, + "grad_norm": 0.29477276011153525, + "learning_rate": 5.694766354380637e-05, + "loss": 0.4122, + "step": 5631 + }, + { + "epoch": 2.1236802413273, + "grad_norm": 0.28354059519428004, + "learning_rate": 5.6938124913836276e-05, + "loss": 0.4273, + "step": 5632 + }, + { + "epoch": 2.124057315233786, + "grad_norm": 0.21976586937175385, + "learning_rate": 5.6928585110069155e-05, + "loss": 0.4181, + "step": 5633 + }, + { + "epoch": 2.1244343891402715, + "grad_norm": 0.23473835552438124, + "learning_rate": 5.691904413316609e-05, + "loss": 0.4119, + "step": 5634 + }, + { + "epoch": 2.124811463046757, + "grad_norm": 0.21720925716828685, + "learning_rate": 5.690950198378828e-05, + "loss": 0.4132, + "step": 5635 + }, + { + "epoch": 2.125188536953243, + "grad_norm": 0.30103730414695196, + "learning_rate": 5.689995866259699e-05, + "loss": 0.4254, + "step": 5636 + }, + { + "epoch": 2.1255656108597285, + "grad_norm": 0.3559456494587747, + "learning_rate": 5.689041417025354e-05, + "loss": 0.4187, + "step": 5637 + }, + { + "epoch": 2.125942684766214, + "grad_norm": 0.35072785199673967, + "learning_rate": 5.688086850741937e-05, + "loss": 0.4165, + "step": 5638 + }, + { + "epoch": 2.1263197586727, + "grad_norm": 0.30072311363238924, + "learning_rate": 5.6871321674756e-05, + "loss": 0.4193, + "step": 5639 + }, + { + "epoch": 2.1266968325791855, + "grad_norm": 0.2750971269089357, + "learning_rate": 5.686177367292499e-05, + "loss": 0.4242, + "step": 5640 + }, + { + "epoch": 2.127073906485671, + "grad_norm": 0.29485890805417014, + "learning_rate": 5.685222450258802e-05, + "loss": 0.4186, + "step": 5641 + }, + { + "epoch": 2.127450980392157, + "grad_norm": 0.30220260008355293, + "learning_rate": 5.684267416440685e-05, + "loss": 0.4287, + "step": 5642 + }, + { + "epoch": 2.1278280542986425, + "grad_norm": 0.3047394698368501, + "learning_rate": 5.683312265904329e-05, + "loss": 0.4144, + "step": 5643 + }, + { + "epoch": 2.128205128205128, + "grad_norm": 0.32501003524586586, + "learning_rate": 5.6823569987159266e-05, + "loss": 0.4259, + "step": 5644 + }, + { + "epoch": 2.128582202111614, + "grad_norm": 0.3327485553558704, + "learning_rate": 5.681401614941677e-05, + "loss": 0.4233, + "step": 5645 + }, + { + "epoch": 2.1289592760180995, + "grad_norm": 0.33120997491053944, + "learning_rate": 5.6804461146477856e-05, + "loss": 0.4285, + "step": 5646 + }, + { + "epoch": 2.129336349924585, + "grad_norm": 0.2843931812318769, + "learning_rate": 5.679490497900469e-05, + "loss": 0.4162, + "step": 5647 + }, + { + "epoch": 2.129713423831071, + "grad_norm": 0.25527402135109395, + "learning_rate": 5.678534764765951e-05, + "loss": 0.4229, + "step": 5648 + }, + { + "epoch": 2.1300904977375565, + "grad_norm": 0.22649727340305928, + "learning_rate": 5.677578915310464e-05, + "loss": 0.4185, + "step": 5649 + }, + { + "epoch": 2.130467571644042, + "grad_norm": 0.19177258983372483, + "learning_rate": 5.676622949600244e-05, + "loss": 0.4285, + "step": 5650 + }, + { + "epoch": 2.130844645550528, + "grad_norm": 0.25217336393915, + "learning_rate": 5.675666867701542e-05, + "loss": 0.4275, + "step": 5651 + }, + { + "epoch": 2.1312217194570136, + "grad_norm": 0.27660789028570976, + "learning_rate": 5.674710669680613e-05, + "loss": 0.4116, + "step": 5652 + }, + { + "epoch": 2.1315987933634992, + "grad_norm": 0.3112006281726905, + "learning_rate": 5.673754355603719e-05, + "loss": 0.4195, + "step": 5653 + }, + { + "epoch": 2.131975867269985, + "grad_norm": 0.3481372693083451, + "learning_rate": 5.6727979255371325e-05, + "loss": 0.4219, + "step": 5654 + }, + { + "epoch": 2.1323529411764706, + "grad_norm": 0.34094326632421734, + "learning_rate": 5.6718413795471346e-05, + "loss": 0.4419, + "step": 5655 + }, + { + "epoch": 2.1327300150829562, + "grad_norm": 0.27266038741143184, + "learning_rate": 5.670884717700011e-05, + "loss": 0.4197, + "step": 5656 + }, + { + "epoch": 2.133107088989442, + "grad_norm": 0.2532668852952511, + "learning_rate": 5.6699279400620594e-05, + "loss": 0.4225, + "step": 5657 + }, + { + "epoch": 2.1334841628959276, + "grad_norm": 0.2541388648448934, + "learning_rate": 5.668971046699582e-05, + "loss": 0.4156, + "step": 5658 + }, + { + "epoch": 2.1338612368024132, + "grad_norm": 0.2045978516564288, + "learning_rate": 5.668014037678892e-05, + "loss": 0.4165, + "step": 5659 + }, + { + "epoch": 2.134238310708899, + "grad_norm": 0.2209859039128265, + "learning_rate": 5.667056913066308e-05, + "loss": 0.4239, + "step": 5660 + }, + { + "epoch": 2.1346153846153846, + "grad_norm": 0.30074693913470213, + "learning_rate": 5.6660996729281606e-05, + "loss": 0.4227, + "step": 5661 + }, + { + "epoch": 2.1349924585218703, + "grad_norm": 0.35038626863734523, + "learning_rate": 5.665142317330782e-05, + "loss": 0.4263, + "step": 5662 + }, + { + "epoch": 2.135369532428356, + "grad_norm": 0.3293239622992472, + "learning_rate": 5.664184846340517e-05, + "loss": 0.4199, + "step": 5663 + }, + { + "epoch": 2.1357466063348416, + "grad_norm": 0.33871905143722497, + "learning_rate": 5.6632272600237206e-05, + "loss": 0.4233, + "step": 5664 + }, + { + "epoch": 2.1361236802413273, + "grad_norm": 0.31030831423606314, + "learning_rate": 5.662269558446748e-05, + "loss": 0.4124, + "step": 5665 + }, + { + "epoch": 2.136500754147813, + "grad_norm": 0.2664105519858822, + "learning_rate": 5.661311741675971e-05, + "loss": 0.4283, + "step": 5666 + }, + { + "epoch": 2.1368778280542986, + "grad_norm": 0.2653389260569212, + "learning_rate": 5.660353809777764e-05, + "loss": 0.417, + "step": 5667 + }, + { + "epoch": 2.1372549019607843, + "grad_norm": 0.2628147146031069, + "learning_rate": 5.6593957628185095e-05, + "loss": 0.4248, + "step": 5668 + }, + { + "epoch": 2.13763197586727, + "grad_norm": 0.24699446008861425, + "learning_rate": 5.6584376008646005e-05, + "loss": 0.4216, + "step": 5669 + }, + { + "epoch": 2.1380090497737556, + "grad_norm": 0.266025670705307, + "learning_rate": 5.657479323982438e-05, + "loss": 0.4123, + "step": 5670 + }, + { + "epoch": 2.1383861236802413, + "grad_norm": 0.33391916321003023, + "learning_rate": 5.656520932238428e-05, + "loss": 0.421, + "step": 5671 + }, + { + "epoch": 2.138763197586727, + "grad_norm": 0.3156363566444026, + "learning_rate": 5.655562425698987e-05, + "loss": 0.4237, + "step": 5672 + }, + { + "epoch": 2.1391402714932126, + "grad_norm": 0.23802497818527427, + "learning_rate": 5.654603804430539e-05, + "loss": 0.4187, + "step": 5673 + }, + { + "epoch": 2.1395173453996983, + "grad_norm": 0.2260244904937757, + "learning_rate": 5.653645068499513e-05, + "loss": 0.4168, + "step": 5674 + }, + { + "epoch": 2.139894419306184, + "grad_norm": 0.26450631845274764, + "learning_rate": 5.652686217972351e-05, + "loss": 0.4164, + "step": 5675 + }, + { + "epoch": 2.1402714932126696, + "grad_norm": 0.24143988083015988, + "learning_rate": 5.6517272529155006e-05, + "loss": 0.4196, + "step": 5676 + }, + { + "epoch": 2.1406485671191553, + "grad_norm": 0.24671759245063762, + "learning_rate": 5.6507681733954166e-05, + "loss": 0.4084, + "step": 5677 + }, + { + "epoch": 2.141025641025641, + "grad_norm": 0.2597034028782852, + "learning_rate": 5.649808979478563e-05, + "loss": 0.4185, + "step": 5678 + }, + { + "epoch": 2.1414027149321266, + "grad_norm": 0.27668858838787896, + "learning_rate": 5.648849671231409e-05, + "loss": 0.4346, + "step": 5679 + }, + { + "epoch": 2.1417797888386123, + "grad_norm": 0.21713905502703781, + "learning_rate": 5.647890248720437e-05, + "loss": 0.4161, + "step": 5680 + }, + { + "epoch": 2.142156862745098, + "grad_norm": 0.1902326364679508, + "learning_rate": 5.646930712012132e-05, + "loss": 0.4303, + "step": 5681 + }, + { + "epoch": 2.1425339366515836, + "grad_norm": 0.2110420583257873, + "learning_rate": 5.645971061172988e-05, + "loss": 0.4144, + "step": 5682 + }, + { + "epoch": 2.1429110105580693, + "grad_norm": 0.22029646926349536, + "learning_rate": 5.6450112962695116e-05, + "loss": 0.4124, + "step": 5683 + }, + { + "epoch": 2.143288084464555, + "grad_norm": 0.24937671840632764, + "learning_rate": 5.644051417368211e-05, + "loss": 0.4259, + "step": 5684 + }, + { + "epoch": 2.1436651583710407, + "grad_norm": 0.23262715495509678, + "learning_rate": 5.643091424535604e-05, + "loss": 0.41, + "step": 5685 + }, + { + "epoch": 2.1440422322775263, + "grad_norm": 0.22075636921818098, + "learning_rate": 5.6421313178382195e-05, + "loss": 0.4245, + "step": 5686 + }, + { + "epoch": 2.144419306184012, + "grad_norm": 0.24229996322952724, + "learning_rate": 5.6411710973425924e-05, + "loss": 0.4168, + "step": 5687 + }, + { + "epoch": 2.1447963800904977, + "grad_norm": 0.24417549557108953, + "learning_rate": 5.640210763115263e-05, + "loss": 0.4211, + "step": 5688 + }, + { + "epoch": 2.1451734539969833, + "grad_norm": 0.24363467006166833, + "learning_rate": 5.639250315222782e-05, + "loss": 0.4116, + "step": 5689 + }, + { + "epoch": 2.145550527903469, + "grad_norm": 0.30707546731906105, + "learning_rate": 5.638289753731709e-05, + "loss": 0.4256, + "step": 5690 + }, + { + "epoch": 2.1459276018099547, + "grad_norm": 0.350734812303871, + "learning_rate": 5.637329078708609e-05, + "loss": 0.4141, + "step": 5691 + }, + { + "epoch": 2.1463046757164403, + "grad_norm": 0.285346192269109, + "learning_rate": 5.636368290220056e-05, + "loss": 0.4226, + "step": 5692 + }, + { + "epoch": 2.146681749622926, + "grad_norm": 0.17316884184183864, + "learning_rate": 5.635407388332632e-05, + "loss": 0.4123, + "step": 5693 + }, + { + "epoch": 2.1470588235294117, + "grad_norm": 0.2841803375510086, + "learning_rate": 5.634446373112926e-05, + "loss": 0.4208, + "step": 5694 + }, + { + "epoch": 2.1474358974358974, + "grad_norm": 0.37595493217547993, + "learning_rate": 5.6334852446275365e-05, + "loss": 0.4129, + "step": 5695 + }, + { + "epoch": 2.147812971342383, + "grad_norm": 0.3467202110214958, + "learning_rate": 5.632524002943069e-05, + "loss": 0.4139, + "step": 5696 + }, + { + "epoch": 2.1481900452488687, + "grad_norm": 0.2748248663554704, + "learning_rate": 5.631562648126137e-05, + "loss": 0.4253, + "step": 5697 + }, + { + "epoch": 2.1485671191553544, + "grad_norm": 0.24449322452103706, + "learning_rate": 5.630601180243359e-05, + "loss": 0.4371, + "step": 5698 + }, + { + "epoch": 2.14894419306184, + "grad_norm": 0.20865740768896127, + "learning_rate": 5.629639599361366e-05, + "loss": 0.4132, + "step": 5699 + }, + { + "epoch": 2.1493212669683257, + "grad_norm": 0.2406454539799824, + "learning_rate": 5.628677905546795e-05, + "loss": 0.4229, + "step": 5700 + }, + { + "epoch": 2.1496983408748114, + "grad_norm": 0.2650699781033578, + "learning_rate": 5.627716098866288e-05, + "loss": 0.4054, + "step": 5701 + }, + { + "epoch": 2.150075414781297, + "grad_norm": 0.32417589466465463, + "learning_rate": 5.626754179386501e-05, + "loss": 0.4089, + "step": 5702 + }, + { + "epoch": 2.1504524886877827, + "grad_norm": 0.35669759162169085, + "learning_rate": 5.625792147174092e-05, + "loss": 0.4101, + "step": 5703 + }, + { + "epoch": 2.1508295625942684, + "grad_norm": 0.2834180953600253, + "learning_rate": 5.624830002295728e-05, + "loss": 0.4133, + "step": 5704 + }, + { + "epoch": 2.151206636500754, + "grad_norm": 0.2593561297239729, + "learning_rate": 5.6238677448180865e-05, + "loss": 0.4238, + "step": 5705 + }, + { + "epoch": 2.1515837104072397, + "grad_norm": 0.34513175467166946, + "learning_rate": 5.6229053748078525e-05, + "loss": 0.4198, + "step": 5706 + }, + { + "epoch": 2.1519607843137254, + "grad_norm": 0.320690738098826, + "learning_rate": 5.621942892331714e-05, + "loss": 0.4218, + "step": 5707 + }, + { + "epoch": 2.152337858220211, + "grad_norm": 0.27653139997068055, + "learning_rate": 5.62098029745637e-05, + "loss": 0.4111, + "step": 5708 + }, + { + "epoch": 2.1527149321266967, + "grad_norm": 0.296056245368014, + "learning_rate": 5.620017590248531e-05, + "loss": 0.4138, + "step": 5709 + }, + { + "epoch": 2.1530920060331824, + "grad_norm": 0.3259187634208203, + "learning_rate": 5.619054770774911e-05, + "loss": 0.4153, + "step": 5710 + }, + { + "epoch": 2.153469079939668, + "grad_norm": 0.3771410243309578, + "learning_rate": 5.6180918391022296e-05, + "loss": 0.4302, + "step": 5711 + }, + { + "epoch": 2.1538461538461537, + "grad_norm": 0.4151330556859276, + "learning_rate": 5.61712879529722e-05, + "loss": 0.4228, + "step": 5712 + }, + { + "epoch": 2.1542232277526394, + "grad_norm": 0.385281783393806, + "learning_rate": 5.616165639426619e-05, + "loss": 0.4172, + "step": 5713 + }, + { + "epoch": 2.154600301659125, + "grad_norm": 0.3202299862990286, + "learning_rate": 5.615202371557173e-05, + "loss": 0.427, + "step": 5714 + }, + { + "epoch": 2.1549773755656108, + "grad_norm": 0.24882930430330827, + "learning_rate": 5.614238991755635e-05, + "loss": 0.4209, + "step": 5715 + }, + { + "epoch": 2.1553544494720964, + "grad_norm": 0.2860943747937665, + "learning_rate": 5.613275500088767e-05, + "loss": 0.423, + "step": 5716 + }, + { + "epoch": 2.155731523378582, + "grad_norm": 0.3112652710701886, + "learning_rate": 5.612311896623338e-05, + "loss": 0.4114, + "step": 5717 + }, + { + "epoch": 2.1561085972850678, + "grad_norm": 0.3310388282124181, + "learning_rate": 5.611348181426124e-05, + "loss": 0.424, + "step": 5718 + }, + { + "epoch": 2.1564856711915534, + "grad_norm": 0.4298263624321115, + "learning_rate": 5.610384354563911e-05, + "loss": 0.4237, + "step": 5719 + }, + { + "epoch": 2.156862745098039, + "grad_norm": 0.42380830040703305, + "learning_rate": 5.60942041610349e-05, + "loss": 0.4186, + "step": 5720 + }, + { + "epoch": 2.1572398190045248, + "grad_norm": 0.34991216363186184, + "learning_rate": 5.608456366111663e-05, + "loss": 0.4192, + "step": 5721 + }, + { + "epoch": 2.1576168929110104, + "grad_norm": 0.3105761445748344, + "learning_rate": 5.607492204655235e-05, + "loss": 0.4213, + "step": 5722 + }, + { + "epoch": 2.157993966817496, + "grad_norm": 0.3409828687260544, + "learning_rate": 5.6065279318010235e-05, + "loss": 0.441, + "step": 5723 + }, + { + "epoch": 2.158371040723982, + "grad_norm": 0.4191708219726606, + "learning_rate": 5.605563547615851e-05, + "loss": 0.423, + "step": 5724 + }, + { + "epoch": 2.1587481146304675, + "grad_norm": 0.38154268889633675, + "learning_rate": 5.60459905216655e-05, + "loss": 0.429, + "step": 5725 + }, + { + "epoch": 2.159125188536953, + "grad_norm": 0.2672476113269887, + "learning_rate": 5.603634445519957e-05, + "loss": 0.4178, + "step": 5726 + }, + { + "epoch": 2.159502262443439, + "grad_norm": 0.23997769638739389, + "learning_rate": 5.602669727742919e-05, + "loss": 0.4323, + "step": 5727 + }, + { + "epoch": 2.1598793363499245, + "grad_norm": 0.33983864001114633, + "learning_rate": 5.601704898902292e-05, + "loss": 0.4222, + "step": 5728 + }, + { + "epoch": 2.16025641025641, + "grad_norm": 0.3531635959711142, + "learning_rate": 5.600739959064935e-05, + "loss": 0.416, + "step": 5729 + }, + { + "epoch": 2.160633484162896, + "grad_norm": 0.34160080079892946, + "learning_rate": 5.5997749082977185e-05, + "loss": 0.4169, + "step": 5730 + }, + { + "epoch": 2.1610105580693815, + "grad_norm": 0.3108005491269098, + "learning_rate": 5.5988097466675206e-05, + "loss": 0.425, + "step": 5731 + }, + { + "epoch": 2.161387631975867, + "grad_norm": 0.23902200614167468, + "learning_rate": 5.597844474241225e-05, + "loss": 0.4246, + "step": 5732 + }, + { + "epoch": 2.161764705882353, + "grad_norm": 0.19284239815057377, + "learning_rate": 5.596879091085724e-05, + "loss": 0.4091, + "step": 5733 + }, + { + "epoch": 2.1621417797888385, + "grad_norm": 0.20473139808785049, + "learning_rate": 5.595913597267919e-05, + "loss": 0.4151, + "step": 5734 + }, + { + "epoch": 2.162518853695324, + "grad_norm": 0.24602294552447848, + "learning_rate": 5.5949479928547176e-05, + "loss": 0.424, + "step": 5735 + }, + { + "epoch": 2.16289592760181, + "grad_norm": 0.2584639938364679, + "learning_rate": 5.593982277913034e-05, + "loss": 0.4209, + "step": 5736 + }, + { + "epoch": 2.1632730015082955, + "grad_norm": 0.22351829405242027, + "learning_rate": 5.593016452509794e-05, + "loss": 0.4111, + "step": 5737 + }, + { + "epoch": 2.163650075414781, + "grad_norm": 0.21184879992646974, + "learning_rate": 5.5920505167119256e-05, + "loss": 0.4181, + "step": 5738 + }, + { + "epoch": 2.164027149321267, + "grad_norm": 0.24585290729101245, + "learning_rate": 5.591084470586369e-05, + "loss": 0.4201, + "step": 5739 + }, + { + "epoch": 2.1644042232277525, + "grad_norm": 0.2822257658305973, + "learning_rate": 5.59011831420007e-05, + "loss": 0.4134, + "step": 5740 + }, + { + "epoch": 2.164781297134238, + "grad_norm": 0.29185783462724674, + "learning_rate": 5.5891520476199815e-05, + "loss": 0.4421, + "step": 5741 + }, + { + "epoch": 2.165158371040724, + "grad_norm": 0.29665178571725515, + "learning_rate": 5.588185670913066e-05, + "loss": 0.4294, + "step": 5742 + }, + { + "epoch": 2.1655354449472095, + "grad_norm": 0.23350375328387799, + "learning_rate": 5.587219184146293e-05, + "loss": 0.4119, + "step": 5743 + }, + { + "epoch": 2.165912518853695, + "grad_norm": 0.23358253449179683, + "learning_rate": 5.586252587386637e-05, + "loss": 0.4118, + "step": 5744 + }, + { + "epoch": 2.166289592760181, + "grad_norm": 0.2944232934680348, + "learning_rate": 5.585285880701086e-05, + "loss": 0.4195, + "step": 5745 + }, + { + "epoch": 2.1666666666666665, + "grad_norm": 0.34379792046259555, + "learning_rate": 5.584319064156628e-05, + "loss": 0.4164, + "step": 5746 + }, + { + "epoch": 2.167043740573152, + "grad_norm": 0.32181374643553456, + "learning_rate": 5.5833521378202634e-05, + "loss": 0.431, + "step": 5747 + }, + { + "epoch": 2.167420814479638, + "grad_norm": 0.2538226214439761, + "learning_rate": 5.582385101759002e-05, + "loss": 0.4355, + "step": 5748 + }, + { + "epoch": 2.1677978883861235, + "grad_norm": 0.18620093137529833, + "learning_rate": 5.581417956039856e-05, + "loss": 0.42, + "step": 5749 + }, + { + "epoch": 2.168174962292609, + "grad_norm": 0.22425410485874675, + "learning_rate": 5.580450700729847e-05, + "loss": 0.4278, + "step": 5750 + }, + { + "epoch": 2.168552036199095, + "grad_norm": 0.26361175309778917, + "learning_rate": 5.5794833358960075e-05, + "loss": 0.4183, + "step": 5751 + }, + { + "epoch": 2.1689291101055805, + "grad_norm": 0.2734324266866591, + "learning_rate": 5.578515861605374e-05, + "loss": 0.4321, + "step": 5752 + }, + { + "epoch": 2.169306184012066, + "grad_norm": 0.27433596724508175, + "learning_rate": 5.57754827792499e-05, + "loss": 0.4296, + "step": 5753 + }, + { + "epoch": 2.169683257918552, + "grad_norm": 0.2897660009082345, + "learning_rate": 5.576580584921911e-05, + "loss": 0.4293, + "step": 5754 + }, + { + "epoch": 2.1700603318250375, + "grad_norm": 0.27894023899280485, + "learning_rate": 5.575612782663196e-05, + "loss": 0.4169, + "step": 5755 + }, + { + "epoch": 2.170437405731523, + "grad_norm": 0.2520753810335443, + "learning_rate": 5.574644871215912e-05, + "loss": 0.4244, + "step": 5756 + }, + { + "epoch": 2.170814479638009, + "grad_norm": 0.25435635946544743, + "learning_rate": 5.5736768506471354e-05, + "loss": 0.4232, + "step": 5757 + }, + { + "epoch": 2.1711915535444946, + "grad_norm": 0.2937080843610806, + "learning_rate": 5.57270872102395e-05, + "loss": 0.3984, + "step": 5758 + }, + { + "epoch": 2.1715686274509802, + "grad_norm": 0.30880284905724864, + "learning_rate": 5.571740482413443e-05, + "loss": 0.4185, + "step": 5759 + }, + { + "epoch": 2.171945701357466, + "grad_norm": 0.2978765098088201, + "learning_rate": 5.570772134882716e-05, + "loss": 0.4068, + "step": 5760 + }, + { + "epoch": 2.1723227752639516, + "grad_norm": 0.26821446797412063, + "learning_rate": 5.569803678498874e-05, + "loss": 0.415, + "step": 5761 + }, + { + "epoch": 2.1726998491704372, + "grad_norm": 0.2007343079945011, + "learning_rate": 5.568835113329028e-05, + "loss": 0.4156, + "step": 5762 + }, + { + "epoch": 2.173076923076923, + "grad_norm": 0.2072339974453207, + "learning_rate": 5.567866439440301e-05, + "loss": 0.432, + "step": 5763 + }, + { + "epoch": 2.1734539969834086, + "grad_norm": 0.2631234071299512, + "learning_rate": 5.5668976568998204e-05, + "loss": 0.4181, + "step": 5764 + }, + { + "epoch": 2.1738310708898942, + "grad_norm": 0.30013857919557735, + "learning_rate": 5.565928765774722e-05, + "loss": 0.4244, + "step": 5765 + }, + { + "epoch": 2.17420814479638, + "grad_norm": 0.2687264182898233, + "learning_rate": 5.5649597661321484e-05, + "loss": 0.4253, + "step": 5766 + }, + { + "epoch": 2.1745852187028656, + "grad_norm": 0.1974021448351799, + "learning_rate": 5.5639906580392524e-05, + "loss": 0.4254, + "step": 5767 + }, + { + "epoch": 2.1749622926093513, + "grad_norm": 0.19869614171013672, + "learning_rate": 5.56302144156319e-05, + "loss": 0.4187, + "step": 5768 + }, + { + "epoch": 2.175339366515837, + "grad_norm": 0.2480900655953161, + "learning_rate": 5.562052116771129e-05, + "loss": 0.4137, + "step": 5769 + }, + { + "epoch": 2.1757164404223226, + "grad_norm": 0.2919965215656794, + "learning_rate": 5.561082683730241e-05, + "loss": 0.4194, + "step": 5770 + }, + { + "epoch": 2.1760935143288083, + "grad_norm": 0.2827065599986503, + "learning_rate": 5.5601131425077085e-05, + "loss": 0.413, + "step": 5771 + }, + { + "epoch": 2.176470588235294, + "grad_norm": 0.291069687394729, + "learning_rate": 5.5591434931707176e-05, + "loss": 0.4171, + "step": 5772 + }, + { + "epoch": 2.1768476621417796, + "grad_norm": 0.31000586561289606, + "learning_rate": 5.558173735786467e-05, + "loss": 0.4251, + "step": 5773 + }, + { + "epoch": 2.1772247360482653, + "grad_norm": 0.2821413027355293, + "learning_rate": 5.557203870422158e-05, + "loss": 0.4118, + "step": 5774 + }, + { + "epoch": 2.177601809954751, + "grad_norm": 0.24681485702028094, + "learning_rate": 5.556233897145003e-05, + "loss": 0.4038, + "step": 5775 + }, + { + "epoch": 2.1779788838612366, + "grad_norm": 0.23370447122917046, + "learning_rate": 5.5552638160222175e-05, + "loss": 0.4125, + "step": 5776 + }, + { + "epoch": 2.1783559577677223, + "grad_norm": 0.2502296882501288, + "learning_rate": 5.5542936271210305e-05, + "loss": 0.4126, + "step": 5777 + }, + { + "epoch": 2.178733031674208, + "grad_norm": 0.22784762452100665, + "learning_rate": 5.553323330508673e-05, + "loss": 0.4209, + "step": 5778 + }, + { + "epoch": 2.1791101055806936, + "grad_norm": 0.20999042643743282, + "learning_rate": 5.552352926252387e-05, + "loss": 0.4252, + "step": 5779 + }, + { + "epoch": 2.1794871794871793, + "grad_norm": 0.21832239516948368, + "learning_rate": 5.55138241441942e-05, + "loss": 0.4216, + "step": 5780 + }, + { + "epoch": 2.179864253393665, + "grad_norm": 0.25145997048565266, + "learning_rate": 5.550411795077028e-05, + "loss": 0.4094, + "step": 5781 + }, + { + "epoch": 2.1802413273001506, + "grad_norm": 0.25720642829833373, + "learning_rate": 5.5494410682924734e-05, + "loss": 0.4184, + "step": 5782 + }, + { + "epoch": 2.1806184012066363, + "grad_norm": 0.2692480954648125, + "learning_rate": 5.548470234133028e-05, + "loss": 0.4124, + "step": 5783 + }, + { + "epoch": 2.180995475113122, + "grad_norm": 0.273970872196953, + "learning_rate": 5.5474992926659673e-05, + "loss": 0.4129, + "step": 5784 + }, + { + "epoch": 2.1813725490196076, + "grad_norm": 0.2647334641166625, + "learning_rate": 5.5465282439585796e-05, + "loss": 0.4269, + "step": 5785 + }, + { + "epoch": 2.1817496229260933, + "grad_norm": 0.24114953507294606, + "learning_rate": 5.5455570880781555e-05, + "loss": 0.4003, + "step": 5786 + }, + { + "epoch": 2.182126696832579, + "grad_norm": 0.2720546995044713, + "learning_rate": 5.544585825091997e-05, + "loss": 0.4138, + "step": 5787 + }, + { + "epoch": 2.1825037707390647, + "grad_norm": 0.25622445048748793, + "learning_rate": 5.54361445506741e-05, + "loss": 0.4255, + "step": 5788 + }, + { + "epoch": 2.1828808446455503, + "grad_norm": 0.2910493261075018, + "learning_rate": 5.542642978071711e-05, + "loss": 0.4287, + "step": 5789 + }, + { + "epoch": 2.183257918552036, + "grad_norm": 0.30079154837803634, + "learning_rate": 5.5416713941722215e-05, + "loss": 0.4114, + "step": 5790 + }, + { + "epoch": 2.183634992458522, + "grad_norm": 0.21403150160216486, + "learning_rate": 5.5406997034362716e-05, + "loss": 0.4092, + "step": 5791 + }, + { + "epoch": 2.1840120663650078, + "grad_norm": 0.2188799140838283, + "learning_rate": 5.5397279059312e-05, + "loss": 0.4364, + "step": 5792 + }, + { + "epoch": 2.1843891402714934, + "grad_norm": 0.23366912244380056, + "learning_rate": 5.538756001724349e-05, + "loss": 0.4138, + "step": 5793 + }, + { + "epoch": 2.184766214177979, + "grad_norm": 0.22341858628578173, + "learning_rate": 5.537783990883073e-05, + "loss": 0.4166, + "step": 5794 + }, + { + "epoch": 2.185143288084465, + "grad_norm": 0.2262301832632667, + "learning_rate": 5.5368118734747304e-05, + "loss": 0.4151, + "step": 5795 + }, + { + "epoch": 2.1855203619909505, + "grad_norm": 0.21613450246742943, + "learning_rate": 5.535839649566687e-05, + "loss": 0.4055, + "step": 5796 + }, + { + "epoch": 2.185897435897436, + "grad_norm": 0.22833961999008598, + "learning_rate": 5.534867319226319e-05, + "loss": 0.4177, + "step": 5797 + }, + { + "epoch": 2.186274509803922, + "grad_norm": 0.23252296751713536, + "learning_rate": 5.5338948825210076e-05, + "loss": 0.413, + "step": 5798 + }, + { + "epoch": 2.1866515837104075, + "grad_norm": 0.27463382239680095, + "learning_rate": 5.532922339518141e-05, + "loss": 0.4097, + "step": 5799 + }, + { + "epoch": 2.187028657616893, + "grad_norm": 0.29083265975052947, + "learning_rate": 5.531949690285116e-05, + "loss": 0.4141, + "step": 5800 + }, + { + "epoch": 2.187405731523379, + "grad_norm": 0.24970740822377333, + "learning_rate": 5.530976934889336e-05, + "loss": 0.4225, + "step": 5801 + }, + { + "epoch": 2.1877828054298645, + "grad_norm": 0.2756698407177277, + "learning_rate": 5.530004073398212e-05, + "loss": 0.4198, + "step": 5802 + }, + { + "epoch": 2.18815987933635, + "grad_norm": 0.33443756683697257, + "learning_rate": 5.5290311058791634e-05, + "loss": 0.4284, + "step": 5803 + }, + { + "epoch": 2.188536953242836, + "grad_norm": 0.37585938571998667, + "learning_rate": 5.528058032399615e-05, + "loss": 0.411, + "step": 5804 + }, + { + "epoch": 2.1889140271493215, + "grad_norm": 0.36173900076114845, + "learning_rate": 5.527084853027001e-05, + "loss": 0.4332, + "step": 5805 + }, + { + "epoch": 2.189291101055807, + "grad_norm": 0.29957475322371874, + "learning_rate": 5.5261115678287605e-05, + "loss": 0.4266, + "step": 5806 + }, + { + "epoch": 2.189668174962293, + "grad_norm": 0.1973356556274677, + "learning_rate": 5.525138176872343e-05, + "loss": 0.4186, + "step": 5807 + }, + { + "epoch": 2.1900452488687785, + "grad_norm": 0.2016009752263113, + "learning_rate": 5.524164680225202e-05, + "loss": 0.4127, + "step": 5808 + }, + { + "epoch": 2.190422322775264, + "grad_norm": 0.27162779788547686, + "learning_rate": 5.523191077954801e-05, + "loss": 0.428, + "step": 5809 + }, + { + "epoch": 2.19079939668175, + "grad_norm": 0.2958777346919523, + "learning_rate": 5.5222173701286096e-05, + "loss": 0.4258, + "step": 5810 + }, + { + "epoch": 2.1911764705882355, + "grad_norm": 0.20929625005640096, + "learning_rate": 5.5212435568141036e-05, + "loss": 0.4264, + "step": 5811 + }, + { + "epoch": 2.191553544494721, + "grad_norm": 0.22228102277065584, + "learning_rate": 5.520269638078769e-05, + "loss": 0.4105, + "step": 5812 + }, + { + "epoch": 2.191930618401207, + "grad_norm": 0.18882671088873482, + "learning_rate": 5.5192956139900974e-05, + "loss": 0.4132, + "step": 5813 + }, + { + "epoch": 2.1923076923076925, + "grad_norm": 0.15969300418426316, + "learning_rate": 5.5183214846155864e-05, + "loss": 0.4157, + "step": 5814 + }, + { + "epoch": 2.192684766214178, + "grad_norm": 0.18217693761744294, + "learning_rate": 5.5173472500227436e-05, + "loss": 0.4258, + "step": 5815 + }, + { + "epoch": 2.193061840120664, + "grad_norm": 0.18029280425623276, + "learning_rate": 5.516372910279083e-05, + "loss": 0.4238, + "step": 5816 + }, + { + "epoch": 2.1934389140271495, + "grad_norm": 0.1646168391697555, + "learning_rate": 5.515398465452124e-05, + "loss": 0.4261, + "step": 5817 + }, + { + "epoch": 2.193815987933635, + "grad_norm": 0.15700778540466606, + "learning_rate": 5.514423915609395e-05, + "loss": 0.4121, + "step": 5818 + }, + { + "epoch": 2.194193061840121, + "grad_norm": 0.15544458870481428, + "learning_rate": 5.513449260818434e-05, + "loss": 0.4155, + "step": 5819 + }, + { + "epoch": 2.1945701357466065, + "grad_norm": 0.19197828414590343, + "learning_rate": 5.512474501146779e-05, + "loss": 0.4113, + "step": 5820 + }, + { + "epoch": 2.194947209653092, + "grad_norm": 0.2082891005498869, + "learning_rate": 5.511499636661984e-05, + "loss": 0.4101, + "step": 5821 + }, + { + "epoch": 2.195324283559578, + "grad_norm": 0.2361511587932874, + "learning_rate": 5.5105246674316046e-05, + "loss": 0.415, + "step": 5822 + }, + { + "epoch": 2.1957013574660635, + "grad_norm": 0.2605408825762213, + "learning_rate": 5.5095495935232056e-05, + "loss": 0.4225, + "step": 5823 + }, + { + "epoch": 2.196078431372549, + "grad_norm": 0.2924513470076548, + "learning_rate": 5.5085744150043586e-05, + "loss": 0.4224, + "step": 5824 + }, + { + "epoch": 2.196455505279035, + "grad_norm": 0.3412484357221025, + "learning_rate": 5.507599131942643e-05, + "loss": 0.4226, + "step": 5825 + }, + { + "epoch": 2.1968325791855206, + "grad_norm": 0.38585162895212655, + "learning_rate": 5.506623744405645e-05, + "loss": 0.4194, + "step": 5826 + }, + { + "epoch": 2.197209653092006, + "grad_norm": 0.409962358922688, + "learning_rate": 5.505648252460958e-05, + "loss": 0.4227, + "step": 5827 + }, + { + "epoch": 2.197586726998492, + "grad_norm": 0.36320508366386495, + "learning_rate": 5.504672656176182e-05, + "loss": 0.417, + "step": 5828 + }, + { + "epoch": 2.1979638009049776, + "grad_norm": 0.2900168337201339, + "learning_rate": 5.503696955618926e-05, + "loss": 0.4197, + "step": 5829 + }, + { + "epoch": 2.1983408748114632, + "grad_norm": 0.2631810117332293, + "learning_rate": 5.502721150856804e-05, + "loss": 0.4182, + "step": 5830 + }, + { + "epoch": 2.198717948717949, + "grad_norm": 0.25795672803690367, + "learning_rate": 5.50174524195744e-05, + "loss": 0.4086, + "step": 5831 + }, + { + "epoch": 2.1990950226244346, + "grad_norm": 0.2634024942946764, + "learning_rate": 5.5007692289884636e-05, + "loss": 0.4206, + "step": 5832 + }, + { + "epoch": 2.1994720965309202, + "grad_norm": 0.2550316420249682, + "learning_rate": 5.499793112017509e-05, + "loss": 0.4153, + "step": 5833 + }, + { + "epoch": 2.199849170437406, + "grad_norm": 0.24212266054267842, + "learning_rate": 5.498816891112223e-05, + "loss": 0.4106, + "step": 5834 + }, + { + "epoch": 2.2002262443438916, + "grad_norm": 0.23683839744231633, + "learning_rate": 5.497840566340255e-05, + "loss": 0.4236, + "step": 5835 + }, + { + "epoch": 2.2006033182503772, + "grad_norm": 0.25619219931858184, + "learning_rate": 5.4968641377692655e-05, + "loss": 0.4203, + "step": 5836 + }, + { + "epoch": 2.200980392156863, + "grad_norm": 0.24274155864173766, + "learning_rate": 5.495887605466919e-05, + "loss": 0.427, + "step": 5837 + }, + { + "epoch": 2.2013574660633486, + "grad_norm": 0.24173990228003872, + "learning_rate": 5.494910969500887e-05, + "loss": 0.422, + "step": 5838 + }, + { + "epoch": 2.2017345399698343, + "grad_norm": 0.27299793890369106, + "learning_rate": 5.493934229938852e-05, + "loss": 0.4172, + "step": 5839 + }, + { + "epoch": 2.20211161387632, + "grad_norm": 0.2718735338538917, + "learning_rate": 5.492957386848499e-05, + "loss": 0.4177, + "step": 5840 + }, + { + "epoch": 2.2024886877828056, + "grad_norm": 0.25819655521217477, + "learning_rate": 5.491980440297524e-05, + "loss": 0.4111, + "step": 5841 + }, + { + "epoch": 2.2028657616892913, + "grad_norm": 0.3269934691573221, + "learning_rate": 5.4910033903536284e-05, + "loss": 0.4053, + "step": 5842 + }, + { + "epoch": 2.203242835595777, + "grad_norm": 0.38576291975720495, + "learning_rate": 5.4900262370845196e-05, + "loss": 0.4218, + "step": 5843 + }, + { + "epoch": 2.2036199095022626, + "grad_norm": 0.33422206513076425, + "learning_rate": 5.489048980557914e-05, + "loss": 0.4265, + "step": 5844 + }, + { + "epoch": 2.2039969834087483, + "grad_norm": 0.2926242666641696, + "learning_rate": 5.4880716208415345e-05, + "loss": 0.4154, + "step": 5845 + }, + { + "epoch": 2.204374057315234, + "grad_norm": 0.2999215553620021, + "learning_rate": 5.487094158003114e-05, + "loss": 0.422, + "step": 5846 + }, + { + "epoch": 2.2047511312217196, + "grad_norm": 0.3545855879391825, + "learning_rate": 5.486116592110385e-05, + "loss": 0.4248, + "step": 5847 + }, + { + "epoch": 2.2051282051282053, + "grad_norm": 0.32880953914723055, + "learning_rate": 5.485138923231095e-05, + "loss": 0.4164, + "step": 5848 + }, + { + "epoch": 2.205505279034691, + "grad_norm": 0.2948041129725479, + "learning_rate": 5.484161151432995e-05, + "loss": 0.414, + "step": 5849 + }, + { + "epoch": 2.2058823529411766, + "grad_norm": 0.25540370905441045, + "learning_rate": 5.4831832767838436e-05, + "loss": 0.4194, + "step": 5850 + }, + { + "epoch": 2.2062594268476623, + "grad_norm": 0.2436538999943906, + "learning_rate": 5.482205299351407e-05, + "loss": 0.4144, + "step": 5851 + }, + { + "epoch": 2.206636500754148, + "grad_norm": 0.2822707609015956, + "learning_rate": 5.481227219203458e-05, + "loss": 0.4295, + "step": 5852 + }, + { + "epoch": 2.2070135746606336, + "grad_norm": 0.28968654427555657, + "learning_rate": 5.4802490364077754e-05, + "loss": 0.4225, + "step": 5853 + }, + { + "epoch": 2.2073906485671193, + "grad_norm": 0.24800698252148287, + "learning_rate": 5.479270751032149e-05, + "loss": 0.4139, + "step": 5854 + }, + { + "epoch": 2.207767722473605, + "grad_norm": 0.19738524784524356, + "learning_rate": 5.47829236314437e-05, + "loss": 0.423, + "step": 5855 + }, + { + "epoch": 2.2081447963800906, + "grad_norm": 0.19220973911298359, + "learning_rate": 5.477313872812243e-05, + "loss": 0.4247, + "step": 5856 + }, + { + "epoch": 2.2085218702865763, + "grad_norm": 0.21673942872378163, + "learning_rate": 5.4763352801035745e-05, + "loss": 0.4154, + "step": 5857 + }, + { + "epoch": 2.208898944193062, + "grad_norm": 0.2548470665028681, + "learning_rate": 5.47535658508618e-05, + "loss": 0.4211, + "step": 5858 + }, + { + "epoch": 2.2092760180995477, + "grad_norm": 0.24959268032306198, + "learning_rate": 5.474377787827884e-05, + "loss": 0.4225, + "step": 5859 + }, + { + "epoch": 2.2096530920060333, + "grad_norm": 0.21174376001854744, + "learning_rate": 5.4733988883965144e-05, + "loss": 0.4263, + "step": 5860 + }, + { + "epoch": 2.210030165912519, + "grad_norm": 0.2145107672196856, + "learning_rate": 5.472419886859909e-05, + "loss": 0.4242, + "step": 5861 + }, + { + "epoch": 2.2104072398190047, + "grad_norm": 0.23948864563248648, + "learning_rate": 5.4714407832859116e-05, + "loss": 0.4169, + "step": 5862 + }, + { + "epoch": 2.2107843137254903, + "grad_norm": 0.243735869709662, + "learning_rate": 5.470461577742372e-05, + "loss": 0.4073, + "step": 5863 + }, + { + "epoch": 2.211161387631976, + "grad_norm": 0.22758684227604475, + "learning_rate": 5.46948227029715e-05, + "loss": 0.4326, + "step": 5864 + }, + { + "epoch": 2.2115384615384617, + "grad_norm": 0.2343127625367229, + "learning_rate": 5.468502861018111e-05, + "loss": 0.4195, + "step": 5865 + }, + { + "epoch": 2.2119155354449473, + "grad_norm": 0.2274395116497928, + "learning_rate": 5.467523349973125e-05, + "loss": 0.4214, + "step": 5866 + }, + { + "epoch": 2.212292609351433, + "grad_norm": 0.1836701716363539, + "learning_rate": 5.466543737230073e-05, + "loss": 0.4261, + "step": 5867 + }, + { + "epoch": 2.2126696832579187, + "grad_norm": 0.16088506693102705, + "learning_rate": 5.465564022856841e-05, + "loss": 0.4097, + "step": 5868 + }, + { + "epoch": 2.2130467571644044, + "grad_norm": 0.17679718027513766, + "learning_rate": 5.4645842069213216e-05, + "loss": 0.4167, + "step": 5869 + }, + { + "epoch": 2.21342383107089, + "grad_norm": 0.15258179368193572, + "learning_rate": 5.4636042894914166e-05, + "loss": 0.4155, + "step": 5870 + }, + { + "epoch": 2.2138009049773757, + "grad_norm": 0.19856188865967225, + "learning_rate": 5.462624270635033e-05, + "loss": 0.4178, + "step": 5871 + }, + { + "epoch": 2.2141779788838614, + "grad_norm": 0.18513263360568036, + "learning_rate": 5.4616441504200834e-05, + "loss": 0.432, + "step": 5872 + }, + { + "epoch": 2.214555052790347, + "grad_norm": 0.18395855343115183, + "learning_rate": 5.460663928914491e-05, + "loss": 0.4182, + "step": 5873 + }, + { + "epoch": 2.2149321266968327, + "grad_norm": 0.2540483480174123, + "learning_rate": 5.459683606186184e-05, + "loss": 0.4147, + "step": 5874 + }, + { + "epoch": 2.2153092006033184, + "grad_norm": 0.3295192363368116, + "learning_rate": 5.4587031823030984e-05, + "loss": 0.4182, + "step": 5875 + }, + { + "epoch": 2.215686274509804, + "grad_norm": 0.3216638070799762, + "learning_rate": 5.4577226573331754e-05, + "loss": 0.4256, + "step": 5876 + }, + { + "epoch": 2.2160633484162897, + "grad_norm": 0.22885457849202603, + "learning_rate": 5.456742031344366e-05, + "loss": 0.4223, + "step": 5877 + }, + { + "epoch": 2.2164404223227754, + "grad_norm": 0.1809917384892877, + "learning_rate": 5.455761304404625e-05, + "loss": 0.412, + "step": 5878 + }, + { + "epoch": 2.216817496229261, + "grad_norm": 0.21640889369425828, + "learning_rate": 5.454780476581917e-05, + "loss": 0.4265, + "step": 5879 + }, + { + "epoch": 2.2171945701357467, + "grad_norm": 0.23784632830576194, + "learning_rate": 5.4537995479442134e-05, + "loss": 0.4197, + "step": 5880 + }, + { + "epoch": 2.2175716440422324, + "grad_norm": 0.2407361995064906, + "learning_rate": 5.4528185185594897e-05, + "loss": 0.4068, + "step": 5881 + }, + { + "epoch": 2.217948717948718, + "grad_norm": 0.26436606740551716, + "learning_rate": 5.451837388495731e-05, + "loss": 0.4268, + "step": 5882 + }, + { + "epoch": 2.2183257918552037, + "grad_norm": 0.2903823028758362, + "learning_rate": 5.45085615782093e-05, + "loss": 0.4201, + "step": 5883 + }, + { + "epoch": 2.2187028657616894, + "grad_norm": 0.31737844483516064, + "learning_rate": 5.4498748266030835e-05, + "loss": 0.4178, + "step": 5884 + }, + { + "epoch": 2.219079939668175, + "grad_norm": 0.319137599770399, + "learning_rate": 5.448893394910197e-05, + "loss": 0.4226, + "step": 5885 + }, + { + "epoch": 2.2194570135746607, + "grad_norm": 0.3369111098705348, + "learning_rate": 5.4479118628102856e-05, + "loss": 0.4268, + "step": 5886 + }, + { + "epoch": 2.2198340874811464, + "grad_norm": 0.3767567962524537, + "learning_rate": 5.4469302303713654e-05, + "loss": 0.4171, + "step": 5887 + }, + { + "epoch": 2.220211161387632, + "grad_norm": 0.36226565614534945, + "learning_rate": 5.445948497661462e-05, + "loss": 0.4225, + "step": 5888 + }, + { + "epoch": 2.2205882352941178, + "grad_norm": 0.2908644337531624, + "learning_rate": 5.444966664748613e-05, + "loss": 0.421, + "step": 5889 + }, + { + "epoch": 2.2209653092006034, + "grad_norm": 0.21844121395361338, + "learning_rate": 5.4439847317008534e-05, + "loss": 0.4224, + "step": 5890 + }, + { + "epoch": 2.221342383107089, + "grad_norm": 0.19597029150624373, + "learning_rate": 5.4430026985862335e-05, + "loss": 0.4275, + "step": 5891 + }, + { + "epoch": 2.2217194570135748, + "grad_norm": 0.2209330704221334, + "learning_rate": 5.442020565472808e-05, + "loss": 0.4155, + "step": 5892 + }, + { + "epoch": 2.2220965309200604, + "grad_norm": 0.2388777923916335, + "learning_rate": 5.4410383324286344e-05, + "loss": 0.4066, + "step": 5893 + }, + { + "epoch": 2.222473604826546, + "grad_norm": 0.2880906902383787, + "learning_rate": 5.4400559995217834e-05, + "loss": 0.4137, + "step": 5894 + }, + { + "epoch": 2.2228506787330318, + "grad_norm": 0.3124386871612973, + "learning_rate": 5.439073566820328e-05, + "loss": 0.4159, + "step": 5895 + }, + { + "epoch": 2.2232277526395174, + "grad_norm": 0.3022163502412727, + "learning_rate": 5.4380910343923526e-05, + "loss": 0.4311, + "step": 5896 + }, + { + "epoch": 2.223604826546003, + "grad_norm": 0.26738693802901187, + "learning_rate": 5.437108402305943e-05, + "loss": 0.4209, + "step": 5897 + }, + { + "epoch": 2.223981900452489, + "grad_norm": 0.18149981359945686, + "learning_rate": 5.4361256706291956e-05, + "loss": 0.4275, + "step": 5898 + }, + { + "epoch": 2.2243589743589745, + "grad_norm": 0.20123157256897545, + "learning_rate": 5.435142839430214e-05, + "loss": 0.4378, + "step": 5899 + }, + { + "epoch": 2.22473604826546, + "grad_norm": 0.23571109178406796, + "learning_rate": 5.434159908777106e-05, + "loss": 0.4169, + "step": 5900 + }, + { + "epoch": 2.225113122171946, + "grad_norm": 0.24703230326264386, + "learning_rate": 5.433176878737988e-05, + "loss": 0.4221, + "step": 5901 + }, + { + "epoch": 2.2254901960784315, + "grad_norm": 0.2538208015773119, + "learning_rate": 5.432193749380985e-05, + "loss": 0.4186, + "step": 5902 + }, + { + "epoch": 2.225867269984917, + "grad_norm": 0.25176399724991966, + "learning_rate": 5.431210520774224e-05, + "loss": 0.4194, + "step": 5903 + }, + { + "epoch": 2.226244343891403, + "grad_norm": 0.2652498670836849, + "learning_rate": 5.430227192985844e-05, + "loss": 0.4122, + "step": 5904 + }, + { + "epoch": 2.2266214177978885, + "grad_norm": 0.2751755067452248, + "learning_rate": 5.429243766083988e-05, + "loss": 0.4211, + "step": 5905 + }, + { + "epoch": 2.226998491704374, + "grad_norm": 0.2569151985247881, + "learning_rate": 5.4282602401368066e-05, + "loss": 0.4215, + "step": 5906 + }, + { + "epoch": 2.22737556561086, + "grad_norm": 0.31515760206505183, + "learning_rate": 5.427276615212458e-05, + "loss": 0.4124, + "step": 5907 + }, + { + "epoch": 2.2277526395173455, + "grad_norm": 0.3632945978701806, + "learning_rate": 5.4262928913791065e-05, + "loss": 0.4222, + "step": 5908 + }, + { + "epoch": 2.228129713423831, + "grad_norm": 0.4027059418117051, + "learning_rate": 5.4253090687049216e-05, + "loss": 0.4207, + "step": 5909 + }, + { + "epoch": 2.228506787330317, + "grad_norm": 0.4090336189574997, + "learning_rate": 5.4243251472580825e-05, + "loss": 0.4237, + "step": 5910 + }, + { + "epoch": 2.2288838612368025, + "grad_norm": 0.4142612250077069, + "learning_rate": 5.423341127106774e-05, + "loss": 0.4153, + "step": 5911 + }, + { + "epoch": 2.229260935143288, + "grad_norm": 0.36288501172496973, + "learning_rate": 5.422357008319188e-05, + "loss": 0.4108, + "step": 5912 + }, + { + "epoch": 2.229638009049774, + "grad_norm": 0.2708317152216778, + "learning_rate": 5.4213727909635234e-05, + "loss": 0.4257, + "step": 5913 + }, + { + "epoch": 2.2300150829562595, + "grad_norm": 0.25353181884557774, + "learning_rate": 5.4203884751079856e-05, + "loss": 0.4246, + "step": 5914 + }, + { + "epoch": 2.230392156862745, + "grad_norm": 0.2877373624260919, + "learning_rate": 5.419404060820785e-05, + "loss": 0.4228, + "step": 5915 + }, + { + "epoch": 2.230769230769231, + "grad_norm": 0.30769150540411855, + "learning_rate": 5.4184195481701425e-05, + "loss": 0.4178, + "step": 5916 + }, + { + "epoch": 2.2311463046757165, + "grad_norm": 0.26874284493738804, + "learning_rate": 5.417434937224285e-05, + "loss": 0.4204, + "step": 5917 + }, + { + "epoch": 2.231523378582202, + "grad_norm": 0.2393926368886303, + "learning_rate": 5.4164502280514413e-05, + "loss": 0.4223, + "step": 5918 + }, + { + "epoch": 2.231900452488688, + "grad_norm": 0.23711289076376782, + "learning_rate": 5.415465420719854e-05, + "loss": 0.4248, + "step": 5919 + }, + { + "epoch": 2.2322775263951735, + "grad_norm": 0.23744563030023813, + "learning_rate": 5.41448051529777e-05, + "loss": 0.4257, + "step": 5920 + }, + { + "epoch": 2.232654600301659, + "grad_norm": 0.22298466165212133, + "learning_rate": 5.4134955118534386e-05, + "loss": 0.4137, + "step": 5921 + }, + { + "epoch": 2.233031674208145, + "grad_norm": 0.25404171427311006, + "learning_rate": 5.412510410455123e-05, + "loss": 0.4265, + "step": 5922 + }, + { + "epoch": 2.2334087481146305, + "grad_norm": 0.2698542542120985, + "learning_rate": 5.4115252111710895e-05, + "loss": 0.4214, + "step": 5923 + }, + { + "epoch": 2.233785822021116, + "grad_norm": 0.27314435183774144, + "learning_rate": 5.41053991406961e-05, + "loss": 0.4144, + "step": 5924 + }, + { + "epoch": 2.234162895927602, + "grad_norm": 0.26794049451883784, + "learning_rate": 5.409554519218965e-05, + "loss": 0.4243, + "step": 5925 + }, + { + "epoch": 2.2345399698340875, + "grad_norm": 0.21971417298782278, + "learning_rate": 5.4085690266874436e-05, + "loss": 0.4231, + "step": 5926 + }, + { + "epoch": 2.234917043740573, + "grad_norm": 0.2543499162668868, + "learning_rate": 5.407583436543338e-05, + "loss": 0.4302, + "step": 5927 + }, + { + "epoch": 2.235294117647059, + "grad_norm": 0.34567619914784253, + "learning_rate": 5.406597748854947e-05, + "loss": 0.437, + "step": 5928 + }, + { + "epoch": 2.2356711915535445, + "grad_norm": 0.36776368861813885, + "learning_rate": 5.4056119636905816e-05, + "loss": 0.4264, + "step": 5929 + }, + { + "epoch": 2.23604826546003, + "grad_norm": 0.30490898677242334, + "learning_rate": 5.4046260811185525e-05, + "loss": 0.4237, + "step": 5930 + }, + { + "epoch": 2.236425339366516, + "grad_norm": 0.2720558903062018, + "learning_rate": 5.403640101207183e-05, + "loss": 0.4155, + "step": 5931 + }, + { + "epoch": 2.2368024132730016, + "grad_norm": 0.2891470099177509, + "learning_rate": 5.4026540240248e-05, + "loss": 0.4122, + "step": 5932 + }, + { + "epoch": 2.2371794871794872, + "grad_norm": 0.31785927409923254, + "learning_rate": 5.401667849639736e-05, + "loss": 0.4085, + "step": 5933 + }, + { + "epoch": 2.237556561085973, + "grad_norm": 0.3635624424693829, + "learning_rate": 5.400681578120333e-05, + "loss": 0.4235, + "step": 5934 + }, + { + "epoch": 2.2379336349924586, + "grad_norm": 0.3092809788246149, + "learning_rate": 5.39969520953494e-05, + "loss": 0.4197, + "step": 5935 + }, + { + "epoch": 2.2383107088989442, + "grad_norm": 0.2777453951633188, + "learning_rate": 5.3987087439519096e-05, + "loss": 0.4122, + "step": 5936 + }, + { + "epoch": 2.23868778280543, + "grad_norm": 0.24296982273401543, + "learning_rate": 5.397722181439604e-05, + "loss": 0.4227, + "step": 5937 + }, + { + "epoch": 2.2390648567119156, + "grad_norm": 0.18468001989272945, + "learning_rate": 5.396735522066392e-05, + "loss": 0.4286, + "step": 5938 + }, + { + "epoch": 2.2394419306184012, + "grad_norm": 0.23151602112006564, + "learning_rate": 5.3957487659006464e-05, + "loss": 0.4193, + "step": 5939 + }, + { + "epoch": 2.239819004524887, + "grad_norm": 0.274570414912365, + "learning_rate": 5.3947619130107503e-05, + "loss": 0.4261, + "step": 5940 + }, + { + "epoch": 2.2401960784313726, + "grad_norm": 0.32448019839158027, + "learning_rate": 5.39377496346509e-05, + "loss": 0.4227, + "step": 5941 + }, + { + "epoch": 2.2405731523378583, + "grad_norm": 0.3214694336616882, + "learning_rate": 5.3927879173320605e-05, + "loss": 0.4215, + "step": 5942 + }, + { + "epoch": 2.240950226244344, + "grad_norm": 0.25038288646547646, + "learning_rate": 5.3918007746800645e-05, + "loss": 0.4311, + "step": 5943 + }, + { + "epoch": 2.2413273001508296, + "grad_norm": 0.24344964100387412, + "learning_rate": 5.39081353557751e-05, + "loss": 0.4099, + "step": 5944 + }, + { + "epoch": 2.2417043740573153, + "grad_norm": 0.21057417666998332, + "learning_rate": 5.38982620009281e-05, + "loss": 0.4119, + "step": 5945 + }, + { + "epoch": 2.242081447963801, + "grad_norm": 0.1938557583057962, + "learning_rate": 5.388838768294386e-05, + "loss": 0.4219, + "step": 5946 + }, + { + "epoch": 2.2424585218702866, + "grad_norm": 0.18424664909205601, + "learning_rate": 5.387851240250671e-05, + "loss": 0.4211, + "step": 5947 + }, + { + "epoch": 2.2428355957767723, + "grad_norm": 0.22835970291489183, + "learning_rate": 5.386863616030093e-05, + "loss": 0.4206, + "step": 5948 + }, + { + "epoch": 2.243212669683258, + "grad_norm": 0.23139691760937847, + "learning_rate": 5.3858758957010965e-05, + "loss": 0.4143, + "step": 5949 + }, + { + "epoch": 2.2435897435897436, + "grad_norm": 0.31940632239590366, + "learning_rate": 5.384888079332132e-05, + "loss": 0.4245, + "step": 5950 + }, + { + "epoch": 2.2439668174962293, + "grad_norm": 0.3326046185712551, + "learning_rate": 5.38390016699165e-05, + "loss": 0.4102, + "step": 5951 + }, + { + "epoch": 2.244343891402715, + "grad_norm": 0.32772031446205024, + "learning_rate": 5.382912158748114e-05, + "loss": 0.4172, + "step": 5952 + }, + { + "epoch": 2.2447209653092006, + "grad_norm": 0.323382632782367, + "learning_rate": 5.381924054669993e-05, + "loss": 0.4244, + "step": 5953 + }, + { + "epoch": 2.2450980392156863, + "grad_norm": 0.2807550022388086, + "learning_rate": 5.38093585482576e-05, + "loss": 0.415, + "step": 5954 + }, + { + "epoch": 2.245475113122172, + "grad_norm": 0.22887268829360854, + "learning_rate": 5.3799475592838973e-05, + "loss": 0.4233, + "step": 5955 + }, + { + "epoch": 2.2458521870286576, + "grad_norm": 0.22862935038040408, + "learning_rate": 5.378959168112893e-05, + "loss": 0.4193, + "step": 5956 + }, + { + "epoch": 2.2462292609351433, + "grad_norm": 0.25697834452213064, + "learning_rate": 5.377970681381241e-05, + "loss": 0.4083, + "step": 5957 + }, + { + "epoch": 2.246606334841629, + "grad_norm": 0.2563947197313282, + "learning_rate": 5.376982099157443e-05, + "loss": 0.4265, + "step": 5958 + }, + { + "epoch": 2.2469834087481146, + "grad_norm": 0.25242579776381985, + "learning_rate": 5.375993421510007e-05, + "loss": 0.4189, + "step": 5959 + }, + { + "epoch": 2.2473604826546003, + "grad_norm": 0.2096170887903466, + "learning_rate": 5.375004648507449e-05, + "loss": 0.4199, + "step": 5960 + }, + { + "epoch": 2.247737556561086, + "grad_norm": 0.18867036638016252, + "learning_rate": 5.374015780218287e-05, + "loss": 0.4243, + "step": 5961 + }, + { + "epoch": 2.2481146304675717, + "grad_norm": 0.2208468149347059, + "learning_rate": 5.3730268167110495e-05, + "loss": 0.4288, + "step": 5962 + }, + { + "epoch": 2.2484917043740573, + "grad_norm": 0.22691951548832326, + "learning_rate": 5.3720377580542726e-05, + "loss": 0.4287, + "step": 5963 + }, + { + "epoch": 2.248868778280543, + "grad_norm": 0.24445635082144349, + "learning_rate": 5.371048604316496e-05, + "loss": 0.4257, + "step": 5964 + }, + { + "epoch": 2.2492458521870287, + "grad_norm": 0.22775681936066863, + "learning_rate": 5.370059355566267e-05, + "loss": 0.4194, + "step": 5965 + }, + { + "epoch": 2.2496229260935143, + "grad_norm": 0.25952736465194376, + "learning_rate": 5.369070011872141e-05, + "loss": 0.4107, + "step": 5966 + }, + { + "epoch": 2.25, + "grad_norm": 0.226147889076388, + "learning_rate": 5.368080573302676e-05, + "loss": 0.4144, + "step": 5967 + }, + { + "epoch": 2.2503770739064857, + "grad_norm": 0.20632365299833672, + "learning_rate": 5.3670910399264405e-05, + "loss": 0.42, + "step": 5968 + }, + { + "epoch": 2.2507541478129713, + "grad_norm": 0.26896514290760676, + "learning_rate": 5.3661014118120105e-05, + "loss": 0.4308, + "step": 5969 + }, + { + "epoch": 2.251131221719457, + "grad_norm": 0.24959504251351386, + "learning_rate": 5.3651116890279626e-05, + "loss": 0.4214, + "step": 5970 + }, + { + "epoch": 2.2515082956259427, + "grad_norm": 0.24461007895546008, + "learning_rate": 5.3641218716428856e-05, + "loss": 0.4141, + "step": 5971 + }, + { + "epoch": 2.2518853695324283, + "grad_norm": 0.2710949857423425, + "learning_rate": 5.363131959725374e-05, + "loss": 0.4117, + "step": 5972 + }, + { + "epoch": 2.252262443438914, + "grad_norm": 0.31252780045741174, + "learning_rate": 5.3621419533440256e-05, + "loss": 0.4181, + "step": 5973 + }, + { + "epoch": 2.2526395173453997, + "grad_norm": 0.3137655683993381, + "learning_rate": 5.3611518525674484e-05, + "loss": 0.4087, + "step": 5974 + }, + { + "epoch": 2.2530165912518854, + "grad_norm": 0.30698847528248074, + "learning_rate": 5.3601616574642556e-05, + "loss": 0.4192, + "step": 5975 + }, + { + "epoch": 2.253393665158371, + "grad_norm": 0.24284113351777945, + "learning_rate": 5.359171368103066e-05, + "loss": 0.4255, + "step": 5976 + }, + { + "epoch": 2.2537707390648567, + "grad_norm": 0.22818960119065668, + "learning_rate": 5.358180984552506e-05, + "loss": 0.4442, + "step": 5977 + }, + { + "epoch": 2.2541478129713424, + "grad_norm": 0.22108735464013432, + "learning_rate": 5.35719050688121e-05, + "loss": 0.4158, + "step": 5978 + }, + { + "epoch": 2.254524886877828, + "grad_norm": 0.21313340767588881, + "learning_rate": 5.356199935157813e-05, + "loss": 0.4171, + "step": 5979 + }, + { + "epoch": 2.2549019607843137, + "grad_norm": 0.2491928866717181, + "learning_rate": 5.355209269450965e-05, + "loss": 0.4183, + "step": 5980 + }, + { + "epoch": 2.2552790346907994, + "grad_norm": 0.23200171001916875, + "learning_rate": 5.3542185098293175e-05, + "loss": 0.4133, + "step": 5981 + }, + { + "epoch": 2.255656108597285, + "grad_norm": 0.26810111551054616, + "learning_rate": 5.3532276563615266e-05, + "loss": 0.4158, + "step": 5982 + }, + { + "epoch": 2.2560331825037707, + "grad_norm": 0.3360577516472834, + "learning_rate": 5.352236709116261e-05, + "loss": 0.4212, + "step": 5983 + }, + { + "epoch": 2.2564102564102564, + "grad_norm": 0.2639473781737787, + "learning_rate": 5.351245668162189e-05, + "loss": 0.4115, + "step": 5984 + }, + { + "epoch": 2.256787330316742, + "grad_norm": 0.21218582229099803, + "learning_rate": 5.3502545335679924e-05, + "loss": 0.4256, + "step": 5985 + }, + { + "epoch": 2.2571644042232277, + "grad_norm": 0.2731251685358493, + "learning_rate": 5.3492633054023545e-05, + "loss": 0.4163, + "step": 5986 + }, + { + "epoch": 2.2575414781297134, + "grad_norm": 0.24455354878820482, + "learning_rate": 5.3482719837339645e-05, + "loss": 0.4175, + "step": 5987 + }, + { + "epoch": 2.257918552036199, + "grad_norm": 0.1918133624127353, + "learning_rate": 5.347280568631524e-05, + "loss": 0.4113, + "step": 5988 + }, + { + "epoch": 2.2582956259426847, + "grad_norm": 0.23366068895559167, + "learning_rate": 5.3462890601637343e-05, + "loss": 0.4179, + "step": 5989 + }, + { + "epoch": 2.2586726998491704, + "grad_norm": 0.24166609241460532, + "learning_rate": 5.345297458399306e-05, + "loss": 0.4213, + "step": 5990 + }, + { + "epoch": 2.259049773755656, + "grad_norm": 0.19124566023541886, + "learning_rate": 5.344305763406959e-05, + "loss": 0.4137, + "step": 5991 + }, + { + "epoch": 2.2594268476621417, + "grad_norm": 0.19432321895084942, + "learning_rate": 5.343313975255413e-05, + "loss": 0.4272, + "step": 5992 + }, + { + "epoch": 2.2598039215686274, + "grad_norm": 0.19803817531807807, + "learning_rate": 5.342322094013401e-05, + "loss": 0.424, + "step": 5993 + }, + { + "epoch": 2.260180995475113, + "grad_norm": 0.21201464156331626, + "learning_rate": 5.341330119749659e-05, + "loss": 0.4167, + "step": 5994 + }, + { + "epoch": 2.2605580693815988, + "grad_norm": 0.23564270861786402, + "learning_rate": 5.340338052532929e-05, + "loss": 0.4156, + "step": 5995 + }, + { + "epoch": 2.2609351432880844, + "grad_norm": 0.26767256984274007, + "learning_rate": 5.33934589243196e-05, + "loss": 0.4248, + "step": 5996 + }, + { + "epoch": 2.26131221719457, + "grad_norm": 0.27881224808859456, + "learning_rate": 5.33835363951551e-05, + "loss": 0.431, + "step": 5997 + }, + { + "epoch": 2.2616892911010558, + "grad_norm": 0.2795740380248877, + "learning_rate": 5.33736129385234e-05, + "loss": 0.4122, + "step": 5998 + }, + { + "epoch": 2.2620663650075414, + "grad_norm": 0.32444983989355236, + "learning_rate": 5.336368855511217e-05, + "loss": 0.4225, + "step": 5999 + }, + { + "epoch": 2.262443438914027, + "grad_norm": 0.36944971116334463, + "learning_rate": 5.335376324560919e-05, + "loss": 0.4217, + "step": 6000 + }, + { + "epoch": 2.2628205128205128, + "grad_norm": 0.3536398671903163, + "learning_rate": 5.3343837010702265e-05, + "loss": 0.404, + "step": 6001 + }, + { + "epoch": 2.2631975867269984, + "grad_norm": 0.3109980483476074, + "learning_rate": 5.3333909851079257e-05, + "loss": 0.4127, + "step": 6002 + }, + { + "epoch": 2.263574660633484, + "grad_norm": 0.2682689968604249, + "learning_rate": 5.332398176742812e-05, + "loss": 0.4274, + "step": 6003 + }, + { + "epoch": 2.26395173453997, + "grad_norm": 0.25130631867701203, + "learning_rate": 5.331405276043688e-05, + "loss": 0.4302, + "step": 6004 + }, + { + "epoch": 2.2643288084464555, + "grad_norm": 0.2587976353889382, + "learning_rate": 5.3304122830793595e-05, + "loss": 0.4145, + "step": 6005 + }, + { + "epoch": 2.264705882352941, + "grad_norm": 0.2848036069600828, + "learning_rate": 5.329419197918639e-05, + "loss": 0.4224, + "step": 6006 + }, + { + "epoch": 2.265082956259427, + "grad_norm": 0.26480807685344315, + "learning_rate": 5.328426020630347e-05, + "loss": 0.4266, + "step": 6007 + }, + { + "epoch": 2.2654600301659125, + "grad_norm": 0.2522298701361535, + "learning_rate": 5.3274327512833116e-05, + "loss": 0.4327, + "step": 6008 + }, + { + "epoch": 2.265837104072398, + "grad_norm": 0.2298849849686556, + "learning_rate": 5.326439389946362e-05, + "loss": 0.4105, + "step": 6009 + }, + { + "epoch": 2.266214177978884, + "grad_norm": 0.19918854030954547, + "learning_rate": 5.32544593668834e-05, + "loss": 0.4179, + "step": 6010 + }, + { + "epoch": 2.2665912518853695, + "grad_norm": 0.19933507590031652, + "learning_rate": 5.324452391578091e-05, + "loss": 0.4276, + "step": 6011 + }, + { + "epoch": 2.266968325791855, + "grad_norm": 0.19518488174679735, + "learning_rate": 5.323458754684465e-05, + "loss": 0.4196, + "step": 6012 + }, + { + "epoch": 2.267345399698341, + "grad_norm": 0.24043027340279166, + "learning_rate": 5.322465026076322e-05, + "loss": 0.4152, + "step": 6013 + }, + { + "epoch": 2.2677224736048265, + "grad_norm": 0.26869581918421814, + "learning_rate": 5.3214712058225265e-05, + "loss": 0.4077, + "step": 6014 + }, + { + "epoch": 2.268099547511312, + "grad_norm": 0.3131664020840422, + "learning_rate": 5.320477293991946e-05, + "loss": 0.4207, + "step": 6015 + }, + { + "epoch": 2.268476621417798, + "grad_norm": 0.36742581233711846, + "learning_rate": 5.319483290653463e-05, + "loss": 0.4313, + "step": 6016 + }, + { + "epoch": 2.2688536953242835, + "grad_norm": 0.37345211001252787, + "learning_rate": 5.318489195875957e-05, + "loss": 0.4163, + "step": 6017 + }, + { + "epoch": 2.269230769230769, + "grad_norm": 0.3115618315419247, + "learning_rate": 5.317495009728319e-05, + "loss": 0.4158, + "step": 6018 + }, + { + "epoch": 2.269607843137255, + "grad_norm": 0.2266617793837777, + "learning_rate": 5.316500732279446e-05, + "loss": 0.4184, + "step": 6019 + }, + { + "epoch": 2.2699849170437405, + "grad_norm": 0.21307445612499884, + "learning_rate": 5.315506363598239e-05, + "loss": 0.4257, + "step": 6020 + }, + { + "epoch": 2.270361990950226, + "grad_norm": 0.22085961570317414, + "learning_rate": 5.314511903753609e-05, + "loss": 0.4146, + "step": 6021 + }, + { + "epoch": 2.270739064856712, + "grad_norm": 0.2277768778916667, + "learning_rate": 5.31351735281447e-05, + "loss": 0.4237, + "step": 6022 + }, + { + "epoch": 2.2711161387631975, + "grad_norm": 0.23463224768800886, + "learning_rate": 5.3125227108497424e-05, + "loss": 0.4216, + "step": 6023 + }, + { + "epoch": 2.271493212669683, + "grad_norm": 0.24121447455771614, + "learning_rate": 5.311527977928357e-05, + "loss": 0.423, + "step": 6024 + }, + { + "epoch": 2.271870286576169, + "grad_norm": 0.2705315903714634, + "learning_rate": 5.310533154119244e-05, + "loss": 0.4314, + "step": 6025 + }, + { + "epoch": 2.2722473604826545, + "grad_norm": 0.3125531442220011, + "learning_rate": 5.309538239491347e-05, + "loss": 0.4229, + "step": 6026 + }, + { + "epoch": 2.27262443438914, + "grad_norm": 0.3124656103408714, + "learning_rate": 5.3085432341136115e-05, + "loss": 0.4402, + "step": 6027 + }, + { + "epoch": 2.273001508295626, + "grad_norm": 0.23587458801774164, + "learning_rate": 5.3075481380549906e-05, + "loss": 0.4299, + "step": 6028 + }, + { + "epoch": 2.2733785822021115, + "grad_norm": 0.21300721069824702, + "learning_rate": 5.3065529513844434e-05, + "loss": 0.4264, + "step": 6029 + }, + { + "epoch": 2.273755656108597, + "grad_norm": 0.22668668362490982, + "learning_rate": 5.3055576741709355e-05, + "loss": 0.4247, + "step": 6030 + }, + { + "epoch": 2.274132730015083, + "grad_norm": 0.19685216141972206, + "learning_rate": 5.304562306483439e-05, + "loss": 0.4284, + "step": 6031 + }, + { + "epoch": 2.2745098039215685, + "grad_norm": 0.196490014058972, + "learning_rate": 5.303566848390931e-05, + "loss": 0.4112, + "step": 6032 + }, + { + "epoch": 2.274886877828054, + "grad_norm": 0.20100567339774558, + "learning_rate": 5.302571299962399e-05, + "loss": 0.4077, + "step": 6033 + }, + { + "epoch": 2.27526395173454, + "grad_norm": 0.19593002588032918, + "learning_rate": 5.301575661266829e-05, + "loss": 0.4104, + "step": 6034 + }, + { + "epoch": 2.2756410256410255, + "grad_norm": 0.2767045392127205, + "learning_rate": 5.300579932373221e-05, + "loss": 0.4144, + "step": 6035 + }, + { + "epoch": 2.276018099547511, + "grad_norm": 0.301999041412693, + "learning_rate": 5.2995841133505766e-05, + "loss": 0.4158, + "step": 6036 + }, + { + "epoch": 2.276395173453997, + "grad_norm": 0.28155791076657866, + "learning_rate": 5.2985882042679064e-05, + "loss": 0.412, + "step": 6037 + }, + { + "epoch": 2.2767722473604826, + "grad_norm": 0.25769739088647814, + "learning_rate": 5.297592205194225e-05, + "loss": 0.423, + "step": 6038 + }, + { + "epoch": 2.2771493212669682, + "grad_norm": 0.21140411712656884, + "learning_rate": 5.2965961161985565e-05, + "loss": 0.4227, + "step": 6039 + }, + { + "epoch": 2.277526395173454, + "grad_norm": 0.22610001289434375, + "learning_rate": 5.2955999373499256e-05, + "loss": 0.4126, + "step": 6040 + }, + { + "epoch": 2.2779034690799396, + "grad_norm": 0.24066296782387794, + "learning_rate": 5.294603668717368e-05, + "loss": 0.4307, + "step": 6041 + }, + { + "epoch": 2.2782805429864252, + "grad_norm": 0.20565676246476405, + "learning_rate": 5.293607310369926e-05, + "loss": 0.4002, + "step": 6042 + }, + { + "epoch": 2.278657616892911, + "grad_norm": 0.2110212620463486, + "learning_rate": 5.292610862376643e-05, + "loss": 0.4323, + "step": 6043 + }, + { + "epoch": 2.2790346907993966, + "grad_norm": 0.23095246497997787, + "learning_rate": 5.291614324806575e-05, + "loss": 0.4228, + "step": 6044 + }, + { + "epoch": 2.2794117647058822, + "grad_norm": 0.28798207222994293, + "learning_rate": 5.29061769772878e-05, + "loss": 0.4228, + "step": 6045 + }, + { + "epoch": 2.279788838612368, + "grad_norm": 0.32482497380896347, + "learning_rate": 5.2896209812123224e-05, + "loss": 0.4217, + "step": 6046 + }, + { + "epoch": 2.2801659125188536, + "grad_norm": 0.2634154051597625, + "learning_rate": 5.2886241753262755e-05, + "loss": 0.4147, + "step": 6047 + }, + { + "epoch": 2.2805429864253393, + "grad_norm": 0.28425770208478107, + "learning_rate": 5.287627280139717e-05, + "loss": 0.4142, + "step": 6048 + }, + { + "epoch": 2.280920060331825, + "grad_norm": 0.3231049552343587, + "learning_rate": 5.286630295721728e-05, + "loss": 0.4161, + "step": 6049 + }, + { + "epoch": 2.2812971342383106, + "grad_norm": 0.2863633891601406, + "learning_rate": 5.285633222141402e-05, + "loss": 0.4245, + "step": 6050 + }, + { + "epoch": 2.2816742081447963, + "grad_norm": 0.22972763514901057, + "learning_rate": 5.2846360594678346e-05, + "loss": 0.4243, + "step": 6051 + }, + { + "epoch": 2.282051282051282, + "grad_norm": 0.2460161709604459, + "learning_rate": 5.283638807770127e-05, + "loss": 0.4154, + "step": 6052 + }, + { + "epoch": 2.2824283559577676, + "grad_norm": 0.25258791935276115, + "learning_rate": 5.282641467117389e-05, + "loss": 0.42, + "step": 6053 + }, + { + "epoch": 2.2828054298642533, + "grad_norm": 0.20751161780686384, + "learning_rate": 5.281644037578734e-05, + "loss": 0.4218, + "step": 6054 + }, + { + "epoch": 2.283182503770739, + "grad_norm": 0.20115226986695178, + "learning_rate": 5.2806465192232854e-05, + "loss": 0.4125, + "step": 6055 + }, + { + "epoch": 2.2835595776772246, + "grad_norm": 0.19680816846206028, + "learning_rate": 5.279648912120167e-05, + "loss": 0.4141, + "step": 6056 + }, + { + "epoch": 2.2839366515837103, + "grad_norm": 0.2088598197667914, + "learning_rate": 5.278651216338516e-05, + "loss": 0.4178, + "step": 6057 + }, + { + "epoch": 2.284313725490196, + "grad_norm": 0.2364034039289844, + "learning_rate": 5.277653431947469e-05, + "loss": 0.4213, + "step": 6058 + }, + { + "epoch": 2.2846907993966816, + "grad_norm": 0.271831998489396, + "learning_rate": 5.276655559016172e-05, + "loss": 0.4284, + "step": 6059 + }, + { + "epoch": 2.2850678733031673, + "grad_norm": 0.29295737523546894, + "learning_rate": 5.275657597613779e-05, + "loss": 0.417, + "step": 6060 + }, + { + "epoch": 2.285444947209653, + "grad_norm": 0.29531080190567827, + "learning_rate": 5.274659547809444e-05, + "loss": 0.4281, + "step": 6061 + }, + { + "epoch": 2.2858220211161386, + "grad_norm": 0.24758197023838538, + "learning_rate": 5.273661409672335e-05, + "loss": 0.4207, + "step": 6062 + }, + { + "epoch": 2.2861990950226243, + "grad_norm": 0.21456518923308238, + "learning_rate": 5.272663183271618e-05, + "loss": 0.4339, + "step": 6063 + }, + { + "epoch": 2.28657616892911, + "grad_norm": 0.18798619124125687, + "learning_rate": 5.271664868676472e-05, + "loss": 0.4213, + "step": 6064 + }, + { + "epoch": 2.2869532428355956, + "grad_norm": 0.2556652040241336, + "learning_rate": 5.270666465956078e-05, + "loss": 0.4319, + "step": 6065 + }, + { + "epoch": 2.2873303167420813, + "grad_norm": 0.27975926137287643, + "learning_rate": 5.2696679751796265e-05, + "loss": 0.4181, + "step": 6066 + }, + { + "epoch": 2.287707390648567, + "grad_norm": 0.23853101394339518, + "learning_rate": 5.268669396416309e-05, + "loss": 0.4434, + "step": 6067 + }, + { + "epoch": 2.2880844645550527, + "grad_norm": 0.20851794870554, + "learning_rate": 5.2676707297353284e-05, + "loss": 0.4394, + "step": 6068 + }, + { + "epoch": 2.2884615384615383, + "grad_norm": 0.20376865743130493, + "learning_rate": 5.266671975205891e-05, + "loss": 0.4098, + "step": 6069 + }, + { + "epoch": 2.288838612368024, + "grad_norm": 0.19046107374916266, + "learning_rate": 5.265673132897208e-05, + "loss": 0.4208, + "step": 6070 + }, + { + "epoch": 2.2892156862745097, + "grad_norm": 0.19878603734887001, + "learning_rate": 5.2646742028785005e-05, + "loss": 0.4254, + "step": 6071 + }, + { + "epoch": 2.2895927601809953, + "grad_norm": 0.26127635978622044, + "learning_rate": 5.2636751852189924e-05, + "loss": 0.4242, + "step": 6072 + }, + { + "epoch": 2.289969834087481, + "grad_norm": 0.3100063001913142, + "learning_rate": 5.2626760799879144e-05, + "loss": 0.4284, + "step": 6073 + }, + { + "epoch": 2.2903469079939667, + "grad_norm": 0.2953403511571385, + "learning_rate": 5.2616768872545044e-05, + "loss": 0.4134, + "step": 6074 + }, + { + "epoch": 2.2907239819004523, + "grad_norm": 0.30377887136564724, + "learning_rate": 5.260677607088006e-05, + "loss": 0.4071, + "step": 6075 + }, + { + "epoch": 2.291101055806938, + "grad_norm": 0.27438047330287363, + "learning_rate": 5.259678239557667e-05, + "loss": 0.4187, + "step": 6076 + }, + { + "epoch": 2.2914781297134237, + "grad_norm": 0.2681151743895565, + "learning_rate": 5.258678784732743e-05, + "loss": 0.4284, + "step": 6077 + }, + { + "epoch": 2.2918552036199094, + "grad_norm": 0.24628894836604992, + "learning_rate": 5.257679242682496e-05, + "loss": 0.418, + "step": 6078 + }, + { + "epoch": 2.292232277526395, + "grad_norm": 0.22262145719944246, + "learning_rate": 5.256679613476192e-05, + "loss": 0.4067, + "step": 6079 + }, + { + "epoch": 2.2926093514328807, + "grad_norm": 0.20681064745251698, + "learning_rate": 5.255679897183107e-05, + "loss": 0.4191, + "step": 6080 + }, + { + "epoch": 2.2929864253393664, + "grad_norm": 0.22159432455982683, + "learning_rate": 5.2546800938725175e-05, + "loss": 0.4226, + "step": 6081 + }, + { + "epoch": 2.293363499245852, + "grad_norm": 0.21239337317664164, + "learning_rate": 5.253680203613712e-05, + "loss": 0.4153, + "step": 6082 + }, + { + "epoch": 2.2937405731523377, + "grad_norm": 0.1982825424036337, + "learning_rate": 5.25268022647598e-05, + "loss": 0.4206, + "step": 6083 + }, + { + "epoch": 2.2941176470588234, + "grad_norm": 0.2309317345316547, + "learning_rate": 5.251680162528618e-05, + "loss": 0.4206, + "step": 6084 + }, + { + "epoch": 2.294494720965309, + "grad_norm": 0.24066790256638848, + "learning_rate": 5.250680011840932e-05, + "loss": 0.4183, + "step": 6085 + }, + { + "epoch": 2.2948717948717947, + "grad_norm": 0.240666610661916, + "learning_rate": 5.24967977448223e-05, + "loss": 0.4202, + "step": 6086 + }, + { + "epoch": 2.2952488687782804, + "grad_norm": 0.21164184164574024, + "learning_rate": 5.2486794505218294e-05, + "loss": 0.424, + "step": 6087 + }, + { + "epoch": 2.295625942684766, + "grad_norm": 0.21339315874189244, + "learning_rate": 5.2476790400290496e-05, + "loss": 0.4062, + "step": 6088 + }, + { + "epoch": 2.2960030165912517, + "grad_norm": 0.21838660654272632, + "learning_rate": 5.246678543073219e-05, + "loss": 0.4189, + "step": 6089 + }, + { + "epoch": 2.2963800904977374, + "grad_norm": 0.22671117412072458, + "learning_rate": 5.24567795972367e-05, + "loss": 0.4153, + "step": 6090 + }, + { + "epoch": 2.296757164404223, + "grad_norm": 0.23314478914212622, + "learning_rate": 5.244677290049746e-05, + "loss": 0.4109, + "step": 6091 + }, + { + "epoch": 2.2971342383107087, + "grad_norm": 0.2486228039279442, + "learning_rate": 5.243676534120788e-05, + "loss": 0.4262, + "step": 6092 + }, + { + "epoch": 2.2975113122171944, + "grad_norm": 0.2205539356977048, + "learning_rate": 5.242675692006149e-05, + "loss": 0.4211, + "step": 6093 + }, + { + "epoch": 2.29788838612368, + "grad_norm": 0.1985562767509226, + "learning_rate": 5.241674763775188e-05, + "loss": 0.4289, + "step": 6094 + }, + { + "epoch": 2.2982654600301657, + "grad_norm": 0.21355461224845532, + "learning_rate": 5.240673749497266e-05, + "loss": 0.4186, + "step": 6095 + }, + { + "epoch": 2.2986425339366514, + "grad_norm": 0.2693301423434791, + "learning_rate": 5.2396726492417546e-05, + "loss": 0.4165, + "step": 6096 + }, + { + "epoch": 2.299019607843137, + "grad_norm": 0.2755124238243525, + "learning_rate": 5.238671463078029e-05, + "loss": 0.4232, + "step": 6097 + }, + { + "epoch": 2.2993966817496228, + "grad_norm": 0.20868345540601893, + "learning_rate": 5.2376701910754686e-05, + "loss": 0.4191, + "step": 6098 + }, + { + "epoch": 2.2997737556561084, + "grad_norm": 0.21607484765261242, + "learning_rate": 5.2366688333034613e-05, + "loss": 0.4162, + "step": 6099 + }, + { + "epoch": 2.300150829562594, + "grad_norm": 0.25327551038052587, + "learning_rate": 5.2356673898314015e-05, + "loss": 0.4146, + "step": 6100 + }, + { + "epoch": 2.3005279034690798, + "grad_norm": 0.2298038904739611, + "learning_rate": 5.2346658607286875e-05, + "loss": 0.4141, + "step": 6101 + }, + { + "epoch": 2.3009049773755654, + "grad_norm": 0.15749519548909266, + "learning_rate": 5.233664246064724e-05, + "loss": 0.4156, + "step": 6102 + }, + { + "epoch": 2.301282051282051, + "grad_norm": 0.22512825870885508, + "learning_rate": 5.232662545908923e-05, + "loss": 0.4163, + "step": 6103 + }, + { + "epoch": 2.3016591251885368, + "grad_norm": 0.28421398348585236, + "learning_rate": 5.231660760330701e-05, + "loss": 0.4206, + "step": 6104 + }, + { + "epoch": 2.3020361990950224, + "grad_norm": 0.2725888474254515, + "learning_rate": 5.230658889399481e-05, + "loss": 0.4281, + "step": 6105 + }, + { + "epoch": 2.302413273001508, + "grad_norm": 0.24870249561474175, + "learning_rate": 5.2296569331846914e-05, + "loss": 0.4222, + "step": 6106 + }, + { + "epoch": 2.302790346907994, + "grad_norm": 0.2782428670542092, + "learning_rate": 5.228654891755766e-05, + "loss": 0.4197, + "step": 6107 + }, + { + "epoch": 2.3031674208144794, + "grad_norm": 0.3228059519662814, + "learning_rate": 5.227652765182147e-05, + "loss": 0.4213, + "step": 6108 + }, + { + "epoch": 2.303544494720965, + "grad_norm": 0.34398833348796987, + "learning_rate": 5.226650553533281e-05, + "loss": 0.4235, + "step": 6109 + }, + { + "epoch": 2.303921568627451, + "grad_norm": 0.32587020512258735, + "learning_rate": 5.2256482568786196e-05, + "loss": 0.4204, + "step": 6110 + }, + { + "epoch": 2.3042986425339365, + "grad_norm": 0.3481894599900936, + "learning_rate": 5.22464587528762e-05, + "loss": 0.4177, + "step": 6111 + }, + { + "epoch": 2.304675716440422, + "grad_norm": 0.35316568332595155, + "learning_rate": 5.223643408829749e-05, + "loss": 0.4158, + "step": 6112 + }, + { + "epoch": 2.305052790346908, + "grad_norm": 0.41330501591095375, + "learning_rate": 5.222640857574474e-05, + "loss": 0.4223, + "step": 6113 + }, + { + "epoch": 2.3054298642533935, + "grad_norm": 0.4223004293684973, + "learning_rate": 5.2216382215912724e-05, + "loss": 0.4251, + "step": 6114 + }, + { + "epoch": 2.305806938159879, + "grad_norm": 0.3405085653797372, + "learning_rate": 5.220635500949627e-05, + "loss": 0.4181, + "step": 6115 + }, + { + "epoch": 2.306184012066365, + "grad_norm": 0.2432007731627995, + "learning_rate": 5.2196326957190224e-05, + "loss": 0.4298, + "step": 6116 + }, + { + "epoch": 2.3065610859728505, + "grad_norm": 0.17301783221336822, + "learning_rate": 5.2186298059689544e-05, + "loss": 0.4275, + "step": 6117 + }, + { + "epoch": 2.306938159879336, + "grad_norm": 0.29771283063420195, + "learning_rate": 5.217626831768922e-05, + "loss": 0.4204, + "step": 6118 + }, + { + "epoch": 2.307315233785822, + "grad_norm": 0.40356829614006345, + "learning_rate": 5.216623773188428e-05, + "loss": 0.4163, + "step": 6119 + }, + { + "epoch": 2.3076923076923075, + "grad_norm": 0.41001800392533233, + "learning_rate": 5.215620630296988e-05, + "loss": 0.4231, + "step": 6120 + }, + { + "epoch": 2.308069381598793, + "grad_norm": 0.36380915609103165, + "learning_rate": 5.214617403164116e-05, + "loss": 0.4247, + "step": 6121 + }, + { + "epoch": 2.308446455505279, + "grad_norm": 0.34802131606372066, + "learning_rate": 5.213614091859335e-05, + "loss": 0.4168, + "step": 6122 + }, + { + "epoch": 2.3088235294117645, + "grad_norm": 0.2847800779406889, + "learning_rate": 5.212610696452174e-05, + "loss": 0.4225, + "step": 6123 + }, + { + "epoch": 2.30920060331825, + "grad_norm": 0.21207102341546416, + "learning_rate": 5.2116072170121686e-05, + "loss": 0.424, + "step": 6124 + }, + { + "epoch": 2.309577677224736, + "grad_norm": 0.22080253573850486, + "learning_rate": 5.210603653608857e-05, + "loss": 0.4223, + "step": 6125 + }, + { + "epoch": 2.3099547511312215, + "grad_norm": 0.22151945704822915, + "learning_rate": 5.209600006311785e-05, + "loss": 0.4305, + "step": 6126 + }, + { + "epoch": 2.310331825037707, + "grad_norm": 0.23144272682217884, + "learning_rate": 5.2085962751905074e-05, + "loss": 0.4192, + "step": 6127 + }, + { + "epoch": 2.310708898944193, + "grad_norm": 0.3121040460366219, + "learning_rate": 5.207592460314578e-05, + "loss": 0.4216, + "step": 6128 + }, + { + "epoch": 2.3110859728506785, + "grad_norm": 0.3702532349119877, + "learning_rate": 5.206588561753563e-05, + "loss": 0.4167, + "step": 6129 + }, + { + "epoch": 2.311463046757164, + "grad_norm": 0.3540512027545107, + "learning_rate": 5.205584579577031e-05, + "loss": 0.4271, + "step": 6130 + }, + { + "epoch": 2.31184012066365, + "grad_norm": 0.2751830327176547, + "learning_rate": 5.204580513854556e-05, + "loss": 0.4198, + "step": 6131 + }, + { + "epoch": 2.3122171945701355, + "grad_norm": 0.2777336203357115, + "learning_rate": 5.203576364655721e-05, + "loss": 0.418, + "step": 6132 + }, + { + "epoch": 2.312594268476621, + "grad_norm": 0.3189026454755855, + "learning_rate": 5.2025721320501116e-05, + "loss": 0.4114, + "step": 6133 + }, + { + "epoch": 2.312971342383107, + "grad_norm": 0.29584208903882836, + "learning_rate": 5.20156781610732e-05, + "loss": 0.4279, + "step": 6134 + }, + { + "epoch": 2.3133484162895925, + "grad_norm": 0.2663663557243376, + "learning_rate": 5.200563416896944e-05, + "loss": 0.4199, + "step": 6135 + }, + { + "epoch": 2.313725490196078, + "grad_norm": 0.26771321180250807, + "learning_rate": 5.199558934488588e-05, + "loss": 0.4227, + "step": 6136 + }, + { + "epoch": 2.314102564102564, + "grad_norm": 0.2689336125379201, + "learning_rate": 5.198554368951863e-05, + "loss": 0.4221, + "step": 6137 + }, + { + "epoch": 2.3144796380090495, + "grad_norm": 0.26163427807581086, + "learning_rate": 5.1975497203563826e-05, + "loss": 0.4292, + "step": 6138 + }, + { + "epoch": 2.314856711915535, + "grad_norm": 0.2921397016802203, + "learning_rate": 5.1965449887717684e-05, + "loss": 0.4362, + "step": 6139 + }, + { + "epoch": 2.315233785822021, + "grad_norm": 0.30524997695765577, + "learning_rate": 5.195540174267649e-05, + "loss": 0.4117, + "step": 6140 + }, + { + "epoch": 2.3156108597285066, + "grad_norm": 0.2593189277735523, + "learning_rate": 5.1945352769136546e-05, + "loss": 0.4154, + "step": 6141 + }, + { + "epoch": 2.3159879336349922, + "grad_norm": 0.22149360531004275, + "learning_rate": 5.1935302967794254e-05, + "loss": 0.4231, + "step": 6142 + }, + { + "epoch": 2.316365007541478, + "grad_norm": 0.24206064672382618, + "learning_rate": 5.1925252339346055e-05, + "loss": 0.426, + "step": 6143 + }, + { + "epoch": 2.3167420814479636, + "grad_norm": 0.20909870599060662, + "learning_rate": 5.191520088448845e-05, + "loss": 0.4176, + "step": 6144 + }, + { + "epoch": 2.3171191553544492, + "grad_norm": 0.26205293973024374, + "learning_rate": 5.190514860391799e-05, + "loss": 0.4213, + "step": 6145 + }, + { + "epoch": 2.317496229260935, + "grad_norm": 0.3011994665718318, + "learning_rate": 5.189509549833128e-05, + "loss": 0.4199, + "step": 6146 + }, + { + "epoch": 2.3178733031674206, + "grad_norm": 0.2494926350494056, + "learning_rate": 5.188504156842502e-05, + "loss": 0.4218, + "step": 6147 + }, + { + "epoch": 2.3182503770739062, + "grad_norm": 0.2010291054908089, + "learning_rate": 5.187498681489592e-05, + "loss": 0.4199, + "step": 6148 + }, + { + "epoch": 2.318627450980392, + "grad_norm": 0.21744424120132885, + "learning_rate": 5.1864931238440754e-05, + "loss": 0.436, + "step": 6149 + }, + { + "epoch": 2.3190045248868776, + "grad_norm": 0.2605233055091198, + "learning_rate": 5.185487483975639e-05, + "loss": 0.4191, + "step": 6150 + }, + { + "epoch": 2.3193815987933633, + "grad_norm": 0.31636963787343314, + "learning_rate": 5.184481761953971e-05, + "loss": 0.4258, + "step": 6151 + }, + { + "epoch": 2.319758672699849, + "grad_norm": 0.32586413766874217, + "learning_rate": 5.183475957848767e-05, + "loss": 0.4261, + "step": 6152 + }, + { + "epoch": 2.3201357466063346, + "grad_norm": 0.28256373799454837, + "learning_rate": 5.1824700717297294e-05, + "loss": 0.4183, + "step": 6153 + }, + { + "epoch": 2.3205128205128207, + "grad_norm": 0.27087935928662626, + "learning_rate": 5.181464103666565e-05, + "loss": 0.4098, + "step": 6154 + }, + { + "epoch": 2.3208898944193064, + "grad_norm": 0.2691663631557036, + "learning_rate": 5.180458053728986e-05, + "loss": 0.4197, + "step": 6155 + }, + { + "epoch": 2.321266968325792, + "grad_norm": 0.22463069563695442, + "learning_rate": 5.1794519219867106e-05, + "loss": 0.4286, + "step": 6156 + }, + { + "epoch": 2.3216440422322777, + "grad_norm": 0.21507893773753273, + "learning_rate": 5.178445708509464e-05, + "loss": 0.4141, + "step": 6157 + }, + { + "epoch": 2.3220211161387634, + "grad_norm": 0.25015307740686565, + "learning_rate": 5.177439413366975e-05, + "loss": 0.4246, + "step": 6158 + }, + { + "epoch": 2.322398190045249, + "grad_norm": 0.2414688060820557, + "learning_rate": 5.176433036628977e-05, + "loss": 0.4135, + "step": 6159 + }, + { + "epoch": 2.3227752639517347, + "grad_norm": 0.20153466368027503, + "learning_rate": 5.1754265783652146e-05, + "loss": 0.4119, + "step": 6160 + }, + { + "epoch": 2.3231523378582204, + "grad_norm": 0.22830914320392579, + "learning_rate": 5.1744200386454335e-05, + "loss": 0.4145, + "step": 6161 + }, + { + "epoch": 2.323529411764706, + "grad_norm": 0.2968449523842248, + "learning_rate": 5.173413417539385e-05, + "loss": 0.4165, + "step": 6162 + }, + { + "epoch": 2.3239064856711917, + "grad_norm": 0.30251041661498834, + "learning_rate": 5.1724067151168274e-05, + "loss": 0.4128, + "step": 6163 + }, + { + "epoch": 2.3242835595776774, + "grad_norm": 0.24938884277501253, + "learning_rate": 5.171399931447524e-05, + "loss": 0.4106, + "step": 6164 + }, + { + "epoch": 2.324660633484163, + "grad_norm": 0.2292497536606942, + "learning_rate": 5.1703930666012445e-05, + "loss": 0.4317, + "step": 6165 + }, + { + "epoch": 2.3250377073906487, + "grad_norm": 0.19997844451815056, + "learning_rate": 5.169386120647763e-05, + "loss": 0.418, + "step": 6166 + }, + { + "epoch": 2.3254147812971344, + "grad_norm": 0.2239623972652854, + "learning_rate": 5.168379093656862e-05, + "loss": 0.4103, + "step": 6167 + }, + { + "epoch": 2.32579185520362, + "grad_norm": 0.25823258802654037, + "learning_rate": 5.167371985698325e-05, + "loss": 0.4342, + "step": 6168 + }, + { + "epoch": 2.3261689291101058, + "grad_norm": 0.2965798394319464, + "learning_rate": 5.1663647968419456e-05, + "loss": 0.4116, + "step": 6169 + }, + { + "epoch": 2.3265460030165914, + "grad_norm": 0.28715909132221884, + "learning_rate": 5.165357527157521e-05, + "loss": 0.4258, + "step": 6170 + }, + { + "epoch": 2.326923076923077, + "grad_norm": 0.26335032319408314, + "learning_rate": 5.164350176714852e-05, + "loss": 0.4234, + "step": 6171 + }, + { + "epoch": 2.3273001508295628, + "grad_norm": 0.25172936348086633, + "learning_rate": 5.163342745583749e-05, + "loss": 0.4195, + "step": 6172 + }, + { + "epoch": 2.3276772247360484, + "grad_norm": 0.2717142128189568, + "learning_rate": 5.162335233834026e-05, + "loss": 0.4346, + "step": 6173 + }, + { + "epoch": 2.328054298642534, + "grad_norm": 0.2383491960149159, + "learning_rate": 5.1613276415355024e-05, + "loss": 0.4247, + "step": 6174 + }, + { + "epoch": 2.3284313725490198, + "grad_norm": 0.20967661413386296, + "learning_rate": 5.1603199687580016e-05, + "loss": 0.4201, + "step": 6175 + }, + { + "epoch": 2.3288084464555054, + "grad_norm": 0.23186551218294674, + "learning_rate": 5.159312215571358e-05, + "loss": 0.408, + "step": 6176 + }, + { + "epoch": 2.329185520361991, + "grad_norm": 0.24096969588632583, + "learning_rate": 5.158304382045406e-05, + "loss": 0.4217, + "step": 6177 + }, + { + "epoch": 2.329562594268477, + "grad_norm": 0.22777373923129363, + "learning_rate": 5.157296468249987e-05, + "loss": 0.4272, + "step": 6178 + }, + { + "epoch": 2.3299396681749625, + "grad_norm": 0.22485239703196835, + "learning_rate": 5.1562884742549506e-05, + "loss": 0.4216, + "step": 6179 + }, + { + "epoch": 2.330316742081448, + "grad_norm": 0.18124329134303335, + "learning_rate": 5.155280400130147e-05, + "loss": 0.4206, + "step": 6180 + }, + { + "epoch": 2.330693815987934, + "grad_norm": 0.22480434914447853, + "learning_rate": 5.1542722459454364e-05, + "loss": 0.4145, + "step": 6181 + }, + { + "epoch": 2.3310708898944195, + "grad_norm": 0.27587710170748425, + "learning_rate": 5.153264011770685e-05, + "loss": 0.4098, + "step": 6182 + }, + { + "epoch": 2.331447963800905, + "grad_norm": 0.28043740762206965, + "learning_rate": 5.152255697675759e-05, + "loss": 0.4201, + "step": 6183 + }, + { + "epoch": 2.331825037707391, + "grad_norm": 0.30453252337527004, + "learning_rate": 5.151247303730534e-05, + "loss": 0.4253, + "step": 6184 + }, + { + "epoch": 2.3322021116138765, + "grad_norm": 0.30805211587716774, + "learning_rate": 5.1502388300048944e-05, + "loss": 0.4215, + "step": 6185 + }, + { + "epoch": 2.332579185520362, + "grad_norm": 0.321966948241494, + "learning_rate": 5.149230276568723e-05, + "loss": 0.4116, + "step": 6186 + }, + { + "epoch": 2.332956259426848, + "grad_norm": 0.33645232067253156, + "learning_rate": 5.148221643491913e-05, + "loss": 0.4152, + "step": 6187 + }, + { + "epoch": 2.3333333333333335, + "grad_norm": 0.319933285761503, + "learning_rate": 5.1472129308443616e-05, + "loss": 0.4086, + "step": 6188 + }, + { + "epoch": 2.333710407239819, + "grad_norm": 0.23848713736396276, + "learning_rate": 5.146204138695971e-05, + "loss": 0.4167, + "step": 6189 + }, + { + "epoch": 2.334087481146305, + "grad_norm": 0.19025163340409593, + "learning_rate": 5.145195267116651e-05, + "loss": 0.4142, + "step": 6190 + }, + { + "epoch": 2.3344645550527905, + "grad_norm": 0.18435881331227943, + "learning_rate": 5.144186316176315e-05, + "loss": 0.4115, + "step": 6191 + }, + { + "epoch": 2.334841628959276, + "grad_norm": 0.1805565786963333, + "learning_rate": 5.143177285944882e-05, + "loss": 0.4244, + "step": 6192 + }, + { + "epoch": 2.335218702865762, + "grad_norm": 0.1933596620164421, + "learning_rate": 5.142168176492276e-05, + "loss": 0.422, + "step": 6193 + }, + { + "epoch": 2.3355957767722475, + "grad_norm": 0.21835131410278477, + "learning_rate": 5.14115898788843e-05, + "loss": 0.4221, + "step": 6194 + }, + { + "epoch": 2.335972850678733, + "grad_norm": 0.24738517180569897, + "learning_rate": 5.140149720203278e-05, + "loss": 0.4191, + "step": 6195 + }, + { + "epoch": 2.336349924585219, + "grad_norm": 0.2504648839356429, + "learning_rate": 5.1391403735067605e-05, + "loss": 0.4219, + "step": 6196 + }, + { + "epoch": 2.3367269984917045, + "grad_norm": 0.2571044735920502, + "learning_rate": 5.1381309478688276e-05, + "loss": 0.4164, + "step": 6197 + }, + { + "epoch": 2.33710407239819, + "grad_norm": 0.25971565343092734, + "learning_rate": 5.1371214433594274e-05, + "loss": 0.4191, + "step": 6198 + }, + { + "epoch": 2.337481146304676, + "grad_norm": 0.19939729390674943, + "learning_rate": 5.1361118600485205e-05, + "loss": 0.4024, + "step": 6199 + }, + { + "epoch": 2.3378582202111615, + "grad_norm": 0.1863919790194985, + "learning_rate": 5.1351021980060706e-05, + "loss": 0.4264, + "step": 6200 + }, + { + "epoch": 2.338235294117647, + "grad_norm": 0.23530197278808038, + "learning_rate": 5.134092457302044e-05, + "loss": 0.4284, + "step": 6201 + }, + { + "epoch": 2.338612368024133, + "grad_norm": 0.2847676805804324, + "learning_rate": 5.133082638006416e-05, + "loss": 0.4281, + "step": 6202 + }, + { + "epoch": 2.3389894419306185, + "grad_norm": 0.2947830026908422, + "learning_rate": 5.1320727401891674e-05, + "loss": 0.4269, + "step": 6203 + }, + { + "epoch": 2.339366515837104, + "grad_norm": 0.36128552592158997, + "learning_rate": 5.1310627639202806e-05, + "loss": 0.4293, + "step": 6204 + }, + { + "epoch": 2.33974358974359, + "grad_norm": 0.36437245295569215, + "learning_rate": 5.1300527092697485e-05, + "loss": 0.4213, + "step": 6205 + }, + { + "epoch": 2.3401206636500755, + "grad_norm": 0.3024122026752855, + "learning_rate": 5.1290425763075646e-05, + "loss": 0.4168, + "step": 6206 + }, + { + "epoch": 2.340497737556561, + "grad_norm": 0.2290087519638166, + "learning_rate": 5.128032365103734e-05, + "loss": 0.4133, + "step": 6207 + }, + { + "epoch": 2.340874811463047, + "grad_norm": 0.19845563592054874, + "learning_rate": 5.127022075728259e-05, + "loss": 0.4193, + "step": 6208 + }, + { + "epoch": 2.3412518853695325, + "grad_norm": 0.21844419586280678, + "learning_rate": 5.126011708251154e-05, + "loss": 0.4247, + "step": 6209 + }, + { + "epoch": 2.341628959276018, + "grad_norm": 0.2444040899517727, + "learning_rate": 5.1250012627424374e-05, + "loss": 0.4213, + "step": 6210 + }, + { + "epoch": 2.342006033182504, + "grad_norm": 0.21396573343895192, + "learning_rate": 5.123990739272131e-05, + "loss": 0.4025, + "step": 6211 + }, + { + "epoch": 2.3423831070889896, + "grad_norm": 0.18525430791257136, + "learning_rate": 5.1229801379102624e-05, + "loss": 0.4275, + "step": 6212 + }, + { + "epoch": 2.3427601809954752, + "grad_norm": 0.1908805716838377, + "learning_rate": 5.1219694587268675e-05, + "loss": 0.4185, + "step": 6213 + }, + { + "epoch": 2.343137254901961, + "grad_norm": 0.25504674396736315, + "learning_rate": 5.120958701791983e-05, + "loss": 0.4218, + "step": 6214 + }, + { + "epoch": 2.3435143288084466, + "grad_norm": 0.270279655229604, + "learning_rate": 5.119947867175656e-05, + "loss": 0.4244, + "step": 6215 + }, + { + "epoch": 2.3438914027149322, + "grad_norm": 0.22987734364775692, + "learning_rate": 5.118936954947935e-05, + "loss": 0.4268, + "step": 6216 + }, + { + "epoch": 2.344268476621418, + "grad_norm": 0.22615016378994982, + "learning_rate": 5.117925965178876e-05, + "loss": 0.4358, + "step": 6217 + }, + { + "epoch": 2.3446455505279036, + "grad_norm": 0.22252387659024955, + "learning_rate": 5.116914897938538e-05, + "loss": 0.4138, + "step": 6218 + }, + { + "epoch": 2.3450226244343892, + "grad_norm": 0.22026279061387816, + "learning_rate": 5.1159037532969885e-05, + "loss": 0.4236, + "step": 6219 + }, + { + "epoch": 2.345399698340875, + "grad_norm": 0.2056529130953894, + "learning_rate": 5.114892531324298e-05, + "loss": 0.417, + "step": 6220 + }, + { + "epoch": 2.3457767722473606, + "grad_norm": 0.20052276015801584, + "learning_rate": 5.113881232090546e-05, + "loss": 0.4095, + "step": 6221 + }, + { + "epoch": 2.3461538461538463, + "grad_norm": 0.2566952620211213, + "learning_rate": 5.112869855665811e-05, + "loss": 0.4315, + "step": 6222 + }, + { + "epoch": 2.346530920060332, + "grad_norm": 0.2680808069792327, + "learning_rate": 5.111858402120183e-05, + "loss": 0.4182, + "step": 6223 + }, + { + "epoch": 2.3469079939668176, + "grad_norm": 0.28344353029269775, + "learning_rate": 5.1108468715237534e-05, + "loss": 0.4226, + "step": 6224 + }, + { + "epoch": 2.3472850678733033, + "grad_norm": 0.3245392097548651, + "learning_rate": 5.1098352639466216e-05, + "loss": 0.4311, + "step": 6225 + }, + { + "epoch": 2.347662141779789, + "grad_norm": 0.3534113054839329, + "learning_rate": 5.1088235794588894e-05, + "loss": 0.4187, + "step": 6226 + }, + { + "epoch": 2.3480392156862746, + "grad_norm": 0.2909131744750243, + "learning_rate": 5.1078118181306675e-05, + "loss": 0.4108, + "step": 6227 + }, + { + "epoch": 2.3484162895927603, + "grad_norm": 0.21798674652497158, + "learning_rate": 5.106799980032069e-05, + "loss": 0.4151, + "step": 6228 + }, + { + "epoch": 2.348793363499246, + "grad_norm": 0.2132897426452603, + "learning_rate": 5.1057880652332134e-05, + "loss": 0.4258, + "step": 6229 + }, + { + "epoch": 2.3491704374057316, + "grad_norm": 0.18391694547231793, + "learning_rate": 5.104776073804226e-05, + "loss": 0.4111, + "step": 6230 + }, + { + "epoch": 2.3495475113122173, + "grad_norm": 0.21582978015948084, + "learning_rate": 5.1037640058152366e-05, + "loss": 0.4255, + "step": 6231 + }, + { + "epoch": 2.349924585218703, + "grad_norm": 0.28598320243664666, + "learning_rate": 5.10275186133638e-05, + "loss": 0.4073, + "step": 6232 + }, + { + "epoch": 2.3503016591251886, + "grad_norm": 0.29854859317474064, + "learning_rate": 5.1017396404377974e-05, + "loss": 0.4192, + "step": 6233 + }, + { + "epoch": 2.3506787330316743, + "grad_norm": 0.25968824682728997, + "learning_rate": 5.100727343189637e-05, + "loss": 0.4156, + "step": 6234 + }, + { + "epoch": 2.35105580693816, + "grad_norm": 0.20097749470104093, + "learning_rate": 5.0997149696620456e-05, + "loss": 0.4164, + "step": 6235 + }, + { + "epoch": 2.3514328808446456, + "grad_norm": 0.17967331302279482, + "learning_rate": 5.0987025199251836e-05, + "loss": 0.4254, + "step": 6236 + }, + { + "epoch": 2.3518099547511313, + "grad_norm": 0.2248134117707814, + "learning_rate": 5.0976899940492124e-05, + "loss": 0.4196, + "step": 6237 + }, + { + "epoch": 2.352187028657617, + "grad_norm": 0.23765045281626726, + "learning_rate": 5.096677392104297e-05, + "loss": 0.4328, + "step": 6238 + }, + { + "epoch": 2.3525641025641026, + "grad_norm": 0.23867953547366505, + "learning_rate": 5.09566471416061e-05, + "loss": 0.4145, + "step": 6239 + }, + { + "epoch": 2.3529411764705883, + "grad_norm": 0.24134730099975538, + "learning_rate": 5.0946519602883326e-05, + "loss": 0.4116, + "step": 6240 + }, + { + "epoch": 2.353318250377074, + "grad_norm": 0.220355568731958, + "learning_rate": 5.093639130557644e-05, + "loss": 0.4081, + "step": 6241 + }, + { + "epoch": 2.3536953242835597, + "grad_norm": 0.2026169550296106, + "learning_rate": 5.0926262250387334e-05, + "loss": 0.4175, + "step": 6242 + }, + { + "epoch": 2.3540723981900453, + "grad_norm": 0.20701131905576176, + "learning_rate": 5.0916132438017955e-05, + "loss": 0.4102, + "step": 6243 + }, + { + "epoch": 2.354449472096531, + "grad_norm": 0.23355542473943564, + "learning_rate": 5.0906001869170275e-05, + "loss": 0.4315, + "step": 6244 + }, + { + "epoch": 2.3548265460030167, + "grad_norm": 0.2067944684460148, + "learning_rate": 5.089587054454633e-05, + "loss": 0.4151, + "step": 6245 + }, + { + "epoch": 2.3552036199095023, + "grad_norm": 0.21107408952852316, + "learning_rate": 5.088573846484824e-05, + "loss": 0.4229, + "step": 6246 + }, + { + "epoch": 2.355580693815988, + "grad_norm": 0.2512581537843379, + "learning_rate": 5.0875605630778116e-05, + "loss": 0.4153, + "step": 6247 + }, + { + "epoch": 2.3559577677224737, + "grad_norm": 0.24696851565553543, + "learning_rate": 5.0865472043038164e-05, + "loss": 0.4165, + "step": 6248 + }, + { + "epoch": 2.3563348416289593, + "grad_norm": 0.2998448511137827, + "learning_rate": 5.085533770233066e-05, + "loss": 0.4241, + "step": 6249 + }, + { + "epoch": 2.356711915535445, + "grad_norm": 0.3473676178294892, + "learning_rate": 5.0845202609357865e-05, + "loss": 0.4101, + "step": 6250 + }, + { + "epoch": 2.3570889894419307, + "grad_norm": 0.3504419172614909, + "learning_rate": 5.083506676482214e-05, + "loss": 0.4241, + "step": 6251 + }, + { + "epoch": 2.3574660633484164, + "grad_norm": 0.2982066037163341, + "learning_rate": 5.082493016942592e-05, + "loss": 0.4284, + "step": 6252 + }, + { + "epoch": 2.357843137254902, + "grad_norm": 0.30677113664506506, + "learning_rate": 5.081479282387163e-05, + "loss": 0.422, + "step": 6253 + }, + { + "epoch": 2.3582202111613877, + "grad_norm": 0.3013210583356914, + "learning_rate": 5.080465472886179e-05, + "loss": 0.4134, + "step": 6254 + }, + { + "epoch": 2.3585972850678734, + "grad_norm": 0.2915618159266526, + "learning_rate": 5.079451588509897e-05, + "loss": 0.4218, + "step": 6255 + }, + { + "epoch": 2.358974358974359, + "grad_norm": 0.3047804884236238, + "learning_rate": 5.078437629328576e-05, + "loss": 0.423, + "step": 6256 + }, + { + "epoch": 2.3593514328808447, + "grad_norm": 0.28618464928988996, + "learning_rate": 5.0774235954124835e-05, + "loss": 0.4286, + "step": 6257 + }, + { + "epoch": 2.3597285067873304, + "grad_norm": 0.2278283347879529, + "learning_rate": 5.076409486831894e-05, + "loss": 0.4203, + "step": 6258 + }, + { + "epoch": 2.360105580693816, + "grad_norm": 0.22210412310563923, + "learning_rate": 5.07539530365708e-05, + "loss": 0.4083, + "step": 6259 + }, + { + "epoch": 2.3604826546003017, + "grad_norm": 0.24504818454363755, + "learning_rate": 5.074381045958325e-05, + "loss": 0.4252, + "step": 6260 + }, + { + "epoch": 2.3608597285067874, + "grad_norm": 0.23837279194360336, + "learning_rate": 5.073366713805918e-05, + "loss": 0.4177, + "step": 6261 + }, + { + "epoch": 2.361236802413273, + "grad_norm": 0.226171810991608, + "learning_rate": 5.072352307270149e-05, + "loss": 0.4203, + "step": 6262 + }, + { + "epoch": 2.3616138763197587, + "grad_norm": 0.218717955403284, + "learning_rate": 5.071337826421316e-05, + "loss": 0.4162, + "step": 6263 + }, + { + "epoch": 2.3619909502262444, + "grad_norm": 0.25175117948063186, + "learning_rate": 5.070323271329722e-05, + "loss": 0.4176, + "step": 6264 + }, + { + "epoch": 2.36236802413273, + "grad_norm": 0.29733965707020255, + "learning_rate": 5.0693086420656754e-05, + "loss": 0.4192, + "step": 6265 + }, + { + "epoch": 2.3627450980392157, + "grad_norm": 0.3238094796897895, + "learning_rate": 5.068293938699487e-05, + "loss": 0.4269, + "step": 6266 + }, + { + "epoch": 2.3631221719457014, + "grad_norm": 0.2953241161558985, + "learning_rate": 5.0672791613014775e-05, + "loss": 0.4201, + "step": 6267 + }, + { + "epoch": 2.363499245852187, + "grad_norm": 0.32585126908206163, + "learning_rate": 5.066264309941968e-05, + "loss": 0.4177, + "step": 6268 + }, + { + "epoch": 2.3638763197586727, + "grad_norm": 0.3168805039370545, + "learning_rate": 5.0652493846912876e-05, + "loss": 0.4179, + "step": 6269 + }, + { + "epoch": 2.3642533936651584, + "grad_norm": 0.24807079508835492, + "learning_rate": 5.06423438561977e-05, + "loss": 0.4205, + "step": 6270 + }, + { + "epoch": 2.364630467571644, + "grad_norm": 0.2502278030314659, + "learning_rate": 5.0632193127977544e-05, + "loss": 0.4266, + "step": 6271 + }, + { + "epoch": 2.3650075414781297, + "grad_norm": 0.26739221357666565, + "learning_rate": 5.062204166295582e-05, + "loss": 0.4124, + "step": 6272 + }, + { + "epoch": 2.3653846153846154, + "grad_norm": 0.2891416517876238, + "learning_rate": 5.061188946183603e-05, + "loss": 0.4221, + "step": 6273 + }, + { + "epoch": 2.365761689291101, + "grad_norm": 0.33017070540628074, + "learning_rate": 5.0601736525321735e-05, + "loss": 0.4096, + "step": 6274 + }, + { + "epoch": 2.3661387631975868, + "grad_norm": 0.3241370412281608, + "learning_rate": 5.0591582854116476e-05, + "loss": 0.4095, + "step": 6275 + }, + { + "epoch": 2.3665158371040724, + "grad_norm": 0.3156956616645148, + "learning_rate": 5.058142844892393e-05, + "loss": 0.4094, + "step": 6276 + }, + { + "epoch": 2.366892911010558, + "grad_norm": 0.29269483381787265, + "learning_rate": 5.057127331044779e-05, + "loss": 0.4372, + "step": 6277 + }, + { + "epoch": 2.3672699849170438, + "grad_norm": 0.21181009983861462, + "learning_rate": 5.056111743939177e-05, + "loss": 0.4314, + "step": 6278 + }, + { + "epoch": 2.3676470588235294, + "grad_norm": 0.21449933053323947, + "learning_rate": 5.0550960836459674e-05, + "loss": 0.4149, + "step": 6279 + }, + { + "epoch": 2.368024132730015, + "grad_norm": 0.247193149870605, + "learning_rate": 5.054080350235537e-05, + "loss": 0.4212, + "step": 6280 + }, + { + "epoch": 2.3684012066365008, + "grad_norm": 0.27286465391484416, + "learning_rate": 5.053064543778273e-05, + "loss": 0.4122, + "step": 6281 + }, + { + "epoch": 2.3687782805429864, + "grad_norm": 0.2715806936428017, + "learning_rate": 5.0520486643445685e-05, + "loss": 0.4374, + "step": 6282 + }, + { + "epoch": 2.369155354449472, + "grad_norm": 0.23241986450556182, + "learning_rate": 5.051032712004826e-05, + "loss": 0.4245, + "step": 6283 + }, + { + "epoch": 2.369532428355958, + "grad_norm": 0.21974512391036688, + "learning_rate": 5.050016686829449e-05, + "loss": 0.4163, + "step": 6284 + }, + { + "epoch": 2.3699095022624435, + "grad_norm": 0.2637522759298059, + "learning_rate": 5.0490005888888463e-05, + "loss": 0.4274, + "step": 6285 + }, + { + "epoch": 2.370286576168929, + "grad_norm": 0.27849903904858964, + "learning_rate": 5.047984418253434e-05, + "loss": 0.4157, + "step": 6286 + }, + { + "epoch": 2.370663650075415, + "grad_norm": 0.27763742621104176, + "learning_rate": 5.0469681749936304e-05, + "loss": 0.416, + "step": 6287 + }, + { + "epoch": 2.3710407239819005, + "grad_norm": 0.22378583632108914, + "learning_rate": 5.0459518591798606e-05, + "loss": 0.4135, + "step": 6288 + }, + { + "epoch": 2.371417797888386, + "grad_norm": 0.17949730986532206, + "learning_rate": 5.044935470882556e-05, + "loss": 0.4153, + "step": 6289 + }, + { + "epoch": 2.371794871794872, + "grad_norm": 0.19480057830611017, + "learning_rate": 5.04391901017215e-05, + "loss": 0.4155, + "step": 6290 + }, + { + "epoch": 2.3721719457013575, + "grad_norm": 0.261590950245675, + "learning_rate": 5.042902477119081e-05, + "loss": 0.4231, + "step": 6291 + }, + { + "epoch": 2.372549019607843, + "grad_norm": 0.2782149607102941, + "learning_rate": 5.041885871793796e-05, + "loss": 0.4049, + "step": 6292 + }, + { + "epoch": 2.372926093514329, + "grad_norm": 0.1975196924887061, + "learning_rate": 5.040869194266745e-05, + "loss": 0.4198, + "step": 6293 + }, + { + "epoch": 2.3733031674208145, + "grad_norm": 0.17412266275607854, + "learning_rate": 5.0398524446083824e-05, + "loss": 0.4257, + "step": 6294 + }, + { + "epoch": 2.3736802413273, + "grad_norm": 0.21156718636263738, + "learning_rate": 5.038835622889166e-05, + "loss": 0.4154, + "step": 6295 + }, + { + "epoch": 2.374057315233786, + "grad_norm": 0.29643491667398114, + "learning_rate": 5.0378187291795636e-05, + "loss": 0.415, + "step": 6296 + }, + { + "epoch": 2.3744343891402715, + "grad_norm": 0.35227391263119223, + "learning_rate": 5.036801763550044e-05, + "loss": 0.4314, + "step": 6297 + }, + { + "epoch": 2.374811463046757, + "grad_norm": 0.3392783660955526, + "learning_rate": 5.035784726071081e-05, + "loss": 0.4273, + "step": 6298 + }, + { + "epoch": 2.375188536953243, + "grad_norm": 0.31467745401934727, + "learning_rate": 5.0347676168131554e-05, + "loss": 0.4186, + "step": 6299 + }, + { + "epoch": 2.3755656108597285, + "grad_norm": 0.2804315651804067, + "learning_rate": 5.033750435846752e-05, + "loss": 0.4165, + "step": 6300 + }, + { + "epoch": 2.375942684766214, + "grad_norm": 0.26046091446493236, + "learning_rate": 5.032733183242361e-05, + "loss": 0.4223, + "step": 6301 + }, + { + "epoch": 2.3763197586727, + "grad_norm": 0.20341856321299737, + "learning_rate": 5.031715859070476e-05, + "loss": 0.411, + "step": 6302 + }, + { + "epoch": 2.3766968325791855, + "grad_norm": 0.23653933955337428, + "learning_rate": 5.0306984634015966e-05, + "loss": 0.4297, + "step": 6303 + }, + { + "epoch": 2.377073906485671, + "grad_norm": 0.27241371883709103, + "learning_rate": 5.029680996306228e-05, + "loss": 0.4223, + "step": 6304 + }, + { + "epoch": 2.377450980392157, + "grad_norm": 0.2673146390155438, + "learning_rate": 5.02866345785488e-05, + "loss": 0.4093, + "step": 6305 + }, + { + "epoch": 2.3778280542986425, + "grad_norm": 0.23391091928767963, + "learning_rate": 5.0276458481180665e-05, + "loss": 0.418, + "step": 6306 + }, + { + "epoch": 2.378205128205128, + "grad_norm": 0.2219311651718214, + "learning_rate": 5.0266281671663075e-05, + "loss": 0.417, + "step": 6307 + }, + { + "epoch": 2.378582202111614, + "grad_norm": 0.26171969999244094, + "learning_rate": 5.025610415070127e-05, + "loss": 0.4049, + "step": 6308 + }, + { + "epoch": 2.3789592760180995, + "grad_norm": 0.32063601876269227, + "learning_rate": 5.024592591900054e-05, + "loss": 0.4225, + "step": 6309 + }, + { + "epoch": 2.379336349924585, + "grad_norm": 0.34321528388914513, + "learning_rate": 5.0235746977266245e-05, + "loss": 0.426, + "step": 6310 + }, + { + "epoch": 2.379713423831071, + "grad_norm": 0.2622825224161943, + "learning_rate": 5.022556732620375e-05, + "loss": 0.4228, + "step": 6311 + }, + { + "epoch": 2.3800904977375565, + "grad_norm": 0.1743160651548171, + "learning_rate": 5.021538696651852e-05, + "loss": 0.4134, + "step": 6312 + }, + { + "epoch": 2.380467571644042, + "grad_norm": 0.25007473525323637, + "learning_rate": 5.020520589891603e-05, + "loss": 0.4044, + "step": 6313 + }, + { + "epoch": 2.380844645550528, + "grad_norm": 0.3479808372340342, + "learning_rate": 5.019502412410182e-05, + "loss": 0.4244, + "step": 6314 + }, + { + "epoch": 2.3812217194570136, + "grad_norm": 0.4027653949528489, + "learning_rate": 5.018484164278148e-05, + "loss": 0.4311, + "step": 6315 + }, + { + "epoch": 2.3815987933634992, + "grad_norm": 0.3556070459009654, + "learning_rate": 5.017465845566066e-05, + "loss": 0.4173, + "step": 6316 + }, + { + "epoch": 2.381975867269985, + "grad_norm": 0.27957001440699863, + "learning_rate": 5.016447456344502e-05, + "loss": 0.4124, + "step": 6317 + }, + { + "epoch": 2.3823529411764706, + "grad_norm": 0.29455179509547563, + "learning_rate": 5.0154289966840315e-05, + "loss": 0.4198, + "step": 6318 + }, + { + "epoch": 2.3827300150829562, + "grad_norm": 0.3317019202269376, + "learning_rate": 5.014410466655233e-05, + "loss": 0.4172, + "step": 6319 + }, + { + "epoch": 2.383107088989442, + "grad_norm": 0.28126534616795473, + "learning_rate": 5.0133918663286877e-05, + "loss": 0.4099, + "step": 6320 + }, + { + "epoch": 2.3834841628959276, + "grad_norm": 0.2517162086859197, + "learning_rate": 5.012373195774985e-05, + "loss": 0.42, + "step": 6321 + }, + { + "epoch": 2.3838612368024132, + "grad_norm": 0.27184057099756787, + "learning_rate": 5.011354455064719e-05, + "loss": 0.43, + "step": 6322 + }, + { + "epoch": 2.384238310708899, + "grad_norm": 0.3022583934828976, + "learning_rate": 5.010335644268485e-05, + "loss": 0.4191, + "step": 6323 + }, + { + "epoch": 2.3846153846153846, + "grad_norm": 0.2906151344388002, + "learning_rate": 5.0093167634568874e-05, + "loss": 0.4232, + "step": 6324 + }, + { + "epoch": 2.3849924585218703, + "grad_norm": 0.283369069451947, + "learning_rate": 5.0082978127005334e-05, + "loss": 0.4191, + "step": 6325 + }, + { + "epoch": 2.385369532428356, + "grad_norm": 0.2713769928133523, + "learning_rate": 5.007278792070036e-05, + "loss": 0.4295, + "step": 6326 + }, + { + "epoch": 2.3857466063348416, + "grad_norm": 0.20634717000397032, + "learning_rate": 5.00625970163601e-05, + "loss": 0.4243, + "step": 6327 + }, + { + "epoch": 2.3861236802413273, + "grad_norm": 0.20047397174698806, + "learning_rate": 5.005240541469081e-05, + "loss": 0.4156, + "step": 6328 + }, + { + "epoch": 2.386500754147813, + "grad_norm": 0.17612207714352252, + "learning_rate": 5.0042213116398735e-05, + "loss": 0.421, + "step": 6329 + }, + { + "epoch": 2.3868778280542986, + "grad_norm": 0.16797267854549652, + "learning_rate": 5.00320201221902e-05, + "loss": 0.4045, + "step": 6330 + }, + { + "epoch": 2.3872549019607843, + "grad_norm": 0.21331562189779496, + "learning_rate": 5.0021826432771556e-05, + "loss": 0.4144, + "step": 6331 + }, + { + "epoch": 2.38763197586727, + "grad_norm": 0.25469709908733884, + "learning_rate": 5.001163204884925e-05, + "loss": 0.4209, + "step": 6332 + }, + { + "epoch": 2.3880090497737556, + "grad_norm": 0.2549160260090877, + "learning_rate": 5.0001436971129715e-05, + "loss": 0.4168, + "step": 6333 + }, + { + "epoch": 2.3883861236802413, + "grad_norm": 0.2337227681775282, + "learning_rate": 4.9991241200319466e-05, + "loss": 0.4269, + "step": 6334 + }, + { + "epoch": 2.388763197586727, + "grad_norm": 0.18264315402815176, + "learning_rate": 4.998104473712508e-05, + "loss": 0.4137, + "step": 6335 + }, + { + "epoch": 2.3891402714932126, + "grad_norm": 0.19378467427241688, + "learning_rate": 4.9970847582253126e-05, + "loss": 0.4204, + "step": 6336 + }, + { + "epoch": 2.3895173453996983, + "grad_norm": 0.21288203965959313, + "learning_rate": 4.996064973641028e-05, + "loss": 0.4188, + "step": 6337 + }, + { + "epoch": 2.389894419306184, + "grad_norm": 0.27227926093592936, + "learning_rate": 4.9950451200303256e-05, + "loss": 0.4259, + "step": 6338 + }, + { + "epoch": 2.3902714932126696, + "grad_norm": 0.3193710616078185, + "learning_rate": 4.9940251974638786e-05, + "loss": 0.4192, + "step": 6339 + }, + { + "epoch": 2.3906485671191553, + "grad_norm": 0.2936871449448269, + "learning_rate": 4.993005206012366e-05, + "loss": 0.42, + "step": 6340 + }, + { + "epoch": 2.391025641025641, + "grad_norm": 0.28928567221340445, + "learning_rate": 4.991985145746474e-05, + "loss": 0.4102, + "step": 6341 + }, + { + "epoch": 2.3914027149321266, + "grad_norm": 0.2727574086947253, + "learning_rate": 4.9909650167368924e-05, + "loss": 0.4246, + "step": 6342 + }, + { + "epoch": 2.3917797888386123, + "grad_norm": 0.2503240745330249, + "learning_rate": 4.989944819054312e-05, + "loss": 0.4155, + "step": 6343 + }, + { + "epoch": 2.392156862745098, + "grad_norm": 0.21435513638684756, + "learning_rate": 4.988924552769436e-05, + "loss": 0.4076, + "step": 6344 + }, + { + "epoch": 2.3925339366515836, + "grad_norm": 0.22856513609968657, + "learning_rate": 4.987904217952964e-05, + "loss": 0.4211, + "step": 6345 + }, + { + "epoch": 2.3929110105580693, + "grad_norm": 0.2157301565839361, + "learning_rate": 4.9868838146756064e-05, + "loss": 0.426, + "step": 6346 + }, + { + "epoch": 2.393288084464555, + "grad_norm": 0.182029398445113, + "learning_rate": 4.985863343008077e-05, + "loss": 0.4172, + "step": 6347 + }, + { + "epoch": 2.3936651583710407, + "grad_norm": 0.17584145720840094, + "learning_rate": 4.9848428030210906e-05, + "loss": 0.419, + "step": 6348 + }, + { + "epoch": 2.3940422322775263, + "grad_norm": 0.28408803560846657, + "learning_rate": 4.9838221947853725e-05, + "loss": 0.4318, + "step": 6349 + }, + { + "epoch": 2.394419306184012, + "grad_norm": 0.30635187189035157, + "learning_rate": 4.9828015183716485e-05, + "loss": 0.406, + "step": 6350 + }, + { + "epoch": 2.3947963800904977, + "grad_norm": 0.22973666926411498, + "learning_rate": 4.9817807738506514e-05, + "loss": 0.4231, + "step": 6351 + }, + { + "epoch": 2.3951734539969833, + "grad_norm": 0.1859336109693651, + "learning_rate": 4.980759961293117e-05, + "loss": 0.4175, + "step": 6352 + }, + { + "epoch": 2.395550527903469, + "grad_norm": 0.1748913047693247, + "learning_rate": 4.979739080769788e-05, + "loss": 0.4255, + "step": 6353 + }, + { + "epoch": 2.3959276018099547, + "grad_norm": 0.1777313487772577, + "learning_rate": 4.978718132351409e-05, + "loss": 0.4152, + "step": 6354 + }, + { + "epoch": 2.3963046757164403, + "grad_norm": 0.20874153575591586, + "learning_rate": 4.977697116108732e-05, + "loss": 0.4251, + "step": 6355 + }, + { + "epoch": 2.396681749622926, + "grad_norm": 0.1754238657834431, + "learning_rate": 4.976676032112512e-05, + "loss": 0.4049, + "step": 6356 + }, + { + "epoch": 2.3970588235294117, + "grad_norm": 0.20274274270189918, + "learning_rate": 4.975654880433509e-05, + "loss": 0.4239, + "step": 6357 + }, + { + "epoch": 2.3974358974358974, + "grad_norm": 0.22470931493093552, + "learning_rate": 4.974633661142488e-05, + "loss": 0.4054, + "step": 6358 + }, + { + "epoch": 2.397812971342383, + "grad_norm": 0.21676016811755983, + "learning_rate": 4.97361237431022e-05, + "loss": 0.4214, + "step": 6359 + }, + { + "epoch": 2.3981900452488687, + "grad_norm": 0.24408936365806502, + "learning_rate": 4.9725910200074764e-05, + "loss": 0.418, + "step": 6360 + }, + { + "epoch": 2.3985671191553544, + "grad_norm": 0.22814875194714962, + "learning_rate": 4.971569598305039e-05, + "loss": 0.4032, + "step": 6361 + }, + { + "epoch": 2.39894419306184, + "grad_norm": 0.2530168264507292, + "learning_rate": 4.97054810927369e-05, + "loss": 0.4195, + "step": 6362 + }, + { + "epoch": 2.3993212669683257, + "grad_norm": 0.282622961062175, + "learning_rate": 4.969526552984218e-05, + "loss": 0.414, + "step": 6363 + }, + { + "epoch": 2.3996983408748114, + "grad_norm": 0.25535804667025025, + "learning_rate": 4.968504929507415e-05, + "loss": 0.4261, + "step": 6364 + }, + { + "epoch": 2.400075414781297, + "grad_norm": 0.23542597566589044, + "learning_rate": 4.967483238914081e-05, + "loss": 0.4126, + "step": 6365 + }, + { + "epoch": 2.4004524886877827, + "grad_norm": 0.21186297699407558, + "learning_rate": 4.9664614812750146e-05, + "loss": 0.4239, + "step": 6366 + }, + { + "epoch": 2.4008295625942684, + "grad_norm": 0.16873660300455626, + "learning_rate": 4.965439656661026e-05, + "loss": 0.4187, + "step": 6367 + }, + { + "epoch": 2.401206636500754, + "grad_norm": 0.1878951356813118, + "learning_rate": 4.964417765142927e-05, + "loss": 0.4232, + "step": 6368 + }, + { + "epoch": 2.4015837104072397, + "grad_norm": 0.23205904759598536, + "learning_rate": 4.9633958067915296e-05, + "loss": 0.4269, + "step": 6369 + }, + { + "epoch": 2.4019607843137254, + "grad_norm": 0.2687159242766847, + "learning_rate": 4.962373781677658e-05, + "loss": 0.4239, + "step": 6370 + }, + { + "epoch": 2.402337858220211, + "grad_norm": 0.29778286113831176, + "learning_rate": 4.9613516898721384e-05, + "loss": 0.4262, + "step": 6371 + }, + { + "epoch": 2.4027149321266967, + "grad_norm": 0.2935227995921288, + "learning_rate": 4.960329531445798e-05, + "loss": 0.4261, + "step": 6372 + }, + { + "epoch": 2.4030920060331824, + "grad_norm": 0.25880440406322375, + "learning_rate": 4.9593073064694734e-05, + "loss": 0.4056, + "step": 6373 + }, + { + "epoch": 2.403469079939668, + "grad_norm": 0.19288112445055147, + "learning_rate": 4.958285015014004e-05, + "loss": 0.4327, + "step": 6374 + }, + { + "epoch": 2.4038461538461537, + "grad_norm": 0.18766074049571763, + "learning_rate": 4.9572626571502316e-05, + "loss": 0.4083, + "step": 6375 + }, + { + "epoch": 2.4042232277526394, + "grad_norm": 0.23411552381391243, + "learning_rate": 4.956240232949007e-05, + "loss": 0.4204, + "step": 6376 + }, + { + "epoch": 2.404600301659125, + "grad_norm": 0.24530084343325928, + "learning_rate": 4.955217742481183e-05, + "loss": 0.4145, + "step": 6377 + }, + { + "epoch": 2.4049773755656108, + "grad_norm": 0.24205419850123905, + "learning_rate": 4.9541951858176154e-05, + "loss": 0.4267, + "step": 6378 + }, + { + "epoch": 2.4053544494720964, + "grad_norm": 0.28692922172697655, + "learning_rate": 4.953172563029168e-05, + "loss": 0.421, + "step": 6379 + }, + { + "epoch": 2.405731523378582, + "grad_norm": 0.2750312586451725, + "learning_rate": 4.952149874186709e-05, + "loss": 0.4099, + "step": 6380 + }, + { + "epoch": 2.4061085972850678, + "grad_norm": 0.221607634055086, + "learning_rate": 4.9511271193611066e-05, + "loss": 0.4252, + "step": 6381 + }, + { + "epoch": 2.4064856711915534, + "grad_norm": 0.22603281128340746, + "learning_rate": 4.950104298623239e-05, + "loss": 0.4187, + "step": 6382 + }, + { + "epoch": 2.406862745098039, + "grad_norm": 0.1907553590291438, + "learning_rate": 4.949081412043987e-05, + "loss": 0.417, + "step": 6383 + }, + { + "epoch": 2.4072398190045248, + "grad_norm": 0.21159799909942673, + "learning_rate": 4.948058459694234e-05, + "loss": 0.4129, + "step": 6384 + }, + { + "epoch": 2.4076168929110104, + "grad_norm": 0.23515845980680813, + "learning_rate": 4.947035441644871e-05, + "loss": 0.4272, + "step": 6385 + }, + { + "epoch": 2.407993966817496, + "grad_norm": 0.29044414932139395, + "learning_rate": 4.946012357966792e-05, + "loss": 0.4233, + "step": 6386 + }, + { + "epoch": 2.408371040723982, + "grad_norm": 0.30307129837344726, + "learning_rate": 4.944989208730896e-05, + "loss": 0.4178, + "step": 6387 + }, + { + "epoch": 2.4087481146304675, + "grad_norm": 0.3020249295739278, + "learning_rate": 4.9439659940080846e-05, + "loss": 0.42, + "step": 6388 + }, + { + "epoch": 2.409125188536953, + "grad_norm": 0.26494943161688406, + "learning_rate": 4.9429427138692685e-05, + "loss": 0.4102, + "step": 6389 + }, + { + "epoch": 2.409502262443439, + "grad_norm": 0.21463427968552773, + "learning_rate": 4.9419193683853585e-05, + "loss": 0.4231, + "step": 6390 + }, + { + "epoch": 2.4098793363499245, + "grad_norm": 0.2248963206640657, + "learning_rate": 4.9408959576272715e-05, + "loss": 0.4292, + "step": 6391 + }, + { + "epoch": 2.41025641025641, + "grad_norm": 0.2288276357860903, + "learning_rate": 4.939872481665929e-05, + "loss": 0.4234, + "step": 6392 + }, + { + "epoch": 2.410633484162896, + "grad_norm": 0.252465650606723, + "learning_rate": 4.938848940572258e-05, + "loss": 0.4101, + "step": 6393 + }, + { + "epoch": 2.4110105580693815, + "grad_norm": 0.25343641786471804, + "learning_rate": 4.937825334417187e-05, + "loss": 0.4092, + "step": 6394 + }, + { + "epoch": 2.411387631975867, + "grad_norm": 0.2462308352662853, + "learning_rate": 4.936801663271653e-05, + "loss": 0.4145, + "step": 6395 + }, + { + "epoch": 2.411764705882353, + "grad_norm": 0.2268755642401999, + "learning_rate": 4.935777927206595e-05, + "loss": 0.4129, + "step": 6396 + }, + { + "epoch": 2.4121417797888385, + "grad_norm": 0.26619966065110773, + "learning_rate": 4.934754126292956e-05, + "loss": 0.4163, + "step": 6397 + }, + { + "epoch": 2.412518853695324, + "grad_norm": 0.3055042329571373, + "learning_rate": 4.9337302606016856e-05, + "loss": 0.4306, + "step": 6398 + }, + { + "epoch": 2.41289592760181, + "grad_norm": 0.3085926885321285, + "learning_rate": 4.932706330203736e-05, + "loss": 0.4189, + "step": 6399 + }, + { + "epoch": 2.4132730015082955, + "grad_norm": 0.266294459024356, + "learning_rate": 4.931682335170065e-05, + "loss": 0.4256, + "step": 6400 + }, + { + "epoch": 2.413650075414781, + "grad_norm": 0.2540773912354375, + "learning_rate": 4.930658275571635e-05, + "loss": 0.4281, + "step": 6401 + }, + { + "epoch": 2.414027149321267, + "grad_norm": 0.2840309849065162, + "learning_rate": 4.9296341514794116e-05, + "loss": 0.4115, + "step": 6402 + }, + { + "epoch": 2.4144042232277525, + "grad_norm": 0.25367739026013425, + "learning_rate": 4.928609962964366e-05, + "loss": 0.4231, + "step": 6403 + }, + { + "epoch": 2.414781297134238, + "grad_norm": 0.22324606584301626, + "learning_rate": 4.927585710097473e-05, + "loss": 0.4189, + "step": 6404 + }, + { + "epoch": 2.415158371040724, + "grad_norm": 0.25363259133914146, + "learning_rate": 4.9265613929497155e-05, + "loss": 0.4329, + "step": 6405 + }, + { + "epoch": 2.4155354449472095, + "grad_norm": 0.2795987050887978, + "learning_rate": 4.9255370115920735e-05, + "loss": 0.4151, + "step": 6406 + }, + { + "epoch": 2.415912518853695, + "grad_norm": 0.28177180362649973, + "learning_rate": 4.924512566095537e-05, + "loss": 0.4121, + "step": 6407 + }, + { + "epoch": 2.416289592760181, + "grad_norm": 0.2699399839795004, + "learning_rate": 4.923488056531101e-05, + "loss": 0.4152, + "step": 6408 + }, + { + "epoch": 2.4166666666666665, + "grad_norm": 0.27120912222571925, + "learning_rate": 4.922463482969761e-05, + "loss": 0.4083, + "step": 6409 + }, + { + "epoch": 2.417043740573152, + "grad_norm": 0.23119451472140815, + "learning_rate": 4.9214388454825205e-05, + "loss": 0.4165, + "step": 6410 + }, + { + "epoch": 2.417420814479638, + "grad_norm": 0.16820780099207722, + "learning_rate": 4.9204141441403855e-05, + "loss": 0.4164, + "step": 6411 + }, + { + "epoch": 2.4177978883861235, + "grad_norm": 0.21499674907340646, + "learning_rate": 4.919389379014366e-05, + "loss": 0.4204, + "step": 6412 + }, + { + "epoch": 2.418174962292609, + "grad_norm": 0.2233250170073784, + "learning_rate": 4.918364550175478e-05, + "loss": 0.404, + "step": 6413 + }, + { + "epoch": 2.418552036199095, + "grad_norm": 0.2211190431680652, + "learning_rate": 4.917339657694742e-05, + "loss": 0.4213, + "step": 6414 + }, + { + "epoch": 2.4189291101055805, + "grad_norm": 0.20980467208567638, + "learning_rate": 4.91631470164318e-05, + "loss": 0.4104, + "step": 6415 + }, + { + "epoch": 2.419306184012066, + "grad_norm": 0.19612581807724552, + "learning_rate": 4.915289682091823e-05, + "loss": 0.4259, + "step": 6416 + }, + { + "epoch": 2.419683257918552, + "grad_norm": 0.26220995254906104, + "learning_rate": 4.914264599111704e-05, + "loss": 0.421, + "step": 6417 + }, + { + "epoch": 2.4200603318250375, + "grad_norm": 0.2566741972728176, + "learning_rate": 4.9132394527738566e-05, + "loss": 0.415, + "step": 6418 + }, + { + "epoch": 2.420437405731523, + "grad_norm": 0.1972704492553392, + "learning_rate": 4.912214243149327e-05, + "loss": 0.4195, + "step": 6419 + }, + { + "epoch": 2.420814479638009, + "grad_norm": 0.1971010906287752, + "learning_rate": 4.91118897030916e-05, + "loss": 0.4185, + "step": 6420 + }, + { + "epoch": 2.4211915535444946, + "grad_norm": 0.23272596725522887, + "learning_rate": 4.910163634324404e-05, + "loss": 0.4264, + "step": 6421 + }, + { + "epoch": 2.4215686274509802, + "grad_norm": 0.2918843725432675, + "learning_rate": 4.909138235266116e-05, + "loss": 0.4197, + "step": 6422 + }, + { + "epoch": 2.421945701357466, + "grad_norm": 0.3734457360597904, + "learning_rate": 4.908112773205355e-05, + "loss": 0.4148, + "step": 6423 + }, + { + "epoch": 2.4223227752639516, + "grad_norm": 0.3396587780501761, + "learning_rate": 4.907087248213183e-05, + "loss": 0.4146, + "step": 6424 + }, + { + "epoch": 2.4226998491704372, + "grad_norm": 0.22038589856369142, + "learning_rate": 4.906061660360671e-05, + "loss": 0.4258, + "step": 6425 + }, + { + "epoch": 2.423076923076923, + "grad_norm": 0.20151783667898324, + "learning_rate": 4.9050360097188904e-05, + "loss": 0.4139, + "step": 6426 + }, + { + "epoch": 2.4234539969834086, + "grad_norm": 0.2569993600647681, + "learning_rate": 4.9040102963589143e-05, + "loss": 0.4135, + "step": 6427 + }, + { + "epoch": 2.4238310708898942, + "grad_norm": 0.31907408097902923, + "learning_rate": 4.902984520351828e-05, + "loss": 0.4251, + "step": 6428 + }, + { + "epoch": 2.42420814479638, + "grad_norm": 0.36710733455750105, + "learning_rate": 4.9019586817687154e-05, + "loss": 0.4189, + "step": 6429 + }, + { + "epoch": 2.4245852187028656, + "grad_norm": 0.37339151381412605, + "learning_rate": 4.900932780680666e-05, + "loss": 0.4114, + "step": 6430 + }, + { + "epoch": 2.4249622926093513, + "grad_norm": 0.34348622659297706, + "learning_rate": 4.8999068171587736e-05, + "loss": 0.4178, + "step": 6431 + }, + { + "epoch": 2.425339366515837, + "grad_norm": 0.25736553719978433, + "learning_rate": 4.8988807912741386e-05, + "loss": 0.4196, + "step": 6432 + }, + { + "epoch": 2.4257164404223226, + "grad_norm": 0.2000945941206443, + "learning_rate": 4.89785470309786e-05, + "loss": 0.429, + "step": 6433 + }, + { + "epoch": 2.4260935143288083, + "grad_norm": 0.22691869340592272, + "learning_rate": 4.896828552701048e-05, + "loss": 0.4119, + "step": 6434 + }, + { + "epoch": 2.426470588235294, + "grad_norm": 0.23265816522224486, + "learning_rate": 4.895802340154813e-05, + "loss": 0.4292, + "step": 6435 + }, + { + "epoch": 2.4268476621417796, + "grad_norm": 0.22417380971053963, + "learning_rate": 4.89477606553027e-05, + "loss": 0.4194, + "step": 6436 + }, + { + "epoch": 2.4272247360482653, + "grad_norm": 0.25504723292492343, + "learning_rate": 4.893749728898539e-05, + "loss": 0.4145, + "step": 6437 + }, + { + "epoch": 2.427601809954751, + "grad_norm": 0.24208675846046598, + "learning_rate": 4.892723330330745e-05, + "loss": 0.419, + "step": 6438 + }, + { + "epoch": 2.4279788838612366, + "grad_norm": 0.2047757719388184, + "learning_rate": 4.891696869898016e-05, + "loss": 0.4081, + "step": 6439 + }, + { + "epoch": 2.4283559577677223, + "grad_norm": 0.23422243013480523, + "learning_rate": 4.890670347671485e-05, + "loss": 0.423, + "step": 6440 + }, + { + "epoch": 2.428733031674208, + "grad_norm": 0.2548862599972601, + "learning_rate": 4.889643763722288e-05, + "loss": 0.4121, + "step": 6441 + }, + { + "epoch": 2.4291101055806936, + "grad_norm": 0.2920210766290575, + "learning_rate": 4.8886171181215686e-05, + "loss": 0.441, + "step": 6442 + }, + { + "epoch": 2.4294871794871793, + "grad_norm": 0.3397904448376867, + "learning_rate": 4.887590410940471e-05, + "loss": 0.4349, + "step": 6443 + }, + { + "epoch": 2.4298642533936654, + "grad_norm": 0.3223547111597941, + "learning_rate": 4.886563642250144e-05, + "loss": 0.4226, + "step": 6444 + }, + { + "epoch": 2.430241327300151, + "grad_norm": 0.22754228754170386, + "learning_rate": 4.8855368121217434e-05, + "loss": 0.4134, + "step": 6445 + }, + { + "epoch": 2.4306184012066367, + "grad_norm": 0.20108120303153845, + "learning_rate": 4.884509920626428e-05, + "loss": 0.414, + "step": 6446 + }, + { + "epoch": 2.4309954751131224, + "grad_norm": 0.23662575886793702, + "learning_rate": 4.883482967835359e-05, + "loss": 0.4229, + "step": 6447 + }, + { + "epoch": 2.431372549019608, + "grad_norm": 0.288656842740706, + "learning_rate": 4.882455953819704e-05, + "loss": 0.4148, + "step": 6448 + }, + { + "epoch": 2.4317496229260938, + "grad_norm": 0.2730969754670376, + "learning_rate": 4.881428878650635e-05, + "loss": 0.4207, + "step": 6449 + }, + { + "epoch": 2.4321266968325794, + "grad_norm": 0.2822043716722475, + "learning_rate": 4.880401742399325e-05, + "loss": 0.4351, + "step": 6450 + }, + { + "epoch": 2.432503770739065, + "grad_norm": 0.30728159826675155, + "learning_rate": 4.8793745451369556e-05, + "loss": 0.4212, + "step": 6451 + }, + { + "epoch": 2.4328808446455508, + "grad_norm": 0.2848672556899169, + "learning_rate": 4.878347286934711e-05, + "loss": 0.409, + "step": 6452 + }, + { + "epoch": 2.4332579185520364, + "grad_norm": 0.2264419163235649, + "learning_rate": 4.8773199678637776e-05, + "loss": 0.4114, + "step": 6453 + }, + { + "epoch": 2.433634992458522, + "grad_norm": 0.20842961524236933, + "learning_rate": 4.876292587995349e-05, + "loss": 0.4168, + "step": 6454 + }, + { + "epoch": 2.4340120663650078, + "grad_norm": 0.21738338357820702, + "learning_rate": 4.875265147400623e-05, + "loss": 0.4085, + "step": 6455 + }, + { + "epoch": 2.4343891402714934, + "grad_norm": 0.250952753667668, + "learning_rate": 4.8742376461507965e-05, + "loss": 0.4173, + "step": 6456 + }, + { + "epoch": 2.434766214177979, + "grad_norm": 0.2485740622760889, + "learning_rate": 4.873210084317077e-05, + "loss": 0.426, + "step": 6457 + }, + { + "epoch": 2.435143288084465, + "grad_norm": 0.16123946576251222, + "learning_rate": 4.872182461970672e-05, + "loss": 0.4286, + "step": 6458 + }, + { + "epoch": 2.4355203619909505, + "grad_norm": 0.20338027248851434, + "learning_rate": 4.8711547791827986e-05, + "loss": 0.4184, + "step": 6459 + }, + { + "epoch": 2.435897435897436, + "grad_norm": 0.2481287672429003, + "learning_rate": 4.87012703602467e-05, + "loss": 0.4217, + "step": 6460 + }, + { + "epoch": 2.436274509803922, + "grad_norm": 0.28435879433490185, + "learning_rate": 4.86909923256751e-05, + "loss": 0.4221, + "step": 6461 + }, + { + "epoch": 2.4366515837104075, + "grad_norm": 0.30816791667170323, + "learning_rate": 4.8680713688825446e-05, + "loss": 0.4285, + "step": 6462 + }, + { + "epoch": 2.437028657616893, + "grad_norm": 0.27068185085259755, + "learning_rate": 4.867043445041003e-05, + "loss": 0.4151, + "step": 6463 + }, + { + "epoch": 2.437405731523379, + "grad_norm": 0.29484583459699837, + "learning_rate": 4.866015461114119e-05, + "loss": 0.4131, + "step": 6464 + }, + { + "epoch": 2.4377828054298645, + "grad_norm": 0.2916880379817028, + "learning_rate": 4.864987417173133e-05, + "loss": 0.4144, + "step": 6465 + }, + { + "epoch": 2.43815987933635, + "grad_norm": 0.2649082576634845, + "learning_rate": 4.863959313289285e-05, + "loss": 0.4136, + "step": 6466 + }, + { + "epoch": 2.438536953242836, + "grad_norm": 0.2775388366596971, + "learning_rate": 4.8629311495338236e-05, + "loss": 0.4192, + "step": 6467 + }, + { + "epoch": 2.4389140271493215, + "grad_norm": 0.22906594074598868, + "learning_rate": 4.8619029259779986e-05, + "loss": 0.4169, + "step": 6468 + }, + { + "epoch": 2.439291101055807, + "grad_norm": 0.22012542314300385, + "learning_rate": 4.860874642693066e-05, + "loss": 0.4096, + "step": 6469 + }, + { + "epoch": 2.439668174962293, + "grad_norm": 0.21646232020808237, + "learning_rate": 4.859846299750283e-05, + "loss": 0.4108, + "step": 6470 + }, + { + "epoch": 2.4400452488687785, + "grad_norm": 0.2572807796352717, + "learning_rate": 4.858817897220915e-05, + "loss": 0.4224, + "step": 6471 + }, + { + "epoch": 2.440422322775264, + "grad_norm": 0.25651929827338144, + "learning_rate": 4.857789435176229e-05, + "loss": 0.4171, + "step": 6472 + }, + { + "epoch": 2.44079939668175, + "grad_norm": 0.2113383006011333, + "learning_rate": 4.856760913687495e-05, + "loss": 0.4114, + "step": 6473 + }, + { + "epoch": 2.4411764705882355, + "grad_norm": 0.2622384553195195, + "learning_rate": 4.85573233282599e-05, + "loss": 0.416, + "step": 6474 + }, + { + "epoch": 2.441553544494721, + "grad_norm": 0.2909493544613145, + "learning_rate": 4.854703692662994e-05, + "loss": 0.428, + "step": 6475 + }, + { + "epoch": 2.441930618401207, + "grad_norm": 0.27797263819986284, + "learning_rate": 4.8536749932697895e-05, + "loss": 0.4226, + "step": 6476 + }, + { + "epoch": 2.4423076923076925, + "grad_norm": 0.23238890361002693, + "learning_rate": 4.8526462347176646e-05, + "loss": 0.4111, + "step": 6477 + }, + { + "epoch": 2.442684766214178, + "grad_norm": 0.2096248192428882, + "learning_rate": 4.851617417077913e-05, + "loss": 0.4243, + "step": 6478 + }, + { + "epoch": 2.443061840120664, + "grad_norm": 0.22523441452320872, + "learning_rate": 4.8505885404218297e-05, + "loss": 0.4212, + "step": 6479 + }, + { + "epoch": 2.4434389140271495, + "grad_norm": 0.2401140404305465, + "learning_rate": 4.849559604820714e-05, + "loss": 0.4185, + "step": 6480 + }, + { + "epoch": 2.443815987933635, + "grad_norm": 0.19676919060322037, + "learning_rate": 4.848530610345872e-05, + "loss": 0.4177, + "step": 6481 + }, + { + "epoch": 2.444193061840121, + "grad_norm": 0.18717034998050333, + "learning_rate": 4.847501557068612e-05, + "loss": 0.4155, + "step": 6482 + }, + { + "epoch": 2.4445701357466065, + "grad_norm": 0.20450057269845168, + "learning_rate": 4.846472445060244e-05, + "loss": 0.4133, + "step": 6483 + }, + { + "epoch": 2.444947209653092, + "grad_norm": 0.26594717650993266, + "learning_rate": 4.8454432743920885e-05, + "loss": 0.4215, + "step": 6484 + }, + { + "epoch": 2.445324283559578, + "grad_norm": 0.3490283072731826, + "learning_rate": 4.844414045135463e-05, + "loss": 0.4285, + "step": 6485 + }, + { + "epoch": 2.4457013574660635, + "grad_norm": 0.4406227880361649, + "learning_rate": 4.8433847573616925e-05, + "loss": 0.4346, + "step": 6486 + }, + { + "epoch": 2.446078431372549, + "grad_norm": 0.4870184446557686, + "learning_rate": 4.842355411142108e-05, + "loss": 0.416, + "step": 6487 + }, + { + "epoch": 2.446455505279035, + "grad_norm": 0.4269291313552149, + "learning_rate": 4.8413260065480394e-05, + "loss": 0.4211, + "step": 6488 + }, + { + "epoch": 2.4468325791855206, + "grad_norm": 0.2807125779864558, + "learning_rate": 4.8402965436508246e-05, + "loss": 0.4204, + "step": 6489 + }, + { + "epoch": 2.447209653092006, + "grad_norm": 0.2465852749054641, + "learning_rate": 4.839267022521806e-05, + "loss": 0.4096, + "step": 6490 + }, + { + "epoch": 2.447586726998492, + "grad_norm": 0.2969932882082842, + "learning_rate": 4.838237443232326e-05, + "loss": 0.412, + "step": 6491 + }, + { + "epoch": 2.4479638009049776, + "grad_norm": 0.31293041799386506, + "learning_rate": 4.8372078058537346e-05, + "loss": 0.4136, + "step": 6492 + }, + { + "epoch": 2.4483408748114632, + "grad_norm": 0.32264045916223727, + "learning_rate": 4.8361781104573864e-05, + "loss": 0.4156, + "step": 6493 + }, + { + "epoch": 2.448717948717949, + "grad_norm": 0.286282364221444, + "learning_rate": 4.8351483571146354e-05, + "loss": 0.4137, + "step": 6494 + }, + { + "epoch": 2.4490950226244346, + "grad_norm": 0.26086223987970836, + "learning_rate": 4.834118545896844e-05, + "loss": 0.4153, + "step": 6495 + }, + { + "epoch": 2.4494720965309202, + "grad_norm": 0.25862981560349046, + "learning_rate": 4.8330886768753784e-05, + "loss": 0.4219, + "step": 6496 + }, + { + "epoch": 2.449849170437406, + "grad_norm": 0.26285660695699037, + "learning_rate": 4.832058750121605e-05, + "loss": 0.4195, + "step": 6497 + }, + { + "epoch": 2.4502262443438916, + "grad_norm": 0.2613062523839984, + "learning_rate": 4.831028765706899e-05, + "loss": 0.4172, + "step": 6498 + }, + { + "epoch": 2.4506033182503772, + "grad_norm": 0.24744656955546498, + "learning_rate": 4.8299987237026374e-05, + "loss": 0.4171, + "step": 6499 + }, + { + "epoch": 2.450980392156863, + "grad_norm": 0.2980565526236778, + "learning_rate": 4.828968624180199e-05, + "loss": 0.4284, + "step": 6500 + }, + { + "epoch": 2.4513574660633486, + "grad_norm": 0.2788854401274065, + "learning_rate": 4.827938467210971e-05, + "loss": 0.4207, + "step": 6501 + }, + { + "epoch": 2.4517345399698343, + "grad_norm": 0.22259765063196388, + "learning_rate": 4.826908252866342e-05, + "loss": 0.418, + "step": 6502 + }, + { + "epoch": 2.45211161387632, + "grad_norm": 0.27640762476888586, + "learning_rate": 4.8258779812177035e-05, + "loss": 0.4292, + "step": 6503 + }, + { + "epoch": 2.4524886877828056, + "grad_norm": 0.3032677785268395, + "learning_rate": 4.824847652336454e-05, + "loss": 0.4249, + "step": 6504 + }, + { + "epoch": 2.4528657616892913, + "grad_norm": 0.2513773825668195, + "learning_rate": 4.823817266293994e-05, + "loss": 0.4289, + "step": 6505 + }, + { + "epoch": 2.453242835595777, + "grad_norm": 0.2500989756426252, + "learning_rate": 4.8227868231617275e-05, + "loss": 0.4092, + "step": 6506 + }, + { + "epoch": 2.4536199095022626, + "grad_norm": 0.286134594534111, + "learning_rate": 4.8217563230110645e-05, + "loss": 0.4232, + "step": 6507 + }, + { + "epoch": 2.4539969834087483, + "grad_norm": 0.27594718449520855, + "learning_rate": 4.8207257659134176e-05, + "loss": 0.421, + "step": 6508 + }, + { + "epoch": 2.454374057315234, + "grad_norm": 0.24008150259639838, + "learning_rate": 4.8196951519402025e-05, + "loss": 0.4203, + "step": 6509 + }, + { + "epoch": 2.4547511312217196, + "grad_norm": 0.21475551774021145, + "learning_rate": 4.8186644811628404e-05, + "loss": 0.4313, + "step": 6510 + }, + { + "epoch": 2.4551282051282053, + "grad_norm": 0.19061078996656183, + "learning_rate": 4.8176337536527565e-05, + "loss": 0.4113, + "step": 6511 + }, + { + "epoch": 2.455505279034691, + "grad_norm": 0.22680364449147952, + "learning_rate": 4.816602969481378e-05, + "loss": 0.4205, + "step": 6512 + }, + { + "epoch": 2.4558823529411766, + "grad_norm": 0.29019767706232774, + "learning_rate": 4.815572128720138e-05, + "loss": 0.4193, + "step": 6513 + }, + { + "epoch": 2.4562594268476623, + "grad_norm": 0.29355629790071824, + "learning_rate": 4.8145412314404746e-05, + "loss": 0.4222, + "step": 6514 + }, + { + "epoch": 2.456636500754148, + "grad_norm": 0.29925452917079876, + "learning_rate": 4.8135102777138246e-05, + "loss": 0.4218, + "step": 6515 + }, + { + "epoch": 2.4570135746606336, + "grad_norm": 0.26715970512224313, + "learning_rate": 4.812479267611634e-05, + "loss": 0.4249, + "step": 6516 + }, + { + "epoch": 2.4573906485671193, + "grad_norm": 0.250813890523731, + "learning_rate": 4.811448201205351e-05, + "loss": 0.4216, + "step": 6517 + }, + { + "epoch": 2.457767722473605, + "grad_norm": 0.23227001135924003, + "learning_rate": 4.8104170785664276e-05, + "loss": 0.4145, + "step": 6518 + }, + { + "epoch": 2.4581447963800906, + "grad_norm": 0.22161987893711885, + "learning_rate": 4.80938589976632e-05, + "loss": 0.4221, + "step": 6519 + }, + { + "epoch": 2.4585218702865763, + "grad_norm": 0.22924231502060918, + "learning_rate": 4.808354664876486e-05, + "loss": 0.4272, + "step": 6520 + }, + { + "epoch": 2.458898944193062, + "grad_norm": 0.3379299429309756, + "learning_rate": 4.807323373968393e-05, + "loss": 0.4174, + "step": 6521 + }, + { + "epoch": 2.4592760180995477, + "grad_norm": 0.41067597710924975, + "learning_rate": 4.806292027113504e-05, + "loss": 0.426, + "step": 6522 + }, + { + "epoch": 2.4596530920060333, + "grad_norm": 0.37793449506813576, + "learning_rate": 4.805260624383293e-05, + "loss": 0.4106, + "step": 6523 + }, + { + "epoch": 2.460030165912519, + "grad_norm": 0.3030210909378805, + "learning_rate": 4.8042291658492366e-05, + "loss": 0.4151, + "step": 6524 + }, + { + "epoch": 2.4604072398190047, + "grad_norm": 0.22906550112516155, + "learning_rate": 4.8031976515828114e-05, + "loss": 0.4078, + "step": 6525 + }, + { + "epoch": 2.4607843137254903, + "grad_norm": 0.2600301236407732, + "learning_rate": 4.802166081655501e-05, + "loss": 0.422, + "step": 6526 + }, + { + "epoch": 2.461161387631976, + "grad_norm": 0.29396601740018874, + "learning_rate": 4.801134456138794e-05, + "loss": 0.4044, + "step": 6527 + }, + { + "epoch": 2.4615384615384617, + "grad_norm": 0.2696244352513587, + "learning_rate": 4.8001027751041784e-05, + "loss": 0.4154, + "step": 6528 + }, + { + "epoch": 2.4619155354449473, + "grad_norm": 0.27290328548871184, + "learning_rate": 4.79907103862315e-05, + "loss": 0.4265, + "step": 6529 + }, + { + "epoch": 2.462292609351433, + "grad_norm": 0.3019355898371536, + "learning_rate": 4.798039246767209e-05, + "loss": 0.4113, + "step": 6530 + }, + { + "epoch": 2.4626696832579187, + "grad_norm": 0.31368982499061376, + "learning_rate": 4.797007399607855e-05, + "loss": 0.4001, + "step": 6531 + }, + { + "epoch": 2.4630467571644044, + "grad_norm": 0.2986781584883192, + "learning_rate": 4.7959754972165946e-05, + "loss": 0.4275, + "step": 6532 + }, + { + "epoch": 2.46342383107089, + "grad_norm": 0.2690799995947203, + "learning_rate": 4.7949435396649396e-05, + "loss": 0.4065, + "step": 6533 + }, + { + "epoch": 2.4638009049773757, + "grad_norm": 0.2351877705867045, + "learning_rate": 4.793911527024401e-05, + "loss": 0.4149, + "step": 6534 + }, + { + "epoch": 2.4641779788838614, + "grad_norm": 0.2177555695509782, + "learning_rate": 4.792879459366498e-05, + "loss": 0.4245, + "step": 6535 + }, + { + "epoch": 2.464555052790347, + "grad_norm": 0.21817677328630866, + "learning_rate": 4.791847336762754e-05, + "loss": 0.4196, + "step": 6536 + }, + { + "epoch": 2.4649321266968327, + "grad_norm": 0.2509207048874999, + "learning_rate": 4.7908151592846894e-05, + "loss": 0.4195, + "step": 6537 + }, + { + "epoch": 2.4653092006033184, + "grad_norm": 0.26810795123542175, + "learning_rate": 4.7897829270038366e-05, + "loss": 0.4141, + "step": 6538 + }, + { + "epoch": 2.465686274509804, + "grad_norm": 0.2085112737527727, + "learning_rate": 4.788750639991728e-05, + "loss": 0.4188, + "step": 6539 + }, + { + "epoch": 2.4660633484162897, + "grad_norm": 0.20312576523464623, + "learning_rate": 4.787718298319898e-05, + "loss": 0.4169, + "step": 6540 + }, + { + "epoch": 2.4664404223227754, + "grad_norm": 0.19904527795693985, + "learning_rate": 4.7866859020598895e-05, + "loss": 0.4223, + "step": 6541 + }, + { + "epoch": 2.466817496229261, + "grad_norm": 0.251880118580247, + "learning_rate": 4.7856534512832464e-05, + "loss": 0.419, + "step": 6542 + }, + { + "epoch": 2.4671945701357467, + "grad_norm": 0.271700541050047, + "learning_rate": 4.784620946061516e-05, + "loss": 0.423, + "step": 6543 + }, + { + "epoch": 2.4675716440422324, + "grad_norm": 0.2181153498288086, + "learning_rate": 4.783588386466249e-05, + "loss": 0.424, + "step": 6544 + }, + { + "epoch": 2.467948717948718, + "grad_norm": 0.1563775270936423, + "learning_rate": 4.782555772569003e-05, + "loss": 0.4067, + "step": 6545 + }, + { + "epoch": 2.4683257918552037, + "grad_norm": 0.2124954431119987, + "learning_rate": 4.781523104441335e-05, + "loss": 0.4324, + "step": 6546 + }, + { + "epoch": 2.4687028657616894, + "grad_norm": 0.2218865786020322, + "learning_rate": 4.78049038215481e-05, + "loss": 0.4255, + "step": 6547 + }, + { + "epoch": 2.469079939668175, + "grad_norm": 0.21563475887403027, + "learning_rate": 4.779457605780994e-05, + "loss": 0.415, + "step": 6548 + }, + { + "epoch": 2.4694570135746607, + "grad_norm": 0.2173828761075134, + "learning_rate": 4.7784247753914554e-05, + "loss": 0.4124, + "step": 6549 + }, + { + "epoch": 2.4698340874811464, + "grad_norm": 0.22139100198392747, + "learning_rate": 4.7773918910577714e-05, + "loss": 0.417, + "step": 6550 + }, + { + "epoch": 2.470211161387632, + "grad_norm": 0.2623047110208815, + "learning_rate": 4.77635895285152e-05, + "loss": 0.4335, + "step": 6551 + }, + { + "epoch": 2.4705882352941178, + "grad_norm": 0.2819884915642992, + "learning_rate": 4.7753259608442804e-05, + "loss": 0.4277, + "step": 6552 + }, + { + "epoch": 2.4709653092006034, + "grad_norm": 0.23901632657070548, + "learning_rate": 4.7742929151076396e-05, + "loss": 0.4143, + "step": 6553 + }, + { + "epoch": 2.471342383107089, + "grad_norm": 0.2576734509356794, + "learning_rate": 4.773259815713188e-05, + "loss": 0.4307, + "step": 6554 + }, + { + "epoch": 2.4717194570135748, + "grad_norm": 0.2708803779592886, + "learning_rate": 4.772226662732515e-05, + "loss": 0.4214, + "step": 6555 + }, + { + "epoch": 2.4720965309200604, + "grad_norm": 0.2666885142081512, + "learning_rate": 4.771193456237221e-05, + "loss": 0.4208, + "step": 6556 + }, + { + "epoch": 2.472473604826546, + "grad_norm": 0.25628028680592113, + "learning_rate": 4.770160196298904e-05, + "loss": 0.4137, + "step": 6557 + }, + { + "epoch": 2.4728506787330318, + "grad_norm": 0.25603591585430824, + "learning_rate": 4.7691268829891685e-05, + "loss": 0.4152, + "step": 6558 + }, + { + "epoch": 2.4732277526395174, + "grad_norm": 0.29904250921342573, + "learning_rate": 4.7680935163796204e-05, + "loss": 0.42, + "step": 6559 + }, + { + "epoch": 2.473604826546003, + "grad_norm": 0.3150450645420006, + "learning_rate": 4.767060096541876e-05, + "loss": 0.4217, + "step": 6560 + }, + { + "epoch": 2.473981900452489, + "grad_norm": 0.2813982964955141, + "learning_rate": 4.766026623547546e-05, + "loss": 0.4134, + "step": 6561 + }, + { + "epoch": 2.4743589743589745, + "grad_norm": 0.21201388008251204, + "learning_rate": 4.7649930974682494e-05, + "loss": 0.4108, + "step": 6562 + }, + { + "epoch": 2.47473604826546, + "grad_norm": 0.22893238220625564, + "learning_rate": 4.763959518375611e-05, + "loss": 0.415, + "step": 6563 + }, + { + "epoch": 2.475113122171946, + "grad_norm": 0.27029034592663304, + "learning_rate": 4.762925886341255e-05, + "loss": 0.4168, + "step": 6564 + }, + { + "epoch": 2.4754901960784315, + "grad_norm": 0.2344542947235699, + "learning_rate": 4.761892201436812e-05, + "loss": 0.4263, + "step": 6565 + }, + { + "epoch": 2.475867269984917, + "grad_norm": 0.17893366986288267, + "learning_rate": 4.760858463733915e-05, + "loss": 0.4042, + "step": 6566 + }, + { + "epoch": 2.476244343891403, + "grad_norm": 0.17264663045495562, + "learning_rate": 4.7598246733042025e-05, + "loss": 0.4081, + "step": 6567 + }, + { + "epoch": 2.4766214177978885, + "grad_norm": 0.17973251394205855, + "learning_rate": 4.758790830219312e-05, + "loss": 0.4378, + "step": 6568 + }, + { + "epoch": 2.476998491704374, + "grad_norm": 0.1738529524241517, + "learning_rate": 4.757756934550892e-05, + "loss": 0.4137, + "step": 6569 + }, + { + "epoch": 2.47737556561086, + "grad_norm": 0.19386673813536914, + "learning_rate": 4.756722986370587e-05, + "loss": 0.4247, + "step": 6570 + }, + { + "epoch": 2.4777526395173455, + "grad_norm": 0.20156646418398372, + "learning_rate": 4.755688985750051e-05, + "loss": 0.4328, + "step": 6571 + }, + { + "epoch": 2.478129713423831, + "grad_norm": 0.1772286928618014, + "learning_rate": 4.7546549327609396e-05, + "loss": 0.4201, + "step": 6572 + }, + { + "epoch": 2.478506787330317, + "grad_norm": 0.19396041586332446, + "learning_rate": 4.753620827474909e-05, + "loss": 0.4122, + "step": 6573 + }, + { + "epoch": 2.4788838612368025, + "grad_norm": 0.2137387452355479, + "learning_rate": 4.752586669963624e-05, + "loss": 0.4179, + "step": 6574 + }, + { + "epoch": 2.479260935143288, + "grad_norm": 0.18707274491765088, + "learning_rate": 4.751552460298751e-05, + "loss": 0.4306, + "step": 6575 + }, + { + "epoch": 2.479638009049774, + "grad_norm": 0.2285290078790322, + "learning_rate": 4.7505181985519577e-05, + "loss": 0.4221, + "step": 6576 + }, + { + "epoch": 2.4800150829562595, + "grad_norm": 0.24645951462552645, + "learning_rate": 4.74948388479492e-05, + "loss": 0.4253, + "step": 6577 + }, + { + "epoch": 2.480392156862745, + "grad_norm": 0.23668620476786983, + "learning_rate": 4.748449519099314e-05, + "loss": 0.4224, + "step": 6578 + }, + { + "epoch": 2.480769230769231, + "grad_norm": 0.29059457624959484, + "learning_rate": 4.7474151015368206e-05, + "loss": 0.4216, + "step": 6579 + }, + { + "epoch": 2.4811463046757165, + "grad_norm": 0.309055204226762, + "learning_rate": 4.746380632179122e-05, + "loss": 0.4177, + "step": 6580 + }, + { + "epoch": 2.481523378582202, + "grad_norm": 0.21477618293945835, + "learning_rate": 4.7453461110979086e-05, + "loss": 0.4043, + "step": 6581 + }, + { + "epoch": 2.481900452488688, + "grad_norm": 0.18512267163584573, + "learning_rate": 4.744311538364871e-05, + "loss": 0.4105, + "step": 6582 + }, + { + "epoch": 2.4822775263951735, + "grad_norm": 0.19595692814925705, + "learning_rate": 4.743276914051704e-05, + "loss": 0.4228, + "step": 6583 + }, + { + "epoch": 2.482654600301659, + "grad_norm": 0.2756220883965436, + "learning_rate": 4.742242238230105e-05, + "loss": 0.4239, + "step": 6584 + }, + { + "epoch": 2.483031674208145, + "grad_norm": 0.3347425311029598, + "learning_rate": 4.7412075109717795e-05, + "loss": 0.4145, + "step": 6585 + }, + { + "epoch": 2.4834087481146305, + "grad_norm": 0.29734217360433746, + "learning_rate": 4.74017273234843e-05, + "loss": 0.4088, + "step": 6586 + }, + { + "epoch": 2.483785822021116, + "grad_norm": 0.26931911465904773, + "learning_rate": 4.7391379024317656e-05, + "loss": 0.414, + "step": 6587 + }, + { + "epoch": 2.484162895927602, + "grad_norm": 0.21868429410910664, + "learning_rate": 4.738103021293501e-05, + "loss": 0.4257, + "step": 6588 + }, + { + "epoch": 2.4845399698340875, + "grad_norm": 0.18096544713730536, + "learning_rate": 4.7370680890053516e-05, + "loss": 0.4204, + "step": 6589 + }, + { + "epoch": 2.484917043740573, + "grad_norm": 0.19413113050367722, + "learning_rate": 4.736033105639037e-05, + "loss": 0.4187, + "step": 6590 + }, + { + "epoch": 2.485294117647059, + "grad_norm": 0.18911664453939103, + "learning_rate": 4.734998071266282e-05, + "loss": 0.4181, + "step": 6591 + }, + { + "epoch": 2.4856711915535445, + "grad_norm": 0.1866299463791562, + "learning_rate": 4.733962985958812e-05, + "loss": 0.4117, + "step": 6592 + }, + { + "epoch": 2.48604826546003, + "grad_norm": 0.20596221142708415, + "learning_rate": 4.732927849788358e-05, + "loss": 0.421, + "step": 6593 + }, + { + "epoch": 2.486425339366516, + "grad_norm": 0.2342221011771758, + "learning_rate": 4.731892662826655e-05, + "loss": 0.4188, + "step": 6594 + }, + { + "epoch": 2.4868024132730016, + "grad_norm": 0.22021851616701021, + "learning_rate": 4.730857425145439e-05, + "loss": 0.4185, + "step": 6595 + }, + { + "epoch": 2.4871794871794872, + "grad_norm": 0.21901930452189355, + "learning_rate": 4.7298221368164515e-05, + "loss": 0.4035, + "step": 6596 + }, + { + "epoch": 2.487556561085973, + "grad_norm": 0.23227099574082066, + "learning_rate": 4.7287867979114374e-05, + "loss": 0.424, + "step": 6597 + }, + { + "epoch": 2.4879336349924586, + "grad_norm": 0.21823780629903114, + "learning_rate": 4.727751408502145e-05, + "loss": 0.4182, + "step": 6598 + }, + { + "epoch": 2.4883107088989442, + "grad_norm": 0.19750494605798447, + "learning_rate": 4.726715968660324e-05, + "loss": 0.4133, + "step": 6599 + }, + { + "epoch": 2.48868778280543, + "grad_norm": 0.21109313111275094, + "learning_rate": 4.725680478457733e-05, + "loss": 0.4205, + "step": 6600 + }, + { + "epoch": 2.4890648567119156, + "grad_norm": 0.281896962886296, + "learning_rate": 4.724644937966128e-05, + "loss": 0.4231, + "step": 6601 + }, + { + "epoch": 2.4894419306184012, + "grad_norm": 0.308737397627319, + "learning_rate": 4.723609347257271e-05, + "loss": 0.4254, + "step": 6602 + }, + { + "epoch": 2.489819004524887, + "grad_norm": 0.33077802456616523, + "learning_rate": 4.722573706402928e-05, + "loss": 0.417, + "step": 6603 + }, + { + "epoch": 2.4901960784313726, + "grad_norm": 0.28952673082974945, + "learning_rate": 4.7215380154748685e-05, + "loss": 0.4196, + "step": 6604 + }, + { + "epoch": 2.4905731523378583, + "grad_norm": 0.28724015848638607, + "learning_rate": 4.7205022745448645e-05, + "loss": 0.405, + "step": 6605 + }, + { + "epoch": 2.490950226244344, + "grad_norm": 0.2402180253201705, + "learning_rate": 4.7194664836846914e-05, + "loss": 0.4237, + "step": 6606 + }, + { + "epoch": 2.4913273001508296, + "grad_norm": 0.23095766889988195, + "learning_rate": 4.7184306429661296e-05, + "loss": 0.4179, + "step": 6607 + }, + { + "epoch": 2.4917043740573153, + "grad_norm": 0.22768272485811533, + "learning_rate": 4.717394752460962e-05, + "loss": 0.4183, + "step": 6608 + }, + { + "epoch": 2.492081447963801, + "grad_norm": 0.2990889271210464, + "learning_rate": 4.716358812240974e-05, + "loss": 0.4217, + "step": 6609 + }, + { + "epoch": 2.4924585218702866, + "grad_norm": 0.36065802770844974, + "learning_rate": 4.715322822377955e-05, + "loss": 0.4266, + "step": 6610 + }, + { + "epoch": 2.4928355957767723, + "grad_norm": 0.34630198809987484, + "learning_rate": 4.7142867829437e-05, + "loss": 0.4357, + "step": 6611 + }, + { + "epoch": 2.493212669683258, + "grad_norm": 0.2897836191817711, + "learning_rate": 4.713250694010004e-05, + "loss": 0.4179, + "step": 6612 + }, + { + "epoch": 2.4935897435897436, + "grad_norm": 0.21476684650958694, + "learning_rate": 4.712214555648667e-05, + "loss": 0.4168, + "step": 6613 + }, + { + "epoch": 2.4939668174962293, + "grad_norm": 0.2073782664928248, + "learning_rate": 4.711178367931494e-05, + "loss": 0.4089, + "step": 6614 + }, + { + "epoch": 2.494343891402715, + "grad_norm": 0.3080236564373175, + "learning_rate": 4.7101421309302904e-05, + "loss": 0.4235, + "step": 6615 + }, + { + "epoch": 2.4947209653092006, + "grad_norm": 0.3965390103483159, + "learning_rate": 4.709105844716867e-05, + "loss": 0.4136, + "step": 6616 + }, + { + "epoch": 2.4950980392156863, + "grad_norm": 0.38721201631884294, + "learning_rate": 4.7080695093630376e-05, + "loss": 0.4191, + "step": 6617 + }, + { + "epoch": 2.495475113122172, + "grad_norm": 0.26653356756968194, + "learning_rate": 4.70703312494062e-05, + "loss": 0.4177, + "step": 6618 + }, + { + "epoch": 2.4958521870286576, + "grad_norm": 0.24249322056403128, + "learning_rate": 4.705996691521433e-05, + "loss": 0.4214, + "step": 6619 + }, + { + "epoch": 2.4962292609351433, + "grad_norm": 0.22282546668723457, + "learning_rate": 4.7049602091773014e-05, + "loss": 0.4155, + "step": 6620 + }, + { + "epoch": 2.496606334841629, + "grad_norm": 0.2386760002400646, + "learning_rate": 4.7039236779800534e-05, + "loss": 0.4262, + "step": 6621 + }, + { + "epoch": 2.4969834087481146, + "grad_norm": 0.28969411458113087, + "learning_rate": 4.7028870980015186e-05, + "loss": 0.4384, + "step": 6622 + }, + { + "epoch": 2.4973604826546003, + "grad_norm": 0.28786977729358404, + "learning_rate": 4.701850469313531e-05, + "loss": 0.4042, + "step": 6623 + }, + { + "epoch": 2.497737556561086, + "grad_norm": 0.24259224006310992, + "learning_rate": 4.700813791987929e-05, + "loss": 0.4231, + "step": 6624 + }, + { + "epoch": 2.4981146304675717, + "grad_norm": 0.19022327618148993, + "learning_rate": 4.699777066096553e-05, + "loss": 0.4117, + "step": 6625 + }, + { + "epoch": 2.4984917043740573, + "grad_norm": 0.20526268060806938, + "learning_rate": 4.698740291711246e-05, + "loss": 0.4187, + "step": 6626 + }, + { + "epoch": 2.498868778280543, + "grad_norm": 0.2804162837215144, + "learning_rate": 4.697703468903857e-05, + "loss": 0.4227, + "step": 6627 + }, + { + "epoch": 2.4992458521870287, + "grad_norm": 0.28611102399423255, + "learning_rate": 4.696666597746236e-05, + "loss": 0.4225, + "step": 6628 + }, + { + "epoch": 2.4996229260935143, + "grad_norm": 0.26895829660810383, + "learning_rate": 4.695629678310239e-05, + "loss": 0.4211, + "step": 6629 + }, + { + "epoch": 2.5, + "grad_norm": 0.24014729921490302, + "learning_rate": 4.694592710667723e-05, + "loss": 0.4207, + "step": 6630 + }, + { + "epoch": 2.5003770739064857, + "grad_norm": 0.22379655488897324, + "learning_rate": 4.6935556948905455e-05, + "loss": 0.4157, + "step": 6631 + }, + { + "epoch": 2.5007541478129713, + "grad_norm": 0.21577066309928977, + "learning_rate": 4.692518631050575e-05, + "loss": 0.4147, + "step": 6632 + }, + { + "epoch": 2.501131221719457, + "grad_norm": 0.20563633349413474, + "learning_rate": 4.691481519219678e-05, + "loss": 0.4135, + "step": 6633 + }, + { + "epoch": 2.5015082956259427, + "grad_norm": 0.23239684241212458, + "learning_rate": 4.690444359469724e-05, + "loss": 0.4279, + "step": 6634 + }, + { + "epoch": 2.5018853695324283, + "grad_norm": 0.22808450475503578, + "learning_rate": 4.689407151872589e-05, + "loss": 0.41, + "step": 6635 + }, + { + "epoch": 2.502262443438914, + "grad_norm": 0.23138380541666223, + "learning_rate": 4.688369896500151e-05, + "loss": 0.4205, + "step": 6636 + }, + { + "epoch": 2.5026395173453997, + "grad_norm": 0.20800182094837316, + "learning_rate": 4.6873325934242876e-05, + "loss": 0.4157, + "step": 6637 + }, + { + "epoch": 2.5030165912518854, + "grad_norm": 0.22213842348608887, + "learning_rate": 4.6862952427168856e-05, + "loss": 0.4204, + "step": 6638 + }, + { + "epoch": 2.503393665158371, + "grad_norm": 0.2076853719241175, + "learning_rate": 4.685257844449833e-05, + "loss": 0.4076, + "step": 6639 + }, + { + "epoch": 2.5037707390648567, + "grad_norm": 0.20127754039920387, + "learning_rate": 4.684220398695019e-05, + "loss": 0.4248, + "step": 6640 + }, + { + "epoch": 2.5041478129713424, + "grad_norm": 0.17894644178546903, + "learning_rate": 4.683182905524338e-05, + "loss": 0.4156, + "step": 6641 + }, + { + "epoch": 2.504524886877828, + "grad_norm": 0.18247336377996268, + "learning_rate": 4.682145365009688e-05, + "loss": 0.4127, + "step": 6642 + }, + { + "epoch": 2.5049019607843137, + "grad_norm": 0.2588506366871685, + "learning_rate": 4.68110777722297e-05, + "loss": 0.4307, + "step": 6643 + }, + { + "epoch": 2.5052790346907994, + "grad_norm": 0.30757510859751835, + "learning_rate": 4.680070142236086e-05, + "loss": 0.4303, + "step": 6644 + }, + { + "epoch": 2.505656108597285, + "grad_norm": 0.26415420656776145, + "learning_rate": 4.679032460120945e-05, + "loss": 0.4259, + "step": 6645 + }, + { + "epoch": 2.5060331825037707, + "grad_norm": 0.236607325815986, + "learning_rate": 4.677994730949457e-05, + "loss": 0.4227, + "step": 6646 + }, + { + "epoch": 2.5064102564102564, + "grad_norm": 0.23863341221877182, + "learning_rate": 4.676956954793535e-05, + "loss": 0.43, + "step": 6647 + }, + { + "epoch": 2.506787330316742, + "grad_norm": 0.2494381697130313, + "learning_rate": 4.6759191317250966e-05, + "loss": 0.4187, + "step": 6648 + }, + { + "epoch": 2.5071644042232277, + "grad_norm": 0.24911824649480296, + "learning_rate": 4.6748812618160615e-05, + "loss": 0.415, + "step": 6649 + }, + { + "epoch": 2.5075414781297134, + "grad_norm": 0.24694326626874288, + "learning_rate": 4.673843345138354e-05, + "loss": 0.4059, + "step": 6650 + }, + { + "epoch": 2.507918552036199, + "grad_norm": 0.238536977562718, + "learning_rate": 4.6728053817639e-05, + "loss": 0.4221, + "step": 6651 + }, + { + "epoch": 2.5082956259426847, + "grad_norm": 0.24919545790022768, + "learning_rate": 4.671767371764631e-05, + "loss": 0.4066, + "step": 6652 + }, + { + "epoch": 2.5086726998491704, + "grad_norm": 0.22986330674798247, + "learning_rate": 4.6707293152124786e-05, + "loss": 0.4152, + "step": 6653 + }, + { + "epoch": 2.509049773755656, + "grad_norm": 0.23374720971448815, + "learning_rate": 4.6696912121793786e-05, + "loss": 0.4182, + "step": 6654 + }, + { + "epoch": 2.5094268476621417, + "grad_norm": 0.22468454415486105, + "learning_rate": 4.668653062737272e-05, + "loss": 0.4148, + "step": 6655 + }, + { + "epoch": 2.5098039215686274, + "grad_norm": 0.20102867100192046, + "learning_rate": 4.6676148669581025e-05, + "loss": 0.4085, + "step": 6656 + }, + { + "epoch": 2.510180995475113, + "grad_norm": 0.198154160458221, + "learning_rate": 4.6665766249138136e-05, + "loss": 0.4165, + "step": 6657 + }, + { + "epoch": 2.5105580693815988, + "grad_norm": 0.24483600074056472, + "learning_rate": 4.665538336676356e-05, + "loss": 0.4281, + "step": 6658 + }, + { + "epoch": 2.5109351432880844, + "grad_norm": 0.2518958603167447, + "learning_rate": 4.6645000023176825e-05, + "loss": 0.4158, + "step": 6659 + }, + { + "epoch": 2.51131221719457, + "grad_norm": 0.28858197277332737, + "learning_rate": 4.663461621909747e-05, + "loss": 0.4198, + "step": 6660 + }, + { + "epoch": 2.5116892911010558, + "grad_norm": 0.3002305047437629, + "learning_rate": 4.662423195524511e-05, + "loss": 0.4157, + "step": 6661 + }, + { + "epoch": 2.5120663650075414, + "grad_norm": 0.2526193931966004, + "learning_rate": 4.6613847232339355e-05, + "loss": 0.4332, + "step": 6662 + }, + { + "epoch": 2.512443438914027, + "grad_norm": 0.2544969801305273, + "learning_rate": 4.6603462051099835e-05, + "loss": 0.4163, + "step": 6663 + }, + { + "epoch": 2.5128205128205128, + "grad_norm": 0.23401936558506808, + "learning_rate": 4.659307641224627e-05, + "loss": 0.4223, + "step": 6664 + }, + { + "epoch": 2.5131975867269984, + "grad_norm": 0.2085013836818146, + "learning_rate": 4.658269031649834e-05, + "loss": 0.4062, + "step": 6665 + }, + { + "epoch": 2.513574660633484, + "grad_norm": 0.16478999107041842, + "learning_rate": 4.6572303764575824e-05, + "loss": 0.4303, + "step": 6666 + }, + { + "epoch": 2.51395173453997, + "grad_norm": 0.20041237656575225, + "learning_rate": 4.656191675719848e-05, + "loss": 0.4129, + "step": 6667 + }, + { + "epoch": 2.5143288084464555, + "grad_norm": 0.24583476125678394, + "learning_rate": 4.655152929508613e-05, + "loss": 0.4159, + "step": 6668 + }, + { + "epoch": 2.514705882352941, + "grad_norm": 0.2799638151673776, + "learning_rate": 4.65411413789586e-05, + "loss": 0.429, + "step": 6669 + }, + { + "epoch": 2.515082956259427, + "grad_norm": 0.25078537287284164, + "learning_rate": 4.653075300953579e-05, + "loss": 0.411, + "step": 6670 + }, + { + "epoch": 2.5154600301659125, + "grad_norm": 0.2012901040309205, + "learning_rate": 4.652036418753758e-05, + "loss": 0.4272, + "step": 6671 + }, + { + "epoch": 2.515837104072398, + "grad_norm": 0.23969906656464668, + "learning_rate": 4.650997491368391e-05, + "loss": 0.4175, + "step": 6672 + }, + { + "epoch": 2.516214177978884, + "grad_norm": 0.2818627237112057, + "learning_rate": 4.649958518869477e-05, + "loss": 0.4321, + "step": 6673 + }, + { + "epoch": 2.5165912518853695, + "grad_norm": 0.2704544032253139, + "learning_rate": 4.6489195013290125e-05, + "loss": 0.4251, + "step": 6674 + }, + { + "epoch": 2.516968325791855, + "grad_norm": 0.27481140776494656, + "learning_rate": 4.6478804388190027e-05, + "loss": 0.4194, + "step": 6675 + }, + { + "epoch": 2.517345399698341, + "grad_norm": 0.25334107708045916, + "learning_rate": 4.646841331411454e-05, + "loss": 0.4298, + "step": 6676 + }, + { + "epoch": 2.5177224736048265, + "grad_norm": 0.23635684912502863, + "learning_rate": 4.645802179178374e-05, + "loss": 0.421, + "step": 6677 + }, + { + "epoch": 2.518099547511312, + "grad_norm": 0.18965218175502865, + "learning_rate": 4.6447629821917755e-05, + "loss": 0.421, + "step": 6678 + }, + { + "epoch": 2.518476621417798, + "grad_norm": 0.18954389584066494, + "learning_rate": 4.643723740523674e-05, + "loss": 0.4247, + "step": 6679 + }, + { + "epoch": 2.5188536953242835, + "grad_norm": 0.2282363585780992, + "learning_rate": 4.6426844542460893e-05, + "loss": 0.4154, + "step": 6680 + }, + { + "epoch": 2.519230769230769, + "grad_norm": 0.20592678416133114, + "learning_rate": 4.641645123431041e-05, + "loss": 0.4184, + "step": 6681 + }, + { + "epoch": 2.519607843137255, + "grad_norm": 0.20734455550541306, + "learning_rate": 4.640605748150556e-05, + "loss": 0.4197, + "step": 6682 + }, + { + "epoch": 2.5199849170437405, + "grad_norm": 0.27197921112064205, + "learning_rate": 4.63956632847666e-05, + "loss": 0.4211, + "step": 6683 + }, + { + "epoch": 2.520361990950226, + "grad_norm": 0.292120938505348, + "learning_rate": 4.638526864481384e-05, + "loss": 0.421, + "step": 6684 + }, + { + "epoch": 2.520739064856712, + "grad_norm": 0.20764020028366487, + "learning_rate": 4.637487356236765e-05, + "loss": 0.4202, + "step": 6685 + }, + { + "epoch": 2.5211161387631975, + "grad_norm": 0.1755810020441375, + "learning_rate": 4.636447803814836e-05, + "loss": 0.4215, + "step": 6686 + }, + { + "epoch": 2.521493212669683, + "grad_norm": 0.25559282857139637, + "learning_rate": 4.635408207287638e-05, + "loss": 0.4263, + "step": 6687 + }, + { + "epoch": 2.521870286576169, + "grad_norm": 0.23409422582781922, + "learning_rate": 4.6343685667272167e-05, + "loss": 0.4126, + "step": 6688 + }, + { + "epoch": 2.5222473604826545, + "grad_norm": 0.2270256546289165, + "learning_rate": 4.6333288822056154e-05, + "loss": 0.4143, + "step": 6689 + }, + { + "epoch": 2.52262443438914, + "grad_norm": 0.21379428709503615, + "learning_rate": 4.632289153794884e-05, + "loss": 0.4201, + "step": 6690 + }, + { + "epoch": 2.523001508295626, + "grad_norm": 0.24380346547567883, + "learning_rate": 4.631249381567077e-05, + "loss": 0.4232, + "step": 6691 + }, + { + "epoch": 2.5233785822021115, + "grad_norm": 0.27803746135070667, + "learning_rate": 4.6302095655942455e-05, + "loss": 0.4151, + "step": 6692 + }, + { + "epoch": 2.523755656108597, + "grad_norm": 0.25344739404008054, + "learning_rate": 4.62916970594845e-05, + "loss": 0.4109, + "step": 6693 + }, + { + "epoch": 2.524132730015083, + "grad_norm": 0.1616324148288429, + "learning_rate": 4.628129802701754e-05, + "loss": 0.4276, + "step": 6694 + }, + { + "epoch": 2.5245098039215685, + "grad_norm": 0.19601076239003457, + "learning_rate": 4.627089855926218e-05, + "loss": 0.4186, + "step": 6695 + }, + { + "epoch": 2.524886877828054, + "grad_norm": 0.2250412819609501, + "learning_rate": 4.626049865693912e-05, + "loss": 0.4102, + "step": 6696 + }, + { + "epoch": 2.52526395173454, + "grad_norm": 0.29552260377081446, + "learning_rate": 4.625009832076906e-05, + "loss": 0.4246, + "step": 6697 + }, + { + "epoch": 2.5256410256410255, + "grad_norm": 0.26857475008406373, + "learning_rate": 4.623969755147272e-05, + "loss": 0.4204, + "step": 6698 + }, + { + "epoch": 2.526018099547511, + "grad_norm": 0.23736925696885153, + "learning_rate": 4.6229296349770875e-05, + "loss": 0.4126, + "step": 6699 + }, + { + "epoch": 2.526395173453997, + "grad_norm": 0.19136519051381465, + "learning_rate": 4.621889471638433e-05, + "loss": 0.421, + "step": 6700 + }, + { + "epoch": 2.5267722473604826, + "grad_norm": 0.19223513836169853, + "learning_rate": 4.620849265203388e-05, + "loss": 0.4181, + "step": 6701 + }, + { + "epoch": 2.5271493212669682, + "grad_norm": 0.22579703536564782, + "learning_rate": 4.619809015744041e-05, + "loss": 0.4196, + "step": 6702 + }, + { + "epoch": 2.527526395173454, + "grad_norm": 0.22975131287249612, + "learning_rate": 4.618768723332478e-05, + "loss": 0.4101, + "step": 6703 + }, + { + "epoch": 2.5279034690799396, + "grad_norm": 0.26013527697218, + "learning_rate": 4.6177283880407925e-05, + "loss": 0.4252, + "step": 6704 + }, + { + "epoch": 2.5282805429864252, + "grad_norm": 0.2358565039017829, + "learning_rate": 4.616688009941077e-05, + "loss": 0.4195, + "step": 6705 + }, + { + "epoch": 2.528657616892911, + "grad_norm": 0.2927382492546677, + "learning_rate": 4.615647589105429e-05, + "loss": 0.4143, + "step": 6706 + }, + { + "epoch": 2.5290346907993966, + "grad_norm": 0.32056063197225004, + "learning_rate": 4.6146071256059506e-05, + "loss": 0.407, + "step": 6707 + }, + { + "epoch": 2.5294117647058822, + "grad_norm": 0.24164929011899244, + "learning_rate": 4.6135666195147426e-05, + "loss": 0.4057, + "step": 6708 + }, + { + "epoch": 2.529788838612368, + "grad_norm": 0.17429875692860963, + "learning_rate": 4.612526070903912e-05, + "loss": 0.4259, + "step": 6709 + }, + { + "epoch": 2.5301659125188536, + "grad_norm": 0.2084122413522638, + "learning_rate": 4.6114854798455684e-05, + "loss": 0.4089, + "step": 6710 + }, + { + "epoch": 2.5305429864253393, + "grad_norm": 0.26176614865480957, + "learning_rate": 4.610444846411822e-05, + "loss": 0.4195, + "step": 6711 + }, + { + "epoch": 2.530920060331825, + "grad_norm": 0.20915159877544504, + "learning_rate": 4.609404170674791e-05, + "loss": 0.4183, + "step": 6712 + }, + { + "epoch": 2.5312971342383106, + "grad_norm": 0.23946190829342953, + "learning_rate": 4.6083634527065906e-05, + "loss": 0.4147, + "step": 6713 + }, + { + "epoch": 2.5316742081447963, + "grad_norm": 0.26578386174223284, + "learning_rate": 4.607322692579343e-05, + "loss": 0.415, + "step": 6714 + }, + { + "epoch": 2.532051282051282, + "grad_norm": 0.30094091781587806, + "learning_rate": 4.6062818903651704e-05, + "loss": 0.4138, + "step": 6715 + }, + { + "epoch": 2.5324283559577676, + "grad_norm": 0.2815686077190627, + "learning_rate": 4.605241046136202e-05, + "loss": 0.4137, + "step": 6716 + }, + { + "epoch": 2.5328054298642533, + "grad_norm": 0.2291083167906134, + "learning_rate": 4.604200159964565e-05, + "loss": 0.4116, + "step": 6717 + }, + { + "epoch": 2.533182503770739, + "grad_norm": 0.2087367268281716, + "learning_rate": 4.6031592319223936e-05, + "loss": 0.4229, + "step": 6718 + }, + { + "epoch": 2.5335595776772246, + "grad_norm": 0.1935051260828311, + "learning_rate": 4.6021182620818226e-05, + "loss": 0.4251, + "step": 6719 + }, + { + "epoch": 2.5339366515837103, + "grad_norm": 0.236240217614527, + "learning_rate": 4.601077250514989e-05, + "loss": 0.4248, + "step": 6720 + }, + { + "epoch": 2.534313725490196, + "grad_norm": 0.25735643377833817, + "learning_rate": 4.600036197294035e-05, + "loss": 0.4191, + "step": 6721 + }, + { + "epoch": 2.5346907993966816, + "grad_norm": 0.25963504604491944, + "learning_rate": 4.598995102491106e-05, + "loss": 0.4297, + "step": 6722 + }, + { + "epoch": 2.5350678733031673, + "grad_norm": 0.223332119537489, + "learning_rate": 4.597953966178347e-05, + "loss": 0.4268, + "step": 6723 + }, + { + "epoch": 2.535444947209653, + "grad_norm": 0.16379054097038193, + "learning_rate": 4.596912788427909e-05, + "loss": 0.4117, + "step": 6724 + }, + { + "epoch": 2.5358220211161386, + "grad_norm": 0.18302648551115702, + "learning_rate": 4.595871569311944e-05, + "loss": 0.426, + "step": 6725 + }, + { + "epoch": 2.5361990950226243, + "grad_norm": 0.2094560003705389, + "learning_rate": 4.5948303089026066e-05, + "loss": 0.4155, + "step": 6726 + }, + { + "epoch": 2.53657616892911, + "grad_norm": 0.2415525051581865, + "learning_rate": 4.593789007272058e-05, + "loss": 0.4212, + "step": 6727 + }, + { + "epoch": 2.5369532428355956, + "grad_norm": 0.2323008485497909, + "learning_rate": 4.592747664492458e-05, + "loss": 0.4116, + "step": 6728 + }, + { + "epoch": 2.5373303167420813, + "grad_norm": 0.1875127325949073, + "learning_rate": 4.5917062806359696e-05, + "loss": 0.4301, + "step": 6729 + }, + { + "epoch": 2.537707390648567, + "grad_norm": 0.213641665333491, + "learning_rate": 4.590664855774761e-05, + "loss": 0.4269, + "step": 6730 + }, + { + "epoch": 2.5380844645550527, + "grad_norm": 0.2730280240632213, + "learning_rate": 4.589623389981002e-05, + "loss": 0.4238, + "step": 6731 + }, + { + "epoch": 2.5384615384615383, + "grad_norm": 0.36209066479765456, + "learning_rate": 4.588581883326865e-05, + "loss": 0.4165, + "step": 6732 + }, + { + "epoch": 2.538838612368024, + "grad_norm": 0.3549723677672145, + "learning_rate": 4.587540335884526e-05, + "loss": 0.4137, + "step": 6733 + }, + { + "epoch": 2.5392156862745097, + "grad_norm": 0.24720571869188065, + "learning_rate": 4.586498747726163e-05, + "loss": 0.4071, + "step": 6734 + }, + { + "epoch": 2.5395927601809953, + "grad_norm": 0.17118105593171307, + "learning_rate": 4.5854571189239555e-05, + "loss": 0.4153, + "step": 6735 + }, + { + "epoch": 2.539969834087481, + "grad_norm": 0.18940254853780503, + "learning_rate": 4.5844154495500896e-05, + "loss": 0.4245, + "step": 6736 + }, + { + "epoch": 2.5403469079939667, + "grad_norm": 0.2366771599834102, + "learning_rate": 4.583373739676752e-05, + "loss": 0.4206, + "step": 6737 + }, + { + "epoch": 2.5407239819004523, + "grad_norm": 0.22068821949128417, + "learning_rate": 4.582331989376131e-05, + "loss": 0.3936, + "step": 6738 + }, + { + "epoch": 2.541101055806938, + "grad_norm": 0.1676548141779952, + "learning_rate": 4.581290198720421e-05, + "loss": 0.4117, + "step": 6739 + }, + { + "epoch": 2.5414781297134237, + "grad_norm": 0.2224413644430634, + "learning_rate": 4.5802483677818154e-05, + "loss": 0.4135, + "step": 6740 + }, + { + "epoch": 2.5418552036199094, + "grad_norm": 0.3038458524043167, + "learning_rate": 4.579206496632512e-05, + "loss": 0.4247, + "step": 6741 + }, + { + "epoch": 2.542232277526395, + "grad_norm": 0.23658291114727237, + "learning_rate": 4.578164585344712e-05, + "loss": 0.4146, + "step": 6742 + }, + { + "epoch": 2.5426093514328807, + "grad_norm": 0.17181253694778642, + "learning_rate": 4.5771226339906195e-05, + "loss": 0.4167, + "step": 6743 + }, + { + "epoch": 2.5429864253393664, + "grad_norm": 0.15475888104566488, + "learning_rate": 4.57608064264244e-05, + "loss": 0.4149, + "step": 6744 + }, + { + "epoch": 2.543363499245852, + "grad_norm": 0.2634998260694311, + "learning_rate": 4.5750386113723826e-05, + "loss": 0.4231, + "step": 6745 + }, + { + "epoch": 2.5437405731523377, + "grad_norm": 0.28904354351114847, + "learning_rate": 4.57399654025266e-05, + "loss": 0.4208, + "step": 6746 + }, + { + "epoch": 2.5441176470588234, + "grad_norm": 0.20828493693134276, + "learning_rate": 4.572954429355487e-05, + "loss": 0.4138, + "step": 6747 + }, + { + "epoch": 2.544494720965309, + "grad_norm": 0.15806426195117812, + "learning_rate": 4.571912278753079e-05, + "loss": 0.4177, + "step": 6748 + }, + { + "epoch": 2.5448717948717947, + "grad_norm": 0.2717630151015085, + "learning_rate": 4.570870088517658e-05, + "loss": 0.4231, + "step": 6749 + }, + { + "epoch": 2.5452488687782804, + "grad_norm": 0.28765554922126657, + "learning_rate": 4.569827858721446e-05, + "loss": 0.4125, + "step": 6750 + }, + { + "epoch": 2.545625942684766, + "grad_norm": 0.20805349585068558, + "learning_rate": 4.56878558943667e-05, + "loss": 0.4101, + "step": 6751 + }, + { + "epoch": 2.5460030165912517, + "grad_norm": 0.26316467554255196, + "learning_rate": 4.567743280735555e-05, + "loss": 0.4172, + "step": 6752 + }, + { + "epoch": 2.5463800904977374, + "grad_norm": 0.2921188376325663, + "learning_rate": 4.566700932690337e-05, + "loss": 0.4234, + "step": 6753 + }, + { + "epoch": 2.546757164404223, + "grad_norm": 0.2370216307958384, + "learning_rate": 4.565658545373245e-05, + "loss": 0.4198, + "step": 6754 + }, + { + "epoch": 2.5471342383107087, + "grad_norm": 0.1788078263503027, + "learning_rate": 4.564616118856517e-05, + "loss": 0.4253, + "step": 6755 + }, + { + "epoch": 2.5475113122171944, + "grad_norm": 0.2867288125495997, + "learning_rate": 4.563573653212394e-05, + "loss": 0.4103, + "step": 6756 + }, + { + "epoch": 2.54788838612368, + "grad_norm": 0.3509122385580405, + "learning_rate": 4.562531148513117e-05, + "loss": 0.4247, + "step": 6757 + }, + { + "epoch": 2.5482654600301657, + "grad_norm": 0.2859678038538176, + "learning_rate": 4.5614886048309295e-05, + "loss": 0.4128, + "step": 6758 + }, + { + "epoch": 2.5486425339366514, + "grad_norm": 0.2433879819573446, + "learning_rate": 4.56044602223808e-05, + "loss": 0.4094, + "step": 6759 + }, + { + "epoch": 2.549019607843137, + "grad_norm": 0.26617640457153147, + "learning_rate": 4.559403400806819e-05, + "loss": 0.4153, + "step": 6760 + }, + { + "epoch": 2.5493966817496228, + "grad_norm": 0.25398453388658276, + "learning_rate": 4.5583607406093976e-05, + "loss": 0.4103, + "step": 6761 + }, + { + "epoch": 2.5497737556561084, + "grad_norm": 0.20199839948360016, + "learning_rate": 4.557318041718071e-05, + "loss": 0.4127, + "step": 6762 + }, + { + "epoch": 2.550150829562594, + "grad_norm": 0.16558342826184827, + "learning_rate": 4.556275304205102e-05, + "loss": 0.4233, + "step": 6763 + }, + { + "epoch": 2.5505279034690798, + "grad_norm": 0.20180981084068741, + "learning_rate": 4.5552325281427445e-05, + "loss": 0.4192, + "step": 6764 + }, + { + "epoch": 2.5509049773755654, + "grad_norm": 0.2653972025181499, + "learning_rate": 4.554189713603267e-05, + "loss": 0.4198, + "step": 6765 + }, + { + "epoch": 2.551282051282051, + "grad_norm": 0.27233582687457447, + "learning_rate": 4.553146860658933e-05, + "loss": 0.4236, + "step": 6766 + }, + { + "epoch": 2.5516591251885368, + "grad_norm": 0.2559410951296992, + "learning_rate": 4.552103969382013e-05, + "loss": 0.4142, + "step": 6767 + }, + { + "epoch": 2.5520361990950224, + "grad_norm": 0.21450904511156282, + "learning_rate": 4.5510610398447766e-05, + "loss": 0.4096, + "step": 6768 + }, + { + "epoch": 2.552413273001508, + "grad_norm": 0.18321106936380635, + "learning_rate": 4.5500180721194995e-05, + "loss": 0.4236, + "step": 6769 + }, + { + "epoch": 2.552790346907994, + "grad_norm": 0.20028086254218627, + "learning_rate": 4.548975066278458e-05, + "loss": 0.4175, + "step": 6770 + }, + { + "epoch": 2.5531674208144794, + "grad_norm": 0.1916011228136714, + "learning_rate": 4.547932022393932e-05, + "loss": 0.4208, + "step": 6771 + }, + { + "epoch": 2.553544494720965, + "grad_norm": 0.1963233553805468, + "learning_rate": 4.546888940538202e-05, + "loss": 0.42, + "step": 6772 + }, + { + "epoch": 2.553921568627451, + "grad_norm": 0.2910022385533348, + "learning_rate": 4.545845820783552e-05, + "loss": 0.4154, + "step": 6773 + }, + { + "epoch": 2.5542986425339365, + "grad_norm": 0.22537119319926055, + "learning_rate": 4.5448026632022734e-05, + "loss": 0.4056, + "step": 6774 + }, + { + "epoch": 2.554675716440422, + "grad_norm": 0.23598128163396917, + "learning_rate": 4.543759467866652e-05, + "loss": 0.4253, + "step": 6775 + }, + { + "epoch": 2.555052790346908, + "grad_norm": 0.22628356705601507, + "learning_rate": 4.5427162348489826e-05, + "loss": 0.4122, + "step": 6776 + }, + { + "epoch": 2.5554298642533935, + "grad_norm": 0.18259024838480026, + "learning_rate": 4.54167296422156e-05, + "loss": 0.413, + "step": 6777 + }, + { + "epoch": 2.555806938159879, + "grad_norm": 0.2061115261576958, + "learning_rate": 4.5406296560566806e-05, + "loss": 0.4148, + "step": 6778 + }, + { + "epoch": 2.556184012066365, + "grad_norm": 0.18460443801983603, + "learning_rate": 4.5395863104266445e-05, + "loss": 0.4258, + "step": 6779 + }, + { + "epoch": 2.5565610859728505, + "grad_norm": 0.17072488158695837, + "learning_rate": 4.538542927403758e-05, + "loss": 0.4201, + "step": 6780 + }, + { + "epoch": 2.556938159879336, + "grad_norm": 0.21828097225461096, + "learning_rate": 4.537499507060322e-05, + "loss": 0.4188, + "step": 6781 + }, + { + "epoch": 2.557315233785822, + "grad_norm": 0.24681165234915356, + "learning_rate": 4.5364560494686486e-05, + "loss": 0.4196, + "step": 6782 + }, + { + "epoch": 2.5576923076923075, + "grad_norm": 0.2586704704947006, + "learning_rate": 4.535412554701047e-05, + "loss": 0.4116, + "step": 6783 + }, + { + "epoch": 2.558069381598793, + "grad_norm": 0.25491356158410905, + "learning_rate": 4.534369022829829e-05, + "loss": 0.4179, + "step": 6784 + }, + { + "epoch": 2.558446455505279, + "grad_norm": 0.2463271536077851, + "learning_rate": 4.533325453927312e-05, + "loss": 0.4127, + "step": 6785 + }, + { + "epoch": 2.5588235294117645, + "grad_norm": 0.23528407599667456, + "learning_rate": 4.532281848065816e-05, + "loss": 0.417, + "step": 6786 + }, + { + "epoch": 2.55920060331825, + "grad_norm": 0.21577524227721176, + "learning_rate": 4.5312382053176573e-05, + "loss": 0.4158, + "step": 6787 + }, + { + "epoch": 2.559577677224736, + "grad_norm": 0.2507260717302762, + "learning_rate": 4.5301945257551634e-05, + "loss": 0.4209, + "step": 6788 + }, + { + "epoch": 2.5599547511312215, + "grad_norm": 0.24711630901399897, + "learning_rate": 4.529150809450659e-05, + "loss": 0.4053, + "step": 6789 + }, + { + "epoch": 2.560331825037707, + "grad_norm": 0.23237763926927849, + "learning_rate": 4.5281070564764734e-05, + "loss": 0.4252, + "step": 6790 + }, + { + "epoch": 2.560708898944193, + "grad_norm": 0.25828920212612494, + "learning_rate": 4.527063266904936e-05, + "loss": 0.425, + "step": 6791 + }, + { + "epoch": 2.5610859728506785, + "grad_norm": 0.29079778001635626, + "learning_rate": 4.5260194408083826e-05, + "loss": 0.4096, + "step": 6792 + }, + { + "epoch": 2.561463046757164, + "grad_norm": 0.2439792006613296, + "learning_rate": 4.524975578259149e-05, + "loss": 0.4144, + "step": 6793 + }, + { + "epoch": 2.56184012066365, + "grad_norm": 0.20178837482673498, + "learning_rate": 4.5239316793295716e-05, + "loss": 0.409, + "step": 6794 + }, + { + "epoch": 2.5622171945701355, + "grad_norm": 0.19970491683946417, + "learning_rate": 4.5228877440919965e-05, + "loss": 0.4188, + "step": 6795 + }, + { + "epoch": 2.562594268476621, + "grad_norm": 0.22105191233316213, + "learning_rate": 4.521843772618762e-05, + "loss": 0.4116, + "step": 6796 + }, + { + "epoch": 2.562971342383107, + "grad_norm": 0.20717393492259897, + "learning_rate": 4.520799764982218e-05, + "loss": 0.4331, + "step": 6797 + }, + { + "epoch": 2.5633484162895925, + "grad_norm": 0.19456025584432207, + "learning_rate": 4.519755721254712e-05, + "loss": 0.4182, + "step": 6798 + }, + { + "epoch": 2.563725490196078, + "grad_norm": 0.17956968793266476, + "learning_rate": 4.518711641508596e-05, + "loss": 0.4172, + "step": 6799 + }, + { + "epoch": 2.564102564102564, + "grad_norm": 0.23148445907654078, + "learning_rate": 4.5176675258162214e-05, + "loss": 0.4121, + "step": 6800 + }, + { + "epoch": 2.5644796380090495, + "grad_norm": 0.21599528957167852, + "learning_rate": 4.5166233742499485e-05, + "loss": 0.4213, + "step": 6801 + }, + { + "epoch": 2.564856711915535, + "grad_norm": 0.15805837495836392, + "learning_rate": 4.515579186882134e-05, + "loss": 0.414, + "step": 6802 + }, + { + "epoch": 2.565233785822021, + "grad_norm": 0.19983833613495117, + "learning_rate": 4.5145349637851375e-05, + "loss": 0.4235, + "step": 6803 + }, + { + "epoch": 2.5656108597285066, + "grad_norm": 0.2931570451468816, + "learning_rate": 4.513490705031326e-05, + "loss": 0.4218, + "step": 6804 + }, + { + "epoch": 2.5659879336349922, + "grad_norm": 0.2930712688111307, + "learning_rate": 4.512446410693064e-05, + "loss": 0.4218, + "step": 6805 + }, + { + "epoch": 2.566365007541478, + "grad_norm": 0.21296741833532812, + "learning_rate": 4.511402080842719e-05, + "loss": 0.4184, + "step": 6806 + }, + { + "epoch": 2.5667420814479636, + "grad_norm": 0.2189880266554568, + "learning_rate": 4.510357715552666e-05, + "loss": 0.4022, + "step": 6807 + }, + { + "epoch": 2.5671191553544492, + "grad_norm": 0.203635080889411, + "learning_rate": 4.5093133148952737e-05, + "loss": 0.4096, + "step": 6808 + }, + { + "epoch": 2.567496229260935, + "grad_norm": 0.1757786633969993, + "learning_rate": 4.50826887894292e-05, + "loss": 0.4215, + "step": 6809 + }, + { + "epoch": 2.5678733031674206, + "grad_norm": 0.1815927769188312, + "learning_rate": 4.507224407767986e-05, + "loss": 0.4226, + "step": 6810 + }, + { + "epoch": 2.5682503770739062, + "grad_norm": 0.2412910737374382, + "learning_rate": 4.506179901442849e-05, + "loss": 0.4093, + "step": 6811 + }, + { + "epoch": 2.568627450980392, + "grad_norm": 0.2778199869301783, + "learning_rate": 4.5051353600398934e-05, + "loss": 0.4198, + "step": 6812 + }, + { + "epoch": 2.5690045248868776, + "grad_norm": 0.291034507863486, + "learning_rate": 4.5040907836315063e-05, + "loss": 0.4069, + "step": 6813 + }, + { + "epoch": 2.5693815987933633, + "grad_norm": 0.2916520485257482, + "learning_rate": 4.503046172290075e-05, + "loss": 0.4228, + "step": 6814 + }, + { + "epoch": 2.569758672699849, + "grad_norm": 0.22052307149533726, + "learning_rate": 4.5020015260879894e-05, + "loss": 0.4115, + "step": 6815 + }, + { + "epoch": 2.5701357466063346, + "grad_norm": 0.18612322819495353, + "learning_rate": 4.500956845097645e-05, + "loss": 0.4331, + "step": 6816 + }, + { + "epoch": 2.5705128205128203, + "grad_norm": 0.2529336941209689, + "learning_rate": 4.499912129391434e-05, + "loss": 0.4139, + "step": 6817 + }, + { + "epoch": 2.570889894419306, + "grad_norm": 0.24500633167163646, + "learning_rate": 4.498867379041756e-05, + "loss": 0.4185, + "step": 6818 + }, + { + "epoch": 2.5712669683257916, + "grad_norm": 0.2114498469936754, + "learning_rate": 4.497822594121013e-05, + "loss": 0.4097, + "step": 6819 + }, + { + "epoch": 2.5716440422322773, + "grad_norm": 0.26603263181603515, + "learning_rate": 4.496777774701603e-05, + "loss": 0.4111, + "step": 6820 + }, + { + "epoch": 2.572021116138763, + "grad_norm": 0.3182281059348313, + "learning_rate": 4.4957329208559356e-05, + "loss": 0.4034, + "step": 6821 + }, + { + "epoch": 2.5723981900452486, + "grad_norm": 0.3053355209093718, + "learning_rate": 4.494688032656417e-05, + "loss": 0.4205, + "step": 6822 + }, + { + "epoch": 2.5727752639517343, + "grad_norm": 0.24103094224217683, + "learning_rate": 4.4936431101754546e-05, + "loss": 0.4203, + "step": 6823 + }, + { + "epoch": 2.57315233785822, + "grad_norm": 0.18231474401338849, + "learning_rate": 4.492598153485463e-05, + "loss": 0.4164, + "step": 6824 + }, + { + "epoch": 2.5735294117647056, + "grad_norm": 0.21953650748610565, + "learning_rate": 4.491553162658857e-05, + "loss": 0.404, + "step": 6825 + }, + { + "epoch": 2.5739064856711913, + "grad_norm": 0.2484414266137513, + "learning_rate": 4.4905081377680525e-05, + "loss": 0.4181, + "step": 6826 + }, + { + "epoch": 2.574283559577677, + "grad_norm": 0.2093260488739752, + "learning_rate": 4.489463078885469e-05, + "loss": 0.4021, + "step": 6827 + }, + { + "epoch": 2.5746606334841626, + "grad_norm": 0.2184786360935271, + "learning_rate": 4.4884179860835284e-05, + "loss": 0.4238, + "step": 6828 + }, + { + "epoch": 2.5750377073906483, + "grad_norm": 0.21710285180021743, + "learning_rate": 4.487372859434655e-05, + "loss": 0.4114, + "step": 6829 + }, + { + "epoch": 2.575414781297134, + "grad_norm": 0.19976943827304425, + "learning_rate": 4.4863276990112745e-05, + "loss": 0.4136, + "step": 6830 + }, + { + "epoch": 2.5757918552036196, + "grad_norm": 0.22762663585232368, + "learning_rate": 4.485282504885816e-05, + "loss": 0.4247, + "step": 6831 + }, + { + "epoch": 2.5761689291101053, + "grad_norm": 0.21308675988379883, + "learning_rate": 4.4842372771307104e-05, + "loss": 0.4061, + "step": 6832 + }, + { + "epoch": 2.576546003016591, + "grad_norm": 0.20897357014597825, + "learning_rate": 4.4831920158183906e-05, + "loss": 0.4261, + "step": 6833 + }, + { + "epoch": 2.5769230769230766, + "grad_norm": 0.2138509539738782, + "learning_rate": 4.4821467210212924e-05, + "loss": 0.4148, + "step": 6834 + }, + { + "epoch": 2.5773001508295623, + "grad_norm": 0.18785263968096316, + "learning_rate": 4.4811013928118555e-05, + "loss": 0.415, + "step": 6835 + }, + { + "epoch": 2.577677224736048, + "grad_norm": 0.19945015513641182, + "learning_rate": 4.480056031262517e-05, + "loss": 0.4175, + "step": 6836 + }, + { + "epoch": 2.5780542986425337, + "grad_norm": 0.25082185676637925, + "learning_rate": 4.479010636445722e-05, + "loss": 0.4095, + "step": 6837 + }, + { + "epoch": 2.5784313725490198, + "grad_norm": 0.231281524078383, + "learning_rate": 4.477965208433916e-05, + "loss": 0.422, + "step": 6838 + }, + { + "epoch": 2.5788084464555054, + "grad_norm": 0.20591042912553836, + "learning_rate": 4.476919747299543e-05, + "loss": 0.4154, + "step": 6839 + }, + { + "epoch": 2.579185520361991, + "grad_norm": 0.20155580277120752, + "learning_rate": 4.475874253115055e-05, + "loss": 0.4208, + "step": 6840 + }, + { + "epoch": 2.579562594268477, + "grad_norm": 0.22362365515671337, + "learning_rate": 4.474828725952905e-05, + "loss": 0.4215, + "step": 6841 + }, + { + "epoch": 2.5799396681749625, + "grad_norm": 0.21586541035580367, + "learning_rate": 4.473783165885544e-05, + "loss": 0.4182, + "step": 6842 + }, + { + "epoch": 2.580316742081448, + "grad_norm": 0.2665046360829441, + "learning_rate": 4.472737572985431e-05, + "loss": 0.4219, + "step": 6843 + }, + { + "epoch": 2.580693815987934, + "grad_norm": 0.2729734002301851, + "learning_rate": 4.471691947325022e-05, + "loss": 0.4187, + "step": 6844 + }, + { + "epoch": 2.5810708898944195, + "grad_norm": 0.28254060360227073, + "learning_rate": 4.47064628897678e-05, + "loss": 0.423, + "step": 6845 + }, + { + "epoch": 2.581447963800905, + "grad_norm": 0.32434076603515993, + "learning_rate": 4.469600598013168e-05, + "loss": 0.4271, + "step": 6846 + }, + { + "epoch": 2.581825037707391, + "grad_norm": 0.23431577771887976, + "learning_rate": 4.4685548745066514e-05, + "loss": 0.4192, + "step": 6847 + }, + { + "epoch": 2.5822021116138765, + "grad_norm": 0.17820139540615515, + "learning_rate": 4.4675091185296965e-05, + "loss": 0.4125, + "step": 6848 + }, + { + "epoch": 2.582579185520362, + "grad_norm": 0.21925852488578182, + "learning_rate": 4.466463330154775e-05, + "loss": 0.4098, + "step": 6849 + }, + { + "epoch": 2.582956259426848, + "grad_norm": 0.24090181958909113, + "learning_rate": 4.465417509454358e-05, + "loss": 0.4182, + "step": 6850 + }, + { + "epoch": 2.5833333333333335, + "grad_norm": 0.2818256483762787, + "learning_rate": 4.464371656500921e-05, + "loss": 0.4132, + "step": 6851 + }, + { + "epoch": 2.583710407239819, + "grad_norm": 0.2890412115779988, + "learning_rate": 4.46332577136694e-05, + "loss": 0.4228, + "step": 6852 + }, + { + "epoch": 2.584087481146305, + "grad_norm": 0.2220633615641471, + "learning_rate": 4.462279854124894e-05, + "loss": 0.4287, + "step": 6853 + }, + { + "epoch": 2.5844645550527905, + "grad_norm": 0.18412649901340097, + "learning_rate": 4.461233904847265e-05, + "loss": 0.4302, + "step": 6854 + }, + { + "epoch": 2.584841628959276, + "grad_norm": 0.21888985404333958, + "learning_rate": 4.4601879236065345e-05, + "loss": 0.4113, + "step": 6855 + }, + { + "epoch": 2.585218702865762, + "grad_norm": 0.2591100424542057, + "learning_rate": 4.4591419104751894e-05, + "loss": 0.4091, + "step": 6856 + }, + { + "epoch": 2.5855957767722475, + "grad_norm": 0.26363433434561784, + "learning_rate": 4.4580958655257164e-05, + "loss": 0.4245, + "step": 6857 + }, + { + "epoch": 2.585972850678733, + "grad_norm": 0.26292463051420645, + "learning_rate": 4.457049788830607e-05, + "loss": 0.4181, + "step": 6858 + }, + { + "epoch": 2.586349924585219, + "grad_norm": 0.2530387085081901, + "learning_rate": 4.4560036804623524e-05, + "loss": 0.4269, + "step": 6859 + }, + { + "epoch": 2.5867269984917045, + "grad_norm": 0.23174893309755557, + "learning_rate": 4.454957540493447e-05, + "loss": 0.412, + "step": 6860 + }, + { + "epoch": 2.58710407239819, + "grad_norm": 0.1896694824379452, + "learning_rate": 4.4539113689963885e-05, + "loss": 0.4167, + "step": 6861 + }, + { + "epoch": 2.587481146304676, + "grad_norm": 0.23437424083956918, + "learning_rate": 4.4528651660436744e-05, + "loss": 0.4141, + "step": 6862 + }, + { + "epoch": 2.5878582202111615, + "grad_norm": 0.25796233428171006, + "learning_rate": 4.451818931707805e-05, + "loss": 0.435, + "step": 6863 + }, + { + "epoch": 2.588235294117647, + "grad_norm": 0.20845890051049767, + "learning_rate": 4.450772666061285e-05, + "loss": 0.4166, + "step": 6864 + }, + { + "epoch": 2.588612368024133, + "grad_norm": 0.1994645214328763, + "learning_rate": 4.4497263691766194e-05, + "loss": 0.424, + "step": 6865 + }, + { + "epoch": 2.5889894419306185, + "grad_norm": 0.2165498521809401, + "learning_rate": 4.448680041126315e-05, + "loss": 0.4216, + "step": 6866 + }, + { + "epoch": 2.589366515837104, + "grad_norm": 0.2546813418659189, + "learning_rate": 4.4476336819828815e-05, + "loss": 0.4224, + "step": 6867 + }, + { + "epoch": 2.58974358974359, + "grad_norm": 0.25164840429985935, + "learning_rate": 4.4465872918188317e-05, + "loss": 0.4245, + "step": 6868 + }, + { + "epoch": 2.5901206636500755, + "grad_norm": 0.21565046368024782, + "learning_rate": 4.445540870706678e-05, + "loss": 0.4177, + "step": 6869 + }, + { + "epoch": 2.590497737556561, + "grad_norm": 0.19184764274322683, + "learning_rate": 4.444494418718937e-05, + "loss": 0.4193, + "step": 6870 + }, + { + "epoch": 2.590874811463047, + "grad_norm": 0.2357499771953814, + "learning_rate": 4.443447935928128e-05, + "loss": 0.4115, + "step": 6871 + }, + { + "epoch": 2.5912518853695325, + "grad_norm": 0.2947075251661475, + "learning_rate": 4.44240142240677e-05, + "loss": 0.4168, + "step": 6872 + }, + { + "epoch": 2.591628959276018, + "grad_norm": 0.31391454201979835, + "learning_rate": 4.441354878227385e-05, + "loss": 0.422, + "step": 6873 + }, + { + "epoch": 2.592006033182504, + "grad_norm": 0.3652771278968796, + "learning_rate": 4.4403083034625e-05, + "loss": 0.4197, + "step": 6874 + }, + { + "epoch": 2.5923831070889896, + "grad_norm": 0.36433247833772275, + "learning_rate": 4.439261698184639e-05, + "loss": 0.4225, + "step": 6875 + }, + { + "epoch": 2.5927601809954752, + "grad_norm": 0.25712161874198197, + "learning_rate": 4.438215062466332e-05, + "loss": 0.4125, + "step": 6876 + }, + { + "epoch": 2.593137254901961, + "grad_norm": 0.1663815527356112, + "learning_rate": 4.43716839638011e-05, + "loss": 0.4209, + "step": 6877 + }, + { + "epoch": 2.5935143288084466, + "grad_norm": 0.22613811497302172, + "learning_rate": 4.4361216999985056e-05, + "loss": 0.4239, + "step": 6878 + }, + { + "epoch": 2.5938914027149322, + "grad_norm": 0.31749000434259855, + "learning_rate": 4.435074973394054e-05, + "loss": 0.4166, + "step": 6879 + }, + { + "epoch": 2.594268476621418, + "grad_norm": 0.31332073100620345, + "learning_rate": 4.434028216639294e-05, + "loss": 0.4153, + "step": 6880 + }, + { + "epoch": 2.5946455505279036, + "grad_norm": 0.2273473949551053, + "learning_rate": 4.432981429806763e-05, + "loss": 0.4131, + "step": 6881 + }, + { + "epoch": 2.5950226244343892, + "grad_norm": 0.17922215235285902, + "learning_rate": 4.431934612969002e-05, + "loss": 0.4119, + "step": 6882 + }, + { + "epoch": 2.595399698340875, + "grad_norm": 0.2625697249994332, + "learning_rate": 4.4308877661985565e-05, + "loss": 0.4109, + "step": 6883 + }, + { + "epoch": 2.5957767722473606, + "grad_norm": 0.26836653981438413, + "learning_rate": 4.42984088956797e-05, + "loss": 0.422, + "step": 6884 + }, + { + "epoch": 2.5961538461538463, + "grad_norm": 0.22085330501876643, + "learning_rate": 4.428793983149792e-05, + "loss": 0.4218, + "step": 6885 + }, + { + "epoch": 2.596530920060332, + "grad_norm": 0.23286483416920373, + "learning_rate": 4.427747047016572e-05, + "loss": 0.4292, + "step": 6886 + }, + { + "epoch": 2.5969079939668176, + "grad_norm": 0.20341452458584727, + "learning_rate": 4.426700081240859e-05, + "loss": 0.4134, + "step": 6887 + }, + { + "epoch": 2.5972850678733033, + "grad_norm": 0.15540829091799083, + "learning_rate": 4.42565308589521e-05, + "loss": 0.4182, + "step": 6888 + }, + { + "epoch": 2.597662141779789, + "grad_norm": 0.1468834774355667, + "learning_rate": 4.424606061052179e-05, + "loss": 0.4221, + "step": 6889 + }, + { + "epoch": 2.5980392156862746, + "grad_norm": 0.16009086433636457, + "learning_rate": 4.4235590067843266e-05, + "loss": 0.4168, + "step": 6890 + }, + { + "epoch": 2.5984162895927603, + "grad_norm": 0.23454832223310604, + "learning_rate": 4.422511923164209e-05, + "loss": 0.4201, + "step": 6891 + }, + { + "epoch": 2.598793363499246, + "grad_norm": 0.3293459264693608, + "learning_rate": 4.421464810264389e-05, + "loss": 0.4215, + "step": 6892 + }, + { + "epoch": 2.5991704374057316, + "grad_norm": 0.3235105416561178, + "learning_rate": 4.4204176681574345e-05, + "loss": 0.4197, + "step": 6893 + }, + { + "epoch": 2.5995475113122173, + "grad_norm": 0.22643437267252808, + "learning_rate": 4.419370496915907e-05, + "loss": 0.409, + "step": 6894 + }, + { + "epoch": 2.599924585218703, + "grad_norm": 0.18722504036472953, + "learning_rate": 4.418323296612375e-05, + "loss": 0.4176, + "step": 6895 + }, + { + "epoch": 2.6003016591251886, + "grad_norm": 0.2026555508048777, + "learning_rate": 4.417276067319411e-05, + "loss": 0.4155, + "step": 6896 + }, + { + "epoch": 2.6006787330316743, + "grad_norm": 0.2647487823648774, + "learning_rate": 4.416228809109584e-05, + "loss": 0.4188, + "step": 6897 + }, + { + "epoch": 2.60105580693816, + "grad_norm": 0.3125101314341434, + "learning_rate": 4.4151815220554715e-05, + "loss": 0.4135, + "step": 6898 + }, + { + "epoch": 2.6014328808446456, + "grad_norm": 0.3103625949217333, + "learning_rate": 4.414134206229648e-05, + "loss": 0.4266, + "step": 6899 + }, + { + "epoch": 2.6018099547511313, + "grad_norm": 0.21613416443080755, + "learning_rate": 4.41308686170469e-05, + "loss": 0.4172, + "step": 6900 + }, + { + "epoch": 2.602187028657617, + "grad_norm": 0.18168692184786112, + "learning_rate": 4.412039488553179e-05, + "loss": 0.4226, + "step": 6901 + }, + { + "epoch": 2.6025641025641026, + "grad_norm": 0.1949679936312082, + "learning_rate": 4.410992086847698e-05, + "loss": 0.42, + "step": 6902 + }, + { + "epoch": 2.6029411764705883, + "grad_norm": 0.22233411107619563, + "learning_rate": 4.409944656660828e-05, + "loss": 0.4194, + "step": 6903 + }, + { + "epoch": 2.603318250377074, + "grad_norm": 0.1945787149973013, + "learning_rate": 4.408897198065158e-05, + "loss": 0.4192, + "step": 6904 + }, + { + "epoch": 2.6036953242835597, + "grad_norm": 0.20988911921430217, + "learning_rate": 4.407849711133275e-05, + "loss": 0.4181, + "step": 6905 + }, + { + "epoch": 2.6040723981900453, + "grad_norm": 0.24629422516672675, + "learning_rate": 4.4068021959377695e-05, + "loss": 0.4151, + "step": 6906 + }, + { + "epoch": 2.604449472096531, + "grad_norm": 0.18618786643296814, + "learning_rate": 4.4057546525512314e-05, + "loss": 0.4198, + "step": 6907 + }, + { + "epoch": 2.6048265460030167, + "grad_norm": 0.1941595145928008, + "learning_rate": 4.404707081046256e-05, + "loss": 0.4097, + "step": 6908 + }, + { + "epoch": 2.6052036199095023, + "grad_norm": 0.21626171872256658, + "learning_rate": 4.4036594814954395e-05, + "loss": 0.4195, + "step": 6909 + }, + { + "epoch": 2.605580693815988, + "grad_norm": 0.1979243393555614, + "learning_rate": 4.4026118539713786e-05, + "loss": 0.4152, + "step": 6910 + }, + { + "epoch": 2.6059577677224737, + "grad_norm": 0.1866741890311282, + "learning_rate": 4.401564198546673e-05, + "loss": 0.4153, + "step": 6911 + }, + { + "epoch": 2.6063348416289593, + "grad_norm": 0.2063850293762783, + "learning_rate": 4.400516515293925e-05, + "loss": 0.4206, + "step": 6912 + }, + { + "epoch": 2.606711915535445, + "grad_norm": 0.20796592665567493, + "learning_rate": 4.3994688042857375e-05, + "loss": 0.4295, + "step": 6913 + }, + { + "epoch": 2.6070889894419307, + "grad_norm": 0.1722649943571839, + "learning_rate": 4.3984210655947165e-05, + "loss": 0.4181, + "step": 6914 + }, + { + "epoch": 2.6074660633484164, + "grad_norm": 0.20530321561448453, + "learning_rate": 4.397373299293469e-05, + "loss": 0.4158, + "step": 6915 + }, + { + "epoch": 2.607843137254902, + "grad_norm": 0.26212651198708364, + "learning_rate": 4.3963255054546045e-05, + "loss": 0.4242, + "step": 6916 + }, + { + "epoch": 2.6082202111613877, + "grad_norm": 0.21404196723974808, + "learning_rate": 4.3952776841507336e-05, + "loss": 0.4205, + "step": 6917 + }, + { + "epoch": 2.6085972850678734, + "grad_norm": 0.20043374345141793, + "learning_rate": 4.39422983545447e-05, + "loss": 0.4187, + "step": 6918 + }, + { + "epoch": 2.608974358974359, + "grad_norm": 0.22377436564294928, + "learning_rate": 4.393181959438429e-05, + "loss": 0.4183, + "step": 6919 + }, + { + "epoch": 2.6093514328808447, + "grad_norm": 0.23298678270113637, + "learning_rate": 4.392134056175226e-05, + "loss": 0.4139, + "step": 6920 + }, + { + "epoch": 2.6097285067873304, + "grad_norm": 0.26655512021650746, + "learning_rate": 4.391086125737481e-05, + "loss": 0.4245, + "step": 6921 + }, + { + "epoch": 2.610105580693816, + "grad_norm": 0.3078385949508503, + "learning_rate": 4.390038168197815e-05, + "loss": 0.4271, + "step": 6922 + }, + { + "epoch": 2.6104826546003017, + "grad_norm": 0.302045307290354, + "learning_rate": 4.3889901836288494e-05, + "loss": 0.4224, + "step": 6923 + }, + { + "epoch": 2.6108597285067874, + "grad_norm": 0.22898210048217757, + "learning_rate": 4.387942172103209e-05, + "loss": 0.4149, + "step": 6924 + }, + { + "epoch": 2.611236802413273, + "grad_norm": 0.2078158078676012, + "learning_rate": 4.386894133693521e-05, + "loss": 0.4094, + "step": 6925 + }, + { + "epoch": 2.6116138763197587, + "grad_norm": 0.2573761868555376, + "learning_rate": 4.385846068472412e-05, + "loss": 0.4169, + "step": 6926 + }, + { + "epoch": 2.6119909502262444, + "grad_norm": 0.2410555149102359, + "learning_rate": 4.384797976512512e-05, + "loss": 0.4141, + "step": 6927 + }, + { + "epoch": 2.61236802413273, + "grad_norm": 0.19981702107722504, + "learning_rate": 4.383749857886455e-05, + "loss": 0.411, + "step": 6928 + }, + { + "epoch": 2.6127450980392157, + "grad_norm": 0.2264229766132319, + "learning_rate": 4.382701712666873e-05, + "loss": 0.4134, + "step": 6929 + }, + { + "epoch": 2.6131221719457014, + "grad_norm": 0.18299002093677108, + "learning_rate": 4.381653540926401e-05, + "loss": 0.422, + "step": 6930 + }, + { + "epoch": 2.613499245852187, + "grad_norm": 0.17178165351753255, + "learning_rate": 4.380605342737677e-05, + "loss": 0.4307, + "step": 6931 + }, + { + "epoch": 2.6138763197586727, + "grad_norm": 0.19951205528513902, + "learning_rate": 4.379557118173341e-05, + "loss": 0.4281, + "step": 6932 + }, + { + "epoch": 2.6142533936651584, + "grad_norm": 0.236265852235394, + "learning_rate": 4.3785088673060324e-05, + "loss": 0.4168, + "step": 6933 + }, + { + "epoch": 2.614630467571644, + "grad_norm": 0.1941200944749124, + "learning_rate": 4.377460590208396e-05, + "loss": 0.4137, + "step": 6934 + }, + { + "epoch": 2.6150075414781297, + "grad_norm": 0.1941394477866101, + "learning_rate": 4.376412286953076e-05, + "loss": 0.4293, + "step": 6935 + }, + { + "epoch": 2.6153846153846154, + "grad_norm": 0.18504176544221437, + "learning_rate": 4.375363957612717e-05, + "loss": 0.4247, + "step": 6936 + }, + { + "epoch": 2.615761689291101, + "grad_norm": 0.20893975321671712, + "learning_rate": 4.37431560225997e-05, + "loss": 0.4132, + "step": 6937 + }, + { + "epoch": 2.6161387631975868, + "grad_norm": 0.19762502103792673, + "learning_rate": 4.373267220967483e-05, + "loss": 0.4202, + "step": 6938 + }, + { + "epoch": 2.6165158371040724, + "grad_norm": 0.1598381535933921, + "learning_rate": 4.372218813807909e-05, + "loss": 0.4138, + "step": 6939 + }, + { + "epoch": 2.616892911010558, + "grad_norm": 0.17621828596302772, + "learning_rate": 4.371170380853901e-05, + "loss": 0.4177, + "step": 6940 + }, + { + "epoch": 2.6172699849170438, + "grad_norm": 0.18844153383792317, + "learning_rate": 4.370121922178116e-05, + "loss": 0.4079, + "step": 6941 + }, + { + "epoch": 2.6176470588235294, + "grad_norm": 0.14989892393547988, + "learning_rate": 4.369073437853208e-05, + "loss": 0.4125, + "step": 6942 + }, + { + "epoch": 2.618024132730015, + "grad_norm": 0.1672354974664743, + "learning_rate": 4.3680249279518396e-05, + "loss": 0.4216, + "step": 6943 + }, + { + "epoch": 2.6184012066365008, + "grad_norm": 0.18167498198423254, + "learning_rate": 4.3669763925466706e-05, + "loss": 0.424, + "step": 6944 + }, + { + "epoch": 2.6187782805429864, + "grad_norm": 0.19752225407277305, + "learning_rate": 4.365927831710361e-05, + "loss": 0.4041, + "step": 6945 + }, + { + "epoch": 2.619155354449472, + "grad_norm": 0.18244622247802866, + "learning_rate": 4.364879245515579e-05, + "loss": 0.4222, + "step": 6946 + }, + { + "epoch": 2.619532428355958, + "grad_norm": 0.206354891647613, + "learning_rate": 4.363830634034989e-05, + "loss": 0.419, + "step": 6947 + }, + { + "epoch": 2.6199095022624435, + "grad_norm": 0.18953742036549737, + "learning_rate": 4.362781997341258e-05, + "loss": 0.4275, + "step": 6948 + }, + { + "epoch": 2.620286576168929, + "grad_norm": 0.19154712151120082, + "learning_rate": 4.3617333355070564e-05, + "loss": 0.4146, + "step": 6949 + }, + { + "epoch": 2.620663650075415, + "grad_norm": 0.20329987321524065, + "learning_rate": 4.360684648605056e-05, + "loss": 0.4323, + "step": 6950 + }, + { + "epoch": 2.6210407239819005, + "grad_norm": 0.22204313394609082, + "learning_rate": 4.3596359367079295e-05, + "loss": 0.4224, + "step": 6951 + }, + { + "epoch": 2.621417797888386, + "grad_norm": 0.22984617219838865, + "learning_rate": 4.3585871998883504e-05, + "loss": 0.4147, + "step": 6952 + }, + { + "epoch": 2.621794871794872, + "grad_norm": 0.2055442460340532, + "learning_rate": 4.357538438218997e-05, + "loss": 0.4247, + "step": 6953 + }, + { + "epoch": 2.6221719457013575, + "grad_norm": 0.19683113136589608, + "learning_rate": 4.356489651772547e-05, + "loss": 0.4227, + "step": 6954 + }, + { + "epoch": 2.622549019607843, + "grad_norm": 0.15149898790288696, + "learning_rate": 4.355440840621681e-05, + "loss": 0.4275, + "step": 6955 + }, + { + "epoch": 2.622926093514329, + "grad_norm": 0.151479021451428, + "learning_rate": 4.354392004839079e-05, + "loss": 0.4171, + "step": 6956 + }, + { + "epoch": 2.6233031674208145, + "grad_norm": 0.18956788434625046, + "learning_rate": 4.353343144497427e-05, + "loss": 0.4187, + "step": 6957 + }, + { + "epoch": 2.6236802413273, + "grad_norm": 0.23132882502675164, + "learning_rate": 4.352294259669407e-05, + "loss": 0.4124, + "step": 6958 + }, + { + "epoch": 2.624057315233786, + "grad_norm": 0.24266395072265343, + "learning_rate": 4.351245350427707e-05, + "loss": 0.4283, + "step": 6959 + }, + { + "epoch": 2.6244343891402715, + "grad_norm": 0.20417288418816915, + "learning_rate": 4.3501964168450173e-05, + "loss": 0.4184, + "step": 6960 + }, + { + "epoch": 2.624811463046757, + "grad_norm": 0.18712008622594684, + "learning_rate": 4.349147458994026e-05, + "loss": 0.4259, + "step": 6961 + }, + { + "epoch": 2.625188536953243, + "grad_norm": 0.22584446626876462, + "learning_rate": 4.348098476947424e-05, + "loss": 0.4232, + "step": 6962 + }, + { + "epoch": 2.6255656108597285, + "grad_norm": 0.27907742812351405, + "learning_rate": 4.3470494707779084e-05, + "loss": 0.417, + "step": 6963 + }, + { + "epoch": 2.625942684766214, + "grad_norm": 0.2642504675103517, + "learning_rate": 4.3460004405581715e-05, + "loss": 0.4121, + "step": 6964 + }, + { + "epoch": 2.6263197586727, + "grad_norm": 0.21273915915306324, + "learning_rate": 4.3449513863609104e-05, + "loss": 0.4143, + "step": 6965 + }, + { + "epoch": 2.6266968325791855, + "grad_norm": 0.1845933283654853, + "learning_rate": 4.343902308258825e-05, + "loss": 0.4197, + "step": 6966 + }, + { + "epoch": 2.627073906485671, + "grad_norm": 0.22892831654733473, + "learning_rate": 4.342853206324615e-05, + "loss": 0.4377, + "step": 6967 + }, + { + "epoch": 2.627450980392157, + "grad_norm": 0.2619425785636635, + "learning_rate": 4.3418040806309805e-05, + "loss": 0.4287, + "step": 6968 + }, + { + "epoch": 2.6278280542986425, + "grad_norm": 0.20980423920479307, + "learning_rate": 4.3407549312506275e-05, + "loss": 0.4203, + "step": 6969 + }, + { + "epoch": 2.628205128205128, + "grad_norm": 0.18032162897294918, + "learning_rate": 4.33970575825626e-05, + "loss": 0.4266, + "step": 6970 + }, + { + "epoch": 2.628582202111614, + "grad_norm": 0.21299270660074124, + "learning_rate": 4.338656561720584e-05, + "loss": 0.4045, + "step": 6971 + }, + { + "epoch": 2.6289592760180995, + "grad_norm": 0.24380273488522866, + "learning_rate": 4.33760734171631e-05, + "loss": 0.4199, + "step": 6972 + }, + { + "epoch": 2.629336349924585, + "grad_norm": 0.19856363475146005, + "learning_rate": 4.336558098316145e-05, + "loss": 0.4131, + "step": 6973 + }, + { + "epoch": 2.629713423831071, + "grad_norm": 0.17428626097149322, + "learning_rate": 4.335508831592803e-05, + "loss": 0.4194, + "step": 6974 + }, + { + "epoch": 2.6300904977375565, + "grad_norm": 0.15751220023869333, + "learning_rate": 4.3344595416189975e-05, + "loss": 0.4205, + "step": 6975 + }, + { + "epoch": 2.630467571644042, + "grad_norm": 0.1634922623490514, + "learning_rate": 4.333410228467441e-05, + "loss": 0.4009, + "step": 6976 + }, + { + "epoch": 2.630844645550528, + "grad_norm": 0.17142664454397175, + "learning_rate": 4.3323608922108515e-05, + "loss": 0.4308, + "step": 6977 + }, + { + "epoch": 2.6312217194570136, + "grad_norm": 0.20348156834500747, + "learning_rate": 4.331311532921948e-05, + "loss": 0.4334, + "step": 6978 + }, + { + "epoch": 2.6315987933634992, + "grad_norm": 0.20343862536259036, + "learning_rate": 4.3302621506734484e-05, + "loss": 0.4061, + "step": 6979 + }, + { + "epoch": 2.631975867269985, + "grad_norm": 0.19021915584484722, + "learning_rate": 4.3292127455380754e-05, + "loss": 0.4111, + "step": 6980 + }, + { + "epoch": 2.6323529411764706, + "grad_norm": 0.1740685242656198, + "learning_rate": 4.328163317588552e-05, + "loss": 0.4167, + "step": 6981 + }, + { + "epoch": 2.6327300150829562, + "grad_norm": 0.1737526402822493, + "learning_rate": 4.3271138668975994e-05, + "loss": 0.4242, + "step": 6982 + }, + { + "epoch": 2.633107088989442, + "grad_norm": 0.2216351134690876, + "learning_rate": 4.326064393537946e-05, + "loss": 0.4214, + "step": 6983 + }, + { + "epoch": 2.6334841628959276, + "grad_norm": 0.2823926876969329, + "learning_rate": 4.3250148975823215e-05, + "loss": 0.4212, + "step": 6984 + }, + { + "epoch": 2.6338612368024132, + "grad_norm": 0.31268556179480894, + "learning_rate": 4.323965379103451e-05, + "loss": 0.4172, + "step": 6985 + }, + { + "epoch": 2.634238310708899, + "grad_norm": 0.28728324969910896, + "learning_rate": 4.322915838174068e-05, + "loss": 0.4163, + "step": 6986 + }, + { + "epoch": 2.6346153846153846, + "grad_norm": 0.21777373121044322, + "learning_rate": 4.321866274866905e-05, + "loss": 0.4237, + "step": 6987 + }, + { + "epoch": 2.6349924585218703, + "grad_norm": 0.17277530541224873, + "learning_rate": 4.3208166892546925e-05, + "loss": 0.4173, + "step": 6988 + }, + { + "epoch": 2.635369532428356, + "grad_norm": 0.2573837677296097, + "learning_rate": 4.319767081410168e-05, + "loss": 0.4196, + "step": 6989 + }, + { + "epoch": 2.6357466063348416, + "grad_norm": 0.2728194595313269, + "learning_rate": 4.3187174514060705e-05, + "loss": 0.4039, + "step": 6990 + }, + { + "epoch": 2.6361236802413273, + "grad_norm": 0.22065660808793403, + "learning_rate": 4.317667799315135e-05, + "loss": 0.4158, + "step": 6991 + }, + { + "epoch": 2.636500754147813, + "grad_norm": 0.23696726470252732, + "learning_rate": 4.316618125210102e-05, + "loss": 0.4089, + "step": 6992 + }, + { + "epoch": 2.6368778280542986, + "grad_norm": 0.21264946515217437, + "learning_rate": 4.315568429163716e-05, + "loss": 0.4056, + "step": 6993 + }, + { + "epoch": 2.6372549019607843, + "grad_norm": 0.18138105008119274, + "learning_rate": 4.314518711248715e-05, + "loss": 0.4132, + "step": 6994 + }, + { + "epoch": 2.63763197586727, + "grad_norm": 0.17977060167699424, + "learning_rate": 4.313468971537848e-05, + "loss": 0.4047, + "step": 6995 + }, + { + "epoch": 2.6380090497737556, + "grad_norm": 0.23202819387074142, + "learning_rate": 4.312419210103859e-05, + "loss": 0.4128, + "step": 6996 + }, + { + "epoch": 2.6383861236802413, + "grad_norm": 0.2006765939685535, + "learning_rate": 4.3113694270194956e-05, + "loss": 0.4138, + "step": 6997 + }, + { + "epoch": 2.638763197586727, + "grad_norm": 0.196405664297723, + "learning_rate": 4.310319622357507e-05, + "loss": 0.4126, + "step": 6998 + }, + { + "epoch": 2.6391402714932126, + "grad_norm": 0.18739159635187552, + "learning_rate": 4.309269796190645e-05, + "loss": 0.4206, + "step": 6999 + }, + { + "epoch": 2.6395173453996983, + "grad_norm": 0.22676162068750866, + "learning_rate": 4.3082199485916594e-05, + "loss": 0.416, + "step": 7000 + }, + { + "epoch": 2.639894419306184, + "grad_norm": 0.2343643348442126, + "learning_rate": 4.307170079633305e-05, + "loss": 0.4236, + "step": 7001 + }, + { + "epoch": 2.6402714932126696, + "grad_norm": 0.233982730625486, + "learning_rate": 4.3061201893883366e-05, + "loss": 0.4271, + "step": 7002 + }, + { + "epoch": 2.6406485671191553, + "grad_norm": 0.26799873857552653, + "learning_rate": 4.305070277929511e-05, + "loss": 0.4118, + "step": 7003 + }, + { + "epoch": 2.641025641025641, + "grad_norm": 0.30865985223429226, + "learning_rate": 4.3040203453295845e-05, + "loss": 0.413, + "step": 7004 + }, + { + "epoch": 2.6414027149321266, + "grad_norm": 0.2647354252786327, + "learning_rate": 4.302970391661319e-05, + "loss": 0.4232, + "step": 7005 + }, + { + "epoch": 2.6417797888386123, + "grad_norm": 0.22407984804522266, + "learning_rate": 4.301920416997473e-05, + "loss": 0.4366, + "step": 7006 + }, + { + "epoch": 2.642156862745098, + "grad_norm": 0.17917349228587837, + "learning_rate": 4.30087042141081e-05, + "loss": 0.4282, + "step": 7007 + }, + { + "epoch": 2.6425339366515836, + "grad_norm": 0.19078799664619756, + "learning_rate": 4.299820404974095e-05, + "loss": 0.4049, + "step": 7008 + }, + { + "epoch": 2.6429110105580693, + "grad_norm": 0.2531894793891468, + "learning_rate": 4.29877036776009e-05, + "loss": 0.4071, + "step": 7009 + }, + { + "epoch": 2.643288084464555, + "grad_norm": 0.2279247831343238, + "learning_rate": 4.2977203098415654e-05, + "loss": 0.419, + "step": 7010 + }, + { + "epoch": 2.6436651583710407, + "grad_norm": 0.18496257918855905, + "learning_rate": 4.2966702312912865e-05, + "loss": 0.4151, + "step": 7011 + }, + { + "epoch": 2.6440422322775263, + "grad_norm": 0.18730386931289322, + "learning_rate": 4.295620132182024e-05, + "loss": 0.4098, + "step": 7012 + }, + { + "epoch": 2.644419306184012, + "grad_norm": 0.20769680910729763, + "learning_rate": 4.2945700125865475e-05, + "loss": 0.402, + "step": 7013 + }, + { + "epoch": 2.6447963800904977, + "grad_norm": 0.1769517303239376, + "learning_rate": 4.293519872577631e-05, + "loss": 0.4196, + "step": 7014 + }, + { + "epoch": 2.6451734539969833, + "grad_norm": 0.2003778881833254, + "learning_rate": 4.292469712228048e-05, + "loss": 0.4205, + "step": 7015 + }, + { + "epoch": 2.645550527903469, + "grad_norm": 0.19353793248611376, + "learning_rate": 4.291419531610573e-05, + "loss": 0.4316, + "step": 7016 + }, + { + "epoch": 2.6459276018099547, + "grad_norm": 0.21772487219708345, + "learning_rate": 4.290369330797983e-05, + "loss": 0.427, + "step": 7017 + }, + { + "epoch": 2.6463046757164403, + "grad_norm": 0.2693825945728099, + "learning_rate": 4.289319109863057e-05, + "loss": 0.4135, + "step": 7018 + }, + { + "epoch": 2.646681749622926, + "grad_norm": 0.23670920399785647, + "learning_rate": 4.288268868878573e-05, + "loss": 0.4063, + "step": 7019 + }, + { + "epoch": 2.6470588235294117, + "grad_norm": 0.17694806961922413, + "learning_rate": 4.2872186079173106e-05, + "loss": 0.4174, + "step": 7020 + }, + { + "epoch": 2.6474358974358974, + "grad_norm": 0.17060642083875457, + "learning_rate": 4.286168327052055e-05, + "loss": 0.4276, + "step": 7021 + }, + { + "epoch": 2.647812971342383, + "grad_norm": 0.20897739006964894, + "learning_rate": 4.285118026355587e-05, + "loss": 0.4211, + "step": 7022 + }, + { + "epoch": 2.6481900452488687, + "grad_norm": 0.19599057101353323, + "learning_rate": 4.284067705900693e-05, + "loss": 0.4175, + "step": 7023 + }, + { + "epoch": 2.6485671191553544, + "grad_norm": 0.18281629449591283, + "learning_rate": 4.283017365760159e-05, + "loss": 0.4178, + "step": 7024 + }, + { + "epoch": 2.64894419306184, + "grad_norm": 0.224556639404082, + "learning_rate": 4.281967006006773e-05, + "loss": 0.4166, + "step": 7025 + }, + { + "epoch": 2.6493212669683257, + "grad_norm": 0.21662774969760826, + "learning_rate": 4.280916626713323e-05, + "loss": 0.4086, + "step": 7026 + }, + { + "epoch": 2.6496983408748114, + "grad_norm": 0.1844905937179884, + "learning_rate": 4.279866227952601e-05, + "loss": 0.4277, + "step": 7027 + }, + { + "epoch": 2.650075414781297, + "grad_norm": 0.196221305756656, + "learning_rate": 4.2788158097973966e-05, + "loss": 0.4279, + "step": 7028 + }, + { + "epoch": 2.6504524886877827, + "grad_norm": 0.17062086370364912, + "learning_rate": 4.277765372320504e-05, + "loss": 0.4122, + "step": 7029 + }, + { + "epoch": 2.6508295625942684, + "grad_norm": 0.19672423217044355, + "learning_rate": 4.276714915594718e-05, + "loss": 0.4109, + "step": 7030 + }, + { + "epoch": 2.651206636500754, + "grad_norm": 0.24178874072452744, + "learning_rate": 4.275664439692834e-05, + "loss": 0.4266, + "step": 7031 + }, + { + "epoch": 2.6515837104072397, + "grad_norm": 0.23783789402929276, + "learning_rate": 4.274613944687647e-05, + "loss": 0.41, + "step": 7032 + }, + { + "epoch": 2.6519607843137254, + "grad_norm": 0.2076744703215308, + "learning_rate": 4.27356343065196e-05, + "loss": 0.4119, + "step": 7033 + }, + { + "epoch": 2.652337858220211, + "grad_norm": 0.21170606464340092, + "learning_rate": 4.272512897658568e-05, + "loss": 0.4205, + "step": 7034 + }, + { + "epoch": 2.6527149321266967, + "grad_norm": 0.23496277568744986, + "learning_rate": 4.2714623457802736e-05, + "loss": 0.4056, + "step": 7035 + }, + { + "epoch": 2.6530920060331824, + "grad_norm": 0.19620930877247128, + "learning_rate": 4.270411775089881e-05, + "loss": 0.4114, + "step": 7036 + }, + { + "epoch": 2.653469079939668, + "grad_norm": 0.19206101633284564, + "learning_rate": 4.2693611856601916e-05, + "loss": 0.4271, + "step": 7037 + }, + { + "epoch": 2.6538461538461537, + "grad_norm": 0.19880828839374712, + "learning_rate": 4.2683105775640096e-05, + "loss": 0.4332, + "step": 7038 + }, + { + "epoch": 2.6542232277526394, + "grad_norm": 0.19708229116728057, + "learning_rate": 4.2672599508741445e-05, + "loss": 0.4174, + "step": 7039 + }, + { + "epoch": 2.654600301659125, + "grad_norm": 0.2096184889118617, + "learning_rate": 4.2662093056634005e-05, + "loss": 0.4249, + "step": 7040 + }, + { + "epoch": 2.6549773755656108, + "grad_norm": 0.18099045213750062, + "learning_rate": 4.265158642004589e-05, + "loss": 0.4221, + "step": 7041 + }, + { + "epoch": 2.6553544494720964, + "grad_norm": 0.1758901411936113, + "learning_rate": 4.2641079599705185e-05, + "loss": 0.4192, + "step": 7042 + }, + { + "epoch": 2.655731523378582, + "grad_norm": 0.19956045976790115, + "learning_rate": 4.2630572596339995e-05, + "loss": 0.4177, + "step": 7043 + }, + { + "epoch": 2.6561085972850678, + "grad_norm": 0.2372644467968339, + "learning_rate": 4.262006541067846e-05, + "loss": 0.4257, + "step": 7044 + }, + { + "epoch": 2.6564856711915534, + "grad_norm": 0.217685294372246, + "learning_rate": 4.260955804344873e-05, + "loss": 0.4189, + "step": 7045 + }, + { + "epoch": 2.656862745098039, + "grad_norm": 0.18748350975685926, + "learning_rate": 4.259905049537893e-05, + "loss": 0.4256, + "step": 7046 + }, + { + "epoch": 2.6572398190045248, + "grad_norm": 0.19688320081646427, + "learning_rate": 4.258854276719724e-05, + "loss": 0.4177, + "step": 7047 + }, + { + "epoch": 2.6576168929110104, + "grad_norm": 0.19222339472828354, + "learning_rate": 4.257803485963184e-05, + "loss": 0.422, + "step": 7048 + }, + { + "epoch": 2.657993966817496, + "grad_norm": 0.21734633912855317, + "learning_rate": 4.2567526773410896e-05, + "loss": 0.4268, + "step": 7049 + }, + { + "epoch": 2.658371040723982, + "grad_norm": 0.21306643157086844, + "learning_rate": 4.255701850926263e-05, + "loss": 0.4124, + "step": 7050 + }, + { + "epoch": 2.6587481146304675, + "grad_norm": 0.18080941991763252, + "learning_rate": 4.254651006791526e-05, + "loss": 0.4185, + "step": 7051 + }, + { + "epoch": 2.659125188536953, + "grad_norm": 0.15853466898275564, + "learning_rate": 4.253600145009699e-05, + "loss": 0.4092, + "step": 7052 + }, + { + "epoch": 2.659502262443439, + "grad_norm": 0.2381494873936133, + "learning_rate": 4.252549265653606e-05, + "loss": 0.412, + "step": 7053 + }, + { + "epoch": 2.6598793363499245, + "grad_norm": 0.28090376620977964, + "learning_rate": 4.251498368796074e-05, + "loss": 0.4121, + "step": 7054 + }, + { + "epoch": 2.66025641025641, + "grad_norm": 0.2389551225962507, + "learning_rate": 4.250447454509928e-05, + "loss": 0.4296, + "step": 7055 + }, + { + "epoch": 2.660633484162896, + "grad_norm": 0.19074871218422418, + "learning_rate": 4.249396522867996e-05, + "loss": 0.4149, + "step": 7056 + }, + { + "epoch": 2.6610105580693815, + "grad_norm": 0.2204562942288333, + "learning_rate": 4.2483455739431056e-05, + "loss": 0.4248, + "step": 7057 + }, + { + "epoch": 2.661387631975867, + "grad_norm": 0.20382699858447934, + "learning_rate": 4.2472946078080876e-05, + "loss": 0.4188, + "step": 7058 + }, + { + "epoch": 2.661764705882353, + "grad_norm": 0.20217326188073154, + "learning_rate": 4.2462436245357724e-05, + "loss": 0.4146, + "step": 7059 + }, + { + "epoch": 2.6621417797888385, + "grad_norm": 0.22449017223982412, + "learning_rate": 4.245192624198992e-05, + "loss": 0.4077, + "step": 7060 + }, + { + "epoch": 2.662518853695324, + "grad_norm": 0.1945605522875304, + "learning_rate": 4.24414160687058e-05, + "loss": 0.415, + "step": 7061 + }, + { + "epoch": 2.66289592760181, + "grad_norm": 0.1819680381689244, + "learning_rate": 4.2430905726233724e-05, + "loss": 0.4167, + "step": 7062 + }, + { + "epoch": 2.6632730015082955, + "grad_norm": 0.20003706614594946, + "learning_rate": 4.242039521530204e-05, + "loss": 0.4202, + "step": 7063 + }, + { + "epoch": 2.663650075414781, + "grad_norm": 0.20375600538573324, + "learning_rate": 4.240988453663909e-05, + "loss": 0.4209, + "step": 7064 + }, + { + "epoch": 2.664027149321267, + "grad_norm": 0.1932366372242297, + "learning_rate": 4.239937369097331e-05, + "loss": 0.414, + "step": 7065 + }, + { + "epoch": 2.6644042232277525, + "grad_norm": 0.17326425315889732, + "learning_rate": 4.238886267903304e-05, + "loss": 0.4111, + "step": 7066 + }, + { + "epoch": 2.664781297134238, + "grad_norm": 0.17588873395275556, + "learning_rate": 4.237835150154671e-05, + "loss": 0.4073, + "step": 7067 + }, + { + "epoch": 2.665158371040724, + "grad_norm": 0.18122104096894284, + "learning_rate": 4.236784015924274e-05, + "loss": 0.4288, + "step": 7068 + }, + { + "epoch": 2.6655354449472095, + "grad_norm": 0.1759182425151517, + "learning_rate": 4.235732865284954e-05, + "loss": 0.4101, + "step": 7069 + }, + { + "epoch": 2.665912518853695, + "grad_norm": 0.20250130408828101, + "learning_rate": 4.2346816983095546e-05, + "loss": 0.4141, + "step": 7070 + }, + { + "epoch": 2.666289592760181, + "grad_norm": 0.1908538894828157, + "learning_rate": 4.233630515070923e-05, + "loss": 0.4043, + "step": 7071 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 0.20419865034061727, + "learning_rate": 4.2325793156419035e-05, + "loss": 0.4123, + "step": 7072 + }, + { + "epoch": 2.667043740573152, + "grad_norm": 0.2226834117090769, + "learning_rate": 4.231528100095344e-05, + "loss": 0.4042, + "step": 7073 + }, + { + "epoch": 2.667420814479638, + "grad_norm": 0.26507630659254644, + "learning_rate": 4.230476868504091e-05, + "loss": 0.4286, + "step": 7074 + }, + { + "epoch": 2.6677978883861235, + "grad_norm": 0.2565149890272516, + "learning_rate": 4.229425620940998e-05, + "loss": 0.4048, + "step": 7075 + }, + { + "epoch": 2.668174962292609, + "grad_norm": 0.24441972723705602, + "learning_rate": 4.228374357478911e-05, + "loss": 0.4126, + "step": 7076 + }, + { + "epoch": 2.668552036199095, + "grad_norm": 0.22692432501677434, + "learning_rate": 4.227323078190685e-05, + "loss": 0.4289, + "step": 7077 + }, + { + "epoch": 2.6689291101055805, + "grad_norm": 0.17993361867602242, + "learning_rate": 4.226271783149171e-05, + "loss": 0.4245, + "step": 7078 + }, + { + "epoch": 2.669306184012066, + "grad_norm": 0.24694536529614494, + "learning_rate": 4.2252204724272224e-05, + "loss": 0.4197, + "step": 7079 + }, + { + "epoch": 2.669683257918552, + "grad_norm": 0.3116818678925759, + "learning_rate": 4.224169146097695e-05, + "loss": 0.4224, + "step": 7080 + }, + { + "epoch": 2.6700603318250375, + "grad_norm": 0.26484854561672444, + "learning_rate": 4.2231178042334445e-05, + "loss": 0.4191, + "step": 7081 + }, + { + "epoch": 2.670437405731523, + "grad_norm": 0.2027331134594049, + "learning_rate": 4.222066446907329e-05, + "loss": 0.414, + "step": 7082 + }, + { + "epoch": 2.670814479638009, + "grad_norm": 0.17415029927814826, + "learning_rate": 4.221015074192205e-05, + "loss": 0.4208, + "step": 7083 + }, + { + "epoch": 2.6711915535444946, + "grad_norm": 0.23025588658033566, + "learning_rate": 4.219963686160932e-05, + "loss": 0.4155, + "step": 7084 + }, + { + "epoch": 2.6715686274509802, + "grad_norm": 0.26085987768210245, + "learning_rate": 4.218912282886372e-05, + "loss": 0.4197, + "step": 7085 + }, + { + "epoch": 2.6719457013574663, + "grad_norm": 0.21225866364861465, + "learning_rate": 4.2178608644413835e-05, + "loss": 0.4213, + "step": 7086 + }, + { + "epoch": 2.672322775263952, + "grad_norm": 0.1733403553911161, + "learning_rate": 4.2168094308988306e-05, + "loss": 0.4115, + "step": 7087 + }, + { + "epoch": 2.6726998491704377, + "grad_norm": 0.21260950667478454, + "learning_rate": 4.2157579823315774e-05, + "loss": 0.4201, + "step": 7088 + }, + { + "epoch": 2.6730769230769234, + "grad_norm": 0.2155205505836972, + "learning_rate": 4.2147065188124865e-05, + "loss": 0.4202, + "step": 7089 + }, + { + "epoch": 2.673453996983409, + "grad_norm": 0.19436664902210668, + "learning_rate": 4.213655040414424e-05, + "loss": 0.4158, + "step": 7090 + }, + { + "epoch": 2.6738310708898947, + "grad_norm": 0.19019239973023147, + "learning_rate": 4.212603547210259e-05, + "loss": 0.412, + "step": 7091 + }, + { + "epoch": 2.6742081447963804, + "grad_norm": 0.18693381711277215, + "learning_rate": 4.211552039272854e-05, + "loss": 0.4239, + "step": 7092 + }, + { + "epoch": 2.674585218702866, + "grad_norm": 0.1969326508818525, + "learning_rate": 4.2105005166750814e-05, + "loss": 0.4243, + "step": 7093 + }, + { + "epoch": 2.6749622926093517, + "grad_norm": 0.18044459798949125, + "learning_rate": 4.2094489794898095e-05, + "loss": 0.4129, + "step": 7094 + }, + { + "epoch": 2.6753393665158374, + "grad_norm": 0.2084681625137763, + "learning_rate": 4.208397427789909e-05, + "loss": 0.4152, + "step": 7095 + }, + { + "epoch": 2.675716440422323, + "grad_norm": 0.1961751401896715, + "learning_rate": 4.207345861648252e-05, + "loss": 0.4068, + "step": 7096 + }, + { + "epoch": 2.6760935143288087, + "grad_norm": 0.16331076464152522, + "learning_rate": 4.206294281137712e-05, + "loss": 0.4295, + "step": 7097 + }, + { + "epoch": 2.6764705882352944, + "grad_norm": 0.18858899156843092, + "learning_rate": 4.205242686331159e-05, + "loss": 0.4193, + "step": 7098 + }, + { + "epoch": 2.67684766214178, + "grad_norm": 0.1696506546480849, + "learning_rate": 4.2041910773014704e-05, + "loss": 0.4097, + "step": 7099 + }, + { + "epoch": 2.6772247360482657, + "grad_norm": 0.16798613602197174, + "learning_rate": 4.203139454121522e-05, + "loss": 0.409, + "step": 7100 + }, + { + "epoch": 2.6776018099547514, + "grad_norm": 0.18067598391019601, + "learning_rate": 4.20208781686419e-05, + "loss": 0.4274, + "step": 7101 + }, + { + "epoch": 2.677978883861237, + "grad_norm": 0.18848320617537453, + "learning_rate": 4.20103616560235e-05, + "loss": 0.4081, + "step": 7102 + }, + { + "epoch": 2.6783559577677227, + "grad_norm": 0.21115870629423927, + "learning_rate": 4.199984500408883e-05, + "loss": 0.4257, + "step": 7103 + }, + { + "epoch": 2.6787330316742084, + "grad_norm": 0.20872475443888264, + "learning_rate": 4.198932821356668e-05, + "loss": 0.4216, + "step": 7104 + }, + { + "epoch": 2.679110105580694, + "grad_norm": 0.22707081388544229, + "learning_rate": 4.197881128518583e-05, + "loss": 0.4299, + "step": 7105 + }, + { + "epoch": 2.6794871794871797, + "grad_norm": 0.2586131803045491, + "learning_rate": 4.196829421967513e-05, + "loss": 0.418, + "step": 7106 + }, + { + "epoch": 2.6798642533936654, + "grad_norm": 0.26740345961957873, + "learning_rate": 4.195777701776338e-05, + "loss": 0.417, + "step": 7107 + }, + { + "epoch": 2.680241327300151, + "grad_norm": 0.24917333843364328, + "learning_rate": 4.194725968017943e-05, + "loss": 0.4115, + "step": 7108 + }, + { + "epoch": 2.6806184012066367, + "grad_norm": 0.2321640795092577, + "learning_rate": 4.193674220765211e-05, + "loss": 0.4164, + "step": 7109 + }, + { + "epoch": 2.6809954751131224, + "grad_norm": 0.1830039455343796, + "learning_rate": 4.192622460091025e-05, + "loss": 0.4047, + "step": 7110 + }, + { + "epoch": 2.681372549019608, + "grad_norm": 0.17327192412796646, + "learning_rate": 4.1915706860682747e-05, + "loss": 0.4115, + "step": 7111 + }, + { + "epoch": 2.6817496229260938, + "grad_norm": 0.22001263535465387, + "learning_rate": 4.1905188987698476e-05, + "loss": 0.4124, + "step": 7112 + }, + { + "epoch": 2.6821266968325794, + "grad_norm": 0.21290364458700012, + "learning_rate": 4.189467098268628e-05, + "loss": 0.4137, + "step": 7113 + }, + { + "epoch": 2.682503770739065, + "grad_norm": 0.20436554219907982, + "learning_rate": 4.188415284637506e-05, + "loss": 0.4036, + "step": 7114 + }, + { + "epoch": 2.6828808446455508, + "grad_norm": 0.21910310555304524, + "learning_rate": 4.187363457949373e-05, + "loss": 0.4238, + "step": 7115 + }, + { + "epoch": 2.6832579185520364, + "grad_norm": 0.21806986734820108, + "learning_rate": 4.186311618277118e-05, + "loss": 0.4176, + "step": 7116 + }, + { + "epoch": 2.683634992458522, + "grad_norm": 0.18564793114523873, + "learning_rate": 4.1852597656936336e-05, + "loss": 0.422, + "step": 7117 + }, + { + "epoch": 2.6840120663650078, + "grad_norm": 0.17157937952549787, + "learning_rate": 4.184207900271812e-05, + "loss": 0.4109, + "step": 7118 + }, + { + "epoch": 2.6843891402714934, + "grad_norm": 0.1965518899990823, + "learning_rate": 4.1831560220845455e-05, + "loss": 0.4043, + "step": 7119 + }, + { + "epoch": 2.684766214177979, + "grad_norm": 0.17583430917081228, + "learning_rate": 4.182104131204729e-05, + "loss": 0.4231, + "step": 7120 + }, + { + "epoch": 2.685143288084465, + "grad_norm": 0.17771883624329302, + "learning_rate": 4.181052227705259e-05, + "loss": 0.4114, + "step": 7121 + }, + { + "epoch": 2.6855203619909505, + "grad_norm": 0.17451458940988732, + "learning_rate": 4.180000311659029e-05, + "loss": 0.4277, + "step": 7122 + }, + { + "epoch": 2.685897435897436, + "grad_norm": 0.1602362376275546, + "learning_rate": 4.178948383138937e-05, + "loss": 0.4118, + "step": 7123 + }, + { + "epoch": 2.686274509803922, + "grad_norm": 0.17043893835846147, + "learning_rate": 4.177896442217882e-05, + "loss": 0.4195, + "step": 7124 + }, + { + "epoch": 2.6866515837104075, + "grad_norm": 0.1892056002154046, + "learning_rate": 4.17684448896876e-05, + "loss": 0.419, + "step": 7125 + }, + { + "epoch": 2.687028657616893, + "grad_norm": 0.19651468187568938, + "learning_rate": 4.1757925234644725e-05, + "loss": 0.4179, + "step": 7126 + }, + { + "epoch": 2.687405731523379, + "grad_norm": 0.22687623640080704, + "learning_rate": 4.174740545777919e-05, + "loss": 0.4182, + "step": 7127 + }, + { + "epoch": 2.6877828054298645, + "grad_norm": 0.22549179593910945, + "learning_rate": 4.173688555982001e-05, + "loss": 0.4172, + "step": 7128 + }, + { + "epoch": 2.68815987933635, + "grad_norm": 0.22912578995034602, + "learning_rate": 4.172636554149619e-05, + "loss": 0.422, + "step": 7129 + }, + { + "epoch": 2.688536953242836, + "grad_norm": 0.20226669809474287, + "learning_rate": 4.171584540353678e-05, + "loss": 0.4229, + "step": 7130 + }, + { + "epoch": 2.6889140271493215, + "grad_norm": 0.20008722291329317, + "learning_rate": 4.1705325146670806e-05, + "loss": 0.418, + "step": 7131 + }, + { + "epoch": 2.689291101055807, + "grad_norm": 0.2619994996261917, + "learning_rate": 4.16948047716273e-05, + "loss": 0.41, + "step": 7132 + }, + { + "epoch": 2.689668174962293, + "grad_norm": 0.2764716890514154, + "learning_rate": 4.168428427913534e-05, + "loss": 0.4155, + "step": 7133 + }, + { + "epoch": 2.6900452488687785, + "grad_norm": 0.21509878532180626, + "learning_rate": 4.167376366992397e-05, + "loss": 0.4167, + "step": 7134 + }, + { + "epoch": 2.690422322775264, + "grad_norm": 0.2195441404220009, + "learning_rate": 4.166324294472225e-05, + "loss": 0.4167, + "step": 7135 + }, + { + "epoch": 2.69079939668175, + "grad_norm": 0.20740576941022235, + "learning_rate": 4.1652722104259295e-05, + "loss": 0.424, + "step": 7136 + }, + { + "epoch": 2.6911764705882355, + "grad_norm": 0.19677794689393516, + "learning_rate": 4.164220114926414e-05, + "loss": 0.4001, + "step": 7137 + }, + { + "epoch": 2.691553544494721, + "grad_norm": 0.2049464986144602, + "learning_rate": 4.163168008046591e-05, + "loss": 0.4215, + "step": 7138 + }, + { + "epoch": 2.691930618401207, + "grad_norm": 0.18731889234390903, + "learning_rate": 4.16211588985937e-05, + "loss": 0.4267, + "step": 7139 + }, + { + "epoch": 2.6923076923076925, + "grad_norm": 0.2439135525366067, + "learning_rate": 4.1610637604376614e-05, + "loss": 0.4162, + "step": 7140 + }, + { + "epoch": 2.692684766214178, + "grad_norm": 0.2609348565927516, + "learning_rate": 4.160011619854377e-05, + "loss": 0.4173, + "step": 7141 + }, + { + "epoch": 2.693061840120664, + "grad_norm": 0.247502232272177, + "learning_rate": 4.158959468182428e-05, + "loss": 0.411, + "step": 7142 + }, + { + "epoch": 2.6934389140271495, + "grad_norm": 0.19477580929541094, + "learning_rate": 4.157907305494732e-05, + "loss": 0.4176, + "step": 7143 + }, + { + "epoch": 2.693815987933635, + "grad_norm": 0.17693469639715054, + "learning_rate": 4.156855131864198e-05, + "loss": 0.4351, + "step": 7144 + }, + { + "epoch": 2.694193061840121, + "grad_norm": 0.1854389838082834, + "learning_rate": 4.155802947363742e-05, + "loss": 0.4174, + "step": 7145 + }, + { + "epoch": 2.6945701357466065, + "grad_norm": 0.18194713909639507, + "learning_rate": 4.1547507520662805e-05, + "loss": 0.4301, + "step": 7146 + }, + { + "epoch": 2.694947209653092, + "grad_norm": 0.24001589958818922, + "learning_rate": 4.153698546044729e-05, + "loss": 0.4135, + "step": 7147 + }, + { + "epoch": 2.695324283559578, + "grad_norm": 0.2511643222379948, + "learning_rate": 4.152646329372005e-05, + "loss": 0.4269, + "step": 7148 + }, + { + "epoch": 2.6957013574660635, + "grad_norm": 0.19692773242030331, + "learning_rate": 4.151594102121026e-05, + "loss": 0.4156, + "step": 7149 + }, + { + "epoch": 2.696078431372549, + "grad_norm": 0.189376848277135, + "learning_rate": 4.150541864364709e-05, + "loss": 0.4069, + "step": 7150 + }, + { + "epoch": 2.696455505279035, + "grad_norm": 0.22206753407502827, + "learning_rate": 4.149489616175974e-05, + "loss": 0.4135, + "step": 7151 + }, + { + "epoch": 2.6968325791855206, + "grad_norm": 0.2202981375337744, + "learning_rate": 4.1484373576277434e-05, + "loss": 0.4176, + "step": 7152 + }, + { + "epoch": 2.697209653092006, + "grad_norm": 0.1666642578622925, + "learning_rate": 4.1473850887929344e-05, + "loss": 0.4163, + "step": 7153 + }, + { + "epoch": 2.697586726998492, + "grad_norm": 0.21295134993161924, + "learning_rate": 4.146332809744469e-05, + "loss": 0.4196, + "step": 7154 + }, + { + "epoch": 2.6979638009049776, + "grad_norm": 0.22488031256361785, + "learning_rate": 4.145280520555271e-05, + "loss": 0.4203, + "step": 7155 + }, + { + "epoch": 2.6983408748114632, + "grad_norm": 0.22444015625703115, + "learning_rate": 4.14422822129826e-05, + "loss": 0.4197, + "step": 7156 + }, + { + "epoch": 2.698717948717949, + "grad_norm": 0.24008100716674588, + "learning_rate": 4.143175912046362e-05, + "loss": 0.4084, + "step": 7157 + }, + { + "epoch": 2.6990950226244346, + "grad_norm": 0.23920251348472324, + "learning_rate": 4.142123592872502e-05, + "loss": 0.42, + "step": 7158 + }, + { + "epoch": 2.6994720965309202, + "grad_norm": 0.22184690374645827, + "learning_rate": 4.1410712638496e-05, + "loss": 0.4292, + "step": 7159 + }, + { + "epoch": 2.699849170437406, + "grad_norm": 0.2391019473780932, + "learning_rate": 4.140018925050586e-05, + "loss": 0.4065, + "step": 7160 + }, + { + "epoch": 2.7002262443438916, + "grad_norm": 0.24662458390744194, + "learning_rate": 4.138966576548385e-05, + "loss": 0.4193, + "step": 7161 + }, + { + "epoch": 2.7006033182503772, + "grad_norm": 0.22759794613668147, + "learning_rate": 4.137914218415923e-05, + "loss": 0.4277, + "step": 7162 + }, + { + "epoch": 2.700980392156863, + "grad_norm": 0.1836437289956804, + "learning_rate": 4.1368618507261284e-05, + "loss": 0.4115, + "step": 7163 + }, + { + "epoch": 2.7013574660633486, + "grad_norm": 0.19376282075302506, + "learning_rate": 4.1358094735519294e-05, + "loss": 0.4104, + "step": 7164 + }, + { + "epoch": 2.7017345399698343, + "grad_norm": 0.2175847723984157, + "learning_rate": 4.134757086966254e-05, + "loss": 0.4044, + "step": 7165 + }, + { + "epoch": 2.70211161387632, + "grad_norm": 0.21358844652148878, + "learning_rate": 4.1337046910420305e-05, + "loss": 0.4097, + "step": 7166 + }, + { + "epoch": 2.7024886877828056, + "grad_norm": 0.22065800056650214, + "learning_rate": 4.1326522858521924e-05, + "loss": 0.4222, + "step": 7167 + }, + { + "epoch": 2.7028657616892913, + "grad_norm": 0.20316600381912503, + "learning_rate": 4.131599871469668e-05, + "loss": 0.4119, + "step": 7168 + }, + { + "epoch": 2.703242835595777, + "grad_norm": 0.23624986917101823, + "learning_rate": 4.1305474479673886e-05, + "loss": 0.4209, + "step": 7169 + }, + { + "epoch": 2.7036199095022626, + "grad_norm": 0.21272085778605912, + "learning_rate": 4.129495015418288e-05, + "loss": 0.4145, + "step": 7170 + }, + { + "epoch": 2.7039969834087483, + "grad_norm": 0.18183069129465038, + "learning_rate": 4.128442573895297e-05, + "loss": 0.4129, + "step": 7171 + }, + { + "epoch": 2.704374057315234, + "grad_norm": 0.15948998160677788, + "learning_rate": 4.1273901234713495e-05, + "loss": 0.4228, + "step": 7172 + }, + { + "epoch": 2.7047511312217196, + "grad_norm": 0.20562423092665658, + "learning_rate": 4.12633766421938e-05, + "loss": 0.4242, + "step": 7173 + }, + { + "epoch": 2.7051282051282053, + "grad_norm": 0.23218903238934507, + "learning_rate": 4.1252851962123216e-05, + "loss": 0.4406, + "step": 7174 + }, + { + "epoch": 2.705505279034691, + "grad_norm": 0.255907405567962, + "learning_rate": 4.12423271952311e-05, + "loss": 0.4205, + "step": 7175 + }, + { + "epoch": 2.7058823529411766, + "grad_norm": 0.23118369770014088, + "learning_rate": 4.123180234224682e-05, + "loss": 0.4117, + "step": 7176 + }, + { + "epoch": 2.7062594268476623, + "grad_norm": 0.21331840265594768, + "learning_rate": 4.122127740389973e-05, + "loss": 0.4174, + "step": 7177 + }, + { + "epoch": 2.706636500754148, + "grad_norm": 0.21789221188303262, + "learning_rate": 4.121075238091919e-05, + "loss": 0.4173, + "step": 7178 + }, + { + "epoch": 2.7070135746606336, + "grad_norm": 0.18065464066688924, + "learning_rate": 4.12002272740346e-05, + "loss": 0.4133, + "step": 7179 + }, + { + "epoch": 2.7073906485671193, + "grad_norm": 0.19032034589564603, + "learning_rate": 4.1189702083975316e-05, + "loss": 0.4078, + "step": 7180 + }, + { + "epoch": 2.707767722473605, + "grad_norm": 0.2214352261063055, + "learning_rate": 4.1179176811470724e-05, + "loss": 0.4185, + "step": 7181 + }, + { + "epoch": 2.7081447963800906, + "grad_norm": 0.21184242918740892, + "learning_rate": 4.116865145725024e-05, + "loss": 0.4171, + "step": 7182 + }, + { + "epoch": 2.7085218702865763, + "grad_norm": 0.226954516214594, + "learning_rate": 4.115812602204324e-05, + "loss": 0.4165, + "step": 7183 + }, + { + "epoch": 2.708898944193062, + "grad_norm": 0.22358039806221244, + "learning_rate": 4.114760050657913e-05, + "loss": 0.4145, + "step": 7184 + }, + { + "epoch": 2.7092760180995477, + "grad_norm": 0.2058048159063874, + "learning_rate": 4.113707491158733e-05, + "loss": 0.4212, + "step": 7185 + }, + { + "epoch": 2.7096530920060333, + "grad_norm": 0.18235856722275726, + "learning_rate": 4.112654923779724e-05, + "loss": 0.417, + "step": 7186 + }, + { + "epoch": 2.710030165912519, + "grad_norm": 0.1587931579760762, + "learning_rate": 4.111602348593829e-05, + "loss": 0.4123, + "step": 7187 + }, + { + "epoch": 2.7104072398190047, + "grad_norm": 0.16539964836188267, + "learning_rate": 4.110549765673991e-05, + "loss": 0.4179, + "step": 7188 + }, + { + "epoch": 2.7107843137254903, + "grad_norm": 0.2069528791502323, + "learning_rate": 4.109497175093151e-05, + "loss": 0.4075, + "step": 7189 + }, + { + "epoch": 2.711161387631976, + "grad_norm": 0.20178531490279902, + "learning_rate": 4.108444576924255e-05, + "loss": 0.4259, + "step": 7190 + }, + { + "epoch": 2.7115384615384617, + "grad_norm": 0.21179146151015668, + "learning_rate": 4.107391971240246e-05, + "loss": 0.4234, + "step": 7191 + }, + { + "epoch": 2.7119155354449473, + "grad_norm": 0.17845078081279547, + "learning_rate": 4.106339358114068e-05, + "loss": 0.4173, + "step": 7192 + }, + { + "epoch": 2.712292609351433, + "grad_norm": 0.1992133077054457, + "learning_rate": 4.105286737618666e-05, + "loss": 0.4225, + "step": 7193 + }, + { + "epoch": 2.7126696832579187, + "grad_norm": 0.1896973008144068, + "learning_rate": 4.104234109826988e-05, + "loss": 0.4163, + "step": 7194 + }, + { + "epoch": 2.7130467571644044, + "grad_norm": 0.20456138985307035, + "learning_rate": 4.103181474811978e-05, + "loss": 0.4183, + "step": 7195 + }, + { + "epoch": 2.71342383107089, + "grad_norm": 0.20256665940501395, + "learning_rate": 4.1021288326465824e-05, + "loss": 0.4131, + "step": 7196 + }, + { + "epoch": 2.7138009049773757, + "grad_norm": 0.21111765731332657, + "learning_rate": 4.101076183403752e-05, + "loss": 0.4279, + "step": 7197 + }, + { + "epoch": 2.7141779788838614, + "grad_norm": 0.19458977466205213, + "learning_rate": 4.1000235271564296e-05, + "loss": 0.4116, + "step": 7198 + }, + { + "epoch": 2.714555052790347, + "grad_norm": 0.16601062116437673, + "learning_rate": 4.098970863977565e-05, + "loss": 0.4156, + "step": 7199 + }, + { + "epoch": 2.7149321266968327, + "grad_norm": 0.18642991612654994, + "learning_rate": 4.097918193940109e-05, + "loss": 0.4068, + "step": 7200 + }, + { + "epoch": 2.7153092006033184, + "grad_norm": 0.1933145883692803, + "learning_rate": 4.096865517117008e-05, + "loss": 0.4131, + "step": 7201 + }, + { + "epoch": 2.715686274509804, + "grad_norm": 0.1424592668941369, + "learning_rate": 4.095812833581213e-05, + "loss": 0.4075, + "step": 7202 + }, + { + "epoch": 2.7160633484162897, + "grad_norm": 0.20561704448770576, + "learning_rate": 4.0947601434056725e-05, + "loss": 0.4167, + "step": 7203 + }, + { + "epoch": 2.7164404223227754, + "grad_norm": 0.2047635287237235, + "learning_rate": 4.09370744666334e-05, + "loss": 0.4172, + "step": 7204 + }, + { + "epoch": 2.716817496229261, + "grad_norm": 0.18461379350270063, + "learning_rate": 4.0926547434271634e-05, + "loss": 0.4101, + "step": 7205 + }, + { + "epoch": 2.7171945701357467, + "grad_norm": 0.18434150939936894, + "learning_rate": 4.091602033770096e-05, + "loss": 0.4101, + "step": 7206 + }, + { + "epoch": 2.7175716440422324, + "grad_norm": 0.18547805828837113, + "learning_rate": 4.0905493177650874e-05, + "loss": 0.4125, + "step": 7207 + }, + { + "epoch": 2.717948717948718, + "grad_norm": 0.17804368317243596, + "learning_rate": 4.089496595485093e-05, + "loss": 0.4241, + "step": 7208 + }, + { + "epoch": 2.7183257918552037, + "grad_norm": 0.1773152551992774, + "learning_rate": 4.088443867003062e-05, + "loss": 0.4187, + "step": 7209 + }, + { + "epoch": 2.7187028657616894, + "grad_norm": 0.17562735954936315, + "learning_rate": 4.087391132391951e-05, + "loss": 0.4145, + "step": 7210 + }, + { + "epoch": 2.719079939668175, + "grad_norm": 0.1562003149609797, + "learning_rate": 4.086338391724713e-05, + "loss": 0.4147, + "step": 7211 + }, + { + "epoch": 2.7194570135746607, + "grad_norm": 0.1803826067075835, + "learning_rate": 4.085285645074299e-05, + "loss": 0.4274, + "step": 7212 + }, + { + "epoch": 2.7198340874811464, + "grad_norm": 0.20116838339413146, + "learning_rate": 4.084232892513666e-05, + "loss": 0.4074, + "step": 7213 + }, + { + "epoch": 2.720211161387632, + "grad_norm": 0.16909239767345804, + "learning_rate": 4.083180134115769e-05, + "loss": 0.4173, + "step": 7214 + }, + { + "epoch": 2.7205882352941178, + "grad_norm": 0.15997288319674097, + "learning_rate": 4.0821273699535625e-05, + "loss": 0.4143, + "step": 7215 + }, + { + "epoch": 2.7209653092006034, + "grad_norm": 0.18575521664897526, + "learning_rate": 4.081074600100001e-05, + "loss": 0.4219, + "step": 7216 + }, + { + "epoch": 2.721342383107089, + "grad_norm": 0.20253911154598672, + "learning_rate": 4.080021824628043e-05, + "loss": 0.4203, + "step": 7217 + }, + { + "epoch": 2.7217194570135748, + "grad_norm": 0.21340862481354955, + "learning_rate": 4.0789690436106425e-05, + "loss": 0.4151, + "step": 7218 + }, + { + "epoch": 2.7220965309200604, + "grad_norm": 0.22432917144335116, + "learning_rate": 4.077916257120757e-05, + "loss": 0.41, + "step": 7219 + }, + { + "epoch": 2.722473604826546, + "grad_norm": 0.20629632705220768, + "learning_rate": 4.076863465231345e-05, + "loss": 0.4176, + "step": 7220 + }, + { + "epoch": 2.7228506787330318, + "grad_norm": 0.1732448713372434, + "learning_rate": 4.075810668015363e-05, + "loss": 0.4097, + "step": 7221 + }, + { + "epoch": 2.7232277526395174, + "grad_norm": 0.19209307568269654, + "learning_rate": 4.0747578655457685e-05, + "loss": 0.418, + "step": 7222 + }, + { + "epoch": 2.723604826546003, + "grad_norm": 0.2230349533695211, + "learning_rate": 4.07370505789552e-05, + "loss": 0.4113, + "step": 7223 + }, + { + "epoch": 2.723981900452489, + "grad_norm": 0.2699652061655503, + "learning_rate": 4.072652245137577e-05, + "loss": 0.4166, + "step": 7224 + }, + { + "epoch": 2.7243589743589745, + "grad_norm": 0.27309383898120915, + "learning_rate": 4.071599427344897e-05, + "loss": 0.4245, + "step": 7225 + }, + { + "epoch": 2.72473604826546, + "grad_norm": 0.21945794887478945, + "learning_rate": 4.070546604590441e-05, + "loss": 0.4354, + "step": 7226 + }, + { + "epoch": 2.725113122171946, + "grad_norm": 0.18101016543267157, + "learning_rate": 4.0694937769471675e-05, + "loss": 0.4136, + "step": 7227 + }, + { + "epoch": 2.7254901960784315, + "grad_norm": 0.16374779545174556, + "learning_rate": 4.068440944488037e-05, + "loss": 0.4107, + "step": 7228 + }, + { + "epoch": 2.725867269984917, + "grad_norm": 0.17438130267176732, + "learning_rate": 4.06738810728601e-05, + "loss": 0.421, + "step": 7229 + }, + { + "epoch": 2.726244343891403, + "grad_norm": 0.22169854942094275, + "learning_rate": 4.066335265414046e-05, + "loss": 0.4167, + "step": 7230 + }, + { + "epoch": 2.7266214177978885, + "grad_norm": 0.2645773283653406, + "learning_rate": 4.065282418945109e-05, + "loss": 0.4065, + "step": 7231 + }, + { + "epoch": 2.726998491704374, + "grad_norm": 0.26700143412006155, + "learning_rate": 4.064229567952156e-05, + "loss": 0.4185, + "step": 7232 + }, + { + "epoch": 2.72737556561086, + "grad_norm": 0.192081287310772, + "learning_rate": 4.063176712508152e-05, + "loss": 0.42, + "step": 7233 + }, + { + "epoch": 2.7277526395173455, + "grad_norm": 0.2269165990673935, + "learning_rate": 4.062123852686058e-05, + "loss": 0.4126, + "step": 7234 + }, + { + "epoch": 2.728129713423831, + "grad_norm": 0.27761126365820393, + "learning_rate": 4.061070988558836e-05, + "loss": 0.4187, + "step": 7235 + }, + { + "epoch": 2.728506787330317, + "grad_norm": 0.2478310840660672, + "learning_rate": 4.0600181201994474e-05, + "loss": 0.4212, + "step": 7236 + }, + { + "epoch": 2.7288838612368025, + "grad_norm": 0.2004708386200095, + "learning_rate": 4.058965247680858e-05, + "loss": 0.4204, + "step": 7237 + }, + { + "epoch": 2.729260935143288, + "grad_norm": 0.15112180476908535, + "learning_rate": 4.057912371076029e-05, + "loss": 0.4102, + "step": 7238 + }, + { + "epoch": 2.729638009049774, + "grad_norm": 0.17751805281803715, + "learning_rate": 4.056859490457923e-05, + "loss": 0.4353, + "step": 7239 + }, + { + "epoch": 2.7300150829562595, + "grad_norm": 0.17373272869061088, + "learning_rate": 4.0558066058995066e-05, + "loss": 0.4102, + "step": 7240 + }, + { + "epoch": 2.730392156862745, + "grad_norm": 0.17154899328001524, + "learning_rate": 4.05475371747374e-05, + "loss": 0.4207, + "step": 7241 + }, + { + "epoch": 2.730769230769231, + "grad_norm": 0.19292316520294062, + "learning_rate": 4.0537008252535904e-05, + "loss": 0.4212, + "step": 7242 + }, + { + "epoch": 2.7311463046757165, + "grad_norm": 0.22004365003730275, + "learning_rate": 4.052647929312021e-05, + "loss": 0.4078, + "step": 7243 + }, + { + "epoch": 2.731523378582202, + "grad_norm": 0.22523645861304056, + "learning_rate": 4.0515950297219964e-05, + "loss": 0.4095, + "step": 7244 + }, + { + "epoch": 2.731900452488688, + "grad_norm": 0.17835280189740108, + "learning_rate": 4.050542126556482e-05, + "loss": 0.4187, + "step": 7245 + }, + { + "epoch": 2.7322775263951735, + "grad_norm": 0.15553839870774516, + "learning_rate": 4.049489219888444e-05, + "loss": 0.4196, + "step": 7246 + }, + { + "epoch": 2.732654600301659, + "grad_norm": 0.20974392540745876, + "learning_rate": 4.0484363097908464e-05, + "loss": 0.4224, + "step": 7247 + }, + { + "epoch": 2.733031674208145, + "grad_norm": 0.15978155260272678, + "learning_rate": 4.0473833963366555e-05, + "loss": 0.4151, + "step": 7248 + }, + { + "epoch": 2.7334087481146305, + "grad_norm": 0.17307746673265326, + "learning_rate": 4.046330479598838e-05, + "loss": 0.4201, + "step": 7249 + }, + { + "epoch": 2.733785822021116, + "grad_norm": 0.19776435739508014, + "learning_rate": 4.045277559650358e-05, + "loss": 0.4234, + "step": 7250 + }, + { + "epoch": 2.734162895927602, + "grad_norm": 0.18066354149994587, + "learning_rate": 4.0442246365641826e-05, + "loss": 0.4224, + "step": 7251 + }, + { + "epoch": 2.7345399698340875, + "grad_norm": 0.18784165614220294, + "learning_rate": 4.0431717104132815e-05, + "loss": 0.409, + "step": 7252 + }, + { + "epoch": 2.734917043740573, + "grad_norm": 0.20792687455017891, + "learning_rate": 4.0421187812706174e-05, + "loss": 0.409, + "step": 7253 + }, + { + "epoch": 2.735294117647059, + "grad_norm": 0.17727911572398586, + "learning_rate": 4.04106584920916e-05, + "loss": 0.4113, + "step": 7254 + }, + { + "epoch": 2.7356711915535445, + "grad_norm": 0.17567681948265582, + "learning_rate": 4.040012914301876e-05, + "loss": 0.4202, + "step": 7255 + }, + { + "epoch": 2.73604826546003, + "grad_norm": 0.186609568810349, + "learning_rate": 4.038959976621732e-05, + "loss": 0.4112, + "step": 7256 + }, + { + "epoch": 2.736425339366516, + "grad_norm": 0.19048972408053763, + "learning_rate": 4.037907036241695e-05, + "loss": 0.4184, + "step": 7257 + }, + { + "epoch": 2.7368024132730016, + "grad_norm": 0.18208822113975312, + "learning_rate": 4.036854093234736e-05, + "loss": 0.4154, + "step": 7258 + }, + { + "epoch": 2.7371794871794872, + "grad_norm": 0.16270707492940933, + "learning_rate": 4.0358011476738195e-05, + "loss": 0.4006, + "step": 7259 + }, + { + "epoch": 2.737556561085973, + "grad_norm": 0.20103927532929783, + "learning_rate": 4.0347481996319164e-05, + "loss": 0.4042, + "step": 7260 + }, + { + "epoch": 2.7379336349924586, + "grad_norm": 0.1886466773002198, + "learning_rate": 4.0336952491819935e-05, + "loss": 0.4136, + "step": 7261 + }, + { + "epoch": 2.7383107088989442, + "grad_norm": 0.20676353776185855, + "learning_rate": 4.032642296397019e-05, + "loss": 0.4079, + "step": 7262 + }, + { + "epoch": 2.73868778280543, + "grad_norm": 0.1595762635558245, + "learning_rate": 4.031589341349963e-05, + "loss": 0.4168, + "step": 7263 + }, + { + "epoch": 2.7390648567119156, + "grad_norm": 0.18099859004821722, + "learning_rate": 4.030536384113793e-05, + "loss": 0.4033, + "step": 7264 + }, + { + "epoch": 2.7394419306184012, + "grad_norm": 0.19130188182534386, + "learning_rate": 4.029483424761481e-05, + "loss": 0.4181, + "step": 7265 + }, + { + "epoch": 2.739819004524887, + "grad_norm": 0.1573021539506015, + "learning_rate": 4.028430463365991e-05, + "loss": 0.4102, + "step": 7266 + }, + { + "epoch": 2.7401960784313726, + "grad_norm": 0.17693594636833537, + "learning_rate": 4.027377500000296e-05, + "loss": 0.4021, + "step": 7267 + }, + { + "epoch": 2.7405731523378583, + "grad_norm": 0.15274198946102527, + "learning_rate": 4.026324534737365e-05, + "loss": 0.4089, + "step": 7268 + }, + { + "epoch": 2.740950226244344, + "grad_norm": 0.15623378129431698, + "learning_rate": 4.0252715676501663e-05, + "loss": 0.4297, + "step": 7269 + }, + { + "epoch": 2.7413273001508296, + "grad_norm": 0.18012155739239422, + "learning_rate": 4.024218598811671e-05, + "loss": 0.4067, + "step": 7270 + }, + { + "epoch": 2.7417043740573153, + "grad_norm": 0.2419550651140255, + "learning_rate": 4.023165628294849e-05, + "loss": 0.4124, + "step": 7271 + }, + { + "epoch": 2.742081447963801, + "grad_norm": 0.24809616214652613, + "learning_rate": 4.0221126561726675e-05, + "loss": 0.4147, + "step": 7272 + }, + { + "epoch": 2.7424585218702866, + "grad_norm": 0.21298621814951219, + "learning_rate": 4.0210596825181e-05, + "loss": 0.4154, + "step": 7273 + }, + { + "epoch": 2.7428355957767723, + "grad_norm": 0.21574549883826677, + "learning_rate": 4.0200067074041154e-05, + "loss": 0.4131, + "step": 7274 + }, + { + "epoch": 2.743212669683258, + "grad_norm": 0.1970407445085288, + "learning_rate": 4.0189537309036827e-05, + "loss": 0.4099, + "step": 7275 + }, + { + "epoch": 2.7435897435897436, + "grad_norm": 0.1805626361201273, + "learning_rate": 4.017900753089774e-05, + "loss": 0.4033, + "step": 7276 + }, + { + "epoch": 2.7439668174962293, + "grad_norm": 0.19102156027776562, + "learning_rate": 4.016847774035359e-05, + "loss": 0.4212, + "step": 7277 + }, + { + "epoch": 2.744343891402715, + "grad_norm": 0.2049296052211214, + "learning_rate": 4.015794793813408e-05, + "loss": 0.4146, + "step": 7278 + }, + { + "epoch": 2.7447209653092006, + "grad_norm": 0.1842725451593675, + "learning_rate": 4.0147418124968914e-05, + "loss": 0.4182, + "step": 7279 + }, + { + "epoch": 2.7450980392156863, + "grad_norm": 0.23538430665846735, + "learning_rate": 4.0136888301587815e-05, + "loss": 0.4247, + "step": 7280 + }, + { + "epoch": 2.745475113122172, + "grad_norm": 0.2045895076821099, + "learning_rate": 4.012635846872046e-05, + "loss": 0.4098, + "step": 7281 + }, + { + "epoch": 2.7458521870286576, + "grad_norm": 0.19717225987095563, + "learning_rate": 4.011582862709658e-05, + "loss": 0.4212, + "step": 7282 + }, + { + "epoch": 2.7462292609351433, + "grad_norm": 0.21931231802793658, + "learning_rate": 4.0105298777445896e-05, + "loss": 0.4229, + "step": 7283 + }, + { + "epoch": 2.746606334841629, + "grad_norm": 0.19331953341607558, + "learning_rate": 4.009476892049808e-05, + "loss": 0.4198, + "step": 7284 + }, + { + "epoch": 2.7469834087481146, + "grad_norm": 0.23294240847185552, + "learning_rate": 4.008423905698287e-05, + "loss": 0.412, + "step": 7285 + }, + { + "epoch": 2.7473604826546003, + "grad_norm": 0.2699095007233407, + "learning_rate": 4.0073709187629975e-05, + "loss": 0.4204, + "step": 7286 + }, + { + "epoch": 2.747737556561086, + "grad_norm": 0.22697158984663426, + "learning_rate": 4.006317931316908e-05, + "loss": 0.4115, + "step": 7287 + }, + { + "epoch": 2.7481146304675717, + "grad_norm": 0.13802261693047058, + "learning_rate": 4.005264943432993e-05, + "loss": 0.4131, + "step": 7288 + }, + { + "epoch": 2.7484917043740573, + "grad_norm": 0.18361029303249807, + "learning_rate": 4.0042119551842215e-05, + "loss": 0.4242, + "step": 7289 + }, + { + "epoch": 2.748868778280543, + "grad_norm": 0.23788570461292002, + "learning_rate": 4.003158966643566e-05, + "loss": 0.4196, + "step": 7290 + }, + { + "epoch": 2.7492458521870287, + "grad_norm": 0.20726682638362337, + "learning_rate": 4.0021059778839954e-05, + "loss": 0.4065, + "step": 7291 + }, + { + "epoch": 2.7496229260935143, + "grad_norm": 0.22678497952965743, + "learning_rate": 4.001052988978484e-05, + "loss": 0.4222, + "step": 7292 + }, + { + "epoch": 2.75, + "grad_norm": 0.2232040736741837, + "learning_rate": 4e-05, + "loss": 0.4086, + "step": 7293 + }, + { + "epoch": 2.7503770739064857, + "grad_norm": 0.20138788843525932, + "learning_rate": 3.998947011021518e-05, + "loss": 0.4122, + "step": 7294 + }, + { + "epoch": 2.7507541478129713, + "grad_norm": 0.1988717087430598, + "learning_rate": 3.9978940221160045e-05, + "loss": 0.4174, + "step": 7295 + }, + { + "epoch": 2.751131221719457, + "grad_norm": 0.2156058132215814, + "learning_rate": 3.9968410333564355e-05, + "loss": 0.4211, + "step": 7296 + }, + { + "epoch": 2.7515082956259427, + "grad_norm": 0.1864455914674211, + "learning_rate": 3.99578804481578e-05, + "loss": 0.4185, + "step": 7297 + }, + { + "epoch": 2.7518853695324283, + "grad_norm": 0.1607391842528234, + "learning_rate": 3.9947350565670076e-05, + "loss": 0.4155, + "step": 7298 + }, + { + "epoch": 2.752262443438914, + "grad_norm": 0.2021169657676209, + "learning_rate": 3.9936820686830924e-05, + "loss": 0.424, + "step": 7299 + }, + { + "epoch": 2.7526395173453997, + "grad_norm": 0.19225523230733102, + "learning_rate": 3.992629081237005e-05, + "loss": 0.4121, + "step": 7300 + }, + { + "epoch": 2.7530165912518854, + "grad_norm": 0.18529492454666927, + "learning_rate": 3.9915760943017134e-05, + "loss": 0.4176, + "step": 7301 + }, + { + "epoch": 2.753393665158371, + "grad_norm": 0.24213291161755, + "learning_rate": 3.990523107950193e-05, + "loss": 0.4107, + "step": 7302 + }, + { + "epoch": 2.7537707390648567, + "grad_norm": 0.26079253604509106, + "learning_rate": 3.9894701222554125e-05, + "loss": 0.4105, + "step": 7303 + }, + { + "epoch": 2.7541478129713424, + "grad_norm": 0.24114395338891698, + "learning_rate": 3.988417137290342e-05, + "loss": 0.4085, + "step": 7304 + }, + { + "epoch": 2.754524886877828, + "grad_norm": 0.21518136185832148, + "learning_rate": 3.987364153127955e-05, + "loss": 0.4095, + "step": 7305 + }, + { + "epoch": 2.7549019607843137, + "grad_norm": 0.19395894528267674, + "learning_rate": 3.986311169841221e-05, + "loss": 0.409, + "step": 7306 + }, + { + "epoch": 2.7552790346907994, + "grad_norm": 0.17792023056785075, + "learning_rate": 3.985258187503109e-05, + "loss": 0.407, + "step": 7307 + }, + { + "epoch": 2.755656108597285, + "grad_norm": 0.17333511578948707, + "learning_rate": 3.9842052061865936e-05, + "loss": 0.4166, + "step": 7308 + }, + { + "epoch": 2.7560331825037707, + "grad_norm": 0.2054611562330912, + "learning_rate": 3.983152225964643e-05, + "loss": 0.4132, + "step": 7309 + }, + { + "epoch": 2.7564102564102564, + "grad_norm": 0.16412026977581937, + "learning_rate": 3.9820992469102264e-05, + "loss": 0.4202, + "step": 7310 + }, + { + "epoch": 2.756787330316742, + "grad_norm": 0.18909568286507472, + "learning_rate": 3.981046269096318e-05, + "loss": 0.4249, + "step": 7311 + }, + { + "epoch": 2.7571644042232277, + "grad_norm": 0.26633784183032, + "learning_rate": 3.9799932925958866e-05, + "loss": 0.426, + "step": 7312 + }, + { + "epoch": 2.7575414781297134, + "grad_norm": 0.24769280095224344, + "learning_rate": 3.9789403174819005e-05, + "loss": 0.4128, + "step": 7313 + }, + { + "epoch": 2.757918552036199, + "grad_norm": 0.22303008000313426, + "learning_rate": 3.977887343827333e-05, + "loss": 0.4158, + "step": 7314 + }, + { + "epoch": 2.7582956259426847, + "grad_norm": 0.27810331438374897, + "learning_rate": 3.9768343717051534e-05, + "loss": 0.4211, + "step": 7315 + }, + { + "epoch": 2.7586726998491704, + "grad_norm": 0.2843558335957031, + "learning_rate": 3.9757814011883295e-05, + "loss": 0.4214, + "step": 7316 + }, + { + "epoch": 2.759049773755656, + "grad_norm": 0.2339258622414377, + "learning_rate": 3.974728432349835e-05, + "loss": 0.4109, + "step": 7317 + }, + { + "epoch": 2.7594268476621417, + "grad_norm": 0.18051999189902568, + "learning_rate": 3.9736754652626366e-05, + "loss": 0.4287, + "step": 7318 + }, + { + "epoch": 2.7598039215686274, + "grad_norm": 0.23803601011927872, + "learning_rate": 3.9726224999997047e-05, + "loss": 0.4072, + "step": 7319 + }, + { + "epoch": 2.760180995475113, + "grad_norm": 0.24627296126699325, + "learning_rate": 3.97156953663401e-05, + "loss": 0.4211, + "step": 7320 + }, + { + "epoch": 2.7605580693815988, + "grad_norm": 0.3148815384665872, + "learning_rate": 3.970516575238522e-05, + "loss": 0.4193, + "step": 7321 + }, + { + "epoch": 2.7609351432880844, + "grad_norm": 0.3214934446154317, + "learning_rate": 3.969463615886207e-05, + "loss": 0.4199, + "step": 7322 + }, + { + "epoch": 2.76131221719457, + "grad_norm": 0.27504776918777535, + "learning_rate": 3.9684106586500376e-05, + "loss": 0.4102, + "step": 7323 + }, + { + "epoch": 2.7616892911010558, + "grad_norm": 0.24550667803237042, + "learning_rate": 3.967357703602983e-05, + "loss": 0.4243, + "step": 7324 + }, + { + "epoch": 2.7620663650075414, + "grad_norm": 0.22907195168193356, + "learning_rate": 3.966304750818008e-05, + "loss": 0.4135, + "step": 7325 + }, + { + "epoch": 2.762443438914027, + "grad_norm": 0.1659683804115743, + "learning_rate": 3.9652518003680856e-05, + "loss": 0.418, + "step": 7326 + }, + { + "epoch": 2.7628205128205128, + "grad_norm": 0.18846047386886386, + "learning_rate": 3.964198852326182e-05, + "loss": 0.4169, + "step": 7327 + }, + { + "epoch": 2.7631975867269984, + "grad_norm": 0.2572333888815513, + "learning_rate": 3.9631459067652646e-05, + "loss": 0.4153, + "step": 7328 + }, + { + "epoch": 2.763574660633484, + "grad_norm": 0.2275156206165353, + "learning_rate": 3.9620929637583056e-05, + "loss": 0.4075, + "step": 7329 + }, + { + "epoch": 2.76395173453997, + "grad_norm": 0.18391465092774203, + "learning_rate": 3.96104002337827e-05, + "loss": 0.4127, + "step": 7330 + }, + { + "epoch": 2.7643288084464555, + "grad_norm": 0.1582106270539595, + "learning_rate": 3.959987085698125e-05, + "loss": 0.4036, + "step": 7331 + }, + { + "epoch": 2.764705882352941, + "grad_norm": 0.14935009191793241, + "learning_rate": 3.9589341507908415e-05, + "loss": 0.413, + "step": 7332 + }, + { + "epoch": 2.765082956259427, + "grad_norm": 0.17900259382028655, + "learning_rate": 3.9578812187293846e-05, + "loss": 0.4217, + "step": 7333 + }, + { + "epoch": 2.7654600301659125, + "grad_norm": 0.208877683764546, + "learning_rate": 3.95682828958672e-05, + "loss": 0.4308, + "step": 7334 + }, + { + "epoch": 2.765837104072398, + "grad_norm": 0.21347415778451015, + "learning_rate": 3.955775363435818e-05, + "loss": 0.4041, + "step": 7335 + }, + { + "epoch": 2.766214177978884, + "grad_norm": 0.1675272163943106, + "learning_rate": 3.954722440349644e-05, + "loss": 0.4099, + "step": 7336 + }, + { + "epoch": 2.7665912518853695, + "grad_norm": 0.1390715838796224, + "learning_rate": 3.9536695204011636e-05, + "loss": 0.412, + "step": 7337 + }, + { + "epoch": 2.766968325791855, + "grad_norm": 0.16741273554808347, + "learning_rate": 3.9526166036633465e-05, + "loss": 0.4159, + "step": 7338 + }, + { + "epoch": 2.767345399698341, + "grad_norm": 0.1985913377364947, + "learning_rate": 3.9515636902091556e-05, + "loss": 0.4079, + "step": 7339 + }, + { + "epoch": 2.7677224736048265, + "grad_norm": 0.1761868594498084, + "learning_rate": 3.950510780111557e-05, + "loss": 0.4161, + "step": 7340 + }, + { + "epoch": 2.768099547511312, + "grad_norm": 0.17485838628535316, + "learning_rate": 3.949457873443519e-05, + "loss": 0.4148, + "step": 7341 + }, + { + "epoch": 2.768476621417798, + "grad_norm": 0.17544370040796883, + "learning_rate": 3.948404970278004e-05, + "loss": 0.4168, + "step": 7342 + }, + { + "epoch": 2.7688536953242835, + "grad_norm": 0.17334449478505928, + "learning_rate": 3.94735207068798e-05, + "loss": 0.4126, + "step": 7343 + }, + { + "epoch": 2.769230769230769, + "grad_norm": 0.15739836345218153, + "learning_rate": 3.946299174746411e-05, + "loss": 0.4207, + "step": 7344 + }, + { + "epoch": 2.769607843137255, + "grad_norm": 0.17298328073726316, + "learning_rate": 3.9452462825262606e-05, + "loss": 0.4313, + "step": 7345 + }, + { + "epoch": 2.7699849170437405, + "grad_norm": 0.18458586532311078, + "learning_rate": 3.9441933941004954e-05, + "loss": 0.4141, + "step": 7346 + }, + { + "epoch": 2.770361990950226, + "grad_norm": 0.1746122603700262, + "learning_rate": 3.943140509542078e-05, + "loss": 0.4228, + "step": 7347 + }, + { + "epoch": 2.770739064856712, + "grad_norm": 0.16285862286257055, + "learning_rate": 3.9420876289239724e-05, + "loss": 0.4199, + "step": 7348 + }, + { + "epoch": 2.7711161387631975, + "grad_norm": 0.15992746380325407, + "learning_rate": 3.941034752319143e-05, + "loss": 0.4249, + "step": 7349 + }, + { + "epoch": 2.771493212669683, + "grad_norm": 0.18669006279236205, + "learning_rate": 3.939981879800553e-05, + "loss": 0.4226, + "step": 7350 + }, + { + "epoch": 2.771870286576169, + "grad_norm": 0.15776535599221297, + "learning_rate": 3.938929011441165e-05, + "loss": 0.4209, + "step": 7351 + }, + { + "epoch": 2.7722473604826545, + "grad_norm": 0.16152958305115772, + "learning_rate": 3.937876147313943e-05, + "loss": 0.4251, + "step": 7352 + }, + { + "epoch": 2.77262443438914, + "grad_norm": 0.18801843981825894, + "learning_rate": 3.936823287491849e-05, + "loss": 0.4104, + "step": 7353 + }, + { + "epoch": 2.773001508295626, + "grad_norm": 0.16339994098827945, + "learning_rate": 3.935770432047845e-05, + "loss": 0.4045, + "step": 7354 + }, + { + "epoch": 2.7733785822021115, + "grad_norm": 0.21635660966526052, + "learning_rate": 3.934717581054893e-05, + "loss": 0.417, + "step": 7355 + }, + { + "epoch": 2.773755656108597, + "grad_norm": 0.22793011560262302, + "learning_rate": 3.933664734585955e-05, + "loss": 0.4015, + "step": 7356 + }, + { + "epoch": 2.774132730015083, + "grad_norm": 0.21061786271513183, + "learning_rate": 3.93261189271399e-05, + "loss": 0.4213, + "step": 7357 + }, + { + "epoch": 2.7745098039215685, + "grad_norm": 0.167483822278408, + "learning_rate": 3.9315590555119636e-05, + "loss": 0.4104, + "step": 7358 + }, + { + "epoch": 2.774886877828054, + "grad_norm": 0.17148706940525008, + "learning_rate": 3.930506223052833e-05, + "loss": 0.4047, + "step": 7359 + }, + { + "epoch": 2.77526395173454, + "grad_norm": 0.16131945129324504, + "learning_rate": 3.929453395409559e-05, + "loss": 0.4064, + "step": 7360 + }, + { + "epoch": 2.7756410256410255, + "grad_norm": 0.17253770850064715, + "learning_rate": 3.9284005726551034e-05, + "loss": 0.4329, + "step": 7361 + }, + { + "epoch": 2.776018099547511, + "grad_norm": 0.18533682127966, + "learning_rate": 3.927347754862424e-05, + "loss": 0.4168, + "step": 7362 + }, + { + "epoch": 2.776395173453997, + "grad_norm": 0.21761116753597848, + "learning_rate": 3.9262949421044805e-05, + "loss": 0.4158, + "step": 7363 + }, + { + "epoch": 2.7767722473604826, + "grad_norm": 0.19594599147297048, + "learning_rate": 3.925242134454233e-05, + "loss": 0.4152, + "step": 7364 + }, + { + "epoch": 2.7771493212669682, + "grad_norm": 0.18142383411529228, + "learning_rate": 3.924189331984638e-05, + "loss": 0.4106, + "step": 7365 + }, + { + "epoch": 2.777526395173454, + "grad_norm": 0.21547476507208657, + "learning_rate": 3.9231365347686557e-05, + "loss": 0.4022, + "step": 7366 + }, + { + "epoch": 2.7779034690799396, + "grad_norm": 0.2169288330495943, + "learning_rate": 3.922083742879244e-05, + "loss": 0.4142, + "step": 7367 + }, + { + "epoch": 2.7782805429864252, + "grad_norm": 0.18395210177930904, + "learning_rate": 3.921030956389359e-05, + "loss": 0.4042, + "step": 7368 + }, + { + "epoch": 2.778657616892911, + "grad_norm": 0.18467643687851476, + "learning_rate": 3.919978175371958e-05, + "loss": 0.4131, + "step": 7369 + }, + { + "epoch": 2.7790346907993966, + "grad_norm": 0.16621953092151065, + "learning_rate": 3.9189253999e-05, + "loss": 0.4189, + "step": 7370 + }, + { + "epoch": 2.7794117647058822, + "grad_norm": 0.19097414052596995, + "learning_rate": 3.917872630046439e-05, + "loss": 0.4172, + "step": 7371 + }, + { + "epoch": 2.779788838612368, + "grad_norm": 0.2022621078595119, + "learning_rate": 3.9168198658842315e-05, + "loss": 0.4166, + "step": 7372 + }, + { + "epoch": 2.7801659125188536, + "grad_norm": 0.19718167862715852, + "learning_rate": 3.915767107486335e-05, + "loss": 0.4047, + "step": 7373 + }, + { + "epoch": 2.7805429864253393, + "grad_norm": 0.18796704749007298, + "learning_rate": 3.914714354925702e-05, + "loss": 0.4056, + "step": 7374 + }, + { + "epoch": 2.780920060331825, + "grad_norm": 0.19707081748394603, + "learning_rate": 3.913661608275289e-05, + "loss": 0.4135, + "step": 7375 + }, + { + "epoch": 2.7812971342383106, + "grad_norm": 0.20996138081713828, + "learning_rate": 3.91260886760805e-05, + "loss": 0.422, + "step": 7376 + }, + { + "epoch": 2.7816742081447963, + "grad_norm": 0.157680820769913, + "learning_rate": 3.9115561329969384e-05, + "loss": 0.4179, + "step": 7377 + }, + { + "epoch": 2.782051282051282, + "grad_norm": 0.15841653920660217, + "learning_rate": 3.9105034045149084e-05, + "loss": 0.4133, + "step": 7378 + }, + { + "epoch": 2.7824283559577676, + "grad_norm": 0.178908211307677, + "learning_rate": 3.909450682234913e-05, + "loss": 0.4052, + "step": 7379 + }, + { + "epoch": 2.7828054298642533, + "grad_norm": 0.19229348292805695, + "learning_rate": 3.908397966229906e-05, + "loss": 0.4236, + "step": 7380 + }, + { + "epoch": 2.783182503770739, + "grad_norm": 0.1711161217949341, + "learning_rate": 3.907345256572838e-05, + "loss": 0.4168, + "step": 7381 + }, + { + "epoch": 2.7835595776772246, + "grad_norm": 0.16046323717425925, + "learning_rate": 3.9062925533366615e-05, + "loss": 0.42, + "step": 7382 + }, + { + "epoch": 2.7839366515837103, + "grad_norm": 0.16933181847223003, + "learning_rate": 3.905239856594328e-05, + "loss": 0.4101, + "step": 7383 + }, + { + "epoch": 2.784313725490196, + "grad_norm": 0.19368365062272547, + "learning_rate": 3.9041871664187886e-05, + "loss": 0.407, + "step": 7384 + }, + { + "epoch": 2.7846907993966816, + "grad_norm": 0.201980273930628, + "learning_rate": 3.903134482882993e-05, + "loss": 0.4026, + "step": 7385 + }, + { + "epoch": 2.7850678733031673, + "grad_norm": 0.19412023427036346, + "learning_rate": 3.9020818060598926e-05, + "loss": 0.4241, + "step": 7386 + }, + { + "epoch": 2.785444947209653, + "grad_norm": 0.216765457829254, + "learning_rate": 3.901029136022436e-05, + "loss": 0.4204, + "step": 7387 + }, + { + "epoch": 2.7858220211161386, + "grad_norm": 0.24434869896854183, + "learning_rate": 3.899976472843572e-05, + "loss": 0.407, + "step": 7388 + }, + { + "epoch": 2.7861990950226243, + "grad_norm": 0.1960883752364613, + "learning_rate": 3.8989238165962506e-05, + "loss": 0.4096, + "step": 7389 + }, + { + "epoch": 2.78657616892911, + "grad_norm": 0.13126748904877383, + "learning_rate": 3.897871167353418e-05, + "loss": 0.4058, + "step": 7390 + }, + { + "epoch": 2.7869532428355956, + "grad_norm": 0.13069966595921623, + "learning_rate": 3.896818525188023e-05, + "loss": 0.4127, + "step": 7391 + }, + { + "epoch": 2.7873303167420813, + "grad_norm": 0.15103241845314883, + "learning_rate": 3.895765890173014e-05, + "loss": 0.4134, + "step": 7392 + }, + { + "epoch": 2.787707390648567, + "grad_norm": 0.16862411754559092, + "learning_rate": 3.894713262381335e-05, + "loss": 0.4096, + "step": 7393 + }, + { + "epoch": 2.7880844645550527, + "grad_norm": 0.17762446036999727, + "learning_rate": 3.8936606418859336e-05, + "loss": 0.408, + "step": 7394 + }, + { + "epoch": 2.7884615384615383, + "grad_norm": 0.17810797170336815, + "learning_rate": 3.892608028759756e-05, + "loss": 0.4237, + "step": 7395 + }, + { + "epoch": 2.788838612368024, + "grad_norm": 0.15541889142225512, + "learning_rate": 3.8915554230757455e-05, + "loss": 0.4186, + "step": 7396 + }, + { + "epoch": 2.7892156862745097, + "grad_norm": 0.2093513584557765, + "learning_rate": 3.89050282490685e-05, + "loss": 0.4155, + "step": 7397 + }, + { + "epoch": 2.7895927601809953, + "grad_norm": 0.21705178490953267, + "learning_rate": 3.889450234326011e-05, + "loss": 0.4147, + "step": 7398 + }, + { + "epoch": 2.789969834087481, + "grad_norm": 0.19461143364795008, + "learning_rate": 3.888397651406171e-05, + "loss": 0.4174, + "step": 7399 + }, + { + "epoch": 2.7903469079939667, + "grad_norm": 0.19968800650825178, + "learning_rate": 3.887345076220277e-05, + "loss": 0.4077, + "step": 7400 + }, + { + "epoch": 2.7907239819004523, + "grad_norm": 0.20634643225598603, + "learning_rate": 3.886292508841269e-05, + "loss": 0.4229, + "step": 7401 + }, + { + "epoch": 2.791101055806938, + "grad_norm": 0.23102152387302438, + "learning_rate": 3.8852399493420874e-05, + "loss": 0.4183, + "step": 7402 + }, + { + "epoch": 2.7914781297134237, + "grad_norm": 0.24846148641619994, + "learning_rate": 3.8841873977956774e-05, + "loss": 0.3938, + "step": 7403 + }, + { + "epoch": 2.7918552036199094, + "grad_norm": 0.20664561496167824, + "learning_rate": 3.883134854274978e-05, + "loss": 0.4206, + "step": 7404 + }, + { + "epoch": 2.792232277526395, + "grad_norm": 0.20261245476704845, + "learning_rate": 3.8820823188529275e-05, + "loss": 0.4221, + "step": 7405 + }, + { + "epoch": 2.7926093514328807, + "grad_norm": 0.17905611387567466, + "learning_rate": 3.88102979160247e-05, + "loss": 0.4095, + "step": 7406 + }, + { + "epoch": 2.7929864253393664, + "grad_norm": 0.17284291657247633, + "learning_rate": 3.879977272596542e-05, + "loss": 0.4177, + "step": 7407 + }, + { + "epoch": 2.793363499245852, + "grad_norm": 0.16385509055425623, + "learning_rate": 3.878924761908081e-05, + "loss": 0.4205, + "step": 7408 + }, + { + "epoch": 2.7937405731523377, + "grad_norm": 0.20482682090408053, + "learning_rate": 3.877872259610028e-05, + "loss": 0.4187, + "step": 7409 + }, + { + "epoch": 2.7941176470588234, + "grad_norm": 0.2403379730870233, + "learning_rate": 3.8768197657753194e-05, + "loss": 0.415, + "step": 7410 + }, + { + "epoch": 2.794494720965309, + "grad_norm": 0.1815897803927181, + "learning_rate": 3.8757672804768903e-05, + "loss": 0.4193, + "step": 7411 + }, + { + "epoch": 2.7948717948717947, + "grad_norm": 0.19335691941840527, + "learning_rate": 3.87471480378768e-05, + "loss": 0.4258, + "step": 7412 + }, + { + "epoch": 2.7952488687782804, + "grad_norm": 0.24645931577735192, + "learning_rate": 3.873662335780622e-05, + "loss": 0.4103, + "step": 7413 + }, + { + "epoch": 2.795625942684766, + "grad_norm": 0.2461497663868794, + "learning_rate": 3.872609876528651e-05, + "loss": 0.4232, + "step": 7414 + }, + { + "epoch": 2.7960030165912517, + "grad_norm": 0.19122011110068646, + "learning_rate": 3.8715574261047045e-05, + "loss": 0.4156, + "step": 7415 + }, + { + "epoch": 2.7963800904977374, + "grad_norm": 0.17543122402978192, + "learning_rate": 3.870504984581714e-05, + "loss": 0.4243, + "step": 7416 + }, + { + "epoch": 2.796757164404223, + "grad_norm": 0.23526723162046922, + "learning_rate": 3.869452552032612e-05, + "loss": 0.4257, + "step": 7417 + }, + { + "epoch": 2.7971342383107087, + "grad_norm": 0.23642701151504883, + "learning_rate": 3.868400128530333e-05, + "loss": 0.4235, + "step": 7418 + }, + { + "epoch": 2.7975113122171944, + "grad_norm": 0.22071848206550407, + "learning_rate": 3.8673477141478096e-05, + "loss": 0.4241, + "step": 7419 + }, + { + "epoch": 2.79788838612368, + "grad_norm": 0.21567724430937307, + "learning_rate": 3.8662953089579694e-05, + "loss": 0.4141, + "step": 7420 + }, + { + "epoch": 2.7982654600301657, + "grad_norm": 0.1753618023269018, + "learning_rate": 3.865242913033748e-05, + "loss": 0.418, + "step": 7421 + }, + { + "epoch": 2.7986425339366514, + "grad_norm": 0.17681093864460468, + "learning_rate": 3.8641905264480726e-05, + "loss": 0.4245, + "step": 7422 + }, + { + "epoch": 2.799019607843137, + "grad_norm": 0.22080896182447549, + "learning_rate": 3.863138149273872e-05, + "loss": 0.3962, + "step": 7423 + }, + { + "epoch": 2.7993966817496228, + "grad_norm": 0.2173649548284633, + "learning_rate": 3.862085781584078e-05, + "loss": 0.4206, + "step": 7424 + }, + { + "epoch": 2.7997737556561084, + "grad_norm": 0.17998606748941265, + "learning_rate": 3.861033423451617e-05, + "loss": 0.4179, + "step": 7425 + }, + { + "epoch": 2.800150829562594, + "grad_norm": 0.1685011761715935, + "learning_rate": 3.859981074949414e-05, + "loss": 0.4119, + "step": 7426 + }, + { + "epoch": 2.8005279034690798, + "grad_norm": 0.18857320538710926, + "learning_rate": 3.8589287361504003e-05, + "loss": 0.4025, + "step": 7427 + }, + { + "epoch": 2.8009049773755654, + "grad_norm": 0.15311442853783855, + "learning_rate": 3.857876407127501e-05, + "loss": 0.4203, + "step": 7428 + }, + { + "epoch": 2.801282051282051, + "grad_norm": 0.1855466680314537, + "learning_rate": 3.8568240879536384e-05, + "loss": 0.4113, + "step": 7429 + }, + { + "epoch": 2.8016591251885368, + "grad_norm": 0.19444252653021932, + "learning_rate": 3.855771778701741e-05, + "loss": 0.4226, + "step": 7430 + }, + { + "epoch": 2.8020361990950224, + "grad_norm": 0.1950858481568705, + "learning_rate": 3.854719479444731e-05, + "loss": 0.4113, + "step": 7431 + }, + { + "epoch": 2.802413273001508, + "grad_norm": 0.21466282153644306, + "learning_rate": 3.8536671902555314e-05, + "loss": 0.4187, + "step": 7432 + }, + { + "epoch": 2.802790346907994, + "grad_norm": 0.20326806984402862, + "learning_rate": 3.852614911207067e-05, + "loss": 0.4166, + "step": 7433 + }, + { + "epoch": 2.8031674208144794, + "grad_norm": 0.20507747472291585, + "learning_rate": 3.8515626423722586e-05, + "loss": 0.4245, + "step": 7434 + }, + { + "epoch": 2.803544494720965, + "grad_norm": 0.16778124558294052, + "learning_rate": 3.850510383824026e-05, + "loss": 0.4197, + "step": 7435 + }, + { + "epoch": 2.803921568627451, + "grad_norm": 0.17811221299138613, + "learning_rate": 3.849458135635292e-05, + "loss": 0.4167, + "step": 7436 + }, + { + "epoch": 2.8042986425339365, + "grad_norm": 0.21870302417353507, + "learning_rate": 3.848405897878977e-05, + "loss": 0.4233, + "step": 7437 + }, + { + "epoch": 2.804675716440422, + "grad_norm": 0.23426950111586056, + "learning_rate": 3.847353670627996e-05, + "loss": 0.4134, + "step": 7438 + }, + { + "epoch": 2.805052790346908, + "grad_norm": 0.2252742761506629, + "learning_rate": 3.8463014539552716e-05, + "loss": 0.4195, + "step": 7439 + }, + { + "epoch": 2.8054298642533935, + "grad_norm": 0.22254823034893592, + "learning_rate": 3.8452492479337215e-05, + "loss": 0.4152, + "step": 7440 + }, + { + "epoch": 2.805806938159879, + "grad_norm": 0.19303832246290276, + "learning_rate": 3.8441970526362584e-05, + "loss": 0.4132, + "step": 7441 + }, + { + "epoch": 2.806184012066365, + "grad_norm": 0.17633363767725163, + "learning_rate": 3.843144868135804e-05, + "loss": 0.4204, + "step": 7442 + }, + { + "epoch": 2.8065610859728505, + "grad_norm": 0.20479844893592666, + "learning_rate": 3.8420926945052706e-05, + "loss": 0.4157, + "step": 7443 + }, + { + "epoch": 2.806938159879336, + "grad_norm": 0.2043249749507125, + "learning_rate": 3.841040531817571e-05, + "loss": 0.4079, + "step": 7444 + }, + { + "epoch": 2.807315233785822, + "grad_norm": 0.23069389975205837, + "learning_rate": 3.8399883801456244e-05, + "loss": 0.4158, + "step": 7445 + }, + { + "epoch": 2.8076923076923075, + "grad_norm": 0.22316734653618084, + "learning_rate": 3.8389362395623406e-05, + "loss": 0.4151, + "step": 7446 + }, + { + "epoch": 2.808069381598793, + "grad_norm": 0.1811073623112037, + "learning_rate": 3.837884110140631e-05, + "loss": 0.4126, + "step": 7447 + }, + { + "epoch": 2.808446455505279, + "grad_norm": 0.18490214284126258, + "learning_rate": 3.8368319919534104e-05, + "loss": 0.4125, + "step": 7448 + }, + { + "epoch": 2.8088235294117645, + "grad_norm": 0.18768134064321118, + "learning_rate": 3.835779885073588e-05, + "loss": 0.4144, + "step": 7449 + }, + { + "epoch": 2.80920060331825, + "grad_norm": 0.19844762662255935, + "learning_rate": 3.8347277895740725e-05, + "loss": 0.4155, + "step": 7450 + }, + { + "epoch": 2.809577677224736, + "grad_norm": 0.22118705445472708, + "learning_rate": 3.833675705527776e-05, + "loss": 0.4169, + "step": 7451 + }, + { + "epoch": 2.8099547511312215, + "grad_norm": 0.19737530393368563, + "learning_rate": 3.832623633007606e-05, + "loss": 0.4284, + "step": 7452 + }, + { + "epoch": 2.810331825037707, + "grad_norm": 0.2013071779912892, + "learning_rate": 3.831571572086467e-05, + "loss": 0.4163, + "step": 7453 + }, + { + "epoch": 2.810708898944193, + "grad_norm": 0.22122281659223303, + "learning_rate": 3.830519522837271e-05, + "loss": 0.4167, + "step": 7454 + }, + { + "epoch": 2.8110859728506785, + "grad_norm": 0.22286771502501274, + "learning_rate": 3.829467485332922e-05, + "loss": 0.4037, + "step": 7455 + }, + { + "epoch": 2.811463046757164, + "grad_norm": 0.23000800894332923, + "learning_rate": 3.828415459646323e-05, + "loss": 0.4216, + "step": 7456 + }, + { + "epoch": 2.81184012066365, + "grad_norm": 0.20484899767176293, + "learning_rate": 3.8273634458503825e-05, + "loss": 0.4264, + "step": 7457 + }, + { + "epoch": 2.8122171945701355, + "grad_norm": 0.2283655737699593, + "learning_rate": 3.826311444018001e-05, + "loss": 0.4176, + "step": 7458 + }, + { + "epoch": 2.812594268476621, + "grad_norm": 0.22352536956590688, + "learning_rate": 3.825259454222082e-05, + "loss": 0.4189, + "step": 7459 + }, + { + "epoch": 2.812971342383107, + "grad_norm": 0.2204936003613647, + "learning_rate": 3.824207476535529e-05, + "loss": 0.4058, + "step": 7460 + }, + { + "epoch": 2.8133484162895925, + "grad_norm": 0.23543991126929933, + "learning_rate": 3.823155511031242e-05, + "loss": 0.4192, + "step": 7461 + }, + { + "epoch": 2.813725490196078, + "grad_norm": 0.20566972046256693, + "learning_rate": 3.822103557782119e-05, + "loss": 0.4143, + "step": 7462 + }, + { + "epoch": 2.814102564102564, + "grad_norm": 0.1885580469364348, + "learning_rate": 3.821051616861064e-05, + "loss": 0.4164, + "step": 7463 + }, + { + "epoch": 2.8144796380090495, + "grad_norm": 0.17935501720886204, + "learning_rate": 3.819999688340973e-05, + "loss": 0.4196, + "step": 7464 + }, + { + "epoch": 2.814856711915535, + "grad_norm": 0.17196804023062456, + "learning_rate": 3.8189477722947426e-05, + "loss": 0.415, + "step": 7465 + }, + { + "epoch": 2.815233785822021, + "grad_norm": 0.16850924546995533, + "learning_rate": 3.817895868795272e-05, + "loss": 0.4127, + "step": 7466 + }, + { + "epoch": 2.8156108597285066, + "grad_norm": 0.15809419576622408, + "learning_rate": 3.816843977915457e-05, + "loss": 0.4114, + "step": 7467 + }, + { + "epoch": 2.8159879336349922, + "grad_norm": 0.17364532637303245, + "learning_rate": 3.815792099728189e-05, + "loss": 0.4152, + "step": 7468 + }, + { + "epoch": 2.816365007541478, + "grad_norm": 0.18516234131648732, + "learning_rate": 3.814740234306368e-05, + "loss": 0.4234, + "step": 7469 + }, + { + "epoch": 2.8167420814479636, + "grad_norm": 0.1746801675876003, + "learning_rate": 3.813688381722884e-05, + "loss": 0.4009, + "step": 7470 + }, + { + "epoch": 2.8171191553544492, + "grad_norm": 0.15757395268137966, + "learning_rate": 3.812636542050628e-05, + "loss": 0.4045, + "step": 7471 + }, + { + "epoch": 2.817496229260935, + "grad_norm": 0.1824725522788028, + "learning_rate": 3.811584715362495e-05, + "loss": 0.3998, + "step": 7472 + }, + { + "epoch": 2.8178733031674206, + "grad_norm": 0.16783052759374598, + "learning_rate": 3.810532901731375e-05, + "loss": 0.4216, + "step": 7473 + }, + { + "epoch": 2.8182503770739062, + "grad_norm": 0.16276875868538632, + "learning_rate": 3.8094811012301544e-05, + "loss": 0.417, + "step": 7474 + }, + { + "epoch": 2.818627450980392, + "grad_norm": 0.18106462004819251, + "learning_rate": 3.808429313931726e-05, + "loss": 0.4205, + "step": 7475 + }, + { + "epoch": 2.8190045248868776, + "grad_norm": 0.1878844041068084, + "learning_rate": 3.8073775399089764e-05, + "loss": 0.4122, + "step": 7476 + }, + { + "epoch": 2.8193815987933633, + "grad_norm": 0.209681556039567, + "learning_rate": 3.806325779234791e-05, + "loss": 0.4226, + "step": 7477 + }, + { + "epoch": 2.819758672699849, + "grad_norm": 0.19896754790915053, + "learning_rate": 3.805274031982059e-05, + "loss": 0.406, + "step": 7478 + }, + { + "epoch": 2.8201357466063346, + "grad_norm": 0.20234040238273404, + "learning_rate": 3.804222298223664e-05, + "loss": 0.418, + "step": 7479 + }, + { + "epoch": 2.8205128205128203, + "grad_norm": 0.21124176401419623, + "learning_rate": 3.803170578032488e-05, + "loss": 0.4148, + "step": 7480 + }, + { + "epoch": 2.820889894419306, + "grad_norm": 0.24765205057989606, + "learning_rate": 3.802118871481418e-05, + "loss": 0.4171, + "step": 7481 + }, + { + "epoch": 2.8212669683257916, + "grad_norm": 0.2077449261128279, + "learning_rate": 3.801067178643333e-05, + "loss": 0.4106, + "step": 7482 + }, + { + "epoch": 2.8216440422322773, + "grad_norm": 0.1566266966696841, + "learning_rate": 3.8000154995911175e-05, + "loss": 0.4097, + "step": 7483 + }, + { + "epoch": 2.822021116138763, + "grad_norm": 0.20092672039710094, + "learning_rate": 3.798963834397651e-05, + "loss": 0.4261, + "step": 7484 + }, + { + "epoch": 2.8223981900452486, + "grad_norm": 0.1931197216314952, + "learning_rate": 3.797912183135811e-05, + "loss": 0.4209, + "step": 7485 + }, + { + "epoch": 2.8227752639517343, + "grad_norm": 0.21063998773508824, + "learning_rate": 3.7968605458784786e-05, + "loss": 0.4222, + "step": 7486 + }, + { + "epoch": 2.82315233785822, + "grad_norm": 0.20551384686455001, + "learning_rate": 3.795808922698531e-05, + "loss": 0.4088, + "step": 7487 + }, + { + "epoch": 2.8235294117647056, + "grad_norm": 0.23259828674033511, + "learning_rate": 3.794757313668841e-05, + "loss": 0.4202, + "step": 7488 + }, + { + "epoch": 2.8239064856711913, + "grad_norm": 0.24259282786028552, + "learning_rate": 3.7937057188622896e-05, + "loss": 0.4264, + "step": 7489 + }, + { + "epoch": 2.824283559577677, + "grad_norm": 0.19126741765192604, + "learning_rate": 3.792654138351749e-05, + "loss": 0.4246, + "step": 7490 + }, + { + "epoch": 2.8246606334841626, + "grad_norm": 0.1499734059081834, + "learning_rate": 3.791602572210091e-05, + "loss": 0.4109, + "step": 7491 + }, + { + "epoch": 2.8250377073906483, + "grad_norm": 0.1814800507348789, + "learning_rate": 3.790551020510191e-05, + "loss": 0.429, + "step": 7492 + }, + { + "epoch": 2.825414781297134, + "grad_norm": 0.17455781330042522, + "learning_rate": 3.78949948332492e-05, + "loss": 0.4075, + "step": 7493 + }, + { + "epoch": 2.8257918552036196, + "grad_norm": 0.15528535112306094, + "learning_rate": 3.7884479607271465e-05, + "loss": 0.4135, + "step": 7494 + }, + { + "epoch": 2.8261689291101053, + "grad_norm": 0.16866625814423788, + "learning_rate": 3.787396452789743e-05, + "loss": 0.4177, + "step": 7495 + }, + { + "epoch": 2.826546003016591, + "grad_norm": 0.1916452465335679, + "learning_rate": 3.786344959585577e-05, + "loss": 0.4172, + "step": 7496 + }, + { + "epoch": 2.8269230769230766, + "grad_norm": 0.16828193434127783, + "learning_rate": 3.7852934811875135e-05, + "loss": 0.4168, + "step": 7497 + }, + { + "epoch": 2.8273001508295623, + "grad_norm": 0.21901970707644489, + "learning_rate": 3.784242017668424e-05, + "loss": 0.4143, + "step": 7498 + }, + { + "epoch": 2.827677224736048, + "grad_norm": 0.22892502668837575, + "learning_rate": 3.78319056910117e-05, + "loss": 0.4142, + "step": 7499 + }, + { + "epoch": 2.8280542986425337, + "grad_norm": 0.18040720728613596, + "learning_rate": 3.7821391355586165e-05, + "loss": 0.4166, + "step": 7500 + }, + { + "epoch": 2.8284313725490198, + "grad_norm": 0.1414186983539682, + "learning_rate": 3.7810877171136296e-05, + "loss": 0.4113, + "step": 7501 + }, + { + "epoch": 2.8288084464555054, + "grad_norm": 0.15482471230542497, + "learning_rate": 3.780036313839069e-05, + "loss": 0.4156, + "step": 7502 + }, + { + "epoch": 2.829185520361991, + "grad_norm": 0.1967669907024065, + "learning_rate": 3.778984925807796e-05, + "loss": 0.3996, + "step": 7503 + }, + { + "epoch": 2.829562594268477, + "grad_norm": 0.19567929007453275, + "learning_rate": 3.7779335530926724e-05, + "loss": 0.4218, + "step": 7504 + }, + { + "epoch": 2.8299396681749625, + "grad_norm": 0.1721299410635748, + "learning_rate": 3.776882195766557e-05, + "loss": 0.4166, + "step": 7505 + }, + { + "epoch": 2.830316742081448, + "grad_norm": 0.19002305925235843, + "learning_rate": 3.775830853902305e-05, + "loss": 0.4198, + "step": 7506 + }, + { + "epoch": 2.830693815987934, + "grad_norm": 0.1826207583513599, + "learning_rate": 3.774779527572779e-05, + "loss": 0.416, + "step": 7507 + }, + { + "epoch": 2.8310708898944195, + "grad_norm": 0.24283388496510802, + "learning_rate": 3.77372821685083e-05, + "loss": 0.4181, + "step": 7508 + }, + { + "epoch": 2.831447963800905, + "grad_norm": 0.24635318274599463, + "learning_rate": 3.772676921809316e-05, + "loss": 0.4195, + "step": 7509 + }, + { + "epoch": 2.831825037707391, + "grad_norm": 0.2165074996850674, + "learning_rate": 3.7716256425210895e-05, + "loss": 0.4138, + "step": 7510 + }, + { + "epoch": 2.8322021116138765, + "grad_norm": 0.1831827808104451, + "learning_rate": 3.770574379059003e-05, + "loss": 0.4134, + "step": 7511 + }, + { + "epoch": 2.832579185520362, + "grad_norm": 0.2439297272012443, + "learning_rate": 3.769523131495908e-05, + "loss": 0.4146, + "step": 7512 + }, + { + "epoch": 2.832956259426848, + "grad_norm": 0.27705699724696004, + "learning_rate": 3.7684718999046573e-05, + "loss": 0.4194, + "step": 7513 + }, + { + "epoch": 2.8333333333333335, + "grad_norm": 0.21739909195768284, + "learning_rate": 3.767420684358097e-05, + "loss": 0.4074, + "step": 7514 + }, + { + "epoch": 2.833710407239819, + "grad_norm": 0.17278573807851066, + "learning_rate": 3.766369484929077e-05, + "loss": 0.4149, + "step": 7515 + }, + { + "epoch": 2.834087481146305, + "grad_norm": 0.1842465096337816, + "learning_rate": 3.765318301690446e-05, + "loss": 0.4264, + "step": 7516 + }, + { + "epoch": 2.8344645550527905, + "grad_norm": 0.1643908558007432, + "learning_rate": 3.7642671347150475e-05, + "loss": 0.4047, + "step": 7517 + }, + { + "epoch": 2.834841628959276, + "grad_norm": 0.1678401502538409, + "learning_rate": 3.763215984075727e-05, + "loss": 0.4083, + "step": 7518 + }, + { + "epoch": 2.835218702865762, + "grad_norm": 0.22912943485159842, + "learning_rate": 3.76216484984533e-05, + "loss": 0.4168, + "step": 7519 + }, + { + "epoch": 2.8355957767722475, + "grad_norm": 0.18436446007832308, + "learning_rate": 3.761113732096697e-05, + "loss": 0.4105, + "step": 7520 + }, + { + "epoch": 2.835972850678733, + "grad_norm": 0.15756192300335806, + "learning_rate": 3.76006263090267e-05, + "loss": 0.414, + "step": 7521 + }, + { + "epoch": 2.836349924585219, + "grad_norm": 0.19551136750394774, + "learning_rate": 3.7590115463360916e-05, + "loss": 0.4305, + "step": 7522 + }, + { + "epoch": 2.8367269984917045, + "grad_norm": 0.16933567696633872, + "learning_rate": 3.757960478469798e-05, + "loss": 0.4161, + "step": 7523 + }, + { + "epoch": 2.83710407239819, + "grad_norm": 0.16678454240620316, + "learning_rate": 3.756909427376628e-05, + "loss": 0.4157, + "step": 7524 + }, + { + "epoch": 2.837481146304676, + "grad_norm": 0.17308205944841834, + "learning_rate": 3.755858393129421e-05, + "loss": 0.4108, + "step": 7525 + }, + { + "epoch": 2.8378582202111615, + "grad_norm": 0.18003883534890713, + "learning_rate": 3.7548073758010096e-05, + "loss": 0.4287, + "step": 7526 + }, + { + "epoch": 2.838235294117647, + "grad_norm": 0.20363664777466675, + "learning_rate": 3.753756375464229e-05, + "loss": 0.4217, + "step": 7527 + }, + { + "epoch": 2.838612368024133, + "grad_norm": 0.22231260462166735, + "learning_rate": 3.752705392191914e-05, + "loss": 0.4161, + "step": 7528 + }, + { + "epoch": 2.8389894419306185, + "grad_norm": 0.18853375713223783, + "learning_rate": 3.7516544260568964e-05, + "loss": 0.4185, + "step": 7529 + }, + { + "epoch": 2.839366515837104, + "grad_norm": 0.20906666100015003, + "learning_rate": 3.750603477132005e-05, + "loss": 0.4228, + "step": 7530 + }, + { + "epoch": 2.83974358974359, + "grad_norm": 0.27228599786228747, + "learning_rate": 3.749552545490073e-05, + "loss": 0.4048, + "step": 7531 + }, + { + "epoch": 2.8401206636500755, + "grad_norm": 0.2134493912708313, + "learning_rate": 3.7485016312039275e-05, + "loss": 0.4147, + "step": 7532 + }, + { + "epoch": 2.840497737556561, + "grad_norm": 0.21496126430522655, + "learning_rate": 3.7474507343463946e-05, + "loss": 0.4253, + "step": 7533 + }, + { + "epoch": 2.840874811463047, + "grad_norm": 0.24840946518165818, + "learning_rate": 3.7463998549903026e-05, + "loss": 0.4132, + "step": 7534 + }, + { + "epoch": 2.8412518853695325, + "grad_norm": 0.21340574773585305, + "learning_rate": 3.7453489932084765e-05, + "loss": 0.4189, + "step": 7535 + }, + { + "epoch": 2.841628959276018, + "grad_norm": 0.1901896306060256, + "learning_rate": 3.744298149073738e-05, + "loss": 0.4217, + "step": 7536 + }, + { + "epoch": 2.842006033182504, + "grad_norm": 0.23640950707921266, + "learning_rate": 3.743247322658911e-05, + "loss": 0.4135, + "step": 7537 + }, + { + "epoch": 2.8423831070889896, + "grad_norm": 0.21498228894119195, + "learning_rate": 3.742196514036818e-05, + "loss": 0.4116, + "step": 7538 + }, + { + "epoch": 2.8427601809954752, + "grad_norm": 0.27012526528709185, + "learning_rate": 3.741145723280277e-05, + "loss": 0.4128, + "step": 7539 + }, + { + "epoch": 2.843137254901961, + "grad_norm": 0.27417617270972383, + "learning_rate": 3.7400949504621075e-05, + "loss": 0.4103, + "step": 7540 + }, + { + "epoch": 2.8435143288084466, + "grad_norm": 0.20257761758085926, + "learning_rate": 3.7390441956551284e-05, + "loss": 0.417, + "step": 7541 + }, + { + "epoch": 2.8438914027149322, + "grad_norm": 0.18565637911916835, + "learning_rate": 3.7379934589321544e-05, + "loss": 0.4065, + "step": 7542 + }, + { + "epoch": 2.844268476621418, + "grad_norm": 0.20475614763299965, + "learning_rate": 3.736942740366001e-05, + "loss": 0.4191, + "step": 7543 + }, + { + "epoch": 2.8446455505279036, + "grad_norm": 0.2012647387403796, + "learning_rate": 3.7358920400294835e-05, + "loss": 0.4202, + "step": 7544 + }, + { + "epoch": 2.8450226244343892, + "grad_norm": 0.21695225828989512, + "learning_rate": 3.734841357995412e-05, + "loss": 0.4056, + "step": 7545 + }, + { + "epoch": 2.845399698340875, + "grad_norm": 0.21932123343356238, + "learning_rate": 3.7337906943366e-05, + "loss": 0.4098, + "step": 7546 + }, + { + "epoch": 2.8457767722473606, + "grad_norm": 0.17227471616913642, + "learning_rate": 3.7327400491258575e-05, + "loss": 0.4179, + "step": 7547 + }, + { + "epoch": 2.8461538461538463, + "grad_norm": 0.1502808373864921, + "learning_rate": 3.7316894224359904e-05, + "loss": 0.4193, + "step": 7548 + }, + { + "epoch": 2.846530920060332, + "grad_norm": 0.17767974842453088, + "learning_rate": 3.73063881433981e-05, + "loss": 0.3976, + "step": 7549 + }, + { + "epoch": 2.8469079939668176, + "grad_norm": 0.1865206850665127, + "learning_rate": 3.729588224910121e-05, + "loss": 0.4226, + "step": 7550 + }, + { + "epoch": 2.8472850678733033, + "grad_norm": 0.19832456502533707, + "learning_rate": 3.728537654219726e-05, + "loss": 0.4205, + "step": 7551 + }, + { + "epoch": 2.847662141779789, + "grad_norm": 0.18974472738952688, + "learning_rate": 3.7274871023414334e-05, + "loss": 0.4177, + "step": 7552 + }, + { + "epoch": 2.8480392156862746, + "grad_norm": 0.23319386332357134, + "learning_rate": 3.726436569348042e-05, + "loss": 0.4112, + "step": 7553 + }, + { + "epoch": 2.8484162895927603, + "grad_norm": 0.1789860853616542, + "learning_rate": 3.725386055312353e-05, + "loss": 0.4149, + "step": 7554 + }, + { + "epoch": 2.848793363499246, + "grad_norm": 0.2311601559931308, + "learning_rate": 3.724335560307168e-05, + "loss": 0.4159, + "step": 7555 + }, + { + "epoch": 2.8491704374057316, + "grad_norm": 0.24673752791716622, + "learning_rate": 3.723285084405284e-05, + "loss": 0.4255, + "step": 7556 + }, + { + "epoch": 2.8495475113122173, + "grad_norm": 0.21151598665734234, + "learning_rate": 3.722234627679497e-05, + "loss": 0.4224, + "step": 7557 + }, + { + "epoch": 2.849924585218703, + "grad_norm": 0.1961366306959224, + "learning_rate": 3.721184190202604e-05, + "loss": 0.4228, + "step": 7558 + }, + { + "epoch": 2.8503016591251886, + "grad_norm": 0.25064637942331786, + "learning_rate": 3.7201337720474014e-05, + "loss": 0.413, + "step": 7559 + }, + { + "epoch": 2.8506787330316743, + "grad_norm": 0.21753754562154817, + "learning_rate": 3.719083373286677e-05, + "loss": 0.4181, + "step": 7560 + }, + { + "epoch": 2.85105580693816, + "grad_norm": 0.19856068480700786, + "learning_rate": 3.718032993993228e-05, + "loss": 0.4104, + "step": 7561 + }, + { + "epoch": 2.8514328808446456, + "grad_norm": 0.20680016498768253, + "learning_rate": 3.716982634239843e-05, + "loss": 0.4176, + "step": 7562 + }, + { + "epoch": 2.8518099547511313, + "grad_norm": 0.1581923908066538, + "learning_rate": 3.7159322940993074e-05, + "loss": 0.4252, + "step": 7563 + }, + { + "epoch": 2.852187028657617, + "grad_norm": 0.2815141216788276, + "learning_rate": 3.714881973644414e-05, + "loss": 0.4151, + "step": 7564 + }, + { + "epoch": 2.8525641025641026, + "grad_norm": 0.37597939514391315, + "learning_rate": 3.713831672947947e-05, + "loss": 0.4224, + "step": 7565 + }, + { + "epoch": 2.8529411764705883, + "grad_norm": 0.3069063461213418, + "learning_rate": 3.71278139208269e-05, + "loss": 0.4183, + "step": 7566 + }, + { + "epoch": 2.853318250377074, + "grad_norm": 0.2974616513506547, + "learning_rate": 3.711731131121429e-05, + "loss": 0.4196, + "step": 7567 + }, + { + "epoch": 2.8536953242835597, + "grad_norm": 0.2700360778236087, + "learning_rate": 3.710680890136945e-05, + "loss": 0.4027, + "step": 7568 + }, + { + "epoch": 2.8540723981900453, + "grad_norm": 0.2259294120521848, + "learning_rate": 3.709630669202017e-05, + "loss": 0.431, + "step": 7569 + }, + { + "epoch": 2.854449472096531, + "grad_norm": 0.1834597617176033, + "learning_rate": 3.7085804683894274e-05, + "loss": 0.4081, + "step": 7570 + }, + { + "epoch": 2.8548265460030167, + "grad_norm": 0.2177854829247232, + "learning_rate": 3.7075302877719534e-05, + "loss": 0.4012, + "step": 7571 + }, + { + "epoch": 2.8552036199095023, + "grad_norm": 0.29344627129836226, + "learning_rate": 3.706480127422369e-05, + "loss": 0.4186, + "step": 7572 + }, + { + "epoch": 2.855580693815988, + "grad_norm": 0.20209739789890893, + "learning_rate": 3.705429987413454e-05, + "loss": 0.4093, + "step": 7573 + }, + { + "epoch": 2.8559577677224737, + "grad_norm": 0.22109956499636782, + "learning_rate": 3.704379867817978e-05, + "loss": 0.4153, + "step": 7574 + }, + { + "epoch": 2.8563348416289593, + "grad_norm": 0.21984722033917542, + "learning_rate": 3.703329768708714e-05, + "loss": 0.4272, + "step": 7575 + }, + { + "epoch": 2.856711915535445, + "grad_norm": 0.21357089321454317, + "learning_rate": 3.7022796901584366e-05, + "loss": 0.409, + "step": 7576 + }, + { + "epoch": 2.8570889894419307, + "grad_norm": 0.16490174355725087, + "learning_rate": 3.701229632239912e-05, + "loss": 0.4215, + "step": 7577 + }, + { + "epoch": 2.8574660633484164, + "grad_norm": 0.16171447333321665, + "learning_rate": 3.700179595025906e-05, + "loss": 0.4183, + "step": 7578 + }, + { + "epoch": 2.857843137254902, + "grad_norm": 0.19340541250356053, + "learning_rate": 3.699129578589191e-05, + "loss": 0.412, + "step": 7579 + }, + { + "epoch": 2.8582202111613877, + "grad_norm": 0.19573104939627986, + "learning_rate": 3.6980795830025294e-05, + "loss": 0.422, + "step": 7580 + }, + { + "epoch": 2.8585972850678734, + "grad_norm": 0.17011319580665382, + "learning_rate": 3.6970296083386826e-05, + "loss": 0.4169, + "step": 7581 + }, + { + "epoch": 2.858974358974359, + "grad_norm": 0.1661147057368156, + "learning_rate": 3.695979654670417e-05, + "loss": 0.4152, + "step": 7582 + }, + { + "epoch": 2.8593514328808447, + "grad_norm": 0.1852878299046089, + "learning_rate": 3.694929722070492e-05, + "loss": 0.4196, + "step": 7583 + }, + { + "epoch": 2.8597285067873304, + "grad_norm": 0.16873496142657607, + "learning_rate": 3.693879810611664e-05, + "loss": 0.417, + "step": 7584 + }, + { + "epoch": 2.860105580693816, + "grad_norm": 0.1839844469136517, + "learning_rate": 3.6928299203666965e-05, + "loss": 0.4198, + "step": 7585 + }, + { + "epoch": 2.8604826546003017, + "grad_norm": 0.21265286455494714, + "learning_rate": 3.6917800514083426e-05, + "loss": 0.4156, + "step": 7586 + }, + { + "epoch": 2.8608597285067874, + "grad_norm": 0.22427986445943449, + "learning_rate": 3.690730203809356e-05, + "loss": 0.4103, + "step": 7587 + }, + { + "epoch": 2.861236802413273, + "grad_norm": 0.1973046074538492, + "learning_rate": 3.6896803776424935e-05, + "loss": 0.4175, + "step": 7588 + }, + { + "epoch": 2.8616138763197587, + "grad_norm": 0.2041686553775783, + "learning_rate": 3.6886305729805064e-05, + "loss": 0.4142, + "step": 7589 + }, + { + "epoch": 2.8619909502262444, + "grad_norm": 0.1793512510256523, + "learning_rate": 3.6875807898961416e-05, + "loss": 0.4134, + "step": 7590 + }, + { + "epoch": 2.86236802413273, + "grad_norm": 0.15802501528824406, + "learning_rate": 3.686531028462154e-05, + "loss": 0.4168, + "step": 7591 + }, + { + "epoch": 2.8627450980392157, + "grad_norm": 0.15813541277024407, + "learning_rate": 3.6854812887512866e-05, + "loss": 0.4205, + "step": 7592 + }, + { + "epoch": 2.8631221719457014, + "grad_norm": 0.2646484114054392, + "learning_rate": 3.6844315708362856e-05, + "loss": 0.4255, + "step": 7593 + }, + { + "epoch": 2.863499245852187, + "grad_norm": 0.3192978669123269, + "learning_rate": 3.683381874789899e-05, + "loss": 0.4135, + "step": 7594 + }, + { + "epoch": 2.8638763197586727, + "grad_norm": 0.2477814682951315, + "learning_rate": 3.682332200684867e-05, + "loss": 0.423, + "step": 7595 + }, + { + "epoch": 2.8642533936651584, + "grad_norm": 0.2007911723745997, + "learning_rate": 3.681282548593931e-05, + "loss": 0.4159, + "step": 7596 + }, + { + "epoch": 2.864630467571644, + "grad_norm": 0.32184476025957964, + "learning_rate": 3.6802329185898325e-05, + "loss": 0.4139, + "step": 7597 + }, + { + "epoch": 2.8650075414781297, + "grad_norm": 0.2681184316783548, + "learning_rate": 3.6791833107453095e-05, + "loss": 0.408, + "step": 7598 + }, + { + "epoch": 2.8653846153846154, + "grad_norm": 0.2261068236208369, + "learning_rate": 3.6781337251330967e-05, + "loss": 0.422, + "step": 7599 + }, + { + "epoch": 2.865761689291101, + "grad_norm": 0.26142818889836444, + "learning_rate": 3.677084161825933e-05, + "loss": 0.4157, + "step": 7600 + }, + { + "epoch": 2.8661387631975868, + "grad_norm": 0.21825367959585903, + "learning_rate": 3.67603462089655e-05, + "loss": 0.4101, + "step": 7601 + }, + { + "epoch": 2.8665158371040724, + "grad_norm": 0.18976231160928997, + "learning_rate": 3.674985102417679e-05, + "loss": 0.4026, + "step": 7602 + }, + { + "epoch": 2.866892911010558, + "grad_norm": 0.20810000963257227, + "learning_rate": 3.6739356064620546e-05, + "loss": 0.4295, + "step": 7603 + }, + { + "epoch": 2.8672699849170438, + "grad_norm": 0.235122391393062, + "learning_rate": 3.672886133102401e-05, + "loss": 0.4167, + "step": 7604 + }, + { + "epoch": 2.8676470588235294, + "grad_norm": 0.24683397775990093, + "learning_rate": 3.67183668241145e-05, + "loss": 0.4229, + "step": 7605 + }, + { + "epoch": 2.868024132730015, + "grad_norm": 0.29715020001009784, + "learning_rate": 3.670787254461926e-05, + "loss": 0.4103, + "step": 7606 + }, + { + "epoch": 2.8684012066365008, + "grad_norm": 0.2523171044835644, + "learning_rate": 3.669737849326551e-05, + "loss": 0.4133, + "step": 7607 + }, + { + "epoch": 2.8687782805429864, + "grad_norm": 0.19703169489671468, + "learning_rate": 3.668688467078053e-05, + "loss": 0.4061, + "step": 7608 + }, + { + "epoch": 2.869155354449472, + "grad_norm": 0.20003157166334926, + "learning_rate": 3.667639107789149e-05, + "loss": 0.4235, + "step": 7609 + }, + { + "epoch": 2.869532428355958, + "grad_norm": 0.19704031709169592, + "learning_rate": 3.6665897715325595e-05, + "loss": 0.4188, + "step": 7610 + }, + { + "epoch": 2.8699095022624435, + "grad_norm": 0.19445764351688716, + "learning_rate": 3.665540458381004e-05, + "loss": 0.4169, + "step": 7611 + }, + { + "epoch": 2.870286576168929, + "grad_norm": 0.23625169925886252, + "learning_rate": 3.664491168407198e-05, + "loss": 0.4143, + "step": 7612 + }, + { + "epoch": 2.870663650075415, + "grad_norm": 0.1937594946757765, + "learning_rate": 3.663441901683855e-05, + "loss": 0.4167, + "step": 7613 + }, + { + "epoch": 2.8710407239819005, + "grad_norm": 0.16135993907338145, + "learning_rate": 3.662392658283692e-05, + "loss": 0.4175, + "step": 7614 + }, + { + "epoch": 2.871417797888386, + "grad_norm": 0.17439809493731726, + "learning_rate": 3.661343438279417e-05, + "loss": 0.4134, + "step": 7615 + }, + { + "epoch": 2.871794871794872, + "grad_norm": 0.20965499919068067, + "learning_rate": 3.66029424174374e-05, + "loss": 0.4147, + "step": 7616 + }, + { + "epoch": 2.8721719457013575, + "grad_norm": 0.20264741382891996, + "learning_rate": 3.659245068749373e-05, + "loss": 0.4226, + "step": 7617 + }, + { + "epoch": 2.872549019607843, + "grad_norm": 0.19988362626427125, + "learning_rate": 3.65819591936902e-05, + "loss": 0.4096, + "step": 7618 + }, + { + "epoch": 2.872926093514329, + "grad_norm": 0.18136147769758273, + "learning_rate": 3.657146793675386e-05, + "loss": 0.4168, + "step": 7619 + }, + { + "epoch": 2.8733031674208145, + "grad_norm": 0.18592715503316493, + "learning_rate": 3.6560976917411755e-05, + "loss": 0.4162, + "step": 7620 + }, + { + "epoch": 2.8736802413273, + "grad_norm": 0.18017099860599775, + "learning_rate": 3.655048613639091e-05, + "loss": 0.4145, + "step": 7621 + }, + { + "epoch": 2.874057315233786, + "grad_norm": 0.18120724875511357, + "learning_rate": 3.653999559441829e-05, + "loss": 0.4033, + "step": 7622 + }, + { + "epoch": 2.8744343891402715, + "grad_norm": 0.16626856238830473, + "learning_rate": 3.652950529222092e-05, + "loss": 0.4197, + "step": 7623 + }, + { + "epoch": 2.874811463046757, + "grad_norm": 0.17294778609065012, + "learning_rate": 3.6519015230525764e-05, + "loss": 0.4156, + "step": 7624 + }, + { + "epoch": 2.875188536953243, + "grad_norm": 0.19440897267142435, + "learning_rate": 3.650852541005974e-05, + "loss": 0.4136, + "step": 7625 + }, + { + "epoch": 2.8755656108597285, + "grad_norm": 0.17531011484709993, + "learning_rate": 3.649803583154983e-05, + "loss": 0.411, + "step": 7626 + }, + { + "epoch": 2.875942684766214, + "grad_norm": 0.15249050500922057, + "learning_rate": 3.648754649572294e-05, + "loss": 0.4095, + "step": 7627 + }, + { + "epoch": 2.8763197586727, + "grad_norm": 0.16513301747618633, + "learning_rate": 3.6477057403305935e-05, + "loss": 0.4104, + "step": 7628 + }, + { + "epoch": 2.8766968325791855, + "grad_norm": 0.19551428961300238, + "learning_rate": 3.6466568555025745e-05, + "loss": 0.4338, + "step": 7629 + }, + { + "epoch": 2.877073906485671, + "grad_norm": 0.2012658921071381, + "learning_rate": 3.645607995160922e-05, + "loss": 0.4219, + "step": 7630 + }, + { + "epoch": 2.877450980392157, + "grad_norm": 0.16342617875539653, + "learning_rate": 3.644559159378319e-05, + "loss": 0.3956, + "step": 7631 + }, + { + "epoch": 2.8778280542986425, + "grad_norm": 0.1343628125189391, + "learning_rate": 3.6435103482274534e-05, + "loss": 0.41, + "step": 7632 + }, + { + "epoch": 2.878205128205128, + "grad_norm": 0.17924767857519616, + "learning_rate": 3.642461561781004e-05, + "loss": 0.4159, + "step": 7633 + }, + { + "epoch": 2.878582202111614, + "grad_norm": 0.18377626896122054, + "learning_rate": 3.64141280011165e-05, + "loss": 0.4158, + "step": 7634 + }, + { + "epoch": 2.8789592760180995, + "grad_norm": 0.17292215193837907, + "learning_rate": 3.6403640632920726e-05, + "loss": 0.4214, + "step": 7635 + }, + { + "epoch": 2.879336349924585, + "grad_norm": 0.16842525148038406, + "learning_rate": 3.639315351394946e-05, + "loss": 0.4192, + "step": 7636 + }, + { + "epoch": 2.879713423831071, + "grad_norm": 0.198686926186633, + "learning_rate": 3.6382666644929436e-05, + "loss": 0.4147, + "step": 7637 + }, + { + "epoch": 2.8800904977375565, + "grad_norm": 0.17824420944808475, + "learning_rate": 3.637218002658743e-05, + "loss": 0.4179, + "step": 7638 + }, + { + "epoch": 2.880467571644042, + "grad_norm": 0.1448905390389342, + "learning_rate": 3.636169365965013e-05, + "loss": 0.4142, + "step": 7639 + }, + { + "epoch": 2.880844645550528, + "grad_norm": 0.1758225032072476, + "learning_rate": 3.635120754484422e-05, + "loss": 0.4327, + "step": 7640 + }, + { + "epoch": 2.8812217194570136, + "grad_norm": 0.18684854291484088, + "learning_rate": 3.6340721682896393e-05, + "loss": 0.4067, + "step": 7641 + }, + { + "epoch": 2.8815987933634992, + "grad_norm": 0.19275067543605268, + "learning_rate": 3.633023607453332e-05, + "loss": 0.4147, + "step": 7642 + }, + { + "epoch": 2.881975867269985, + "grad_norm": 0.14409298080817484, + "learning_rate": 3.631975072048161e-05, + "loss": 0.4192, + "step": 7643 + }, + { + "epoch": 2.8823529411764706, + "grad_norm": 0.18503579472129567, + "learning_rate": 3.630926562146792e-05, + "loss": 0.4028, + "step": 7644 + }, + { + "epoch": 2.8827300150829562, + "grad_norm": 0.19308921479561958, + "learning_rate": 3.629878077821886e-05, + "loss": 0.4363, + "step": 7645 + }, + { + "epoch": 2.883107088989442, + "grad_norm": 0.15178911267670644, + "learning_rate": 3.6288296191460995e-05, + "loss": 0.4119, + "step": 7646 + }, + { + "epoch": 2.8834841628959276, + "grad_norm": 0.19345447138000849, + "learning_rate": 3.627781186192092e-05, + "loss": 0.4229, + "step": 7647 + }, + { + "epoch": 2.8838612368024132, + "grad_norm": 0.22033836506137083, + "learning_rate": 3.626732779032519e-05, + "loss": 0.4098, + "step": 7648 + }, + { + "epoch": 2.884238310708899, + "grad_norm": 0.22447063165244946, + "learning_rate": 3.625684397740031e-05, + "loss": 0.4085, + "step": 7649 + }, + { + "epoch": 2.8846153846153846, + "grad_norm": 0.1992672792987634, + "learning_rate": 3.6246360423872834e-05, + "loss": 0.4222, + "step": 7650 + }, + { + "epoch": 2.8849924585218703, + "grad_norm": 0.1862345583307723, + "learning_rate": 3.623587713046926e-05, + "loss": 0.4131, + "step": 7651 + }, + { + "epoch": 2.885369532428356, + "grad_norm": 0.18222743920721793, + "learning_rate": 3.6225394097916044e-05, + "loss": 0.4146, + "step": 7652 + }, + { + "epoch": 2.8857466063348416, + "grad_norm": 0.14863562701511127, + "learning_rate": 3.621491132693968e-05, + "loss": 0.4094, + "step": 7653 + }, + { + "epoch": 2.8861236802413273, + "grad_norm": 0.16207816191584784, + "learning_rate": 3.620442881826661e-05, + "loss": 0.4088, + "step": 7654 + }, + { + "epoch": 2.886500754147813, + "grad_norm": 0.20576070285479026, + "learning_rate": 3.619394657262324e-05, + "loss": 0.4156, + "step": 7655 + }, + { + "epoch": 2.8868778280542986, + "grad_norm": 0.18125914700695145, + "learning_rate": 3.618346459073601e-05, + "loss": 0.4139, + "step": 7656 + }, + { + "epoch": 2.8872549019607843, + "grad_norm": 0.15753585926241975, + "learning_rate": 3.6172982873331286e-05, + "loss": 0.407, + "step": 7657 + }, + { + "epoch": 2.88763197586727, + "grad_norm": 0.1800553251636172, + "learning_rate": 3.616250142113546e-05, + "loss": 0.4149, + "step": 7658 + }, + { + "epoch": 2.8880090497737556, + "grad_norm": 0.18430072360326874, + "learning_rate": 3.615202023487489e-05, + "loss": 0.4237, + "step": 7659 + }, + { + "epoch": 2.8883861236802413, + "grad_norm": 0.15917054107974327, + "learning_rate": 3.6141539315275895e-05, + "loss": 0.4086, + "step": 7660 + }, + { + "epoch": 2.888763197586727, + "grad_norm": 0.17180916903449409, + "learning_rate": 3.61310586630648e-05, + "loss": 0.4248, + "step": 7661 + }, + { + "epoch": 2.8891402714932126, + "grad_norm": 0.19562582852986715, + "learning_rate": 3.612057827896792e-05, + "loss": 0.4038, + "step": 7662 + }, + { + "epoch": 2.8895173453996983, + "grad_norm": 0.14763923136113563, + "learning_rate": 3.611009816371151e-05, + "loss": 0.4281, + "step": 7663 + }, + { + "epoch": 2.889894419306184, + "grad_norm": 0.1649967092924573, + "learning_rate": 3.609961831802186e-05, + "loss": 0.4119, + "step": 7664 + }, + { + "epoch": 2.8902714932126696, + "grad_norm": 0.23198834355984785, + "learning_rate": 3.60891387426252e-05, + "loss": 0.4216, + "step": 7665 + }, + { + "epoch": 2.8906485671191553, + "grad_norm": 0.22326568119748583, + "learning_rate": 3.607865943824775e-05, + "loss": 0.4064, + "step": 7666 + }, + { + "epoch": 2.891025641025641, + "grad_norm": 0.1692601213196732, + "learning_rate": 3.606818040561572e-05, + "loss": 0.4194, + "step": 7667 + }, + { + "epoch": 2.8914027149321266, + "grad_norm": 0.16468707746586136, + "learning_rate": 3.605770164545531e-05, + "loss": 0.4163, + "step": 7668 + }, + { + "epoch": 2.8917797888386123, + "grad_norm": 0.17174988708394917, + "learning_rate": 3.604722315849267e-05, + "loss": 0.4211, + "step": 7669 + }, + { + "epoch": 2.892156862745098, + "grad_norm": 0.16619859757062766, + "learning_rate": 3.603674494545396e-05, + "loss": 0.4215, + "step": 7670 + }, + { + "epoch": 2.8925339366515836, + "grad_norm": 0.20973377948041774, + "learning_rate": 3.602626700706532e-05, + "loss": 0.4131, + "step": 7671 + }, + { + "epoch": 2.8929110105580693, + "grad_norm": 0.2586944507923609, + "learning_rate": 3.601578934405285e-05, + "loss": 0.4071, + "step": 7672 + }, + { + "epoch": 2.893288084464555, + "grad_norm": 0.2097596922518746, + "learning_rate": 3.600531195714263e-05, + "loss": 0.4159, + "step": 7673 + }, + { + "epoch": 2.8936651583710407, + "grad_norm": 0.20287212480474018, + "learning_rate": 3.599483484706076e-05, + "loss": 0.4154, + "step": 7674 + }, + { + "epoch": 2.8940422322775263, + "grad_norm": 0.24038345494914573, + "learning_rate": 3.598435801453328e-05, + "loss": 0.4198, + "step": 7675 + }, + { + "epoch": 2.894419306184012, + "grad_norm": 0.2047737437744957, + "learning_rate": 3.597388146028623e-05, + "loss": 0.4058, + "step": 7676 + }, + { + "epoch": 2.8947963800904977, + "grad_norm": 0.23958218262454484, + "learning_rate": 3.596340518504561e-05, + "loss": 0.4166, + "step": 7677 + }, + { + "epoch": 2.8951734539969833, + "grad_norm": 0.1974885192965493, + "learning_rate": 3.595292918953745e-05, + "loss": 0.4147, + "step": 7678 + }, + { + "epoch": 2.895550527903469, + "grad_norm": 0.14702438295474607, + "learning_rate": 3.594245347448769e-05, + "loss": 0.423, + "step": 7679 + }, + { + "epoch": 2.8959276018099547, + "grad_norm": 0.1903032437171781, + "learning_rate": 3.593197804062232e-05, + "loss": 0.4172, + "step": 7680 + }, + { + "epoch": 2.8963046757164403, + "grad_norm": 0.1722924234023578, + "learning_rate": 3.592150288866726e-05, + "loss": 0.4076, + "step": 7681 + }, + { + "epoch": 2.896681749622926, + "grad_norm": 0.1391252287050137, + "learning_rate": 3.591102801934843e-05, + "loss": 0.4168, + "step": 7682 + }, + { + "epoch": 2.8970588235294117, + "grad_norm": 0.1629845254676682, + "learning_rate": 3.5900553433391724e-05, + "loss": 0.4168, + "step": 7683 + }, + { + "epoch": 2.8974358974358974, + "grad_norm": 0.1717650212011368, + "learning_rate": 3.589007913152304e-05, + "loss": 0.4141, + "step": 7684 + }, + { + "epoch": 2.897812971342383, + "grad_norm": 0.1369980758549551, + "learning_rate": 3.5879605114468224e-05, + "loss": 0.4128, + "step": 7685 + }, + { + "epoch": 2.8981900452488687, + "grad_norm": 0.17673901861558886, + "learning_rate": 3.586913138295311e-05, + "loss": 0.4056, + "step": 7686 + }, + { + "epoch": 2.8985671191553544, + "grad_norm": 0.20855590550425035, + "learning_rate": 3.585865793770355e-05, + "loss": 0.4172, + "step": 7687 + }, + { + "epoch": 2.89894419306184, + "grad_norm": 0.19381710709072617, + "learning_rate": 3.58481847794453e-05, + "loss": 0.4099, + "step": 7688 + }, + { + "epoch": 2.8993212669683257, + "grad_norm": 0.19542960497461992, + "learning_rate": 3.5837711908904165e-05, + "loss": 0.4127, + "step": 7689 + }, + { + "epoch": 2.8996983408748114, + "grad_norm": 0.19063670573447283, + "learning_rate": 3.582723932680591e-05, + "loss": 0.4087, + "step": 7690 + }, + { + "epoch": 2.900075414781297, + "grad_norm": 0.19678526863505222, + "learning_rate": 3.581676703387626e-05, + "loss": 0.4048, + "step": 7691 + }, + { + "epoch": 2.9004524886877827, + "grad_norm": 0.17657663579338356, + "learning_rate": 3.580629503084095e-05, + "loss": 0.4064, + "step": 7692 + }, + { + "epoch": 2.9008295625942684, + "grad_norm": 0.18314062762978586, + "learning_rate": 3.5795823318425675e-05, + "loss": 0.4045, + "step": 7693 + }, + { + "epoch": 2.901206636500754, + "grad_norm": 0.1965624752048595, + "learning_rate": 3.57853518973561e-05, + "loss": 0.4182, + "step": 7694 + }, + { + "epoch": 2.9015837104072397, + "grad_norm": 0.20477153698927175, + "learning_rate": 3.577488076835792e-05, + "loss": 0.4274, + "step": 7695 + }, + { + "epoch": 2.9019607843137254, + "grad_norm": 0.17707062300183068, + "learning_rate": 3.5764409932156754e-05, + "loss": 0.4081, + "step": 7696 + }, + { + "epoch": 2.902337858220211, + "grad_norm": 0.17638980746099833, + "learning_rate": 3.575393938947821e-05, + "loss": 0.4142, + "step": 7697 + }, + { + "epoch": 2.9027149321266967, + "grad_norm": 0.1775458257159119, + "learning_rate": 3.574346914104791e-05, + "loss": 0.4004, + "step": 7698 + }, + { + "epoch": 2.9030920060331824, + "grad_norm": 0.1585575412498117, + "learning_rate": 3.573299918759143e-05, + "loss": 0.4022, + "step": 7699 + }, + { + "epoch": 2.903469079939668, + "grad_norm": 0.17419133473622753, + "learning_rate": 3.5722529529834296e-05, + "loss": 0.4135, + "step": 7700 + }, + { + "epoch": 2.9038461538461537, + "grad_norm": 0.1803390760322743, + "learning_rate": 3.571206016850209e-05, + "loss": 0.4164, + "step": 7701 + }, + { + "epoch": 2.9042232277526394, + "grad_norm": 0.16193532419781906, + "learning_rate": 3.570159110432032e-05, + "loss": 0.4127, + "step": 7702 + }, + { + "epoch": 2.904600301659125, + "grad_norm": 0.1561035388134849, + "learning_rate": 3.569112233801444e-05, + "loss": 0.4097, + "step": 7703 + }, + { + "epoch": 2.9049773755656108, + "grad_norm": 0.1929903145333692, + "learning_rate": 3.568065387030999e-05, + "loss": 0.4211, + "step": 7704 + }, + { + "epoch": 2.9053544494720964, + "grad_norm": 0.18114086523188055, + "learning_rate": 3.56701857019324e-05, + "loss": 0.4102, + "step": 7705 + }, + { + "epoch": 2.905731523378582, + "grad_norm": 0.15885351789239902, + "learning_rate": 3.565971783360707e-05, + "loss": 0.4241, + "step": 7706 + }, + { + "epoch": 2.9061085972850678, + "grad_norm": 0.17302389534197116, + "learning_rate": 3.564925026605947e-05, + "loss": 0.4186, + "step": 7707 + }, + { + "epoch": 2.9064856711915534, + "grad_norm": 0.16628278661428744, + "learning_rate": 3.5638783000014964e-05, + "loss": 0.4067, + "step": 7708 + }, + { + "epoch": 2.906862745098039, + "grad_norm": 0.15725855260188668, + "learning_rate": 3.562831603619891e-05, + "loss": 0.4188, + "step": 7709 + }, + { + "epoch": 2.9072398190045248, + "grad_norm": 0.17107488411692376, + "learning_rate": 3.5617849375336695e-05, + "loss": 0.4134, + "step": 7710 + }, + { + "epoch": 2.9076168929110104, + "grad_norm": 0.16845502936276047, + "learning_rate": 3.560738301815363e-05, + "loss": 0.4142, + "step": 7711 + }, + { + "epoch": 2.907993966817496, + "grad_norm": 0.1469416516160699, + "learning_rate": 3.559691696537501e-05, + "loss": 0.4036, + "step": 7712 + }, + { + "epoch": 2.908371040723982, + "grad_norm": 0.18196645450545965, + "learning_rate": 3.558645121772616e-05, + "loss": 0.413, + "step": 7713 + }, + { + "epoch": 2.9087481146304675, + "grad_norm": 0.14183369836259288, + "learning_rate": 3.557598577593232e-05, + "loss": 0.4046, + "step": 7714 + }, + { + "epoch": 2.909125188536953, + "grad_norm": 0.17693401485658097, + "learning_rate": 3.556552064071873e-05, + "loss": 0.4256, + "step": 7715 + }, + { + "epoch": 2.909502262443439, + "grad_norm": 0.17376959316243434, + "learning_rate": 3.555505581281064e-05, + "loss": 0.424, + "step": 7716 + }, + { + "epoch": 2.9098793363499245, + "grad_norm": 0.1552076099752489, + "learning_rate": 3.554459129293324e-05, + "loss": 0.4191, + "step": 7717 + }, + { + "epoch": 2.91025641025641, + "grad_norm": 0.16118702059919082, + "learning_rate": 3.55341270818117e-05, + "loss": 0.4136, + "step": 7718 + }, + { + "epoch": 2.910633484162896, + "grad_norm": 0.19150376807371844, + "learning_rate": 3.552366318017119e-05, + "loss": 0.4192, + "step": 7719 + }, + { + "epoch": 2.9110105580693815, + "grad_norm": 0.2091074712879234, + "learning_rate": 3.551319958873687e-05, + "loss": 0.4244, + "step": 7720 + }, + { + "epoch": 2.911387631975867, + "grad_norm": 0.2475492955058547, + "learning_rate": 3.550273630823381e-05, + "loss": 0.4193, + "step": 7721 + }, + { + "epoch": 2.911764705882353, + "grad_norm": 0.24730135764360153, + "learning_rate": 3.549227333938716e-05, + "loss": 0.4212, + "step": 7722 + }, + { + "epoch": 2.9121417797888385, + "grad_norm": 0.17931955728839502, + "learning_rate": 3.5481810682921964e-05, + "loss": 0.4115, + "step": 7723 + }, + { + "epoch": 2.912518853695324, + "grad_norm": 0.19896733421100515, + "learning_rate": 3.547134833956327e-05, + "loss": 0.4207, + "step": 7724 + }, + { + "epoch": 2.91289592760181, + "grad_norm": 0.30303349541566915, + "learning_rate": 3.5460886310036135e-05, + "loss": 0.4216, + "step": 7725 + }, + { + "epoch": 2.9132730015082955, + "grad_norm": 0.28458651267406265, + "learning_rate": 3.545042459506554e-05, + "loss": 0.3978, + "step": 7726 + }, + { + "epoch": 2.913650075414781, + "grad_norm": 0.25889293865561763, + "learning_rate": 3.543996319537648e-05, + "loss": 0.4188, + "step": 7727 + }, + { + "epoch": 2.914027149321267, + "grad_norm": 0.2766140150020959, + "learning_rate": 3.542950211169395e-05, + "loss": 0.4065, + "step": 7728 + }, + { + "epoch": 2.9144042232277525, + "grad_norm": 0.21320605627388778, + "learning_rate": 3.5419041344742836e-05, + "loss": 0.4109, + "step": 7729 + }, + { + "epoch": 2.914781297134238, + "grad_norm": 0.2183660644806667, + "learning_rate": 3.540858089524812e-05, + "loss": 0.4123, + "step": 7730 + }, + { + "epoch": 2.915158371040724, + "grad_norm": 0.1863202827976429, + "learning_rate": 3.5398120763934675e-05, + "loss": 0.404, + "step": 7731 + }, + { + "epoch": 2.9155354449472095, + "grad_norm": 0.15105158295164248, + "learning_rate": 3.5387660951527354e-05, + "loss": 0.4156, + "step": 7732 + }, + { + "epoch": 2.915912518853695, + "grad_norm": 0.1638804897569126, + "learning_rate": 3.5377201458751066e-05, + "loss": 0.4263, + "step": 7733 + }, + { + "epoch": 2.916289592760181, + "grad_norm": 0.21107640616074774, + "learning_rate": 3.536674228633061e-05, + "loss": 0.4098, + "step": 7734 + }, + { + "epoch": 2.9166666666666665, + "grad_norm": 0.20345022737296653, + "learning_rate": 3.535628343499079e-05, + "loss": 0.4053, + "step": 7735 + }, + { + "epoch": 2.917043740573152, + "grad_norm": 0.20249984652407235, + "learning_rate": 3.534582490545642e-05, + "loss": 0.4167, + "step": 7736 + }, + { + "epoch": 2.917420814479638, + "grad_norm": 0.19555914196559435, + "learning_rate": 3.533536669845226e-05, + "loss": 0.4247, + "step": 7737 + }, + { + "epoch": 2.9177978883861235, + "grad_norm": 0.18204205374614071, + "learning_rate": 3.5324908814703034e-05, + "loss": 0.4149, + "step": 7738 + }, + { + "epoch": 2.918174962292609, + "grad_norm": 0.16510370837550617, + "learning_rate": 3.53144512549335e-05, + "loss": 0.4202, + "step": 7739 + }, + { + "epoch": 2.918552036199095, + "grad_norm": 0.15745039851614703, + "learning_rate": 3.530399401986833e-05, + "loss": 0.4097, + "step": 7740 + }, + { + "epoch": 2.9189291101055805, + "grad_norm": 0.15034174397616348, + "learning_rate": 3.5293537110232195e-05, + "loss": 0.4075, + "step": 7741 + }, + { + "epoch": 2.919306184012066, + "grad_norm": 0.14769976500647958, + "learning_rate": 3.5283080526749786e-05, + "loss": 0.4185, + "step": 7742 + }, + { + "epoch": 2.919683257918552, + "grad_norm": 0.14564836664029884, + "learning_rate": 3.527262427014571e-05, + "loss": 0.4189, + "step": 7743 + }, + { + "epoch": 2.9200603318250375, + "grad_norm": 0.16902086264692684, + "learning_rate": 3.526216834114456e-05, + "loss": 0.4111, + "step": 7744 + }, + { + "epoch": 2.920437405731523, + "grad_norm": 0.1903341119319637, + "learning_rate": 3.525171274047096e-05, + "loss": 0.4138, + "step": 7745 + }, + { + "epoch": 2.920814479638009, + "grad_norm": 0.1635992783947501, + "learning_rate": 3.524125746884945e-05, + "loss": 0.4173, + "step": 7746 + }, + { + "epoch": 2.9211915535444946, + "grad_norm": 0.2089185567190372, + "learning_rate": 3.523080252700457e-05, + "loss": 0.4342, + "step": 7747 + }, + { + "epoch": 2.9215686274509802, + "grad_norm": 0.20111973544874664, + "learning_rate": 3.522034791566086e-05, + "loss": 0.4181, + "step": 7748 + }, + { + "epoch": 2.9219457013574663, + "grad_norm": 0.17532546406509986, + "learning_rate": 3.520989363554279e-05, + "loss": 0.4172, + "step": 7749 + }, + { + "epoch": 2.922322775263952, + "grad_norm": 0.13211070191325272, + "learning_rate": 3.519943968737483e-05, + "loss": 0.4144, + "step": 7750 + }, + { + "epoch": 2.9226998491704377, + "grad_norm": 0.16187067117833107, + "learning_rate": 3.518898607188146e-05, + "loss": 0.4165, + "step": 7751 + }, + { + "epoch": 2.9230769230769234, + "grad_norm": 0.18398578470676127, + "learning_rate": 3.517853278978708e-05, + "loss": 0.4202, + "step": 7752 + }, + { + "epoch": 2.923453996983409, + "grad_norm": 0.17014803114361113, + "learning_rate": 3.51680798418161e-05, + "loss": 0.4042, + "step": 7753 + }, + { + "epoch": 2.9238310708898947, + "grad_norm": 0.14722485389307444, + "learning_rate": 3.51576272286929e-05, + "loss": 0.4116, + "step": 7754 + }, + { + "epoch": 2.9242081447963804, + "grad_norm": 0.16477295436289052, + "learning_rate": 3.514717495114186e-05, + "loss": 0.4135, + "step": 7755 + }, + { + "epoch": 2.924585218702866, + "grad_norm": 0.14363784277929964, + "learning_rate": 3.5136723009887254e-05, + "loss": 0.4178, + "step": 7756 + }, + { + "epoch": 2.9249622926093517, + "grad_norm": 0.15638436196512015, + "learning_rate": 3.5126271405653454e-05, + "loss": 0.4122, + "step": 7757 + }, + { + "epoch": 2.9253393665158374, + "grad_norm": 0.15449017916323, + "learning_rate": 3.511582013916472e-05, + "loss": 0.4072, + "step": 7758 + }, + { + "epoch": 2.925716440422323, + "grad_norm": 0.1568421334738434, + "learning_rate": 3.510536921114531e-05, + "loss": 0.4096, + "step": 7759 + }, + { + "epoch": 2.9260935143288087, + "grad_norm": 0.20285383140378205, + "learning_rate": 3.509491862231949e-05, + "loss": 0.4093, + "step": 7760 + }, + { + "epoch": 2.9264705882352944, + "grad_norm": 0.18162822005978774, + "learning_rate": 3.5084468373411444e-05, + "loss": 0.4183, + "step": 7761 + }, + { + "epoch": 2.92684766214178, + "grad_norm": 0.17684786483668155, + "learning_rate": 3.507401846514537e-05, + "loss": 0.4261, + "step": 7762 + }, + { + "epoch": 2.9272247360482657, + "grad_norm": 0.18884855124449737, + "learning_rate": 3.506356889824547e-05, + "loss": 0.4051, + "step": 7763 + }, + { + "epoch": 2.9276018099547514, + "grad_norm": 0.16933038127994168, + "learning_rate": 3.505311967343586e-05, + "loss": 0.4112, + "step": 7764 + }, + { + "epoch": 2.927978883861237, + "grad_norm": 0.19755617154807545, + "learning_rate": 3.504267079144065e-05, + "loss": 0.4064, + "step": 7765 + }, + { + "epoch": 2.9283559577677227, + "grad_norm": 0.1664246243897835, + "learning_rate": 3.5032222252983974e-05, + "loss": 0.4057, + "step": 7766 + }, + { + "epoch": 2.9287330316742084, + "grad_norm": 0.20140742368495196, + "learning_rate": 3.50217740587899e-05, + "loss": 0.4251, + "step": 7767 + }, + { + "epoch": 2.929110105580694, + "grad_norm": 0.2256561847945994, + "learning_rate": 3.501132620958244e-05, + "loss": 0.4158, + "step": 7768 + }, + { + "epoch": 2.9294871794871797, + "grad_norm": 0.19334410039846286, + "learning_rate": 3.500087870608567e-05, + "loss": 0.4158, + "step": 7769 + }, + { + "epoch": 2.9298642533936654, + "grad_norm": 0.17465634184279474, + "learning_rate": 3.499043154902357e-05, + "loss": 0.4122, + "step": 7770 + }, + { + "epoch": 2.930241327300151, + "grad_norm": 0.22527369301367894, + "learning_rate": 3.4979984739120106e-05, + "loss": 0.411, + "step": 7771 + }, + { + "epoch": 2.9306184012066367, + "grad_norm": 0.2503024947719022, + "learning_rate": 3.496953827709926e-05, + "loss": 0.4285, + "step": 7772 + }, + { + "epoch": 2.9309954751131224, + "grad_norm": 0.1826759900616651, + "learning_rate": 3.495909216368495e-05, + "loss": 0.4075, + "step": 7773 + }, + { + "epoch": 2.931372549019608, + "grad_norm": 0.18260159752662516, + "learning_rate": 3.4948646399601066e-05, + "loss": 0.4165, + "step": 7774 + }, + { + "epoch": 2.9317496229260938, + "grad_norm": 0.1882977106643369, + "learning_rate": 3.4938200985571525e-05, + "loss": 0.4173, + "step": 7775 + }, + { + "epoch": 2.9321266968325794, + "grad_norm": 0.15810956431535214, + "learning_rate": 3.492775592232016e-05, + "loss": 0.4102, + "step": 7776 + }, + { + "epoch": 2.932503770739065, + "grad_norm": 0.1853461910262828, + "learning_rate": 3.49173112105708e-05, + "loss": 0.4231, + "step": 7777 + }, + { + "epoch": 2.9328808446455508, + "grad_norm": 0.2219253081379432, + "learning_rate": 3.4906866851047284e-05, + "loss": 0.4058, + "step": 7778 + }, + { + "epoch": 2.9332579185520364, + "grad_norm": 0.19553977043969867, + "learning_rate": 3.489642284447337e-05, + "loss": 0.4256, + "step": 7779 + }, + { + "epoch": 2.933634992458522, + "grad_norm": 0.17118194599008132, + "learning_rate": 3.4885979191572805e-05, + "loss": 0.4146, + "step": 7780 + }, + { + "epoch": 2.9340120663650078, + "grad_norm": 0.20213242255065283, + "learning_rate": 3.4875535893069375e-05, + "loss": 0.4195, + "step": 7781 + }, + { + "epoch": 2.9343891402714934, + "grad_norm": 0.2150296076164423, + "learning_rate": 3.486509294968675e-05, + "loss": 0.4148, + "step": 7782 + }, + { + "epoch": 2.934766214177979, + "grad_norm": 0.17408222822317232, + "learning_rate": 3.485465036214862e-05, + "loss": 0.4239, + "step": 7783 + }, + { + "epoch": 2.935143288084465, + "grad_norm": 0.1681627452469725, + "learning_rate": 3.4844208131178674e-05, + "loss": 0.4009, + "step": 7784 + }, + { + "epoch": 2.9355203619909505, + "grad_norm": 0.17587954904421793, + "learning_rate": 3.483376625750053e-05, + "loss": 0.4131, + "step": 7785 + }, + { + "epoch": 2.935897435897436, + "grad_norm": 0.1874823278473511, + "learning_rate": 3.4823324741837786e-05, + "loss": 0.4136, + "step": 7786 + }, + { + "epoch": 2.936274509803922, + "grad_norm": 0.1670573665438567, + "learning_rate": 3.481288358491406e-05, + "loss": 0.412, + "step": 7787 + }, + { + "epoch": 2.9366515837104075, + "grad_norm": 0.1579025544653047, + "learning_rate": 3.48024427874529e-05, + "loss": 0.4113, + "step": 7788 + }, + { + "epoch": 2.937028657616893, + "grad_norm": 0.19929563093053917, + "learning_rate": 3.479200235017783e-05, + "loss": 0.4209, + "step": 7789 + }, + { + "epoch": 2.937405731523379, + "grad_norm": 0.19618335629599573, + "learning_rate": 3.4781562273812394e-05, + "loss": 0.4033, + "step": 7790 + }, + { + "epoch": 2.9377828054298645, + "grad_norm": 0.19471099945220224, + "learning_rate": 3.477112255908006e-05, + "loss": 0.4181, + "step": 7791 + }, + { + "epoch": 2.93815987933635, + "grad_norm": 0.1845629850262611, + "learning_rate": 3.476068320670428e-05, + "loss": 0.4265, + "step": 7792 + }, + { + "epoch": 2.938536953242836, + "grad_norm": 0.18163794512333115, + "learning_rate": 3.475024421740853e-05, + "loss": 0.417, + "step": 7793 + }, + { + "epoch": 2.9389140271493215, + "grad_norm": 0.1819079638584539, + "learning_rate": 3.4739805591916194e-05, + "loss": 0.4187, + "step": 7794 + }, + { + "epoch": 2.939291101055807, + "grad_norm": 0.1737394618838031, + "learning_rate": 3.472936733095064e-05, + "loss": 0.4074, + "step": 7795 + }, + { + "epoch": 2.939668174962293, + "grad_norm": 0.1988369668075885, + "learning_rate": 3.4718929435235286e-05, + "loss": 0.416, + "step": 7796 + }, + { + "epoch": 2.9400452488687785, + "grad_norm": 0.18978308036665656, + "learning_rate": 3.470849190549342e-05, + "loss": 0.4078, + "step": 7797 + }, + { + "epoch": 2.940422322775264, + "grad_norm": 0.17154855849832962, + "learning_rate": 3.469805474244837e-05, + "loss": 0.4164, + "step": 7798 + }, + { + "epoch": 2.94079939668175, + "grad_norm": 0.18939528390865754, + "learning_rate": 3.468761794682344e-05, + "loss": 0.3994, + "step": 7799 + }, + { + "epoch": 2.9411764705882355, + "grad_norm": 0.1988840634766652, + "learning_rate": 3.467718151934187e-05, + "loss": 0.4145, + "step": 7800 + }, + { + "epoch": 2.941553544494721, + "grad_norm": 0.16086597719564785, + "learning_rate": 3.466674546072688e-05, + "loss": 0.4171, + "step": 7801 + }, + { + "epoch": 2.941930618401207, + "grad_norm": 0.1873641952569114, + "learning_rate": 3.465630977170172e-05, + "loss": 0.4156, + "step": 7802 + }, + { + "epoch": 2.9423076923076925, + "grad_norm": 0.1647967423412817, + "learning_rate": 3.464587445298955e-05, + "loss": 0.4166, + "step": 7803 + }, + { + "epoch": 2.942684766214178, + "grad_norm": 0.15747276158737816, + "learning_rate": 3.4635439505313514e-05, + "loss": 0.4162, + "step": 7804 + }, + { + "epoch": 2.943061840120664, + "grad_norm": 0.1771815727590301, + "learning_rate": 3.462500492939678e-05, + "loss": 0.4242, + "step": 7805 + }, + { + "epoch": 2.9434389140271495, + "grad_norm": 0.1700010739156682, + "learning_rate": 3.461457072596243e-05, + "loss": 0.4138, + "step": 7806 + }, + { + "epoch": 2.943815987933635, + "grad_norm": 0.16953160881345783, + "learning_rate": 3.460413689573355e-05, + "loss": 0.4143, + "step": 7807 + }, + { + "epoch": 2.944193061840121, + "grad_norm": 0.17235093050202305, + "learning_rate": 3.459370343943321e-05, + "loss": 0.4038, + "step": 7808 + }, + { + "epoch": 2.9445701357466065, + "grad_norm": 0.1828964078577077, + "learning_rate": 3.4583270357784416e-05, + "loss": 0.4304, + "step": 7809 + }, + { + "epoch": 2.944947209653092, + "grad_norm": 0.15310466784424298, + "learning_rate": 3.457283765151018e-05, + "loss": 0.3992, + "step": 7810 + }, + { + "epoch": 2.945324283559578, + "grad_norm": 0.16909496311887248, + "learning_rate": 3.456240532133349e-05, + "loss": 0.4086, + "step": 7811 + }, + { + "epoch": 2.9457013574660635, + "grad_norm": 0.1920628301252278, + "learning_rate": 3.455197336797727e-05, + "loss": 0.4031, + "step": 7812 + }, + { + "epoch": 2.946078431372549, + "grad_norm": 0.1578214374428637, + "learning_rate": 3.454154179216448e-05, + "loss": 0.4044, + "step": 7813 + }, + { + "epoch": 2.946455505279035, + "grad_norm": 0.18839636171432927, + "learning_rate": 3.4531110594618e-05, + "loss": 0.3937, + "step": 7814 + }, + { + "epoch": 2.9468325791855206, + "grad_norm": 0.1833517605268337, + "learning_rate": 3.45206797760607e-05, + "loss": 0.4245, + "step": 7815 + }, + { + "epoch": 2.947209653092006, + "grad_norm": 0.17076618113111325, + "learning_rate": 3.4510249337215425e-05, + "loss": 0.4024, + "step": 7816 + }, + { + "epoch": 2.947586726998492, + "grad_norm": 0.1703252538394807, + "learning_rate": 3.449981927880502e-05, + "loss": 0.4, + "step": 7817 + }, + { + "epoch": 2.9479638009049776, + "grad_norm": 0.14982559636861717, + "learning_rate": 3.448938960155224e-05, + "loss": 0.4279, + "step": 7818 + }, + { + "epoch": 2.9483408748114632, + "grad_norm": 0.16613921780024968, + "learning_rate": 3.447896030617988e-05, + "loss": 0.4104, + "step": 7819 + }, + { + "epoch": 2.948717948717949, + "grad_norm": 0.15925839349468573, + "learning_rate": 3.446853139341069e-05, + "loss": 0.4189, + "step": 7820 + }, + { + "epoch": 2.9490950226244346, + "grad_norm": 0.1531152171769124, + "learning_rate": 3.445810286396735e-05, + "loss": 0.4213, + "step": 7821 + }, + { + "epoch": 2.9494720965309202, + "grad_norm": 0.15719902766827068, + "learning_rate": 3.444767471857256e-05, + "loss": 0.4242, + "step": 7822 + }, + { + "epoch": 2.949849170437406, + "grad_norm": 0.15506311188307917, + "learning_rate": 3.4437246957949e-05, + "loss": 0.4178, + "step": 7823 + }, + { + "epoch": 2.9502262443438916, + "grad_norm": 0.1567211598260663, + "learning_rate": 3.4426819582819294e-05, + "loss": 0.4097, + "step": 7824 + }, + { + "epoch": 2.9506033182503772, + "grad_norm": 0.16350612207226017, + "learning_rate": 3.441639259390604e-05, + "loss": 0.4278, + "step": 7825 + }, + { + "epoch": 2.950980392156863, + "grad_norm": 0.168604894313827, + "learning_rate": 3.4405965991931824e-05, + "loss": 0.4165, + "step": 7826 + }, + { + "epoch": 2.9513574660633486, + "grad_norm": 0.15861170497706656, + "learning_rate": 3.4395539777619213e-05, + "loss": 0.4199, + "step": 7827 + }, + { + "epoch": 2.9517345399698343, + "grad_norm": 0.16816420149797678, + "learning_rate": 3.438511395169072e-05, + "loss": 0.4201, + "step": 7828 + }, + { + "epoch": 2.95211161387632, + "grad_norm": 0.20820925757097541, + "learning_rate": 3.4374688514868843e-05, + "loss": 0.4165, + "step": 7829 + }, + { + "epoch": 2.9524886877828056, + "grad_norm": 0.2264682563023886, + "learning_rate": 3.4364263467876075e-05, + "loss": 0.4301, + "step": 7830 + }, + { + "epoch": 2.9528657616892913, + "grad_norm": 0.23880784914799197, + "learning_rate": 3.4353838811434836e-05, + "loss": 0.4163, + "step": 7831 + }, + { + "epoch": 2.953242835595777, + "grad_norm": 0.2306485658650106, + "learning_rate": 3.4343414546267566e-05, + "loss": 0.4012, + "step": 7832 + }, + { + "epoch": 2.9536199095022626, + "grad_norm": 0.19436207675593176, + "learning_rate": 3.433299067309665e-05, + "loss": 0.4225, + "step": 7833 + }, + { + "epoch": 2.9539969834087483, + "grad_norm": 0.15607615353482054, + "learning_rate": 3.4322567192644456e-05, + "loss": 0.417, + "step": 7834 + }, + { + "epoch": 2.954374057315234, + "grad_norm": 0.22383787280407372, + "learning_rate": 3.4312144105633316e-05, + "loss": 0.4171, + "step": 7835 + }, + { + "epoch": 2.9547511312217196, + "grad_norm": 0.25538237946617004, + "learning_rate": 3.430172141278555e-05, + "loss": 0.419, + "step": 7836 + }, + { + "epoch": 2.9551282051282053, + "grad_norm": 0.2211942520552908, + "learning_rate": 3.429129911482343e-05, + "loss": 0.4215, + "step": 7837 + }, + { + "epoch": 2.955505279034691, + "grad_norm": 0.20694525700950936, + "learning_rate": 3.428087721246922e-05, + "loss": 0.4257, + "step": 7838 + }, + { + "epoch": 2.9558823529411766, + "grad_norm": 0.1927018376558712, + "learning_rate": 3.427045570644515e-05, + "loss": 0.4247, + "step": 7839 + }, + { + "epoch": 2.9562594268476623, + "grad_norm": 0.23172648236502896, + "learning_rate": 3.426003459747341e-05, + "loss": 0.4111, + "step": 7840 + }, + { + "epoch": 2.956636500754148, + "grad_norm": 0.23101910771873618, + "learning_rate": 3.424961388627618e-05, + "loss": 0.4265, + "step": 7841 + }, + { + "epoch": 2.9570135746606336, + "grad_norm": 0.22847045522715637, + "learning_rate": 3.423919357357562e-05, + "loss": 0.4167, + "step": 7842 + }, + { + "epoch": 2.9573906485671193, + "grad_norm": 0.22828301120703157, + "learning_rate": 3.422877366009381e-05, + "loss": 0.4126, + "step": 7843 + }, + { + "epoch": 2.957767722473605, + "grad_norm": 0.2659873782042032, + "learning_rate": 3.4218354146552893e-05, + "loss": 0.4137, + "step": 7844 + }, + { + "epoch": 2.9581447963800906, + "grad_norm": 0.19149662018110744, + "learning_rate": 3.42079350336749e-05, + "loss": 0.413, + "step": 7845 + }, + { + "epoch": 2.9585218702865763, + "grad_norm": 0.22206517740964474, + "learning_rate": 3.419751632218186e-05, + "loss": 0.4089, + "step": 7846 + }, + { + "epoch": 2.958898944193062, + "grad_norm": 0.2586060057927749, + "learning_rate": 3.4187098012795806e-05, + "loss": 0.4229, + "step": 7847 + }, + { + "epoch": 2.9592760180995477, + "grad_norm": 0.23020467404519168, + "learning_rate": 3.41766801062387e-05, + "loss": 0.415, + "step": 7848 + }, + { + "epoch": 2.9596530920060333, + "grad_norm": 0.23061033722111057, + "learning_rate": 3.4166262603232485e-05, + "loss": 0.4163, + "step": 7849 + }, + { + "epoch": 2.960030165912519, + "grad_norm": 0.14766583386265644, + "learning_rate": 3.415584550449911e-05, + "loss": 0.4015, + "step": 7850 + }, + { + "epoch": 2.9604072398190047, + "grad_norm": 0.18361141368670425, + "learning_rate": 3.414542881076046e-05, + "loss": 0.413, + "step": 7851 + }, + { + "epoch": 2.9607843137254903, + "grad_norm": 0.19546657400684117, + "learning_rate": 3.4135012522738384e-05, + "loss": 0.4067, + "step": 7852 + }, + { + "epoch": 2.961161387631976, + "grad_norm": 0.23665875107324805, + "learning_rate": 3.4124596641154756e-05, + "loss": 0.4169, + "step": 7853 + }, + { + "epoch": 2.9615384615384617, + "grad_norm": 0.27292188856975697, + "learning_rate": 3.4114181166731355e-05, + "loss": 0.4112, + "step": 7854 + }, + { + "epoch": 2.9619155354449473, + "grad_norm": 0.22225409590751782, + "learning_rate": 3.410376610018999e-05, + "loss": 0.4167, + "step": 7855 + }, + { + "epoch": 2.962292609351433, + "grad_norm": 0.18805848574021986, + "learning_rate": 3.40933514422524e-05, + "loss": 0.4309, + "step": 7856 + }, + { + "epoch": 2.9626696832579187, + "grad_norm": 0.22497174855175098, + "learning_rate": 3.408293719364031e-05, + "loss": 0.4096, + "step": 7857 + }, + { + "epoch": 2.9630467571644044, + "grad_norm": 0.22964556985876486, + "learning_rate": 3.407252335507544e-05, + "loss": 0.4138, + "step": 7858 + }, + { + "epoch": 2.96342383107089, + "grad_norm": 0.25297194766215064, + "learning_rate": 3.406210992727944e-05, + "loss": 0.428, + "step": 7859 + }, + { + "epoch": 2.9638009049773757, + "grad_norm": 0.23097796675001564, + "learning_rate": 3.4051696910973927e-05, + "loss": 0.4077, + "step": 7860 + }, + { + "epoch": 2.9641779788838614, + "grad_norm": 0.1953323864264835, + "learning_rate": 3.404128430688057e-05, + "loss": 0.4313, + "step": 7861 + }, + { + "epoch": 2.964555052790347, + "grad_norm": 0.18778184693883526, + "learning_rate": 3.4030872115720926e-05, + "loss": 0.4235, + "step": 7862 + }, + { + "epoch": 2.9649321266968327, + "grad_norm": 0.1753160674795958, + "learning_rate": 3.402046033821653e-05, + "loss": 0.4176, + "step": 7863 + }, + { + "epoch": 2.9653092006033184, + "grad_norm": 0.16202106755447965, + "learning_rate": 3.401004897508895e-05, + "loss": 0.4062, + "step": 7864 + }, + { + "epoch": 2.965686274509804, + "grad_norm": 0.17334228700320845, + "learning_rate": 3.3999638027059656e-05, + "loss": 0.4207, + "step": 7865 + }, + { + "epoch": 2.9660633484162897, + "grad_norm": 0.18326547561541948, + "learning_rate": 3.3989227494850116e-05, + "loss": 0.428, + "step": 7866 + }, + { + "epoch": 2.9664404223227754, + "grad_norm": 0.17165161226002207, + "learning_rate": 3.397881737918179e-05, + "loss": 0.4157, + "step": 7867 + }, + { + "epoch": 2.966817496229261, + "grad_norm": 0.12774901256865895, + "learning_rate": 3.396840768077608e-05, + "loss": 0.4145, + "step": 7868 + }, + { + "epoch": 2.9671945701357467, + "grad_norm": 0.15477288513606036, + "learning_rate": 3.395799840035435e-05, + "loss": 0.4118, + "step": 7869 + }, + { + "epoch": 2.9675716440422324, + "grad_norm": 0.15438947581806123, + "learning_rate": 3.3947589538637985e-05, + "loss": 0.4192, + "step": 7870 + }, + { + "epoch": 2.967948717948718, + "grad_norm": 0.17471576005333495, + "learning_rate": 3.393718109634831e-05, + "loss": 0.4097, + "step": 7871 + }, + { + "epoch": 2.9683257918552037, + "grad_norm": 0.1763844378109978, + "learning_rate": 3.392677307420658e-05, + "loss": 0.4169, + "step": 7872 + }, + { + "epoch": 2.9687028657616894, + "grad_norm": 0.17176124648672342, + "learning_rate": 3.391636547293411e-05, + "loss": 0.4018, + "step": 7873 + }, + { + "epoch": 2.969079939668175, + "grad_norm": 0.15699877378028032, + "learning_rate": 3.3905958293252105e-05, + "loss": 0.4243, + "step": 7874 + }, + { + "epoch": 2.9694570135746607, + "grad_norm": 0.1825325229564802, + "learning_rate": 3.389555153588178e-05, + "loss": 0.4142, + "step": 7875 + }, + { + "epoch": 2.9698340874811464, + "grad_norm": 0.17691938330198784, + "learning_rate": 3.3885145201544336e-05, + "loss": 0.4268, + "step": 7876 + }, + { + "epoch": 2.970211161387632, + "grad_norm": 0.17389793276623983, + "learning_rate": 3.38747392909609e-05, + "loss": 0.4081, + "step": 7877 + }, + { + "epoch": 2.9705882352941178, + "grad_norm": 0.1914765206237473, + "learning_rate": 3.386433380485258e-05, + "loss": 0.4151, + "step": 7878 + }, + { + "epoch": 2.9709653092006034, + "grad_norm": 0.19769639537380754, + "learning_rate": 3.385392874394051e-05, + "loss": 0.4199, + "step": 7879 + }, + { + "epoch": 2.971342383107089, + "grad_norm": 0.17072092034515643, + "learning_rate": 3.384352410894572e-05, + "loss": 0.4141, + "step": 7880 + }, + { + "epoch": 2.9717194570135748, + "grad_norm": 0.17426612842741132, + "learning_rate": 3.383311990058923e-05, + "loss": 0.4228, + "step": 7881 + }, + { + "epoch": 2.9720965309200604, + "grad_norm": 0.17820704196771578, + "learning_rate": 3.382271611959208e-05, + "loss": 0.4133, + "step": 7882 + }, + { + "epoch": 2.972473604826546, + "grad_norm": 0.15571053027505347, + "learning_rate": 3.3812312766675224e-05, + "loss": 0.416, + "step": 7883 + }, + { + "epoch": 2.9728506787330318, + "grad_norm": 0.1869149831085075, + "learning_rate": 3.3801909842559594e-05, + "loss": 0.4124, + "step": 7884 + }, + { + "epoch": 2.9732277526395174, + "grad_norm": 0.16157306993483095, + "learning_rate": 3.3791507347966125e-05, + "loss": 0.4156, + "step": 7885 + }, + { + "epoch": 2.973604826546003, + "grad_norm": 0.1670238611799301, + "learning_rate": 3.378110528361569e-05, + "loss": 0.4208, + "step": 7886 + }, + { + "epoch": 2.973981900452489, + "grad_norm": 0.20325086531014677, + "learning_rate": 3.3770703650229125e-05, + "loss": 0.4149, + "step": 7887 + }, + { + "epoch": 2.9743589743589745, + "grad_norm": 0.21898935511211406, + "learning_rate": 3.376030244852729e-05, + "loss": 0.4078, + "step": 7888 + }, + { + "epoch": 2.97473604826546, + "grad_norm": 0.1735973939031538, + "learning_rate": 3.3749901679230966e-05, + "loss": 0.4206, + "step": 7889 + }, + { + "epoch": 2.975113122171946, + "grad_norm": 0.16854898088067855, + "learning_rate": 3.373950134306089e-05, + "loss": 0.4087, + "step": 7890 + }, + { + "epoch": 2.9754901960784315, + "grad_norm": 0.16079483081673993, + "learning_rate": 3.372910144073783e-05, + "loss": 0.4213, + "step": 7891 + }, + { + "epoch": 2.975867269984917, + "grad_norm": 0.20018114923649993, + "learning_rate": 3.371870197298248e-05, + "loss": 0.4117, + "step": 7892 + }, + { + "epoch": 2.976244343891403, + "grad_norm": 0.1707698243405636, + "learning_rate": 3.37083029405155e-05, + "loss": 0.4115, + "step": 7893 + }, + { + "epoch": 2.9766214177978885, + "grad_norm": 0.15671963872168856, + "learning_rate": 3.369790434405756e-05, + "loss": 0.4201, + "step": 7894 + }, + { + "epoch": 2.976998491704374, + "grad_norm": 0.15158577120865072, + "learning_rate": 3.368750618432926e-05, + "loss": 0.402, + "step": 7895 + }, + { + "epoch": 2.97737556561086, + "grad_norm": 0.1485449653931936, + "learning_rate": 3.3677108462051165e-05, + "loss": 0.4327, + "step": 7896 + }, + { + "epoch": 2.9777526395173455, + "grad_norm": 0.16135553237128172, + "learning_rate": 3.366671117794386e-05, + "loss": 0.4268, + "step": 7897 + }, + { + "epoch": 2.978129713423831, + "grad_norm": 0.1674409529367263, + "learning_rate": 3.3656314332727854e-05, + "loss": 0.414, + "step": 7898 + }, + { + "epoch": 2.978506787330317, + "grad_norm": 0.14500258642692515, + "learning_rate": 3.3645917927123615e-05, + "loss": 0.4389, + "step": 7899 + }, + { + "epoch": 2.9788838612368025, + "grad_norm": 0.1891283630063404, + "learning_rate": 3.3635521961851656e-05, + "loss": 0.4117, + "step": 7900 + }, + { + "epoch": 2.979260935143288, + "grad_norm": 0.1917356888451871, + "learning_rate": 3.3625126437632375e-05, + "loss": 0.408, + "step": 7901 + }, + { + "epoch": 2.979638009049774, + "grad_norm": 0.18451205817272953, + "learning_rate": 3.3614731355186156e-05, + "loss": 0.4115, + "step": 7902 + }, + { + "epoch": 2.9800150829562595, + "grad_norm": 0.1901750574310443, + "learning_rate": 3.360433671523341e-05, + "loss": 0.4284, + "step": 7903 + }, + { + "epoch": 2.980392156862745, + "grad_norm": 0.17035173136872717, + "learning_rate": 3.359394251849445e-05, + "loss": 0.4238, + "step": 7904 + }, + { + "epoch": 2.980769230769231, + "grad_norm": 0.1808876553579747, + "learning_rate": 3.358354876568959e-05, + "loss": 0.4113, + "step": 7905 + }, + { + "epoch": 2.9811463046757165, + "grad_norm": 0.1556111873780182, + "learning_rate": 3.357315545753911e-05, + "loss": 0.4089, + "step": 7906 + }, + { + "epoch": 2.981523378582202, + "grad_norm": 0.15967885183370376, + "learning_rate": 3.356276259476327e-05, + "loss": 0.4301, + "step": 7907 + }, + { + "epoch": 2.981900452488688, + "grad_norm": 0.1728814866293614, + "learning_rate": 3.355237017808225e-05, + "loss": 0.4084, + "step": 7908 + }, + { + "epoch": 2.9822775263951735, + "grad_norm": 0.15984719258980362, + "learning_rate": 3.3541978208216275e-05, + "loss": 0.4068, + "step": 7909 + }, + { + "epoch": 2.982654600301659, + "grad_norm": 0.14835934652705646, + "learning_rate": 3.353158668588548e-05, + "loss": 0.4153, + "step": 7910 + }, + { + "epoch": 2.983031674208145, + "grad_norm": 0.1871531297928058, + "learning_rate": 3.352119561180997e-05, + "loss": 0.41, + "step": 7911 + }, + { + "epoch": 2.9834087481146305, + "grad_norm": 0.17504751997649343, + "learning_rate": 3.351080498670988e-05, + "loss": 0.4222, + "step": 7912 + }, + { + "epoch": 2.983785822021116, + "grad_norm": 0.17718355505426706, + "learning_rate": 3.3500414811305254e-05, + "loss": 0.4104, + "step": 7913 + }, + { + "epoch": 2.984162895927602, + "grad_norm": 0.21132280495424918, + "learning_rate": 3.349002508631609e-05, + "loss": 0.4266, + "step": 7914 + }, + { + "epoch": 2.9845399698340875, + "grad_norm": 0.21734249652011572, + "learning_rate": 3.3479635812462435e-05, + "loss": 0.411, + "step": 7915 + }, + { + "epoch": 2.984917043740573, + "grad_norm": 0.1580136457858344, + "learning_rate": 3.346924699046423e-05, + "loss": 0.4156, + "step": 7916 + }, + { + "epoch": 2.985294117647059, + "grad_norm": 0.15914733773738213, + "learning_rate": 3.34588586210414e-05, + "loss": 0.43, + "step": 7917 + }, + { + "epoch": 2.9856711915535445, + "grad_norm": 0.1716134494887889, + "learning_rate": 3.3448470704913885e-05, + "loss": 0.4275, + "step": 7918 + }, + { + "epoch": 2.98604826546003, + "grad_norm": 0.16159318939735146, + "learning_rate": 3.3438083242801535e-05, + "loss": 0.4103, + "step": 7919 + }, + { + "epoch": 2.986425339366516, + "grad_norm": 0.1293359279029421, + "learning_rate": 3.342769623542418e-05, + "loss": 0.418, + "step": 7920 + }, + { + "epoch": 2.9868024132730016, + "grad_norm": 0.14416158396445483, + "learning_rate": 3.3417309683501664e-05, + "loss": 0.4108, + "step": 7921 + }, + { + "epoch": 2.9871794871794872, + "grad_norm": 0.18636808751729725, + "learning_rate": 3.340692358775375e-05, + "loss": 0.4091, + "step": 7922 + }, + { + "epoch": 2.987556561085973, + "grad_norm": 0.14045885666538135, + "learning_rate": 3.3396537948900164e-05, + "loss": 0.4162, + "step": 7923 + }, + { + "epoch": 2.9879336349924586, + "grad_norm": 0.15092468672472772, + "learning_rate": 3.3386152767660665e-05, + "loss": 0.4158, + "step": 7924 + }, + { + "epoch": 2.9883107088989442, + "grad_norm": 0.16530431687625313, + "learning_rate": 3.33757680447549e-05, + "loss": 0.4066, + "step": 7925 + }, + { + "epoch": 2.98868778280543, + "grad_norm": 0.20029736569031326, + "learning_rate": 3.336538378090253e-05, + "loss": 0.4207, + "step": 7926 + }, + { + "epoch": 2.9890648567119156, + "grad_norm": 0.19735730963300038, + "learning_rate": 3.335499997682319e-05, + "loss": 0.4116, + "step": 7927 + }, + { + "epoch": 2.9894419306184012, + "grad_norm": 0.16044648629200373, + "learning_rate": 3.3344616633236454e-05, + "loss": 0.4141, + "step": 7928 + }, + { + "epoch": 2.989819004524887, + "grad_norm": 0.17650253811800345, + "learning_rate": 3.333423375086187e-05, + "loss": 0.4091, + "step": 7929 + }, + { + "epoch": 2.9901960784313726, + "grad_norm": 0.16211098099382604, + "learning_rate": 3.332385133041899e-05, + "loss": 0.4216, + "step": 7930 + }, + { + "epoch": 2.9905731523378583, + "grad_norm": 0.15609032000382192, + "learning_rate": 3.331346937262729e-05, + "loss": 0.4135, + "step": 7931 + }, + { + "epoch": 2.990950226244344, + "grad_norm": 0.1465676320889167, + "learning_rate": 3.3303087878206214e-05, + "loss": 0.4106, + "step": 7932 + }, + { + "epoch": 2.9913273001508296, + "grad_norm": 0.14566791306254082, + "learning_rate": 3.329270684787523e-05, + "loss": 0.4204, + "step": 7933 + }, + { + "epoch": 2.9917043740573153, + "grad_norm": 0.13423229208130072, + "learning_rate": 3.3282326282353706e-05, + "loss": 0.408, + "step": 7934 + }, + { + "epoch": 2.992081447963801, + "grad_norm": 0.17693812346558435, + "learning_rate": 3.3271946182361e-05, + "loss": 0.43, + "step": 7935 + }, + { + "epoch": 2.9924585218702866, + "grad_norm": 0.18507973426908417, + "learning_rate": 3.3261566548616464e-05, + "loss": 0.4173, + "step": 7936 + }, + { + "epoch": 2.9928355957767723, + "grad_norm": 0.20272234328552455, + "learning_rate": 3.32511873818394e-05, + "loss": 0.4053, + "step": 7937 + }, + { + "epoch": 2.993212669683258, + "grad_norm": 0.17531947093780056, + "learning_rate": 3.324080868274904e-05, + "loss": 0.4058, + "step": 7938 + }, + { + "epoch": 2.9935897435897436, + "grad_norm": 0.15001384811347762, + "learning_rate": 3.3230430452064664e-05, + "loss": 0.4208, + "step": 7939 + }, + { + "epoch": 2.9939668174962293, + "grad_norm": 0.15835959697730903, + "learning_rate": 3.322005269050545e-05, + "loss": 0.4177, + "step": 7940 + }, + { + "epoch": 2.994343891402715, + "grad_norm": 0.19304812641358202, + "learning_rate": 3.320967539879056e-05, + "loss": 0.4239, + "step": 7941 + }, + { + "epoch": 2.9947209653092006, + "grad_norm": 0.20585151297222382, + "learning_rate": 3.319929857763915e-05, + "loss": 0.4259, + "step": 7942 + }, + { + "epoch": 2.9950980392156863, + "grad_norm": 0.16413403669244833, + "learning_rate": 3.318892222777032e-05, + "loss": 0.4152, + "step": 7943 + }, + { + "epoch": 2.995475113122172, + "grad_norm": 0.1496508102806433, + "learning_rate": 3.3178546349903124e-05, + "loss": 0.4089, + "step": 7944 + }, + { + "epoch": 2.9958521870286576, + "grad_norm": 0.17142204460929317, + "learning_rate": 3.3168170944756626e-05, + "loss": 0.4146, + "step": 7945 + }, + { + "epoch": 2.9962292609351433, + "grad_norm": 0.17166238239990486, + "learning_rate": 3.315779601304983e-05, + "loss": 0.4108, + "step": 7946 + }, + { + "epoch": 2.996606334841629, + "grad_norm": 0.17499997470445872, + "learning_rate": 3.3147421555501674e-05, + "loss": 0.4146, + "step": 7947 + }, + { + "epoch": 2.9969834087481146, + "grad_norm": 0.1604752732049505, + "learning_rate": 3.313704757283115e-05, + "loss": 0.4161, + "step": 7948 + }, + { + "epoch": 2.9973604826546003, + "grad_norm": 0.1550829327504555, + "learning_rate": 3.3126674065757144e-05, + "loss": 0.412, + "step": 7949 + }, + { + "epoch": 2.997737556561086, + "grad_norm": 0.20756286965617454, + "learning_rate": 3.311630103499851e-05, + "loss": 0.4109, + "step": 7950 + }, + { + "epoch": 2.9981146304675717, + "grad_norm": 0.2068491715571928, + "learning_rate": 3.3105928481274127e-05, + "loss": 0.409, + "step": 7951 + }, + { + "epoch": 2.9984917043740573, + "grad_norm": 0.16932489184793678, + "learning_rate": 3.309555640530277e-05, + "loss": 0.424, + "step": 7952 + }, + { + "epoch": 2.998868778280543, + "grad_norm": 0.19602363009231188, + "learning_rate": 3.3085184807803236e-05, + "loss": 0.4083, + "step": 7953 + }, + { + "epoch": 2.9992458521870287, + "grad_norm": 0.16122324206646937, + "learning_rate": 3.307481368949426e-05, + "loss": 0.4074, + "step": 7954 + }, + { + "epoch": 2.9996229260935143, + "grad_norm": 0.20268815185617325, + "learning_rate": 3.306444305109455e-05, + "loss": 0.4122, + "step": 7955 + }, + { + "epoch": 3.0, + "grad_norm": 0.20309891055611975, + "learning_rate": 3.305407289332279e-05, + "loss": 0.399, + "step": 7956 + }, + { + "epoch": 3.0003770739064857, + "grad_norm": 0.17424995772141957, + "learning_rate": 3.3043703216897624e-05, + "loss": 0.4017, + "step": 7957 + }, + { + "epoch": 3.0007541478129713, + "grad_norm": 0.20076657917803034, + "learning_rate": 3.303333402253764e-05, + "loss": 0.3923, + "step": 7958 + }, + { + "epoch": 3.001131221719457, + "grad_norm": 0.20662907569024913, + "learning_rate": 3.302296531096144e-05, + "loss": 0.4017, + "step": 7959 + }, + { + "epoch": 3.0015082956259427, + "grad_norm": 0.2323590340465483, + "learning_rate": 3.3012597082887556e-05, + "loss": 0.3999, + "step": 7960 + }, + { + "epoch": 3.0018853695324283, + "grad_norm": 0.1978703378950321, + "learning_rate": 3.300222933903449e-05, + "loss": 0.3903, + "step": 7961 + }, + { + "epoch": 3.002262443438914, + "grad_norm": 0.20494215016283723, + "learning_rate": 3.299186208012072e-05, + "loss": 0.4076, + "step": 7962 + }, + { + "epoch": 3.0026395173453997, + "grad_norm": 0.2163620709766989, + "learning_rate": 3.2981495306864704e-05, + "loss": 0.395, + "step": 7963 + }, + { + "epoch": 3.0030165912518854, + "grad_norm": 0.16283616094101733, + "learning_rate": 3.297112901998483e-05, + "loss": 0.3916, + "step": 7964 + }, + { + "epoch": 3.003393665158371, + "grad_norm": 0.1736982945385893, + "learning_rate": 3.296076322019947e-05, + "loss": 0.4003, + "step": 7965 + }, + { + "epoch": 3.0037707390648567, + "grad_norm": 0.1779381010636698, + "learning_rate": 3.2950397908227e-05, + "loss": 0.3784, + "step": 7966 + }, + { + "epoch": 3.0041478129713424, + "grad_norm": 0.1757175806083518, + "learning_rate": 3.2940033084785684e-05, + "loss": 0.3931, + "step": 7967 + }, + { + "epoch": 3.004524886877828, + "grad_norm": 0.18840078378236352, + "learning_rate": 3.292966875059382e-05, + "loss": 0.4017, + "step": 7968 + }, + { + "epoch": 3.0049019607843137, + "grad_norm": 0.1737351926412328, + "learning_rate": 3.291930490636964e-05, + "loss": 0.4056, + "step": 7969 + }, + { + "epoch": 3.0052790346907994, + "grad_norm": 0.1566128000251608, + "learning_rate": 3.290894155283134e-05, + "loss": 0.4024, + "step": 7970 + }, + { + "epoch": 3.005656108597285, + "grad_norm": 0.15707336528742968, + "learning_rate": 3.28985786906971e-05, + "loss": 0.3902, + "step": 7971 + }, + { + "epoch": 3.0060331825037707, + "grad_norm": 0.16630202841457542, + "learning_rate": 3.2888216320685066e-05, + "loss": 0.3932, + "step": 7972 + }, + { + "epoch": 3.0064102564102564, + "grad_norm": 0.14116525951096753, + "learning_rate": 3.287785444351334e-05, + "loss": 0.377, + "step": 7973 + }, + { + "epoch": 3.006787330316742, + "grad_norm": 0.16946978323360704, + "learning_rate": 3.286749305989997e-05, + "loss": 0.3962, + "step": 7974 + }, + { + "epoch": 3.0071644042232277, + "grad_norm": 0.1855344841342463, + "learning_rate": 3.285713217056301e-05, + "loss": 0.3825, + "step": 7975 + }, + { + "epoch": 3.0075414781297134, + "grad_norm": 0.14773096945199737, + "learning_rate": 3.2846771776220464e-05, + "loss": 0.3815, + "step": 7976 + }, + { + "epoch": 3.007918552036199, + "grad_norm": 0.14140797983242778, + "learning_rate": 3.283641187759028e-05, + "loss": 0.399, + "step": 7977 + }, + { + "epoch": 3.0082956259426847, + "grad_norm": 0.15566415677574907, + "learning_rate": 3.2826052475390394e-05, + "loss": 0.3852, + "step": 7978 + }, + { + "epoch": 3.0086726998491704, + "grad_norm": 0.16961914168819264, + "learning_rate": 3.2815693570338704e-05, + "loss": 0.409, + "step": 7979 + }, + { + "epoch": 3.009049773755656, + "grad_norm": 0.15694630144277796, + "learning_rate": 3.28053351631531e-05, + "loss": 0.4005, + "step": 7980 + }, + { + "epoch": 3.0094268476621417, + "grad_norm": 0.1527250310004573, + "learning_rate": 3.279497725455137e-05, + "loss": 0.4009, + "step": 7981 + }, + { + "epoch": 3.0098039215686274, + "grad_norm": 0.14202731551705247, + "learning_rate": 3.278461984525132e-05, + "loss": 0.3853, + "step": 7982 + }, + { + "epoch": 3.010180995475113, + "grad_norm": 0.14993314281528292, + "learning_rate": 3.2774262935970734e-05, + "loss": 0.3982, + "step": 7983 + }, + { + "epoch": 3.0105580693815988, + "grad_norm": 0.1693163480501089, + "learning_rate": 3.276390652742731e-05, + "loss": 0.3852, + "step": 7984 + }, + { + "epoch": 3.0109351432880844, + "grad_norm": 0.16862019037974596, + "learning_rate": 3.2753550620338735e-05, + "loss": 0.3965, + "step": 7985 + }, + { + "epoch": 3.01131221719457, + "grad_norm": 0.16801471835665943, + "learning_rate": 3.2743195215422686e-05, + "loss": 0.3984, + "step": 7986 + }, + { + "epoch": 3.0116892911010558, + "grad_norm": 0.14672613730342926, + "learning_rate": 3.2732840313396765e-05, + "loss": 0.4065, + "step": 7987 + }, + { + "epoch": 3.0120663650075414, + "grad_norm": 0.13665417405193492, + "learning_rate": 3.2722485914978565e-05, + "loss": 0.4011, + "step": 7988 + }, + { + "epoch": 3.012443438914027, + "grad_norm": 0.1795124306308611, + "learning_rate": 3.2712132020885646e-05, + "loss": 0.4022, + "step": 7989 + }, + { + "epoch": 3.0128205128205128, + "grad_norm": 0.18561261879086194, + "learning_rate": 3.27017786318355e-05, + "loss": 0.4058, + "step": 7990 + }, + { + "epoch": 3.0131975867269984, + "grad_norm": 0.1373186942993209, + "learning_rate": 3.2691425748545626e-05, + "loss": 0.3828, + "step": 7991 + }, + { + "epoch": 3.013574660633484, + "grad_norm": 0.15306498398269874, + "learning_rate": 3.268107337173346e-05, + "loss": 0.3876, + "step": 7992 + }, + { + "epoch": 3.01395173453997, + "grad_norm": 0.1874989625579249, + "learning_rate": 3.267072150211643e-05, + "loss": 0.3925, + "step": 7993 + }, + { + "epoch": 3.0143288084464555, + "grad_norm": 0.17026248134146804, + "learning_rate": 3.266037014041189e-05, + "loss": 0.3972, + "step": 7994 + }, + { + "epoch": 3.014705882352941, + "grad_norm": 0.1625597506439887, + "learning_rate": 3.2650019287337184e-05, + "loss": 0.4009, + "step": 7995 + }, + { + "epoch": 3.015082956259427, + "grad_norm": 0.18040916956567954, + "learning_rate": 3.2639668943609636e-05, + "loss": 0.3914, + "step": 7996 + }, + { + "epoch": 3.0154600301659125, + "grad_norm": 0.16119595636666265, + "learning_rate": 3.262931910994649e-05, + "loss": 0.3918, + "step": 7997 + }, + { + "epoch": 3.015837104072398, + "grad_norm": 0.1920688916561292, + "learning_rate": 3.2618969787064996e-05, + "loss": 0.3934, + "step": 7998 + }, + { + "epoch": 3.016214177978884, + "grad_norm": 0.20543251305852053, + "learning_rate": 3.260862097568236e-05, + "loss": 0.3918, + "step": 7999 + }, + { + "epoch": 3.0165912518853695, + "grad_norm": 0.1818833831547092, + "learning_rate": 3.259827267651572e-05, + "loss": 0.3948, + "step": 8000 + }, + { + "epoch": 3.016968325791855, + "grad_norm": 0.13341534956676415, + "learning_rate": 3.258792489028221e-05, + "loss": 0.3946, + "step": 8001 + }, + { + "epoch": 3.017345399698341, + "grad_norm": 0.18838582494596443, + "learning_rate": 3.2577577617698955e-05, + "loss": 0.3926, + "step": 8002 + }, + { + "epoch": 3.0177224736048265, + "grad_norm": 0.20751488611075808, + "learning_rate": 3.2567230859482974e-05, + "loss": 0.392, + "step": 8003 + }, + { + "epoch": 3.018099547511312, + "grad_norm": 0.15272855732604146, + "learning_rate": 3.25568846163513e-05, + "loss": 0.3922, + "step": 8004 + }, + { + "epoch": 3.018476621417798, + "grad_norm": 0.14893369187676356, + "learning_rate": 3.254653888902093e-05, + "loss": 0.4044, + "step": 8005 + }, + { + "epoch": 3.0188536953242835, + "grad_norm": 0.16954073500789685, + "learning_rate": 3.253619367820879e-05, + "loss": 0.3993, + "step": 8006 + }, + { + "epoch": 3.019230769230769, + "grad_norm": 0.18829474528573675, + "learning_rate": 3.2525848984631814e-05, + "loss": 0.3979, + "step": 8007 + }, + { + "epoch": 3.019607843137255, + "grad_norm": 0.19905474395078276, + "learning_rate": 3.2515504809006875e-05, + "loss": 0.4019, + "step": 8008 + }, + { + "epoch": 3.0199849170437405, + "grad_norm": 0.16853524200310402, + "learning_rate": 3.2505161152050805e-05, + "loss": 0.4106, + "step": 8009 + }, + { + "epoch": 3.020361990950226, + "grad_norm": 0.13901154016704295, + "learning_rate": 3.249481801448043e-05, + "loss": 0.3962, + "step": 8010 + }, + { + "epoch": 3.020739064856712, + "grad_norm": 0.21865430183150958, + "learning_rate": 3.248447539701251e-05, + "loss": 0.3984, + "step": 8011 + }, + { + "epoch": 3.0211161387631975, + "grad_norm": 0.22941949906174386, + "learning_rate": 3.247413330036376e-05, + "loss": 0.4024, + "step": 8012 + }, + { + "epoch": 3.021493212669683, + "grad_norm": 0.17238522118439872, + "learning_rate": 3.246379172525092e-05, + "loss": 0.3874, + "step": 8013 + }, + { + "epoch": 3.021870286576169, + "grad_norm": 0.1636166925907475, + "learning_rate": 3.245345067239063e-05, + "loss": 0.407, + "step": 8014 + }, + { + "epoch": 3.0222473604826545, + "grad_norm": 0.17870500079995139, + "learning_rate": 3.24431101424995e-05, + "loss": 0.3911, + "step": 8015 + }, + { + "epoch": 3.02262443438914, + "grad_norm": 0.17515188478156182, + "learning_rate": 3.243277013629414e-05, + "loss": 0.3855, + "step": 8016 + }, + { + "epoch": 3.023001508295626, + "grad_norm": 0.14855455588084743, + "learning_rate": 3.24224306544911e-05, + "loss": 0.4107, + "step": 8017 + }, + { + "epoch": 3.0233785822021115, + "grad_norm": 0.11724650169050838, + "learning_rate": 3.241209169780688e-05, + "loss": 0.3825, + "step": 8018 + }, + { + "epoch": 3.023755656108597, + "grad_norm": 0.16040603985187368, + "learning_rate": 3.2401753266957995e-05, + "loss": 0.3921, + "step": 8019 + }, + { + "epoch": 3.024132730015083, + "grad_norm": 0.19102551165135986, + "learning_rate": 3.239141536266087e-05, + "loss": 0.4123, + "step": 8020 + }, + { + "epoch": 3.0245098039215685, + "grad_norm": 0.18488228052271696, + "learning_rate": 3.238107798563189e-05, + "loss": 0.4118, + "step": 8021 + }, + { + "epoch": 3.024886877828054, + "grad_norm": 0.14434789724846564, + "learning_rate": 3.237074113658746e-05, + "loss": 0.3893, + "step": 8022 + }, + { + "epoch": 3.02526395173454, + "grad_norm": 0.1639588522232704, + "learning_rate": 3.236040481624391e-05, + "loss": 0.3885, + "step": 8023 + }, + { + "epoch": 3.0256410256410255, + "grad_norm": 0.1874042199571027, + "learning_rate": 3.2350069025317505e-05, + "loss": 0.3934, + "step": 8024 + }, + { + "epoch": 3.026018099547511, + "grad_norm": 0.17765384550274477, + "learning_rate": 3.233973376452456e-05, + "loss": 0.3959, + "step": 8025 + }, + { + "epoch": 3.026395173453997, + "grad_norm": 0.16512529057274547, + "learning_rate": 3.2329399034581265e-05, + "loss": 0.3976, + "step": 8026 + }, + { + "epoch": 3.0267722473604826, + "grad_norm": 0.16696289798172198, + "learning_rate": 3.231906483620379e-05, + "loss": 0.4044, + "step": 8027 + }, + { + "epoch": 3.0271493212669682, + "grad_norm": 0.1993178413375666, + "learning_rate": 3.230873117010833e-05, + "loss": 0.4116, + "step": 8028 + }, + { + "epoch": 3.027526395173454, + "grad_norm": 0.16982795409500817, + "learning_rate": 3.229839803701098e-05, + "loss": 0.3906, + "step": 8029 + }, + { + "epoch": 3.0279034690799396, + "grad_norm": 0.1498674300815518, + "learning_rate": 3.2288065437627796e-05, + "loss": 0.3831, + "step": 8030 + }, + { + "epoch": 3.0282805429864252, + "grad_norm": 0.16854092194779766, + "learning_rate": 3.2277733372674854e-05, + "loss": 0.407, + "step": 8031 + }, + { + "epoch": 3.028657616892911, + "grad_norm": 0.1448983013507956, + "learning_rate": 3.226740184286814e-05, + "loss": 0.3893, + "step": 8032 + }, + { + "epoch": 3.0290346907993966, + "grad_norm": 0.16957060111551267, + "learning_rate": 3.22570708489236e-05, + "loss": 0.3952, + "step": 8033 + }, + { + "epoch": 3.0294117647058822, + "grad_norm": 0.15614407298751537, + "learning_rate": 3.22467403915572e-05, + "loss": 0.4001, + "step": 8034 + }, + { + "epoch": 3.029788838612368, + "grad_norm": 0.16529784840276143, + "learning_rate": 3.223641047148481e-05, + "loss": 0.3812, + "step": 8035 + }, + { + "epoch": 3.0301659125188536, + "grad_norm": 0.1482471360482119, + "learning_rate": 3.2226081089422286e-05, + "loss": 0.3863, + "step": 8036 + }, + { + "epoch": 3.0305429864253393, + "grad_norm": 0.18025469942463626, + "learning_rate": 3.221575224608545e-05, + "loss": 0.3903, + "step": 8037 + }, + { + "epoch": 3.030920060331825, + "grad_norm": 0.18296004273547342, + "learning_rate": 3.220542394219008e-05, + "loss": 0.4005, + "step": 8038 + }, + { + "epoch": 3.0312971342383106, + "grad_norm": 0.16308944948682388, + "learning_rate": 3.219509617845191e-05, + "loss": 0.3929, + "step": 8039 + }, + { + "epoch": 3.0316742081447963, + "grad_norm": 0.1561101176925561, + "learning_rate": 3.218476895558666e-05, + "loss": 0.3969, + "step": 8040 + }, + { + "epoch": 3.032051282051282, + "grad_norm": 0.14848262073182086, + "learning_rate": 3.217444227430999e-05, + "loss": 0.398, + "step": 8041 + }, + { + "epoch": 3.0324283559577676, + "grad_norm": 0.15805851934457563, + "learning_rate": 3.216411613533751e-05, + "loss": 0.3951, + "step": 8042 + }, + { + "epoch": 3.0328054298642533, + "grad_norm": 0.15867926771276544, + "learning_rate": 3.215379053938486e-05, + "loss": 0.4033, + "step": 8043 + }, + { + "epoch": 3.033182503770739, + "grad_norm": 0.1489962191796923, + "learning_rate": 3.2143465487167556e-05, + "loss": 0.3898, + "step": 8044 + }, + { + "epoch": 3.0335595776772246, + "grad_norm": 0.12989048425571043, + "learning_rate": 3.2133140979401105e-05, + "loss": 0.3784, + "step": 8045 + }, + { + "epoch": 3.0339366515837103, + "grad_norm": 0.1276778057127724, + "learning_rate": 3.2122817016801024e-05, + "loss": 0.3976, + "step": 8046 + }, + { + "epoch": 3.034313725490196, + "grad_norm": 0.1501039583287574, + "learning_rate": 3.211249360008274e-05, + "loss": 0.394, + "step": 8047 + }, + { + "epoch": 3.0346907993966816, + "grad_norm": 0.158824794879299, + "learning_rate": 3.210217072996164e-05, + "loss": 0.3932, + "step": 8048 + }, + { + "epoch": 3.0350678733031673, + "grad_norm": 0.1421960593897997, + "learning_rate": 3.209184840715312e-05, + "loss": 0.396, + "step": 8049 + }, + { + "epoch": 3.035444947209653, + "grad_norm": 0.13732056238324378, + "learning_rate": 3.208152663237248e-05, + "loss": 0.3983, + "step": 8050 + }, + { + "epoch": 3.0358220211161386, + "grad_norm": 0.1835489855077451, + "learning_rate": 3.207120540633501e-05, + "loss": 0.4136, + "step": 8051 + }, + { + "epoch": 3.0361990950226243, + "grad_norm": 0.22408788101830013, + "learning_rate": 3.2060884729756e-05, + "loss": 0.3928, + "step": 8052 + }, + { + "epoch": 3.03657616892911, + "grad_norm": 0.18263867541717288, + "learning_rate": 3.2050564603350624e-05, + "loss": 0.3984, + "step": 8053 + }, + { + "epoch": 3.0369532428355956, + "grad_norm": 0.1570479083643806, + "learning_rate": 3.204024502783406e-05, + "loss": 0.3929, + "step": 8054 + }, + { + "epoch": 3.0373303167420813, + "grad_norm": 0.16590950536980387, + "learning_rate": 3.2029926003921465e-05, + "loss": 0.3987, + "step": 8055 + }, + { + "epoch": 3.037707390648567, + "grad_norm": 0.18490286174205744, + "learning_rate": 3.201960753232793e-05, + "loss": 0.408, + "step": 8056 + }, + { + "epoch": 3.0380844645550527, + "grad_norm": 0.2026052890791888, + "learning_rate": 3.20092896137685e-05, + "loss": 0.391, + "step": 8057 + }, + { + "epoch": 3.0384615384615383, + "grad_norm": 0.17370272398441455, + "learning_rate": 3.199897224895823e-05, + "loss": 0.3995, + "step": 8058 + }, + { + "epoch": 3.038838612368024, + "grad_norm": 0.15155629121937284, + "learning_rate": 3.198865543861208e-05, + "loss": 0.3974, + "step": 8059 + }, + { + "epoch": 3.0392156862745097, + "grad_norm": 0.15705523617578399, + "learning_rate": 3.1978339183444993e-05, + "loss": 0.3986, + "step": 8060 + }, + { + "epoch": 3.0395927601809953, + "grad_norm": 0.1593558134071979, + "learning_rate": 3.19680234841719e-05, + "loss": 0.4172, + "step": 8061 + }, + { + "epoch": 3.039969834087481, + "grad_norm": 0.1853792755062897, + "learning_rate": 3.1957708341507654e-05, + "loss": 0.3955, + "step": 8062 + }, + { + "epoch": 3.0403469079939667, + "grad_norm": 0.15726495249014924, + "learning_rate": 3.194739375616707e-05, + "loss": 0.3962, + "step": 8063 + }, + { + "epoch": 3.0407239819004523, + "grad_norm": 0.1548677436675928, + "learning_rate": 3.1937079728864965e-05, + "loss": 0.3953, + "step": 8064 + }, + { + "epoch": 3.041101055806938, + "grad_norm": 0.18255133156558018, + "learning_rate": 3.1926766260316096e-05, + "loss": 0.3912, + "step": 8065 + }, + { + "epoch": 3.0414781297134237, + "grad_norm": 0.21220894675831115, + "learning_rate": 3.1916453351235144e-05, + "loss": 0.4009, + "step": 8066 + }, + { + "epoch": 3.0418552036199094, + "grad_norm": 0.18181762081278222, + "learning_rate": 3.1906141002336816e-05, + "loss": 0.3986, + "step": 8067 + }, + { + "epoch": 3.042232277526395, + "grad_norm": 0.1496050507702184, + "learning_rate": 3.1895829214335744e-05, + "loss": 0.3938, + "step": 8068 + }, + { + "epoch": 3.0426093514328807, + "grad_norm": 0.18638391629950177, + "learning_rate": 3.18855179879465e-05, + "loss": 0.3972, + "step": 8069 + }, + { + "epoch": 3.0429864253393664, + "grad_norm": 0.18040253954894542, + "learning_rate": 3.187520732388367e-05, + "loss": 0.3947, + "step": 8070 + }, + { + "epoch": 3.043363499245852, + "grad_norm": 0.16044557174368235, + "learning_rate": 3.1864897222861774e-05, + "loss": 0.387, + "step": 8071 + }, + { + "epoch": 3.0437405731523377, + "grad_norm": 0.12006366214311662, + "learning_rate": 3.1854587685595274e-05, + "loss": 0.3911, + "step": 8072 + }, + { + "epoch": 3.0441176470588234, + "grad_norm": 0.17409336669860212, + "learning_rate": 3.184427871279863e-05, + "loss": 0.3886, + "step": 8073 + }, + { + "epoch": 3.044494720965309, + "grad_norm": 0.16298301258528394, + "learning_rate": 3.183397030518624e-05, + "loss": 0.3856, + "step": 8074 + }, + { + "epoch": 3.0448717948717947, + "grad_norm": 0.16520699634179115, + "learning_rate": 3.182366246347244e-05, + "loss": 0.3934, + "step": 8075 + }, + { + "epoch": 3.0452488687782804, + "grad_norm": 0.1839445169948564, + "learning_rate": 3.181335518837161e-05, + "loss": 0.3964, + "step": 8076 + }, + { + "epoch": 3.045625942684766, + "grad_norm": 0.15448271301241115, + "learning_rate": 3.1803048480597995e-05, + "loss": 0.3958, + "step": 8077 + }, + { + "epoch": 3.0460030165912517, + "grad_norm": 0.20950248976876681, + "learning_rate": 3.179274234086584e-05, + "loss": 0.4098, + "step": 8078 + }, + { + "epoch": 3.0463800904977374, + "grad_norm": 0.15297438156566415, + "learning_rate": 3.178243676988936e-05, + "loss": 0.39, + "step": 8079 + }, + { + "epoch": 3.046757164404223, + "grad_norm": 0.15421872474607806, + "learning_rate": 3.177213176838274e-05, + "loss": 0.4046, + "step": 8080 + }, + { + "epoch": 3.0471342383107087, + "grad_norm": 0.18310027794125694, + "learning_rate": 3.176182733706007e-05, + "loss": 0.392, + "step": 8081 + }, + { + "epoch": 3.0475113122171944, + "grad_norm": 0.14905785358665438, + "learning_rate": 3.1751523476635474e-05, + "loss": 0.3849, + "step": 8082 + }, + { + "epoch": 3.04788838612368, + "grad_norm": 0.1396376854517869, + "learning_rate": 3.1741220187822985e-05, + "loss": 0.4025, + "step": 8083 + }, + { + "epoch": 3.0482654600301657, + "grad_norm": 0.1653542525536896, + "learning_rate": 3.17309174713366e-05, + "loss": 0.4111, + "step": 8084 + }, + { + "epoch": 3.0486425339366514, + "grad_norm": 0.17117044658714803, + "learning_rate": 3.17206153278903e-05, + "loss": 0.3955, + "step": 8085 + }, + { + "epoch": 3.049019607843137, + "grad_norm": 0.1617421001473641, + "learning_rate": 3.1710313758198025e-05, + "loss": 0.3895, + "step": 8086 + }, + { + "epoch": 3.0493966817496228, + "grad_norm": 0.13748524631506892, + "learning_rate": 3.170001276297364e-05, + "loss": 0.3933, + "step": 8087 + }, + { + "epoch": 3.0497737556561084, + "grad_norm": 0.16236693186561066, + "learning_rate": 3.168971234293102e-05, + "loss": 0.4054, + "step": 8088 + }, + { + "epoch": 3.050150829562594, + "grad_norm": 0.13896455237399402, + "learning_rate": 3.167941249878397e-05, + "loss": 0.3983, + "step": 8089 + }, + { + "epoch": 3.0505279034690798, + "grad_norm": 0.12975238739140024, + "learning_rate": 3.166911323124623e-05, + "loss": 0.3957, + "step": 8090 + }, + { + "epoch": 3.0509049773755654, + "grad_norm": 0.15083574751363577, + "learning_rate": 3.165881454103157e-05, + "loss": 0.39, + "step": 8091 + }, + { + "epoch": 3.051282051282051, + "grad_norm": 0.1703320968868614, + "learning_rate": 3.1648516428853666e-05, + "loss": 0.3823, + "step": 8092 + }, + { + "epoch": 3.0516591251885368, + "grad_norm": 0.14051355432143173, + "learning_rate": 3.163821889542615e-05, + "loss": 0.3978, + "step": 8093 + }, + { + "epoch": 3.0520361990950224, + "grad_norm": 0.18818560481349333, + "learning_rate": 3.162792194146266e-05, + "loss": 0.3959, + "step": 8094 + }, + { + "epoch": 3.052413273001508, + "grad_norm": 0.17818556845485012, + "learning_rate": 3.161762556767676e-05, + "loss": 0.3913, + "step": 8095 + }, + { + "epoch": 3.052790346907994, + "grad_norm": 0.15948792965972497, + "learning_rate": 3.160732977478195e-05, + "loss": 0.402, + "step": 8096 + }, + { + "epoch": 3.0531674208144794, + "grad_norm": 0.21040365580010598, + "learning_rate": 3.159703456349177e-05, + "loss": 0.395, + "step": 8097 + }, + { + "epoch": 3.053544494720965, + "grad_norm": 0.1434646490850979, + "learning_rate": 3.1586739934519626e-05, + "loss": 0.4012, + "step": 8098 + }, + { + "epoch": 3.053921568627451, + "grad_norm": 0.1547745248427168, + "learning_rate": 3.157644588857893e-05, + "loss": 0.3914, + "step": 8099 + }, + { + "epoch": 3.0542986425339365, + "grad_norm": 0.159932187706091, + "learning_rate": 3.156615242638308e-05, + "loss": 0.4043, + "step": 8100 + }, + { + "epoch": 3.054675716440422, + "grad_norm": 0.14816033622418234, + "learning_rate": 3.155585954864539e-05, + "loss": 0.3956, + "step": 8101 + }, + { + "epoch": 3.055052790346908, + "grad_norm": 0.13654342004817538, + "learning_rate": 3.154556725607913e-05, + "loss": 0.3988, + "step": 8102 + }, + { + "epoch": 3.0554298642533935, + "grad_norm": 0.16087724331964115, + "learning_rate": 3.1535275549397565e-05, + "loss": 0.4073, + "step": 8103 + }, + { + "epoch": 3.055806938159879, + "grad_norm": 0.16011440133413188, + "learning_rate": 3.152498442931389e-05, + "loss": 0.3917, + "step": 8104 + }, + { + "epoch": 3.056184012066365, + "grad_norm": 0.13615707590128578, + "learning_rate": 3.1514693896541284e-05, + "loss": 0.3948, + "step": 8105 + }, + { + "epoch": 3.0565610859728505, + "grad_norm": 0.16767900934522448, + "learning_rate": 3.150440395179287e-05, + "loss": 0.3971, + "step": 8106 + }, + { + "epoch": 3.056938159879336, + "grad_norm": 0.1700162421307819, + "learning_rate": 3.149411459578171e-05, + "loss": 0.4091, + "step": 8107 + }, + { + "epoch": 3.057315233785822, + "grad_norm": 0.14473713224774898, + "learning_rate": 3.148382582922088e-05, + "loss": 0.4137, + "step": 8108 + }, + { + "epoch": 3.0576923076923075, + "grad_norm": 0.19707079895417476, + "learning_rate": 3.147353765282337e-05, + "loss": 0.394, + "step": 8109 + }, + { + "epoch": 3.058069381598793, + "grad_norm": 0.15809916158598153, + "learning_rate": 3.146325006730211e-05, + "loss": 0.3936, + "step": 8110 + }, + { + "epoch": 3.058446455505279, + "grad_norm": 0.17221509329682208, + "learning_rate": 3.145296307337007e-05, + "loss": 0.4155, + "step": 8111 + }, + { + "epoch": 3.0588235294117645, + "grad_norm": 0.1769674963083081, + "learning_rate": 3.144267667174011e-05, + "loss": 0.4073, + "step": 8112 + }, + { + "epoch": 3.05920060331825, + "grad_norm": 0.13867681007509702, + "learning_rate": 3.1432390863125055e-05, + "loss": 0.3956, + "step": 8113 + }, + { + "epoch": 3.059577677224736, + "grad_norm": 0.16751050041730228, + "learning_rate": 3.142210564823772e-05, + "loss": 0.3861, + "step": 8114 + }, + { + "epoch": 3.0599547511312215, + "grad_norm": 0.1561744456712476, + "learning_rate": 3.141182102779087e-05, + "loss": 0.4106, + "step": 8115 + }, + { + "epoch": 3.060331825037707, + "grad_norm": 0.1454577154777493, + "learning_rate": 3.140153700249717e-05, + "loss": 0.4111, + "step": 8116 + }, + { + "epoch": 3.060708898944193, + "grad_norm": 0.16097618266725702, + "learning_rate": 3.139125357306935e-05, + "loss": 0.3917, + "step": 8117 + }, + { + "epoch": 3.0610859728506785, + "grad_norm": 0.15035813327785072, + "learning_rate": 3.1380970740220034e-05, + "loss": 0.4036, + "step": 8118 + }, + { + "epoch": 3.061463046757164, + "grad_norm": 0.1322592025017887, + "learning_rate": 3.137068850466177e-05, + "loss": 0.4055, + "step": 8119 + }, + { + "epoch": 3.06184012066365, + "grad_norm": 0.15258186124404094, + "learning_rate": 3.136040686710716e-05, + "loss": 0.404, + "step": 8120 + }, + { + "epoch": 3.0622171945701355, + "grad_norm": 0.1405186548768277, + "learning_rate": 3.1350125828268686e-05, + "loss": 0.4072, + "step": 8121 + }, + { + "epoch": 3.062594268476621, + "grad_norm": 0.13543164328633614, + "learning_rate": 3.133984538885882e-05, + "loss": 0.4002, + "step": 8122 + }, + { + "epoch": 3.062971342383107, + "grad_norm": 0.14358012511084586, + "learning_rate": 3.132956554958998e-05, + "loss": 0.4008, + "step": 8123 + }, + { + "epoch": 3.0633484162895925, + "grad_norm": 0.15290103128985363, + "learning_rate": 3.131928631117457e-05, + "loss": 0.391, + "step": 8124 + }, + { + "epoch": 3.063725490196078, + "grad_norm": 0.13915694513838886, + "learning_rate": 3.13090076743249e-05, + "loss": 0.4072, + "step": 8125 + }, + { + "epoch": 3.064102564102564, + "grad_norm": 0.1536683210209612, + "learning_rate": 3.129872963975331e-05, + "loss": 0.3949, + "step": 8126 + }, + { + "epoch": 3.06447963800905, + "grad_norm": 0.14216314196260865, + "learning_rate": 3.128845220817203e-05, + "loss": 0.3975, + "step": 8127 + }, + { + "epoch": 3.0648567119155357, + "grad_norm": 0.15043277404665673, + "learning_rate": 3.1278175380293277e-05, + "loss": 0.3909, + "step": 8128 + }, + { + "epoch": 3.0652337858220213, + "grad_norm": 0.14521713136807782, + "learning_rate": 3.1267899156829244e-05, + "loss": 0.4008, + "step": 8129 + }, + { + "epoch": 3.065610859728507, + "grad_norm": 0.154775335891028, + "learning_rate": 3.125762353849205e-05, + "loss": 0.402, + "step": 8130 + }, + { + "epoch": 3.0659879336349927, + "grad_norm": 0.1394564184239095, + "learning_rate": 3.124734852599379e-05, + "loss": 0.3946, + "step": 8131 + }, + { + "epoch": 3.0663650075414783, + "grad_norm": 0.13029189240212025, + "learning_rate": 3.123707412004652e-05, + "loss": 0.4014, + "step": 8132 + }, + { + "epoch": 3.066742081447964, + "grad_norm": 0.13737421637058103, + "learning_rate": 3.122680032136223e-05, + "loss": 0.4055, + "step": 8133 + }, + { + "epoch": 3.0671191553544497, + "grad_norm": 0.15087324670908678, + "learning_rate": 3.1216527130652896e-05, + "loss": 0.3965, + "step": 8134 + }, + { + "epoch": 3.0674962292609353, + "grad_norm": 0.13860696471351938, + "learning_rate": 3.120625454863046e-05, + "loss": 0.3919, + "step": 8135 + }, + { + "epoch": 3.067873303167421, + "grad_norm": 0.17003919804336595, + "learning_rate": 3.119598257600676e-05, + "loss": 0.3935, + "step": 8136 + }, + { + "epoch": 3.0682503770739067, + "grad_norm": 0.1503721242498456, + "learning_rate": 3.1185711213493664e-05, + "loss": 0.3913, + "step": 8137 + }, + { + "epoch": 3.0686274509803924, + "grad_norm": 0.15937479310978309, + "learning_rate": 3.1175440461802975e-05, + "loss": 0.3877, + "step": 8138 + }, + { + "epoch": 3.069004524886878, + "grad_norm": 0.16564014366557545, + "learning_rate": 3.1165170321646425e-05, + "loss": 0.3909, + "step": 8139 + }, + { + "epoch": 3.0693815987933637, + "grad_norm": 0.12703023529619312, + "learning_rate": 3.115490079373573e-05, + "loss": 0.3917, + "step": 8140 + }, + { + "epoch": 3.0697586726998494, + "grad_norm": 0.16276221979689728, + "learning_rate": 3.114463187878257e-05, + "loss": 0.3947, + "step": 8141 + }, + { + "epoch": 3.070135746606335, + "grad_norm": 0.1894889182919863, + "learning_rate": 3.113436357749857e-05, + "loss": 0.3921, + "step": 8142 + }, + { + "epoch": 3.0705128205128207, + "grad_norm": 0.16330651715726974, + "learning_rate": 3.1124095890595305e-05, + "loss": 0.3891, + "step": 8143 + }, + { + "epoch": 3.0708898944193064, + "grad_norm": 0.13571583526746614, + "learning_rate": 3.111382881878433e-05, + "loss": 0.402, + "step": 8144 + }, + { + "epoch": 3.071266968325792, + "grad_norm": 0.1367860345817346, + "learning_rate": 3.110356236277713e-05, + "loss": 0.3972, + "step": 8145 + }, + { + "epoch": 3.0716440422322777, + "grad_norm": 0.16552337230352646, + "learning_rate": 3.109329652328516e-05, + "loss": 0.3965, + "step": 8146 + }, + { + "epoch": 3.0720211161387634, + "grad_norm": 0.14570424595515533, + "learning_rate": 3.108303130101985e-05, + "loss": 0.3941, + "step": 8147 + }, + { + "epoch": 3.072398190045249, + "grad_norm": 0.1710353172483106, + "learning_rate": 3.107276669669257e-05, + "loss": 0.3944, + "step": 8148 + }, + { + "epoch": 3.0727752639517347, + "grad_norm": 0.12982317384344788, + "learning_rate": 3.106250271101462e-05, + "loss": 0.3911, + "step": 8149 + }, + { + "epoch": 3.0731523378582204, + "grad_norm": 0.1534306107485871, + "learning_rate": 3.1052239344697314e-05, + "loss": 0.3884, + "step": 8150 + }, + { + "epoch": 3.073529411764706, + "grad_norm": 0.17117779022128776, + "learning_rate": 3.1041976598451884e-05, + "loss": 0.3906, + "step": 8151 + }, + { + "epoch": 3.0739064856711917, + "grad_norm": 0.13191047643957446, + "learning_rate": 3.1031714472989535e-05, + "loss": 0.3946, + "step": 8152 + }, + { + "epoch": 3.0742835595776774, + "grad_norm": 0.15016395128412302, + "learning_rate": 3.102145296902141e-05, + "loss": 0.3963, + "step": 8153 + }, + { + "epoch": 3.074660633484163, + "grad_norm": 0.17945412478508474, + "learning_rate": 3.1011192087258634e-05, + "loss": 0.3924, + "step": 8154 + }, + { + "epoch": 3.0750377073906487, + "grad_norm": 0.14304505807954718, + "learning_rate": 3.100093182841227e-05, + "loss": 0.4019, + "step": 8155 + }, + { + "epoch": 3.0754147812971344, + "grad_norm": 0.1475218024982111, + "learning_rate": 3.099067219319335e-05, + "loss": 0.3946, + "step": 8156 + }, + { + "epoch": 3.07579185520362, + "grad_norm": 0.1554057257864776, + "learning_rate": 3.0980413182312866e-05, + "loss": 0.3952, + "step": 8157 + }, + { + "epoch": 3.0761689291101058, + "grad_norm": 0.15241178405852732, + "learning_rate": 3.097015479648172e-05, + "loss": 0.3904, + "step": 8158 + }, + { + "epoch": 3.0765460030165914, + "grad_norm": 0.1591841688921071, + "learning_rate": 3.095989703641086e-05, + "loss": 0.3946, + "step": 8159 + }, + { + "epoch": 3.076923076923077, + "grad_norm": 0.1721420273405441, + "learning_rate": 3.094963990281112e-05, + "loss": 0.4008, + "step": 8160 + }, + { + "epoch": 3.0773001508295628, + "grad_norm": 0.16639404641079744, + "learning_rate": 3.09393833963933e-05, + "loss": 0.3924, + "step": 8161 + }, + { + "epoch": 3.0776772247360484, + "grad_norm": 0.1315947880202407, + "learning_rate": 3.0929127517868174e-05, + "loss": 0.3935, + "step": 8162 + }, + { + "epoch": 3.078054298642534, + "grad_norm": 0.15690991257398634, + "learning_rate": 3.091887226794647e-05, + "loss": 0.393, + "step": 8163 + }, + { + "epoch": 3.0784313725490198, + "grad_norm": 0.16505490891253227, + "learning_rate": 3.090861764733885e-05, + "loss": 0.4135, + "step": 8164 + }, + { + "epoch": 3.0788084464555054, + "grad_norm": 0.1467361163401909, + "learning_rate": 3.0898363656755975e-05, + "loss": 0.4065, + "step": 8165 + }, + { + "epoch": 3.079185520361991, + "grad_norm": 0.17788793226026653, + "learning_rate": 3.088811029690843e-05, + "loss": 0.4003, + "step": 8166 + }, + { + "epoch": 3.079562594268477, + "grad_norm": 0.14537201511910353, + "learning_rate": 3.0877857568506736e-05, + "loss": 0.404, + "step": 8167 + }, + { + "epoch": 3.0799396681749625, + "grad_norm": 0.14413379254809017, + "learning_rate": 3.086760547226144e-05, + "loss": 0.3942, + "step": 8168 + }, + { + "epoch": 3.080316742081448, + "grad_norm": 0.14521127178164805, + "learning_rate": 3.0857354008882985e-05, + "loss": 0.3991, + "step": 8169 + }, + { + "epoch": 3.080693815987934, + "grad_norm": 0.16625191445211573, + "learning_rate": 3.084710317908177e-05, + "loss": 0.3917, + "step": 8170 + }, + { + "epoch": 3.0810708898944195, + "grad_norm": 0.13232984843473403, + "learning_rate": 3.0836852983568205e-05, + "loss": 0.3877, + "step": 8171 + }, + { + "epoch": 3.081447963800905, + "grad_norm": 0.17171586471857264, + "learning_rate": 3.08266034230526e-05, + "loss": 0.4107, + "step": 8172 + }, + { + "epoch": 3.081825037707391, + "grad_norm": 0.16227029041516527, + "learning_rate": 3.081635449824522e-05, + "loss": 0.3941, + "step": 8173 + }, + { + "epoch": 3.0822021116138765, + "grad_norm": 0.1442595303245897, + "learning_rate": 3.080610620985635e-05, + "loss": 0.3885, + "step": 8174 + }, + { + "epoch": 3.082579185520362, + "grad_norm": 0.16579509164000253, + "learning_rate": 3.0795858558596165e-05, + "loss": 0.4026, + "step": 8175 + }, + { + "epoch": 3.082956259426848, + "grad_norm": 0.1520592494185025, + "learning_rate": 3.07856115451748e-05, + "loss": 0.3955, + "step": 8176 + }, + { + "epoch": 3.0833333333333335, + "grad_norm": 0.16606303927140492, + "learning_rate": 3.07753651703024e-05, + "loss": 0.4041, + "step": 8177 + }, + { + "epoch": 3.083710407239819, + "grad_norm": 0.14960854920682906, + "learning_rate": 3.076511943468901e-05, + "loss": 0.3967, + "step": 8178 + }, + { + "epoch": 3.084087481146305, + "grad_norm": 0.16174699951732818, + "learning_rate": 3.075487433904463e-05, + "loss": 0.4058, + "step": 8179 + }, + { + "epoch": 3.0844645550527905, + "grad_norm": 0.1667515889347053, + "learning_rate": 3.074462988407928e-05, + "loss": 0.394, + "step": 8180 + }, + { + "epoch": 3.084841628959276, + "grad_norm": 0.16453592029548794, + "learning_rate": 3.0734386070502865e-05, + "loss": 0.3909, + "step": 8181 + }, + { + "epoch": 3.085218702865762, + "grad_norm": 0.15352984987202617, + "learning_rate": 3.072414289902527e-05, + "loss": 0.3928, + "step": 8182 + }, + { + "epoch": 3.0855957767722475, + "grad_norm": 0.15135303160916974, + "learning_rate": 3.071390037035635e-05, + "loss": 0.3898, + "step": 8183 + }, + { + "epoch": 3.085972850678733, + "grad_norm": 0.13653085823617966, + "learning_rate": 3.0703658485205904e-05, + "loss": 0.3936, + "step": 8184 + }, + { + "epoch": 3.086349924585219, + "grad_norm": 0.14298747026641315, + "learning_rate": 3.069341724428366e-05, + "loss": 0.398, + "step": 8185 + }, + { + "epoch": 3.0867269984917045, + "grad_norm": 0.15484661174077016, + "learning_rate": 3.068317664829936e-05, + "loss": 0.3904, + "step": 8186 + }, + { + "epoch": 3.08710407239819, + "grad_norm": 0.156854464217817, + "learning_rate": 3.067293669796266e-05, + "loss": 0.3847, + "step": 8187 + }, + { + "epoch": 3.087481146304676, + "grad_norm": 0.14160922799259001, + "learning_rate": 3.066269739398315e-05, + "loss": 0.403, + "step": 8188 + }, + { + "epoch": 3.0878582202111615, + "grad_norm": 0.16354968280072324, + "learning_rate": 3.065245873707045e-05, + "loss": 0.3994, + "step": 8189 + }, + { + "epoch": 3.088235294117647, + "grad_norm": 0.14663822327324205, + "learning_rate": 3.064222072793407e-05, + "loss": 0.407, + "step": 8190 + }, + { + "epoch": 3.088612368024133, + "grad_norm": 0.15145032262920904, + "learning_rate": 3.0631983367283475e-05, + "loss": 0.4001, + "step": 8191 + }, + { + "epoch": 3.0889894419306185, + "grad_norm": 0.19034135098567917, + "learning_rate": 3.062174665582814e-05, + "loss": 0.3985, + "step": 8192 + }, + { + "epoch": 3.089366515837104, + "grad_norm": 0.16276663523887214, + "learning_rate": 3.061151059427744e-05, + "loss": 0.3825, + "step": 8193 + }, + { + "epoch": 3.08974358974359, + "grad_norm": 0.16852436066183205, + "learning_rate": 3.0601275183340715e-05, + "loss": 0.4057, + "step": 8194 + }, + { + "epoch": 3.0901206636500755, + "grad_norm": 0.1422685887803471, + "learning_rate": 3.05910404237273e-05, + "loss": 0.3982, + "step": 8195 + }, + { + "epoch": 3.090497737556561, + "grad_norm": 0.1545339575312335, + "learning_rate": 3.058080631614644e-05, + "loss": 0.3938, + "step": 8196 + }, + { + "epoch": 3.090874811463047, + "grad_norm": 0.19293795034031813, + "learning_rate": 3.057057286130732e-05, + "loss": 0.3935, + "step": 8197 + }, + { + "epoch": 3.0912518853695325, + "grad_norm": 0.13898440571567586, + "learning_rate": 3.056034005991917e-05, + "loss": 0.3913, + "step": 8198 + }, + { + "epoch": 3.091628959276018, + "grad_norm": 0.18803725096014698, + "learning_rate": 3.055010791269107e-05, + "loss": 0.4, + "step": 8199 + }, + { + "epoch": 3.092006033182504, + "grad_norm": 0.22645864613547564, + "learning_rate": 3.0539876420332094e-05, + "loss": 0.3896, + "step": 8200 + }, + { + "epoch": 3.0923831070889896, + "grad_norm": 0.19038639666734972, + "learning_rate": 3.0529645583551304e-05, + "loss": 0.3996, + "step": 8201 + }, + { + "epoch": 3.0927601809954752, + "grad_norm": 0.1544846586725424, + "learning_rate": 3.051941540305768e-05, + "loss": 0.3917, + "step": 8202 + }, + { + "epoch": 3.093137254901961, + "grad_norm": 0.1249219807610815, + "learning_rate": 3.0509185879560142e-05, + "loss": 0.3989, + "step": 8203 + }, + { + "epoch": 3.0935143288084466, + "grad_norm": 0.14895118982787636, + "learning_rate": 3.049895701376762e-05, + "loss": 0.4082, + "step": 8204 + }, + { + "epoch": 3.0938914027149322, + "grad_norm": 0.14935550586940405, + "learning_rate": 3.048872880638895e-05, + "loss": 0.3983, + "step": 8205 + }, + { + "epoch": 3.094268476621418, + "grad_norm": 0.14768436160158774, + "learning_rate": 3.0478501258132924e-05, + "loss": 0.3844, + "step": 8206 + }, + { + "epoch": 3.0946455505279036, + "grad_norm": 0.1210923250627767, + "learning_rate": 3.0468274369708326e-05, + "loss": 0.3917, + "step": 8207 + }, + { + "epoch": 3.0950226244343892, + "grad_norm": 0.16817536412396414, + "learning_rate": 3.0458048141823866e-05, + "loss": 0.3945, + "step": 8208 + }, + { + "epoch": 3.095399698340875, + "grad_norm": 0.1377315006295876, + "learning_rate": 3.0447822575188186e-05, + "loss": 0.4, + "step": 8209 + }, + { + "epoch": 3.0957767722473606, + "grad_norm": 0.1571864700095919, + "learning_rate": 3.0437597670509943e-05, + "loss": 0.4034, + "step": 8210 + }, + { + "epoch": 3.0961538461538463, + "grad_norm": 0.17377351873227498, + "learning_rate": 3.0427373428497704e-05, + "loss": 0.3954, + "step": 8211 + }, + { + "epoch": 3.096530920060332, + "grad_norm": 0.1642539023336738, + "learning_rate": 3.041714984985998e-05, + "loss": 0.3954, + "step": 8212 + }, + { + "epoch": 3.0969079939668176, + "grad_norm": 0.16361041462463724, + "learning_rate": 3.040692693530528e-05, + "loss": 0.4062, + "step": 8213 + }, + { + "epoch": 3.0972850678733033, + "grad_norm": 0.14382865858186009, + "learning_rate": 3.0396704685542037e-05, + "loss": 0.3978, + "step": 8214 + }, + { + "epoch": 3.097662141779789, + "grad_norm": 0.15929185481351904, + "learning_rate": 3.038648310127863e-05, + "loss": 0.3968, + "step": 8215 + }, + { + "epoch": 3.0980392156862746, + "grad_norm": 0.17513512630001632, + "learning_rate": 3.037626218322343e-05, + "loss": 0.409, + "step": 8216 + }, + { + "epoch": 3.0984162895927603, + "grad_norm": 0.14820871279795383, + "learning_rate": 3.036604193208472e-05, + "loss": 0.3959, + "step": 8217 + }, + { + "epoch": 3.098793363499246, + "grad_norm": 0.16135569828667648, + "learning_rate": 3.0355822348570753e-05, + "loss": 0.3977, + "step": 8218 + }, + { + "epoch": 3.0991704374057316, + "grad_norm": 0.1904255385008109, + "learning_rate": 3.0345603433389752e-05, + "loss": 0.3868, + "step": 8219 + }, + { + "epoch": 3.0995475113122173, + "grad_norm": 0.19271727950112444, + "learning_rate": 3.0335385187249867e-05, + "loss": 0.4032, + "step": 8220 + }, + { + "epoch": 3.099924585218703, + "grad_norm": 0.1539482597088116, + "learning_rate": 3.0325167610859205e-05, + "loss": 0.4064, + "step": 8221 + }, + { + "epoch": 3.1003016591251886, + "grad_norm": 0.16459310173320718, + "learning_rate": 3.0314950704925864e-05, + "loss": 0.3906, + "step": 8222 + }, + { + "epoch": 3.1006787330316743, + "grad_norm": 0.14395453960415724, + "learning_rate": 3.0304734470157843e-05, + "loss": 0.404, + "step": 8223 + }, + { + "epoch": 3.10105580693816, + "grad_norm": 0.1679489288438916, + "learning_rate": 3.029451890726311e-05, + "loss": 0.3924, + "step": 8224 + }, + { + "epoch": 3.1014328808446456, + "grad_norm": 0.13217248858348046, + "learning_rate": 3.0284304016949622e-05, + "loss": 0.3888, + "step": 8225 + }, + { + "epoch": 3.1018099547511313, + "grad_norm": 0.15512549897253777, + "learning_rate": 3.0274089799925232e-05, + "loss": 0.402, + "step": 8226 + }, + { + "epoch": 3.102187028657617, + "grad_norm": 0.13207817625896082, + "learning_rate": 3.0263876256897813e-05, + "loss": 0.3892, + "step": 8227 + }, + { + "epoch": 3.1025641025641026, + "grad_norm": 0.18100045111013427, + "learning_rate": 3.0253663388575128e-05, + "loss": 0.3985, + "step": 8228 + }, + { + "epoch": 3.1029411764705883, + "grad_norm": 0.1831040937147972, + "learning_rate": 3.0243451195664914e-05, + "loss": 0.4088, + "step": 8229 + }, + { + "epoch": 3.103318250377074, + "grad_norm": 0.16139522409378912, + "learning_rate": 3.023323967887489e-05, + "loss": 0.3864, + "step": 8230 + }, + { + "epoch": 3.1036953242835597, + "grad_norm": 0.12336482664865302, + "learning_rate": 3.0223028838912695e-05, + "loss": 0.3837, + "step": 8231 + }, + { + "epoch": 3.1040723981900453, + "grad_norm": 0.1666030476317177, + "learning_rate": 3.021281867648591e-05, + "loss": 0.3943, + "step": 8232 + }, + { + "epoch": 3.104449472096531, + "grad_norm": 0.1661871663960149, + "learning_rate": 3.020260919230213e-05, + "loss": 0.3988, + "step": 8233 + }, + { + "epoch": 3.1048265460030167, + "grad_norm": 0.14620519959554704, + "learning_rate": 3.019240038706884e-05, + "loss": 0.3931, + "step": 8234 + }, + { + "epoch": 3.1052036199095023, + "grad_norm": 0.16832521084960514, + "learning_rate": 3.018219226149349e-05, + "loss": 0.4, + "step": 8235 + }, + { + "epoch": 3.105580693815988, + "grad_norm": 0.16671676664769786, + "learning_rate": 3.017198481628352e-05, + "loss": 0.3946, + "step": 8236 + }, + { + "epoch": 3.1059577677224737, + "grad_norm": 0.13012319489082916, + "learning_rate": 3.016177805214629e-05, + "loss": 0.398, + "step": 8237 + }, + { + "epoch": 3.1063348416289593, + "grad_norm": 0.1576042816008004, + "learning_rate": 3.0151571969789094e-05, + "loss": 0.3954, + "step": 8238 + }, + { + "epoch": 3.106711915535445, + "grad_norm": 0.1613356398701305, + "learning_rate": 3.0141366569919243e-05, + "loss": 0.3995, + "step": 8239 + }, + { + "epoch": 3.1070889894419307, + "grad_norm": 0.14264769304484645, + "learning_rate": 3.0131161853243946e-05, + "loss": 0.3894, + "step": 8240 + }, + { + "epoch": 3.1074660633484164, + "grad_norm": 0.17213349300322314, + "learning_rate": 3.0120957820470357e-05, + "loss": 0.3865, + "step": 8241 + }, + { + "epoch": 3.107843137254902, + "grad_norm": 0.21592574520791533, + "learning_rate": 3.0110754472305653e-05, + "loss": 0.4056, + "step": 8242 + }, + { + "epoch": 3.1082202111613877, + "grad_norm": 0.1304004696608728, + "learning_rate": 3.010055180945689e-05, + "loss": 0.3956, + "step": 8243 + }, + { + "epoch": 3.1085972850678734, + "grad_norm": 0.1846697533804546, + "learning_rate": 3.0090349832631083e-05, + "loss": 0.3963, + "step": 8244 + }, + { + "epoch": 3.108974358974359, + "grad_norm": 0.16356549830459696, + "learning_rate": 3.0080148542535267e-05, + "loss": 0.3968, + "step": 8245 + }, + { + "epoch": 3.1093514328808447, + "grad_norm": 0.16117009226054793, + "learning_rate": 3.006994793987635e-05, + "loss": 0.4052, + "step": 8246 + }, + { + "epoch": 3.1097285067873304, + "grad_norm": 0.12815840680222026, + "learning_rate": 3.0059748025361224e-05, + "loss": 0.4027, + "step": 8247 + }, + { + "epoch": 3.110105580693816, + "grad_norm": 0.14190759463245886, + "learning_rate": 3.0049548799696757e-05, + "loss": 0.3911, + "step": 8248 + }, + { + "epoch": 3.1104826546003017, + "grad_norm": 0.16129265007494442, + "learning_rate": 3.0039350263589727e-05, + "loss": 0.3938, + "step": 8249 + }, + { + "epoch": 3.1108597285067874, + "grad_norm": 0.1433671760833719, + "learning_rate": 3.0029152417746877e-05, + "loss": 0.3908, + "step": 8250 + }, + { + "epoch": 3.111236802413273, + "grad_norm": 0.13334715089570676, + "learning_rate": 3.0018955262874936e-05, + "loss": 0.4073, + "step": 8251 + }, + { + "epoch": 3.1116138763197587, + "grad_norm": 0.148590824550482, + "learning_rate": 3.0008758799680544e-05, + "loss": 0.4003, + "step": 8252 + }, + { + "epoch": 3.1119909502262444, + "grad_norm": 0.1689800139905258, + "learning_rate": 2.9998563028870288e-05, + "loss": 0.3978, + "step": 8253 + }, + { + "epoch": 3.11236802413273, + "grad_norm": 0.16513134868378088, + "learning_rate": 2.9988367951150758e-05, + "loss": 0.397, + "step": 8254 + }, + { + "epoch": 3.1127450980392157, + "grad_norm": 0.13704370494955634, + "learning_rate": 2.9978173567228447e-05, + "loss": 0.3876, + "step": 8255 + }, + { + "epoch": 3.1131221719457014, + "grad_norm": 0.19415352805792419, + "learning_rate": 2.9967979877809802e-05, + "loss": 0.4049, + "step": 8256 + }, + { + "epoch": 3.113499245852187, + "grad_norm": 0.18270780334794215, + "learning_rate": 2.9957786883601275e-05, + "loss": 0.3854, + "step": 8257 + }, + { + "epoch": 3.1138763197586727, + "grad_norm": 0.161022275941594, + "learning_rate": 2.9947594585309204e-05, + "loss": 0.3948, + "step": 8258 + }, + { + "epoch": 3.1142533936651584, + "grad_norm": 0.21351492125709362, + "learning_rate": 2.9937402983639895e-05, + "loss": 0.3928, + "step": 8259 + }, + { + "epoch": 3.114630467571644, + "grad_norm": 0.19493185200127283, + "learning_rate": 2.9927212079299655e-05, + "loss": 0.4014, + "step": 8260 + }, + { + "epoch": 3.1150075414781297, + "grad_norm": 0.17207214090166612, + "learning_rate": 2.991702187299468e-05, + "loss": 0.3887, + "step": 8261 + }, + { + "epoch": 3.1153846153846154, + "grad_norm": 0.23511480978686752, + "learning_rate": 2.9906832365431132e-05, + "loss": 0.399, + "step": 8262 + }, + { + "epoch": 3.115761689291101, + "grad_norm": 0.17959215178320653, + "learning_rate": 2.989664355731516e-05, + "loss": 0.4005, + "step": 8263 + }, + { + "epoch": 3.1161387631975868, + "grad_norm": 0.17060737366624734, + "learning_rate": 2.9886455449352828e-05, + "loss": 0.3969, + "step": 8264 + }, + { + "epoch": 3.1165158371040724, + "grad_norm": 0.19042049043973322, + "learning_rate": 2.9876268042250152e-05, + "loss": 0.3965, + "step": 8265 + }, + { + "epoch": 3.116892911010558, + "grad_norm": 0.14985559224382577, + "learning_rate": 2.9866081336713137e-05, + "loss": 0.4026, + "step": 8266 + }, + { + "epoch": 3.1172699849170438, + "grad_norm": 0.2696742490518579, + "learning_rate": 2.9855895333447692e-05, + "loss": 0.3944, + "step": 8267 + }, + { + "epoch": 3.1176470588235294, + "grad_norm": 0.4424942141861576, + "learning_rate": 2.984571003315969e-05, + "loss": 0.3956, + "step": 8268 + }, + { + "epoch": 3.118024132730015, + "grad_norm": 0.49724618792414754, + "learning_rate": 2.9835525436554985e-05, + "loss": 0.4077, + "step": 8269 + }, + { + "epoch": 3.1184012066365008, + "grad_norm": 0.3680805329276421, + "learning_rate": 2.982534154433935e-05, + "loss": 0.4005, + "step": 8270 + }, + { + "epoch": 3.1187782805429864, + "grad_norm": 0.40509885573372845, + "learning_rate": 2.9815158357218518e-05, + "loss": 0.4069, + "step": 8271 + }, + { + "epoch": 3.119155354449472, + "grad_norm": 0.35523099085769083, + "learning_rate": 2.980497587589819e-05, + "loss": 0.3992, + "step": 8272 + }, + { + "epoch": 3.119532428355958, + "grad_norm": 0.320870108006636, + "learning_rate": 2.979479410108398e-05, + "loss": 0.4007, + "step": 8273 + }, + { + "epoch": 3.1199095022624435, + "grad_norm": 0.38084657397136223, + "learning_rate": 2.978461303348149e-05, + "loss": 0.3997, + "step": 8274 + }, + { + "epoch": 3.120286576168929, + "grad_norm": 0.314288275086983, + "learning_rate": 2.9774432673796253e-05, + "loss": 0.4009, + "step": 8275 + }, + { + "epoch": 3.120663650075415, + "grad_norm": 0.29970521835809516, + "learning_rate": 2.976425302273376e-05, + "loss": 0.4008, + "step": 8276 + }, + { + "epoch": 3.1210407239819005, + "grad_norm": 0.3547895734604627, + "learning_rate": 2.975407408099946e-05, + "loss": 0.3929, + "step": 8277 + }, + { + "epoch": 3.121417797888386, + "grad_norm": 0.30712029763019644, + "learning_rate": 2.974389584929874e-05, + "loss": 0.4067, + "step": 8278 + }, + { + "epoch": 3.121794871794872, + "grad_norm": 0.3610015901538955, + "learning_rate": 2.9733718328336932e-05, + "loss": 0.3907, + "step": 8279 + }, + { + "epoch": 3.1221719457013575, + "grad_norm": 0.48589775401385404, + "learning_rate": 2.9723541518819342e-05, + "loss": 0.3963, + "step": 8280 + }, + { + "epoch": 3.122549019607843, + "grad_norm": 0.3306797408235256, + "learning_rate": 2.971336542145121e-05, + "loss": 0.3936, + "step": 8281 + }, + { + "epoch": 3.122926093514329, + "grad_norm": 0.27865152303316276, + "learning_rate": 2.970319003693773e-05, + "loss": 0.4077, + "step": 8282 + }, + { + "epoch": 3.1233031674208145, + "grad_norm": 0.4357990507628747, + "learning_rate": 2.9693015365984044e-05, + "loss": 0.4099, + "step": 8283 + }, + { + "epoch": 3.1236802413273, + "grad_norm": 0.34910125461910896, + "learning_rate": 2.9682841409295257e-05, + "loss": 0.4081, + "step": 8284 + }, + { + "epoch": 3.124057315233786, + "grad_norm": 0.2864186557376678, + "learning_rate": 2.9672668167576402e-05, + "loss": 0.4033, + "step": 8285 + }, + { + "epoch": 3.1244343891402715, + "grad_norm": 0.3948824125377195, + "learning_rate": 2.9662495641532484e-05, + "loss": 0.4058, + "step": 8286 + }, + { + "epoch": 3.124811463046757, + "grad_norm": 0.3621717475294401, + "learning_rate": 2.9652323831868456e-05, + "loss": 0.4055, + "step": 8287 + }, + { + "epoch": 3.125188536953243, + "grad_norm": 0.2793625726459883, + "learning_rate": 2.9642152739289202e-05, + "loss": 0.3933, + "step": 8288 + }, + { + "epoch": 3.1255656108597285, + "grad_norm": 0.3129636218669377, + "learning_rate": 2.963198236449957e-05, + "loss": 0.4043, + "step": 8289 + }, + { + "epoch": 3.125942684766214, + "grad_norm": 0.41066608959473144, + "learning_rate": 2.9621812708204367e-05, + "loss": 0.3843, + "step": 8290 + }, + { + "epoch": 3.1263197586727, + "grad_norm": 0.2057351420331687, + "learning_rate": 2.961164377110835e-05, + "loss": 0.3916, + "step": 8291 + }, + { + "epoch": 3.1266968325791855, + "grad_norm": 0.3494140710486656, + "learning_rate": 2.9601475553916192e-05, + "loss": 0.3947, + "step": 8292 + }, + { + "epoch": 3.127073906485671, + "grad_norm": 0.3152324169472664, + "learning_rate": 2.9591308057332554e-05, + "loss": 0.3995, + "step": 8293 + }, + { + "epoch": 3.127450980392157, + "grad_norm": 0.18713301045176195, + "learning_rate": 2.9581141282062052e-05, + "loss": 0.3947, + "step": 8294 + }, + { + "epoch": 3.1278280542986425, + "grad_norm": 0.31926886897138157, + "learning_rate": 2.9570975228809203e-05, + "loss": 0.4105, + "step": 8295 + }, + { + "epoch": 3.128205128205128, + "grad_norm": 0.25459973208044073, + "learning_rate": 2.956080989827852e-05, + "loss": 0.3914, + "step": 8296 + }, + { + "epoch": 3.128582202111614, + "grad_norm": 0.18873859207619076, + "learning_rate": 2.9550645291174458e-05, + "loss": 0.4067, + "step": 8297 + }, + { + "epoch": 3.1289592760180995, + "grad_norm": 0.28296775365181165, + "learning_rate": 2.95404814082014e-05, + "loss": 0.3933, + "step": 8298 + }, + { + "epoch": 3.129336349924585, + "grad_norm": 0.20946218484972384, + "learning_rate": 2.9530318250063703e-05, + "loss": 0.3904, + "step": 8299 + }, + { + "epoch": 3.129713423831071, + "grad_norm": 0.1954972852690315, + "learning_rate": 2.952015581746568e-05, + "loss": 0.3901, + "step": 8300 + }, + { + "epoch": 3.1300904977375565, + "grad_norm": 0.21764532285190416, + "learning_rate": 2.9509994111111546e-05, + "loss": 0.3996, + "step": 8301 + }, + { + "epoch": 3.130467571644042, + "grad_norm": 0.19527973421791356, + "learning_rate": 2.9499833131705524e-05, + "loss": 0.3865, + "step": 8302 + }, + { + "epoch": 3.130844645550528, + "grad_norm": 0.1808650512133354, + "learning_rate": 2.948967287995175e-05, + "loss": 0.3885, + "step": 8303 + }, + { + "epoch": 3.1312217194570136, + "grad_norm": 0.18855691248547995, + "learning_rate": 2.9479513356554325e-05, + "loss": 0.3937, + "step": 8304 + }, + { + "epoch": 3.1315987933634992, + "grad_norm": 0.1685440960513858, + "learning_rate": 2.946935456221729e-05, + "loss": 0.3914, + "step": 8305 + }, + { + "epoch": 3.131975867269985, + "grad_norm": 0.14855786012849415, + "learning_rate": 2.9459196497644647e-05, + "loss": 0.3913, + "step": 8306 + }, + { + "epoch": 3.1323529411764706, + "grad_norm": 0.16770601395851975, + "learning_rate": 2.944903916354032e-05, + "loss": 0.3955, + "step": 8307 + }, + { + "epoch": 3.1327300150829562, + "grad_norm": 0.14684922434453124, + "learning_rate": 2.9438882560608246e-05, + "loss": 0.406, + "step": 8308 + }, + { + "epoch": 3.133107088989442, + "grad_norm": 0.16170204734828908, + "learning_rate": 2.9428726689552236e-05, + "loss": 0.4003, + "step": 8309 + }, + { + "epoch": 3.1334841628959276, + "grad_norm": 0.1716374143845452, + "learning_rate": 2.9418571551076072e-05, + "loss": 0.3918, + "step": 8310 + }, + { + "epoch": 3.1338612368024132, + "grad_norm": 0.19020888265879574, + "learning_rate": 2.940841714588353e-05, + "loss": 0.3892, + "step": 8311 + }, + { + "epoch": 3.134238310708899, + "grad_norm": 0.1704395255710742, + "learning_rate": 2.9398263474678292e-05, + "loss": 0.4008, + "step": 8312 + }, + { + "epoch": 3.1346153846153846, + "grad_norm": 0.17995089648357446, + "learning_rate": 2.9388110538163974e-05, + "loss": 0.3954, + "step": 8313 + }, + { + "epoch": 3.1349924585218703, + "grad_norm": 0.1864737656974193, + "learning_rate": 2.937795833704419e-05, + "loss": 0.3972, + "step": 8314 + }, + { + "epoch": 3.135369532428356, + "grad_norm": 0.14355256991269402, + "learning_rate": 2.936780687202248e-05, + "loss": 0.4022, + "step": 8315 + }, + { + "epoch": 3.1357466063348416, + "grad_norm": 0.1505609285845415, + "learning_rate": 2.9357656143802303e-05, + "loss": 0.3896, + "step": 8316 + }, + { + "epoch": 3.1361236802413273, + "grad_norm": 0.17869138885075703, + "learning_rate": 2.9347506153087134e-05, + "loss": 0.3809, + "step": 8317 + }, + { + "epoch": 3.136500754147813, + "grad_norm": 0.15069148809166333, + "learning_rate": 2.9337356900580335e-05, + "loss": 0.3903, + "step": 8318 + }, + { + "epoch": 3.1368778280542986, + "grad_norm": 0.18918974015115364, + "learning_rate": 2.9327208386985235e-05, + "loss": 0.4052, + "step": 8319 + }, + { + "epoch": 3.1372549019607843, + "grad_norm": 0.21315171888383425, + "learning_rate": 2.931706061300514e-05, + "loss": 0.4038, + "step": 8320 + }, + { + "epoch": 3.13763197586727, + "grad_norm": 0.15925316863145517, + "learning_rate": 2.9306913579343273e-05, + "loss": 0.3852, + "step": 8321 + }, + { + "epoch": 3.1380090497737556, + "grad_norm": 0.16672278147599903, + "learning_rate": 2.9296767286702787e-05, + "loss": 0.4003, + "step": 8322 + }, + { + "epoch": 3.1383861236802413, + "grad_norm": 0.19097864186792496, + "learning_rate": 2.9286621735786854e-05, + "loss": 0.3927, + "step": 8323 + }, + { + "epoch": 3.138763197586727, + "grad_norm": 0.16430846078368241, + "learning_rate": 2.9276476927298532e-05, + "loss": 0.39, + "step": 8324 + }, + { + "epoch": 3.1391402714932126, + "grad_norm": 0.1553326159467606, + "learning_rate": 2.926633286194083e-05, + "loss": 0.3952, + "step": 8325 + }, + { + "epoch": 3.1395173453996983, + "grad_norm": 0.14785588122873702, + "learning_rate": 2.9256189540416757e-05, + "loss": 0.3972, + "step": 8326 + }, + { + "epoch": 3.139894419306184, + "grad_norm": 0.13364604885503087, + "learning_rate": 2.9246046963429216e-05, + "loss": 0.4129, + "step": 8327 + }, + { + "epoch": 3.1402714932126696, + "grad_norm": 0.14854254138374945, + "learning_rate": 2.9235905131681073e-05, + "loss": 0.3921, + "step": 8328 + }, + { + "epoch": 3.1406485671191553, + "grad_norm": 0.16459756576687914, + "learning_rate": 2.9225764045875168e-05, + "loss": 0.4066, + "step": 8329 + }, + { + "epoch": 3.141025641025641, + "grad_norm": 0.1775227133082571, + "learning_rate": 2.921562370671426e-05, + "loss": 0.3906, + "step": 8330 + }, + { + "epoch": 3.1414027149321266, + "grad_norm": 0.13390027209144975, + "learning_rate": 2.9205484114901046e-05, + "loss": 0.4088, + "step": 8331 + }, + { + "epoch": 3.1417797888386123, + "grad_norm": 0.16746304689225264, + "learning_rate": 2.9195345271138222e-05, + "loss": 0.404, + "step": 8332 + }, + { + "epoch": 3.142156862745098, + "grad_norm": 0.16415916522628898, + "learning_rate": 2.9185207176128388e-05, + "loss": 0.4074, + "step": 8333 + }, + { + "epoch": 3.1425339366515836, + "grad_norm": 0.1598817034807858, + "learning_rate": 2.917506983057409e-05, + "loss": 0.3951, + "step": 8334 + }, + { + "epoch": 3.1429110105580693, + "grad_norm": 0.1784638932247981, + "learning_rate": 2.9164933235177867e-05, + "loss": 0.3826, + "step": 8335 + }, + { + "epoch": 3.143288084464555, + "grad_norm": 0.1410536324626182, + "learning_rate": 2.915479739064216e-05, + "loss": 0.3893, + "step": 8336 + }, + { + "epoch": 3.1436651583710407, + "grad_norm": 0.14886007419593458, + "learning_rate": 2.9144662297669355e-05, + "loss": 0.4052, + "step": 8337 + }, + { + "epoch": 3.1440422322775263, + "grad_norm": 0.12612610676870184, + "learning_rate": 2.9134527956961843e-05, + "loss": 0.3871, + "step": 8338 + }, + { + "epoch": 3.144419306184012, + "grad_norm": 0.15929589808414038, + "learning_rate": 2.91243943692219e-05, + "loss": 0.3894, + "step": 8339 + }, + { + "epoch": 3.1447963800904977, + "grad_norm": 0.12774805446168153, + "learning_rate": 2.9114261535151774e-05, + "loss": 0.3902, + "step": 8340 + }, + { + "epoch": 3.1451734539969833, + "grad_norm": 0.13798433579262287, + "learning_rate": 2.9104129455453675e-05, + "loss": 0.3885, + "step": 8341 + }, + { + "epoch": 3.145550527903469, + "grad_norm": 0.14042990866304, + "learning_rate": 2.9093998130829752e-05, + "loss": 0.3817, + "step": 8342 + }, + { + "epoch": 3.1459276018099547, + "grad_norm": 0.13086756212799144, + "learning_rate": 2.908386756198206e-05, + "loss": 0.3865, + "step": 8343 + }, + { + "epoch": 3.1463046757164403, + "grad_norm": 0.14295506410004655, + "learning_rate": 2.9073737749612676e-05, + "loss": 0.3902, + "step": 8344 + }, + { + "epoch": 3.146681749622926, + "grad_norm": 0.13245055803386774, + "learning_rate": 2.9063608694423586e-05, + "loss": 0.4012, + "step": 8345 + }, + { + "epoch": 3.1470588235294117, + "grad_norm": 0.15537762179835143, + "learning_rate": 2.905348039711669e-05, + "loss": 0.4052, + "step": 8346 + }, + { + "epoch": 3.1474358974358974, + "grad_norm": 0.15988248608488248, + "learning_rate": 2.90433528583939e-05, + "loss": 0.3921, + "step": 8347 + }, + { + "epoch": 3.147812971342383, + "grad_norm": 0.1286035238272328, + "learning_rate": 2.9033226078957053e-05, + "loss": 0.3983, + "step": 8348 + }, + { + "epoch": 3.1481900452488687, + "grad_norm": 0.13113315988843632, + "learning_rate": 2.902310005950789e-05, + "loss": 0.4034, + "step": 8349 + }, + { + "epoch": 3.1485671191553544, + "grad_norm": 0.12379169065745156, + "learning_rate": 2.901297480074817e-05, + "loss": 0.3975, + "step": 8350 + }, + { + "epoch": 3.14894419306184, + "grad_norm": 0.12629963819169812, + "learning_rate": 2.900285030337954e-05, + "loss": 0.3986, + "step": 8351 + }, + { + "epoch": 3.1493212669683257, + "grad_norm": 0.12539784843477414, + "learning_rate": 2.8992726568103645e-05, + "loss": 0.3938, + "step": 8352 + }, + { + "epoch": 3.1496983408748114, + "grad_norm": 0.13090960910789196, + "learning_rate": 2.898260359562203e-05, + "loss": 0.3904, + "step": 8353 + }, + { + "epoch": 3.150075414781297, + "grad_norm": 0.1383571082503038, + "learning_rate": 2.8972481386636197e-05, + "loss": 0.411, + "step": 8354 + }, + { + "epoch": 3.1504524886877827, + "grad_norm": 0.12163320167672428, + "learning_rate": 2.8962359941847644e-05, + "loss": 0.3881, + "step": 8355 + }, + { + "epoch": 3.1508295625942684, + "grad_norm": 0.1312106791175589, + "learning_rate": 2.8952239261957757e-05, + "loss": 0.3928, + "step": 8356 + }, + { + "epoch": 3.151206636500754, + "grad_norm": 0.12713646220973662, + "learning_rate": 2.8942119347667872e-05, + "loss": 0.392, + "step": 8357 + }, + { + "epoch": 3.1515837104072397, + "grad_norm": 0.1572458998679767, + "learning_rate": 2.8932000199679317e-05, + "loss": 0.3963, + "step": 8358 + }, + { + "epoch": 3.1519607843137254, + "grad_norm": 0.131232656947089, + "learning_rate": 2.892188181869334e-05, + "loss": 0.3851, + "step": 8359 + }, + { + "epoch": 3.152337858220211, + "grad_norm": 0.13426939401769858, + "learning_rate": 2.891176420541111e-05, + "loss": 0.3925, + "step": 8360 + }, + { + "epoch": 3.1527149321266967, + "grad_norm": 0.14624979133495544, + "learning_rate": 2.8901647360533797e-05, + "loss": 0.3908, + "step": 8361 + }, + { + "epoch": 3.1530920060331824, + "grad_norm": 0.14362794185523292, + "learning_rate": 2.889153128476248e-05, + "loss": 0.4079, + "step": 8362 + }, + { + "epoch": 3.153469079939668, + "grad_norm": 0.1311048646417151, + "learning_rate": 2.8881415978798174e-05, + "loss": 0.4025, + "step": 8363 + }, + { + "epoch": 3.1538461538461537, + "grad_norm": 0.13451255197239037, + "learning_rate": 2.88713014433419e-05, + "loss": 0.4022, + "step": 8364 + }, + { + "epoch": 3.1542232277526394, + "grad_norm": 0.1231914769032929, + "learning_rate": 2.886118767909456e-05, + "loss": 0.4022, + "step": 8365 + }, + { + "epoch": 3.154600301659125, + "grad_norm": 0.11759675446959042, + "learning_rate": 2.885107468675702e-05, + "loss": 0.3818, + "step": 8366 + }, + { + "epoch": 3.1549773755656108, + "grad_norm": 0.12955211694438848, + "learning_rate": 2.8840962467030125e-05, + "loss": 0.3765, + "step": 8367 + }, + { + "epoch": 3.1553544494720964, + "grad_norm": 0.13105676958939372, + "learning_rate": 2.883085102061464e-05, + "loss": 0.3876, + "step": 8368 + }, + { + "epoch": 3.155731523378582, + "grad_norm": 0.1441445109467752, + "learning_rate": 2.8820740348211253e-05, + "loss": 0.4, + "step": 8369 + }, + { + "epoch": 3.1561085972850678, + "grad_norm": 0.14093311462664743, + "learning_rate": 2.881063045052066e-05, + "loss": 0.3939, + "step": 8370 + }, + { + "epoch": 3.1564856711915534, + "grad_norm": 0.12469479533560027, + "learning_rate": 2.8800521328243455e-05, + "loss": 0.3965, + "step": 8371 + }, + { + "epoch": 3.156862745098039, + "grad_norm": 0.14504862455055764, + "learning_rate": 2.8790412982080168e-05, + "loss": 0.3904, + "step": 8372 + }, + { + "epoch": 3.1572398190045248, + "grad_norm": 0.1388980431911812, + "learning_rate": 2.8780305412731335e-05, + "loss": 0.399, + "step": 8373 + }, + { + "epoch": 3.1576168929110104, + "grad_norm": 0.11524931546879744, + "learning_rate": 2.8770198620897386e-05, + "loss": 0.3916, + "step": 8374 + }, + { + "epoch": 3.157993966817496, + "grad_norm": 0.12516197228712012, + "learning_rate": 2.8760092607278697e-05, + "loss": 0.3927, + "step": 8375 + }, + { + "epoch": 3.158371040723982, + "grad_norm": 0.1433073553728365, + "learning_rate": 2.8749987372575633e-05, + "loss": 0.3867, + "step": 8376 + }, + { + "epoch": 3.1587481146304675, + "grad_norm": 0.11302127140673013, + "learning_rate": 2.8739882917488473e-05, + "loss": 0.3946, + "step": 8377 + }, + { + "epoch": 3.159125188536953, + "grad_norm": 0.13408521069501736, + "learning_rate": 2.8729779242717416e-05, + "loss": 0.3981, + "step": 8378 + }, + { + "epoch": 3.159502262443439, + "grad_norm": 0.14777365484593716, + "learning_rate": 2.871967634896268e-05, + "loss": 0.3932, + "step": 8379 + }, + { + "epoch": 3.1598793363499245, + "grad_norm": 0.12119998269357408, + "learning_rate": 2.870957423692436e-05, + "loss": 0.4016, + "step": 8380 + }, + { + "epoch": 3.16025641025641, + "grad_norm": 0.1209712145673824, + "learning_rate": 2.8699472907302525e-05, + "loss": 0.3881, + "step": 8381 + }, + { + "epoch": 3.160633484162896, + "grad_norm": 0.1324152306713038, + "learning_rate": 2.8689372360797204e-05, + "loss": 0.3966, + "step": 8382 + }, + { + "epoch": 3.1610105580693815, + "grad_norm": 0.13528798733704972, + "learning_rate": 2.8679272598108346e-05, + "loss": 0.3971, + "step": 8383 + }, + { + "epoch": 3.161387631975867, + "grad_norm": 0.16531707783496158, + "learning_rate": 2.8669173619935848e-05, + "loss": 0.3941, + "step": 8384 + }, + { + "epoch": 3.161764705882353, + "grad_norm": 0.15205797984947617, + "learning_rate": 2.865907542697957e-05, + "loss": 0.3879, + "step": 8385 + }, + { + "epoch": 3.1621417797888385, + "grad_norm": 0.12487365008890673, + "learning_rate": 2.8648978019939314e-05, + "loss": 0.3927, + "step": 8386 + }, + { + "epoch": 3.162518853695324, + "grad_norm": 0.16542855130620768, + "learning_rate": 2.863888139951479e-05, + "loss": 0.4031, + "step": 8387 + }, + { + "epoch": 3.16289592760181, + "grad_norm": 0.13686111258688496, + "learning_rate": 2.8628785566405733e-05, + "loss": 0.39, + "step": 8388 + }, + { + "epoch": 3.1632730015082955, + "grad_norm": 0.13332691882805423, + "learning_rate": 2.8618690521311747e-05, + "loss": 0.4089, + "step": 8389 + }, + { + "epoch": 3.163650075414781, + "grad_norm": 0.1716966306739438, + "learning_rate": 2.8608596264932392e-05, + "loss": 0.3818, + "step": 8390 + }, + { + "epoch": 3.164027149321267, + "grad_norm": 0.18686162587778757, + "learning_rate": 2.8598502797967235e-05, + "loss": 0.3968, + "step": 8391 + }, + { + "epoch": 3.1644042232277525, + "grad_norm": 0.1196447714393267, + "learning_rate": 2.8588410121115716e-05, + "loss": 0.3997, + "step": 8392 + }, + { + "epoch": 3.164781297134238, + "grad_norm": 0.1327023875975384, + "learning_rate": 2.857831823507724e-05, + "loss": 0.3967, + "step": 8393 + }, + { + "epoch": 3.165158371040724, + "grad_norm": 0.15743988104122303, + "learning_rate": 2.8568227140551195e-05, + "loss": 0.3877, + "step": 8394 + }, + { + "epoch": 3.1655354449472095, + "grad_norm": 0.15218095369778056, + "learning_rate": 2.8558136838236868e-05, + "loss": 0.3859, + "step": 8395 + }, + { + "epoch": 3.165912518853695, + "grad_norm": 0.1454545002275328, + "learning_rate": 2.8548047328833495e-05, + "loss": 0.3955, + "step": 8396 + }, + { + "epoch": 3.166289592760181, + "grad_norm": 0.12334578207296629, + "learning_rate": 2.85379586130403e-05, + "loss": 0.3977, + "step": 8397 + }, + { + "epoch": 3.1666666666666665, + "grad_norm": 0.1710400933757646, + "learning_rate": 2.8527870691556404e-05, + "loss": 0.3809, + "step": 8398 + }, + { + "epoch": 3.167043740573152, + "grad_norm": 0.16923306532301177, + "learning_rate": 2.8517783565080877e-05, + "loss": 0.4042, + "step": 8399 + }, + { + "epoch": 3.167420814479638, + "grad_norm": 0.14976018452220563, + "learning_rate": 2.8507697234312783e-05, + "loss": 0.3938, + "step": 8400 + }, + { + "epoch": 3.1677978883861235, + "grad_norm": 0.11891579418926015, + "learning_rate": 2.849761169995107e-05, + "loss": 0.3966, + "step": 8401 + }, + { + "epoch": 3.168174962292609, + "grad_norm": 0.1503092818322212, + "learning_rate": 2.8487526962694655e-05, + "loss": 0.4079, + "step": 8402 + }, + { + "epoch": 3.168552036199095, + "grad_norm": 0.17136811838376242, + "learning_rate": 2.847744302324243e-05, + "loss": 0.3997, + "step": 8403 + }, + { + "epoch": 3.1689291101055805, + "grad_norm": 0.14578431642224157, + "learning_rate": 2.846735988229317e-05, + "loss": 0.3998, + "step": 8404 + }, + { + "epoch": 3.169306184012066, + "grad_norm": 0.12924522162928284, + "learning_rate": 2.8457277540545632e-05, + "loss": 0.4077, + "step": 8405 + }, + { + "epoch": 3.169683257918552, + "grad_norm": 0.16507706696792487, + "learning_rate": 2.844719599869854e-05, + "loss": 0.3918, + "step": 8406 + }, + { + "epoch": 3.1700603318250375, + "grad_norm": 0.180375531484722, + "learning_rate": 2.8437115257450515e-05, + "loss": 0.4006, + "step": 8407 + }, + { + "epoch": 3.170437405731523, + "grad_norm": 0.13400574200830628, + "learning_rate": 2.8427035317500132e-05, + "loss": 0.3963, + "step": 8408 + }, + { + "epoch": 3.170814479638009, + "grad_norm": 0.1322850931267744, + "learning_rate": 2.841695617954595e-05, + "loss": 0.392, + "step": 8409 + }, + { + "epoch": 3.1711915535444946, + "grad_norm": 0.1470987188000994, + "learning_rate": 2.840687784428643e-05, + "loss": 0.3996, + "step": 8410 + }, + { + "epoch": 3.1715686274509802, + "grad_norm": 0.13962721062208688, + "learning_rate": 2.839680031241998e-05, + "loss": 0.3995, + "step": 8411 + }, + { + "epoch": 3.171945701357466, + "grad_norm": 0.1171234698047225, + "learning_rate": 2.838672358464499e-05, + "loss": 0.3912, + "step": 8412 + }, + { + "epoch": 3.1723227752639516, + "grad_norm": 0.12720831623363565, + "learning_rate": 2.8376647661659757e-05, + "loss": 0.3886, + "step": 8413 + }, + { + "epoch": 3.1726998491704372, + "grad_norm": 0.11909594429576235, + "learning_rate": 2.8366572544162518e-05, + "loss": 0.3967, + "step": 8414 + }, + { + "epoch": 3.173076923076923, + "grad_norm": 0.12377436743220745, + "learning_rate": 2.8356498232851488e-05, + "loss": 0.3943, + "step": 8415 + }, + { + "epoch": 3.1734539969834086, + "grad_norm": 0.1522580151107673, + "learning_rate": 2.834642472842481e-05, + "loss": 0.4074, + "step": 8416 + }, + { + "epoch": 3.1738310708898942, + "grad_norm": 0.14108811453832248, + "learning_rate": 2.8336352031580543e-05, + "loss": 0.3915, + "step": 8417 + }, + { + "epoch": 3.17420814479638, + "grad_norm": 0.12149000389019861, + "learning_rate": 2.8326280143016755e-05, + "loss": 0.3891, + "step": 8418 + }, + { + "epoch": 3.1745852187028656, + "grad_norm": 0.1513729006204785, + "learning_rate": 2.8316209063431387e-05, + "loss": 0.388, + "step": 8419 + }, + { + "epoch": 3.1749622926093513, + "grad_norm": 0.12437951272686595, + "learning_rate": 2.830613879352237e-05, + "loss": 0.4001, + "step": 8420 + }, + { + "epoch": 3.175339366515837, + "grad_norm": 0.14307092643424452, + "learning_rate": 2.8296069333987568e-05, + "loss": 0.3963, + "step": 8421 + }, + { + "epoch": 3.1757164404223226, + "grad_norm": 0.14546017984220466, + "learning_rate": 2.828600068552477e-05, + "loss": 0.3918, + "step": 8422 + }, + { + "epoch": 3.1760935143288083, + "grad_norm": 0.15482149629404893, + "learning_rate": 2.8275932848831736e-05, + "loss": 0.3925, + "step": 8423 + }, + { + "epoch": 3.176470588235294, + "grad_norm": 0.1283747706317444, + "learning_rate": 2.8265865824606165e-05, + "loss": 0.4001, + "step": 8424 + }, + { + "epoch": 3.1768476621417796, + "grad_norm": 0.13724617522354027, + "learning_rate": 2.8255799613545675e-05, + "loss": 0.3967, + "step": 8425 + }, + { + "epoch": 3.1772247360482653, + "grad_norm": 0.1464244680502675, + "learning_rate": 2.8245734216347857e-05, + "loss": 0.3939, + "step": 8426 + }, + { + "epoch": 3.177601809954751, + "grad_norm": 0.1407790485981763, + "learning_rate": 2.8235669633710237e-05, + "loss": 0.4084, + "step": 8427 + }, + { + "epoch": 3.1779788838612366, + "grad_norm": 0.153121606089701, + "learning_rate": 2.822560586633027e-05, + "loss": 0.4122, + "step": 8428 + }, + { + "epoch": 3.1783559577677223, + "grad_norm": 0.1315968350392767, + "learning_rate": 2.821554291490537e-05, + "loss": 0.4039, + "step": 8429 + }, + { + "epoch": 3.178733031674208, + "grad_norm": 0.138361738716027, + "learning_rate": 2.8205480780132908e-05, + "loss": 0.397, + "step": 8430 + }, + { + "epoch": 3.1791101055806936, + "grad_norm": 0.1601309478139749, + "learning_rate": 2.819541946271015e-05, + "loss": 0.3937, + "step": 8431 + }, + { + "epoch": 3.1794871794871793, + "grad_norm": 0.13473313853773278, + "learning_rate": 2.818535896333436e-05, + "loss": 0.4034, + "step": 8432 + }, + { + "epoch": 3.179864253393665, + "grad_norm": 0.1363474079906072, + "learning_rate": 2.817529928270272e-05, + "loss": 0.4077, + "step": 8433 + }, + { + "epoch": 3.1802413273001506, + "grad_norm": 0.1525069745132617, + "learning_rate": 2.816524042151234e-05, + "loss": 0.4067, + "step": 8434 + }, + { + "epoch": 3.1806184012066363, + "grad_norm": 0.1438878317452361, + "learning_rate": 2.8155182380460303e-05, + "loss": 0.4058, + "step": 8435 + }, + { + "epoch": 3.180995475113122, + "grad_norm": 0.15109889698331372, + "learning_rate": 2.814512516024363e-05, + "loss": 0.3893, + "step": 8436 + }, + { + "epoch": 3.1813725490196076, + "grad_norm": 0.12346889902383143, + "learning_rate": 2.8135068761559256e-05, + "loss": 0.4003, + "step": 8437 + }, + { + "epoch": 3.1817496229260933, + "grad_norm": 0.15774971256896583, + "learning_rate": 2.8125013185104097e-05, + "loss": 0.3993, + "step": 8438 + }, + { + "epoch": 3.182126696832579, + "grad_norm": 0.15419330649652907, + "learning_rate": 2.8114958431574985e-05, + "loss": 0.3919, + "step": 8439 + }, + { + "epoch": 3.1825037707390647, + "grad_norm": 0.13053518104295805, + "learning_rate": 2.8104904501668728e-05, + "loss": 0.4087, + "step": 8440 + }, + { + "epoch": 3.1828808446455503, + "grad_norm": 0.18578960687488386, + "learning_rate": 2.8094851396082025e-05, + "loss": 0.4026, + "step": 8441 + }, + { + "epoch": 3.183257918552036, + "grad_norm": 0.19086163316165575, + "learning_rate": 2.8084799115511563e-05, + "loss": 0.3996, + "step": 8442 + }, + { + "epoch": 3.183634992458522, + "grad_norm": 0.12460201040028127, + "learning_rate": 2.807474766065396e-05, + "loss": 0.391, + "step": 8443 + }, + { + "epoch": 3.1840120663650078, + "grad_norm": 0.2113895656708403, + "learning_rate": 2.8064697032205762e-05, + "loss": 0.4045, + "step": 8444 + }, + { + "epoch": 3.1843891402714934, + "grad_norm": 0.19898720678573947, + "learning_rate": 2.8054647230863464e-05, + "loss": 0.394, + "step": 8445 + }, + { + "epoch": 3.184766214177979, + "grad_norm": 0.1433376065866065, + "learning_rate": 2.8044598257323534e-05, + "loss": 0.4024, + "step": 8446 + }, + { + "epoch": 3.185143288084465, + "grad_norm": 0.1407542049483681, + "learning_rate": 2.8034550112282333e-05, + "loss": 0.4073, + "step": 8447 + }, + { + "epoch": 3.1855203619909505, + "grad_norm": 0.14577705225298232, + "learning_rate": 2.802450279643619e-05, + "loss": 0.3949, + "step": 8448 + }, + { + "epoch": 3.185897435897436, + "grad_norm": 0.17089061940346315, + "learning_rate": 2.801445631048139e-05, + "loss": 0.3851, + "step": 8449 + }, + { + "epoch": 3.186274509803922, + "grad_norm": 0.1483795532282127, + "learning_rate": 2.8004410655114133e-05, + "loss": 0.392, + "step": 8450 + }, + { + "epoch": 3.1866515837104075, + "grad_norm": 0.14749907023553877, + "learning_rate": 2.7994365831030572e-05, + "loss": 0.3938, + "step": 8451 + }, + { + "epoch": 3.187028657616893, + "grad_norm": 0.14008564578613858, + "learning_rate": 2.7984321838926818e-05, + "loss": 0.399, + "step": 8452 + }, + { + "epoch": 3.187405731523379, + "grad_norm": 0.1350859359909763, + "learning_rate": 2.7974278679498898e-05, + "loss": 0.3992, + "step": 8453 + }, + { + "epoch": 3.1877828054298645, + "grad_norm": 0.141983901241961, + "learning_rate": 2.79642363534428e-05, + "loss": 0.3976, + "step": 8454 + }, + { + "epoch": 3.18815987933635, + "grad_norm": 0.13212378781757342, + "learning_rate": 2.795419486145445e-05, + "loss": 0.382, + "step": 8455 + }, + { + "epoch": 3.188536953242836, + "grad_norm": 0.1631294988865846, + "learning_rate": 2.7944154204229703e-05, + "loss": 0.4066, + "step": 8456 + }, + { + "epoch": 3.1889140271493215, + "grad_norm": 0.14895594718920482, + "learning_rate": 2.7934114382464384e-05, + "loss": 0.3979, + "step": 8457 + }, + { + "epoch": 3.189291101055807, + "grad_norm": 0.13584475383857514, + "learning_rate": 2.792407539685424e-05, + "loss": 0.3955, + "step": 8458 + }, + { + "epoch": 3.189668174962293, + "grad_norm": 0.18445731122497847, + "learning_rate": 2.7914037248094943e-05, + "loss": 0.412, + "step": 8459 + }, + { + "epoch": 3.1900452488687785, + "grad_norm": 0.1823414315889785, + "learning_rate": 2.7903999936882163e-05, + "loss": 0.3968, + "step": 8460 + }, + { + "epoch": 3.190422322775264, + "grad_norm": 0.12389460676833772, + "learning_rate": 2.7893963463911454e-05, + "loss": 0.387, + "step": 8461 + }, + { + "epoch": 3.19079939668175, + "grad_norm": 0.17772875687087633, + "learning_rate": 2.788392782987832e-05, + "loss": 0.3835, + "step": 8462 + }, + { + "epoch": 3.1911764705882355, + "grad_norm": 0.12523504910181096, + "learning_rate": 2.7873893035478265e-05, + "loss": 0.4026, + "step": 8463 + }, + { + "epoch": 3.191553544494721, + "grad_norm": 0.17387486689831896, + "learning_rate": 2.786385908140666e-05, + "loss": 0.3996, + "step": 8464 + }, + { + "epoch": 3.191930618401207, + "grad_norm": 0.14546241571039606, + "learning_rate": 2.7853825968358842e-05, + "loss": 0.393, + "step": 8465 + }, + { + "epoch": 3.1923076923076925, + "grad_norm": 0.13924102724424434, + "learning_rate": 2.7843793697030128e-05, + "loss": 0.3891, + "step": 8466 + }, + { + "epoch": 3.192684766214178, + "grad_norm": 0.156368634190254, + "learning_rate": 2.783376226811573e-05, + "loss": 0.4011, + "step": 8467 + }, + { + "epoch": 3.193061840120664, + "grad_norm": 0.13617192393350178, + "learning_rate": 2.7823731682310796e-05, + "loss": 0.3951, + "step": 8468 + }, + { + "epoch": 3.1934389140271495, + "grad_norm": 0.1614163187894352, + "learning_rate": 2.7813701940310472e-05, + "loss": 0.4062, + "step": 8469 + }, + { + "epoch": 3.193815987933635, + "grad_norm": 0.1503792656439362, + "learning_rate": 2.7803673042809796e-05, + "loss": 0.3932, + "step": 8470 + }, + { + "epoch": 3.194193061840121, + "grad_norm": 0.15717187175152184, + "learning_rate": 2.779364499050375e-05, + "loss": 0.4065, + "step": 8471 + }, + { + "epoch": 3.1945701357466065, + "grad_norm": 0.17884768589070962, + "learning_rate": 2.7783617784087282e-05, + "loss": 0.3965, + "step": 8472 + }, + { + "epoch": 3.194947209653092, + "grad_norm": 0.13658618591654673, + "learning_rate": 2.7773591424255273e-05, + "loss": 0.3883, + "step": 8473 + }, + { + "epoch": 3.195324283559578, + "grad_norm": 0.15173423678801465, + "learning_rate": 2.7763565911702514e-05, + "loss": 0.3923, + "step": 8474 + }, + { + "epoch": 3.1957013574660635, + "grad_norm": 0.16917939695398687, + "learning_rate": 2.7753541247123804e-05, + "loss": 0.397, + "step": 8475 + }, + { + "epoch": 3.196078431372549, + "grad_norm": 0.13357248702557448, + "learning_rate": 2.774351743121381e-05, + "loss": 0.4006, + "step": 8476 + }, + { + "epoch": 3.196455505279035, + "grad_norm": 0.14078626422878945, + "learning_rate": 2.7733494464667196e-05, + "loss": 0.3836, + "step": 8477 + }, + { + "epoch": 3.1968325791855206, + "grad_norm": 0.15761356078731847, + "learning_rate": 2.7723472348178537e-05, + "loss": 0.3913, + "step": 8478 + }, + { + "epoch": 3.197209653092006, + "grad_norm": 0.14702238672238396, + "learning_rate": 2.7713451082442334e-05, + "loss": 0.3983, + "step": 8479 + }, + { + "epoch": 3.197586726998492, + "grad_norm": 0.14047793629560215, + "learning_rate": 2.7703430668153096e-05, + "loss": 0.3987, + "step": 8480 + }, + { + "epoch": 3.1979638009049776, + "grad_norm": 0.12844916080149393, + "learning_rate": 2.769341110600521e-05, + "loss": 0.3887, + "step": 8481 + }, + { + "epoch": 3.1983408748114632, + "grad_norm": 0.12351966378158526, + "learning_rate": 2.7683392396692988e-05, + "loss": 0.3901, + "step": 8482 + }, + { + "epoch": 3.198717948717949, + "grad_norm": 0.13520808267277729, + "learning_rate": 2.7673374540910775e-05, + "loss": 0.4055, + "step": 8483 + }, + { + "epoch": 3.1990950226244346, + "grad_norm": 0.13673207469509543, + "learning_rate": 2.766335753935277e-05, + "loss": 0.3928, + "step": 8484 + }, + { + "epoch": 3.1994720965309202, + "grad_norm": 0.13469945615322576, + "learning_rate": 2.765334139271313e-05, + "loss": 0.3948, + "step": 8485 + }, + { + "epoch": 3.199849170437406, + "grad_norm": 0.11522970509572737, + "learning_rate": 2.7643326101685995e-05, + "loss": 0.3934, + "step": 8486 + }, + { + "epoch": 3.2002262443438916, + "grad_norm": 0.1696993001852752, + "learning_rate": 2.76333116669654e-05, + "loss": 0.3976, + "step": 8487 + }, + { + "epoch": 3.2006033182503772, + "grad_norm": 0.11731737658870688, + "learning_rate": 2.7623298089245324e-05, + "loss": 0.389, + "step": 8488 + }, + { + "epoch": 3.200980392156863, + "grad_norm": 0.13948371638884147, + "learning_rate": 2.761328536921972e-05, + "loss": 0.3924, + "step": 8489 + }, + { + "epoch": 3.2013574660633486, + "grad_norm": 0.17317530511176416, + "learning_rate": 2.760327350758246e-05, + "loss": 0.3945, + "step": 8490 + }, + { + "epoch": 3.2017345399698343, + "grad_norm": 0.1289573883994557, + "learning_rate": 2.7593262505027333e-05, + "loss": 0.3954, + "step": 8491 + }, + { + "epoch": 3.20211161387632, + "grad_norm": 0.17906697095464089, + "learning_rate": 2.7583252362248124e-05, + "loss": 0.394, + "step": 8492 + }, + { + "epoch": 3.2024886877828056, + "grad_norm": 0.15367203006647429, + "learning_rate": 2.7573243079938517e-05, + "loss": 0.4081, + "step": 8493 + }, + { + "epoch": 3.2028657616892913, + "grad_norm": 0.13264779738479035, + "learning_rate": 2.7563234658792123e-05, + "loss": 0.3938, + "step": 8494 + }, + { + "epoch": 3.203242835595777, + "grad_norm": 0.15061944190976415, + "learning_rate": 2.7553227099502552e-05, + "loss": 0.3941, + "step": 8495 + }, + { + "epoch": 3.2036199095022626, + "grad_norm": 0.14034563463428992, + "learning_rate": 2.75432204027633e-05, + "loss": 0.3902, + "step": 8496 + }, + { + "epoch": 3.2039969834087483, + "grad_norm": 0.1458951108577711, + "learning_rate": 2.7533214569267816e-05, + "loss": 0.39, + "step": 8497 + }, + { + "epoch": 3.204374057315234, + "grad_norm": 0.15967907422783295, + "learning_rate": 2.7523209599709514e-05, + "loss": 0.3998, + "step": 8498 + }, + { + "epoch": 3.2047511312217196, + "grad_norm": 0.14086825802435182, + "learning_rate": 2.7513205494781723e-05, + "loss": 0.3865, + "step": 8499 + }, + { + "epoch": 3.2051282051282053, + "grad_norm": 0.16956644162765694, + "learning_rate": 2.7503202255177698e-05, + "loss": 0.4043, + "step": 8500 + }, + { + "epoch": 3.205505279034691, + "grad_norm": 0.1214110467884946, + "learning_rate": 2.749319988159069e-05, + "loss": 0.3925, + "step": 8501 + }, + { + "epoch": 3.2058823529411766, + "grad_norm": 0.15126346724442324, + "learning_rate": 2.7483198374713836e-05, + "loss": 0.3923, + "step": 8502 + }, + { + "epoch": 3.2062594268476623, + "grad_norm": 0.13483315599903695, + "learning_rate": 2.7473197735240212e-05, + "loss": 0.401, + "step": 8503 + }, + { + "epoch": 3.206636500754148, + "grad_norm": 0.1319491144954563, + "learning_rate": 2.7463197963862894e-05, + "loss": 0.3966, + "step": 8504 + }, + { + "epoch": 3.2070135746606336, + "grad_norm": 0.15353332406628148, + "learning_rate": 2.745319906127484e-05, + "loss": 0.401, + "step": 8505 + }, + { + "epoch": 3.2073906485671193, + "grad_norm": 0.1284797176978511, + "learning_rate": 2.744320102816894e-05, + "loss": 0.3979, + "step": 8506 + }, + { + "epoch": 3.207767722473605, + "grad_norm": 0.15254678088142076, + "learning_rate": 2.7433203865238086e-05, + "loss": 0.3957, + "step": 8507 + }, + { + "epoch": 3.2081447963800906, + "grad_norm": 0.15682881811023305, + "learning_rate": 2.742320757317506e-05, + "loss": 0.3989, + "step": 8508 + }, + { + "epoch": 3.2085218702865763, + "grad_norm": 0.11991768781054872, + "learning_rate": 2.7413212152672576e-05, + "loss": 0.3929, + "step": 8509 + }, + { + "epoch": 3.208898944193062, + "grad_norm": 0.15536289625259003, + "learning_rate": 2.740321760442335e-05, + "loss": 0.4001, + "step": 8510 + }, + { + "epoch": 3.2092760180995477, + "grad_norm": 0.1412558194234729, + "learning_rate": 2.739322392911996e-05, + "loss": 0.4035, + "step": 8511 + }, + { + "epoch": 3.2096530920060333, + "grad_norm": 0.13693127680527115, + "learning_rate": 2.7383231127454956e-05, + "loss": 0.4025, + "step": 8512 + }, + { + "epoch": 3.210030165912519, + "grad_norm": 0.15653023527282486, + "learning_rate": 2.7373239200120862e-05, + "loss": 0.4027, + "step": 8513 + }, + { + "epoch": 3.2104072398190047, + "grad_norm": 0.13148322073694116, + "learning_rate": 2.736324814781009e-05, + "loss": 0.3796, + "step": 8514 + }, + { + "epoch": 3.2107843137254903, + "grad_norm": 0.1408385152262338, + "learning_rate": 2.7353257971215e-05, + "loss": 0.3994, + "step": 8515 + }, + { + "epoch": 3.211161387631976, + "grad_norm": 0.13588058558423946, + "learning_rate": 2.7343268671027926e-05, + "loss": 0.3995, + "step": 8516 + }, + { + "epoch": 3.2115384615384617, + "grad_norm": 0.12276723795122207, + "learning_rate": 2.7333280247941113e-05, + "loss": 0.4072, + "step": 8517 + }, + { + "epoch": 3.2119155354449473, + "grad_norm": 0.1342219655998041, + "learning_rate": 2.7323292702646723e-05, + "loss": 0.3911, + "step": 8518 + }, + { + "epoch": 3.212292609351433, + "grad_norm": 0.1251689270039307, + "learning_rate": 2.7313306035836915e-05, + "loss": 0.3856, + "step": 8519 + }, + { + "epoch": 3.2126696832579187, + "grad_norm": 0.14098791052443022, + "learning_rate": 2.7303320248203755e-05, + "loss": 0.4044, + "step": 8520 + }, + { + "epoch": 3.2130467571644044, + "grad_norm": 0.13059690373786187, + "learning_rate": 2.729333534043922e-05, + "loss": 0.4066, + "step": 8521 + }, + { + "epoch": 3.21342383107089, + "grad_norm": 0.1448455488693183, + "learning_rate": 2.728335131323529e-05, + "loss": 0.3977, + "step": 8522 + }, + { + "epoch": 3.2138009049773757, + "grad_norm": 0.1431522235349443, + "learning_rate": 2.7273368167283835e-05, + "loss": 0.3703, + "step": 8523 + }, + { + "epoch": 3.2141779788838614, + "grad_norm": 0.11514322002744475, + "learning_rate": 2.7263385903276663e-05, + "loss": 0.4016, + "step": 8524 + }, + { + "epoch": 3.214555052790347, + "grad_norm": 0.12165597103698056, + "learning_rate": 2.7253404521905564e-05, + "loss": 0.3851, + "step": 8525 + }, + { + "epoch": 3.2149321266968327, + "grad_norm": 0.13600211149121602, + "learning_rate": 2.7243424023862232e-05, + "loss": 0.3868, + "step": 8526 + }, + { + "epoch": 3.2153092006033184, + "grad_norm": 0.11933944932644062, + "learning_rate": 2.7233444409838273e-05, + "loss": 0.3887, + "step": 8527 + }, + { + "epoch": 3.215686274509804, + "grad_norm": 0.12266267444425627, + "learning_rate": 2.722346568052532e-05, + "loss": 0.3948, + "step": 8528 + }, + { + "epoch": 3.2160633484162897, + "grad_norm": 0.15396090029561182, + "learning_rate": 2.7213487836614856e-05, + "loss": 0.3982, + "step": 8529 + }, + { + "epoch": 3.2164404223227754, + "grad_norm": 0.13148639486530087, + "learning_rate": 2.7203510878798325e-05, + "loss": 0.3952, + "step": 8530 + }, + { + "epoch": 3.216817496229261, + "grad_norm": 0.14834148798766475, + "learning_rate": 2.7193534807767162e-05, + "loss": 0.3995, + "step": 8531 + }, + { + "epoch": 3.2171945701357467, + "grad_norm": 0.12304373742827673, + "learning_rate": 2.7183559624212673e-05, + "loss": 0.4071, + "step": 8532 + }, + { + "epoch": 3.2175716440422324, + "grad_norm": 0.1522958253062786, + "learning_rate": 2.717358532882612e-05, + "loss": 0.3878, + "step": 8533 + }, + { + "epoch": 3.217948717948718, + "grad_norm": 0.13351759607716998, + "learning_rate": 2.7163611922298742e-05, + "loss": 0.39, + "step": 8534 + }, + { + "epoch": 3.2183257918552037, + "grad_norm": 0.16975949029103934, + "learning_rate": 2.7153639405321674e-05, + "loss": 0.3938, + "step": 8535 + }, + { + "epoch": 3.2187028657616894, + "grad_norm": 0.15781822539139836, + "learning_rate": 2.7143667778585982e-05, + "loss": 0.3937, + "step": 8536 + }, + { + "epoch": 3.219079939668175, + "grad_norm": 0.1455866703815488, + "learning_rate": 2.7133697042782725e-05, + "loss": 0.3923, + "step": 8537 + }, + { + "epoch": 3.2194570135746607, + "grad_norm": 0.12504027168506818, + "learning_rate": 2.7123727198602856e-05, + "loss": 0.3842, + "step": 8538 + }, + { + "epoch": 3.2198340874811464, + "grad_norm": 0.144288285750597, + "learning_rate": 2.7113758246737245e-05, + "loss": 0.4074, + "step": 8539 + }, + { + "epoch": 3.220211161387632, + "grad_norm": 0.14401067193060713, + "learning_rate": 2.710379018787678e-05, + "loss": 0.3839, + "step": 8540 + }, + { + "epoch": 3.2205882352941178, + "grad_norm": 0.13106263298936624, + "learning_rate": 2.7093823022712217e-05, + "loss": 0.4, + "step": 8541 + }, + { + "epoch": 3.2209653092006034, + "grad_norm": 0.16297164497695027, + "learning_rate": 2.708385675193425e-05, + "loss": 0.4097, + "step": 8542 + }, + { + "epoch": 3.221342383107089, + "grad_norm": 0.16002554259252014, + "learning_rate": 2.7073891376233574e-05, + "loss": 0.3971, + "step": 8543 + }, + { + "epoch": 3.2217194570135748, + "grad_norm": 0.13699151111580776, + "learning_rate": 2.7063926896300762e-05, + "loss": 0.4033, + "step": 8544 + }, + { + "epoch": 3.2220965309200604, + "grad_norm": 0.1442453326053698, + "learning_rate": 2.705396331282632e-05, + "loss": 0.3966, + "step": 8545 + }, + { + "epoch": 3.222473604826546, + "grad_norm": 0.1834406047404459, + "learning_rate": 2.7044000626500754e-05, + "loss": 0.4028, + "step": 8546 + }, + { + "epoch": 3.2228506787330318, + "grad_norm": 0.1438262851369589, + "learning_rate": 2.7034038838014455e-05, + "loss": 0.3925, + "step": 8547 + }, + { + "epoch": 3.2232277526395174, + "grad_norm": 0.1739071130323203, + "learning_rate": 2.702407794805775e-05, + "loss": 0.4054, + "step": 8548 + }, + { + "epoch": 3.223604826546003, + "grad_norm": 0.12541866805765559, + "learning_rate": 2.701411795732094e-05, + "loss": 0.3874, + "step": 8549 + }, + { + "epoch": 3.223981900452489, + "grad_norm": 0.14077817011202928, + "learning_rate": 2.7004158866494247e-05, + "loss": 0.3965, + "step": 8550 + }, + { + "epoch": 3.2243589743589745, + "grad_norm": 0.16314243104128068, + "learning_rate": 2.6994200676267798e-05, + "loss": 0.3912, + "step": 8551 + }, + { + "epoch": 3.22473604826546, + "grad_norm": 0.11907796053951913, + "learning_rate": 2.698424338733172e-05, + "loss": 0.3921, + "step": 8552 + }, + { + "epoch": 3.225113122171946, + "grad_norm": 0.13985799047975614, + "learning_rate": 2.6974287000376037e-05, + "loss": 0.3953, + "step": 8553 + }, + { + "epoch": 3.2254901960784315, + "grad_norm": 0.15080303331484923, + "learning_rate": 2.6964331516090683e-05, + "loss": 0.3982, + "step": 8554 + }, + { + "epoch": 3.225867269984917, + "grad_norm": 0.12983415727548753, + "learning_rate": 2.695437693516562e-05, + "loss": 0.3997, + "step": 8555 + }, + { + "epoch": 3.226244343891403, + "grad_norm": 0.10322590791712133, + "learning_rate": 2.6944423258290658e-05, + "loss": 0.3897, + "step": 8556 + }, + { + "epoch": 3.2266214177978885, + "grad_norm": 0.12736997121795285, + "learning_rate": 2.6934470486155572e-05, + "loss": 0.3971, + "step": 8557 + }, + { + "epoch": 3.226998491704374, + "grad_norm": 0.14349546194811588, + "learning_rate": 2.6924518619450104e-05, + "loss": 0.3876, + "step": 8558 + }, + { + "epoch": 3.22737556561086, + "grad_norm": 0.11837571862151919, + "learning_rate": 2.6914567658863898e-05, + "loss": 0.3881, + "step": 8559 + }, + { + "epoch": 3.2277526395173455, + "grad_norm": 0.1254080132087246, + "learning_rate": 2.690461760508653e-05, + "loss": 0.4023, + "step": 8560 + }, + { + "epoch": 3.228129713423831, + "grad_norm": 0.13607367021879235, + "learning_rate": 2.6894668458807567e-05, + "loss": 0.3931, + "step": 8561 + }, + { + "epoch": 3.228506787330317, + "grad_norm": 0.12556385411480694, + "learning_rate": 2.6884720220716452e-05, + "loss": 0.3945, + "step": 8562 + }, + { + "epoch": 3.2288838612368025, + "grad_norm": 0.13358066901230523, + "learning_rate": 2.6874772891502576e-05, + "loss": 0.3951, + "step": 8563 + }, + { + "epoch": 3.229260935143288, + "grad_norm": 0.12583929387548606, + "learning_rate": 2.6864826471855314e-05, + "loss": 0.41, + "step": 8564 + }, + { + "epoch": 3.229638009049774, + "grad_norm": 0.14113213707912928, + "learning_rate": 2.685488096246393e-05, + "loss": 0.3969, + "step": 8565 + }, + { + "epoch": 3.2300150829562595, + "grad_norm": 0.14940385309435686, + "learning_rate": 2.6844936364017616e-05, + "loss": 0.4077, + "step": 8566 + }, + { + "epoch": 3.230392156862745, + "grad_norm": 0.15139832130967115, + "learning_rate": 2.6834992677205556e-05, + "loss": 0.397, + "step": 8567 + }, + { + "epoch": 3.230769230769231, + "grad_norm": 0.1504108938168975, + "learning_rate": 2.682504990271682e-05, + "loss": 0.3904, + "step": 8568 + }, + { + "epoch": 3.2311463046757165, + "grad_norm": 0.14854281230582986, + "learning_rate": 2.6815108041240444e-05, + "loss": 0.3922, + "step": 8569 + }, + { + "epoch": 3.231523378582202, + "grad_norm": 0.13500589196125012, + "learning_rate": 2.6805167093465388e-05, + "loss": 0.3869, + "step": 8570 + }, + { + "epoch": 3.231900452488688, + "grad_norm": 0.11558487906062498, + "learning_rate": 2.6795227060080543e-05, + "loss": 0.3952, + "step": 8571 + }, + { + "epoch": 3.2322775263951735, + "grad_norm": 0.15663749898241547, + "learning_rate": 2.678528794177475e-05, + "loss": 0.3868, + "step": 8572 + }, + { + "epoch": 3.232654600301659, + "grad_norm": 0.12404142449729828, + "learning_rate": 2.6775349739236795e-05, + "loss": 0.401, + "step": 8573 + }, + { + "epoch": 3.233031674208145, + "grad_norm": 0.11926126727616498, + "learning_rate": 2.6765412453155355e-05, + "loss": 0.4056, + "step": 8574 + }, + { + "epoch": 3.2334087481146305, + "grad_norm": 0.12817310956557984, + "learning_rate": 2.6755476084219093e-05, + "loss": 0.3924, + "step": 8575 + }, + { + "epoch": 3.233785822021116, + "grad_norm": 0.14433333173105886, + "learning_rate": 2.6745540633116605e-05, + "loss": 0.392, + "step": 8576 + }, + { + "epoch": 3.234162895927602, + "grad_norm": 0.1441837150838708, + "learning_rate": 2.6735606100536385e-05, + "loss": 0.4039, + "step": 8577 + }, + { + "epoch": 3.2345399698340875, + "grad_norm": 0.15459076178517336, + "learning_rate": 2.67256724871669e-05, + "loss": 0.3871, + "step": 8578 + }, + { + "epoch": 3.234917043740573, + "grad_norm": 0.15271134086624807, + "learning_rate": 2.6715739793696535e-05, + "loss": 0.3967, + "step": 8579 + }, + { + "epoch": 3.235294117647059, + "grad_norm": 0.1301374637325098, + "learning_rate": 2.6705808020813622e-05, + "loss": 0.3933, + "step": 8580 + }, + { + "epoch": 3.2356711915535445, + "grad_norm": 0.1755421273730564, + "learning_rate": 2.6695877169206415e-05, + "loss": 0.3958, + "step": 8581 + }, + { + "epoch": 3.23604826546003, + "grad_norm": 0.14786615003001463, + "learning_rate": 2.668594723956313e-05, + "loss": 0.3926, + "step": 8582 + }, + { + "epoch": 3.236425339366516, + "grad_norm": 0.14753791422542473, + "learning_rate": 2.6676018232571882e-05, + "loss": 0.4057, + "step": 8583 + }, + { + "epoch": 3.2368024132730016, + "grad_norm": 0.19695889659554414, + "learning_rate": 2.6666090148920757e-05, + "loss": 0.4043, + "step": 8584 + }, + { + "epoch": 3.2371794871794872, + "grad_norm": 0.15702782462233839, + "learning_rate": 2.665616298929775e-05, + "loss": 0.3944, + "step": 8585 + }, + { + "epoch": 3.237556561085973, + "grad_norm": 0.14507030759351014, + "learning_rate": 2.6646236754390824e-05, + "loss": 0.4014, + "step": 8586 + }, + { + "epoch": 3.2379336349924586, + "grad_norm": 0.1410425632161172, + "learning_rate": 2.6636311444887834e-05, + "loss": 0.3901, + "step": 8587 + }, + { + "epoch": 3.2383107088989442, + "grad_norm": 0.15550221724230495, + "learning_rate": 2.6626387061476612e-05, + "loss": 0.3966, + "step": 8588 + }, + { + "epoch": 3.23868778280543, + "grad_norm": 0.15138312411793597, + "learning_rate": 2.6616463604844917e-05, + "loss": 0.4006, + "step": 8589 + }, + { + "epoch": 3.2390648567119156, + "grad_norm": 0.14563183837763777, + "learning_rate": 2.6606541075680404e-05, + "loss": 0.4014, + "step": 8590 + }, + { + "epoch": 3.2394419306184012, + "grad_norm": 0.1445707535258157, + "learning_rate": 2.659661947467072e-05, + "loss": 0.397, + "step": 8591 + }, + { + "epoch": 3.239819004524887, + "grad_norm": 0.1791583287082571, + "learning_rate": 2.6586698802503428e-05, + "loss": 0.3896, + "step": 8592 + }, + { + "epoch": 3.2401960784313726, + "grad_norm": 0.1315025003476615, + "learning_rate": 2.6576779059865997e-05, + "loss": 0.3933, + "step": 8593 + }, + { + "epoch": 3.2405731523378583, + "grad_norm": 0.1516448579499765, + "learning_rate": 2.6566860247445875e-05, + "loss": 0.4089, + "step": 8594 + }, + { + "epoch": 3.240950226244344, + "grad_norm": 0.1843579094637099, + "learning_rate": 2.6556942365930432e-05, + "loss": 0.4016, + "step": 8595 + }, + { + "epoch": 3.2413273001508296, + "grad_norm": 0.1439882479759273, + "learning_rate": 2.6547025416006947e-05, + "loss": 0.4137, + "step": 8596 + }, + { + "epoch": 3.2417043740573153, + "grad_norm": 0.14530898604313858, + "learning_rate": 2.653710939836267e-05, + "loss": 0.3899, + "step": 8597 + }, + { + "epoch": 3.242081447963801, + "grad_norm": 0.13563056466130632, + "learning_rate": 2.652719431368478e-05, + "loss": 0.3938, + "step": 8598 + }, + { + "epoch": 3.2424585218702866, + "grad_norm": 0.14471178448175215, + "learning_rate": 2.651728016266036e-05, + "loss": 0.3998, + "step": 8599 + }, + { + "epoch": 3.2428355957767723, + "grad_norm": 0.1371580360883611, + "learning_rate": 2.6507366945976475e-05, + "loss": 0.3956, + "step": 8600 + }, + { + "epoch": 3.243212669683258, + "grad_norm": 0.1324202877705581, + "learning_rate": 2.6497454664320083e-05, + "loss": 0.3929, + "step": 8601 + }, + { + "epoch": 3.2435897435897436, + "grad_norm": 0.13725874752737133, + "learning_rate": 2.648754331837812e-05, + "loss": 0.3945, + "step": 8602 + }, + { + "epoch": 3.2439668174962293, + "grad_norm": 0.1397010587621912, + "learning_rate": 2.6477632908837407e-05, + "loss": 0.3845, + "step": 8603 + }, + { + "epoch": 3.244343891402715, + "grad_norm": 0.11586857681602812, + "learning_rate": 2.646772343638474e-05, + "loss": 0.4007, + "step": 8604 + }, + { + "epoch": 3.2447209653092006, + "grad_norm": 0.14602528042300658, + "learning_rate": 2.645781490170684e-05, + "loss": 0.3799, + "step": 8605 + }, + { + "epoch": 3.2450980392156863, + "grad_norm": 0.15675575074380063, + "learning_rate": 2.644790730549036e-05, + "loss": 0.3963, + "step": 8606 + }, + { + "epoch": 3.245475113122172, + "grad_norm": 0.13989767658260494, + "learning_rate": 2.6438000648421875e-05, + "loss": 0.4027, + "step": 8607 + }, + { + "epoch": 3.2458521870286576, + "grad_norm": 0.1365467895453872, + "learning_rate": 2.642809493118792e-05, + "loss": 0.4036, + "step": 8608 + }, + { + "epoch": 3.2462292609351433, + "grad_norm": 0.16537001489855727, + "learning_rate": 2.641819015447495e-05, + "loss": 0.3932, + "step": 8609 + }, + { + "epoch": 3.246606334841629, + "grad_norm": 0.13049481343073938, + "learning_rate": 2.640828631896935e-05, + "loss": 0.3877, + "step": 8610 + }, + { + "epoch": 3.2469834087481146, + "grad_norm": 0.13719079116667937, + "learning_rate": 2.639838342535745e-05, + "loss": 0.3927, + "step": 8611 + }, + { + "epoch": 3.2473604826546003, + "grad_norm": 0.15872913396714916, + "learning_rate": 2.638848147432553e-05, + "loss": 0.3826, + "step": 8612 + }, + { + "epoch": 3.247737556561086, + "grad_norm": 0.1290798284670069, + "learning_rate": 2.6378580466559754e-05, + "loss": 0.3907, + "step": 8613 + }, + { + "epoch": 3.2481146304675717, + "grad_norm": 0.1497975470851905, + "learning_rate": 2.6368680402746273e-05, + "loss": 0.3996, + "step": 8614 + }, + { + "epoch": 3.2484917043740573, + "grad_norm": 0.1445282808645704, + "learning_rate": 2.6358781283571154e-05, + "loss": 0.39, + "step": 8615 + }, + { + "epoch": 3.248868778280543, + "grad_norm": 0.13851809087184094, + "learning_rate": 2.6348883109720387e-05, + "loss": 0.4023, + "step": 8616 + }, + { + "epoch": 3.2492458521870287, + "grad_norm": 0.14076332067362637, + "learning_rate": 2.6338985881879912e-05, + "loss": 0.3987, + "step": 8617 + }, + { + "epoch": 3.2496229260935143, + "grad_norm": 0.12368826487357096, + "learning_rate": 2.6329089600735605e-05, + "loss": 0.3986, + "step": 8618 + }, + { + "epoch": 3.25, + "grad_norm": 0.1223378161508676, + "learning_rate": 2.6319194266973256e-05, + "loss": 0.4016, + "step": 8619 + }, + { + "epoch": 3.2503770739064857, + "grad_norm": 0.1294349709323713, + "learning_rate": 2.6309299881278607e-05, + "loss": 0.4026, + "step": 8620 + }, + { + "epoch": 3.2507541478129713, + "grad_norm": 0.12333661846103597, + "learning_rate": 2.629940644433734e-05, + "loss": 0.391, + "step": 8621 + }, + { + "epoch": 3.251131221719457, + "grad_norm": 0.12465945302266018, + "learning_rate": 2.6289513956835046e-05, + "loss": 0.4029, + "step": 8622 + }, + { + "epoch": 3.2515082956259427, + "grad_norm": 0.13906438678109168, + "learning_rate": 2.6279622419457277e-05, + "loss": 0.4038, + "step": 8623 + }, + { + "epoch": 3.2518853695324283, + "grad_norm": 0.12834581873046777, + "learning_rate": 2.6269731832889515e-05, + "loss": 0.4023, + "step": 8624 + }, + { + "epoch": 3.252262443438914, + "grad_norm": 0.16092708318591695, + "learning_rate": 2.625984219781714e-05, + "loss": 0.3903, + "step": 8625 + }, + { + "epoch": 3.2526395173453997, + "grad_norm": 0.15475043568228297, + "learning_rate": 2.6249953514925524e-05, + "loss": 0.3975, + "step": 8626 + }, + { + "epoch": 3.2530165912518854, + "grad_norm": 0.13437945607034912, + "learning_rate": 2.6240065784899934e-05, + "loss": 0.4041, + "step": 8627 + }, + { + "epoch": 3.253393665158371, + "grad_norm": 0.12455659953368721, + "learning_rate": 2.6230179008425566e-05, + "loss": 0.3876, + "step": 8628 + }, + { + "epoch": 3.2537707390648567, + "grad_norm": 0.15190765106361348, + "learning_rate": 2.6220293186187594e-05, + "loss": 0.4004, + "step": 8629 + }, + { + "epoch": 3.2541478129713424, + "grad_norm": 0.13214825792206997, + "learning_rate": 2.6210408318871087e-05, + "loss": 0.3899, + "step": 8630 + }, + { + "epoch": 3.254524886877828, + "grad_norm": 0.13241253837412942, + "learning_rate": 2.6200524407161033e-05, + "loss": 0.3951, + "step": 8631 + }, + { + "epoch": 3.2549019607843137, + "grad_norm": 0.14790336001445348, + "learning_rate": 2.6190641451742412e-05, + "loss": 0.3894, + "step": 8632 + }, + { + "epoch": 3.2552790346907994, + "grad_norm": 0.12741210502441724, + "learning_rate": 2.6180759453300093e-05, + "loss": 0.3991, + "step": 8633 + }, + { + "epoch": 3.255656108597285, + "grad_norm": 0.1429590429390404, + "learning_rate": 2.6170878412518862e-05, + "loss": 0.3929, + "step": 8634 + }, + { + "epoch": 3.2560331825037707, + "grad_norm": 0.11893845512042989, + "learning_rate": 2.6160998330083515e-05, + "loss": 0.4031, + "step": 8635 + }, + { + "epoch": 3.2564102564102564, + "grad_norm": 0.1344988852785127, + "learning_rate": 2.615111920667871e-05, + "loss": 0.3941, + "step": 8636 + }, + { + "epoch": 3.256787330316742, + "grad_norm": 0.12947073589041894, + "learning_rate": 2.6141241042989035e-05, + "loss": 0.3899, + "step": 8637 + }, + { + "epoch": 3.2571644042232277, + "grad_norm": 0.12190338057188473, + "learning_rate": 2.6131363839699086e-05, + "loss": 0.3947, + "step": 8638 + }, + { + "epoch": 3.2575414781297134, + "grad_norm": 0.12562428311315627, + "learning_rate": 2.6121487597493316e-05, + "loss": 0.392, + "step": 8639 + }, + { + "epoch": 3.257918552036199, + "grad_norm": 0.12447832579769955, + "learning_rate": 2.6111612317056133e-05, + "loss": 0.391, + "step": 8640 + }, + { + "epoch": 3.2582956259426847, + "grad_norm": 0.1249000004529199, + "learning_rate": 2.610173799907191e-05, + "loss": 0.4087, + "step": 8641 + }, + { + "epoch": 3.2586726998491704, + "grad_norm": 0.11476157475629632, + "learning_rate": 2.609186464422492e-05, + "loss": 0.3963, + "step": 8642 + }, + { + "epoch": 3.259049773755656, + "grad_norm": 0.15004893884388928, + "learning_rate": 2.6081992253199358e-05, + "loss": 0.4044, + "step": 8643 + }, + { + "epoch": 3.2594268476621417, + "grad_norm": 0.12382796607696855, + "learning_rate": 2.6072120826679398e-05, + "loss": 0.3893, + "step": 8644 + }, + { + "epoch": 3.2598039215686274, + "grad_norm": 0.14757711142158977, + "learning_rate": 2.6062250365349115e-05, + "loss": 0.4026, + "step": 8645 + }, + { + "epoch": 3.260180995475113, + "grad_norm": 0.14960511627003628, + "learning_rate": 2.6052380869892507e-05, + "loss": 0.3855, + "step": 8646 + }, + { + "epoch": 3.2605580693815988, + "grad_norm": 0.1355219661959229, + "learning_rate": 2.604251234099354e-05, + "loss": 0.4027, + "step": 8647 + }, + { + "epoch": 3.2609351432880844, + "grad_norm": 0.11545934037897665, + "learning_rate": 2.6032644779336095e-05, + "loss": 0.3991, + "step": 8648 + }, + { + "epoch": 3.26131221719457, + "grad_norm": 0.1632313593405491, + "learning_rate": 2.6022778185603957e-05, + "loss": 0.3944, + "step": 8649 + }, + { + "epoch": 3.2616892911010558, + "grad_norm": 0.15399071547157114, + "learning_rate": 2.601291256048091e-05, + "loss": 0.3985, + "step": 8650 + }, + { + "epoch": 3.2620663650075414, + "grad_norm": 0.1294464428090695, + "learning_rate": 2.600304790465062e-05, + "loss": 0.3982, + "step": 8651 + }, + { + "epoch": 3.262443438914027, + "grad_norm": 0.17097863624678136, + "learning_rate": 2.5993184218796677e-05, + "loss": 0.3948, + "step": 8652 + }, + { + "epoch": 3.2628205128205128, + "grad_norm": 0.15839041562729717, + "learning_rate": 2.598332150360266e-05, + "loss": 0.4055, + "step": 8653 + }, + { + "epoch": 3.2631975867269984, + "grad_norm": 0.13544476345018952, + "learning_rate": 2.597345975975203e-05, + "loss": 0.3939, + "step": 8654 + }, + { + "epoch": 3.263574660633484, + "grad_norm": 0.12184211691892878, + "learning_rate": 2.596359898792818e-05, + "loss": 0.3991, + "step": 8655 + }, + { + "epoch": 3.26395173453997, + "grad_norm": 0.14853371157453468, + "learning_rate": 2.595373918881448e-05, + "loss": 0.3953, + "step": 8656 + }, + { + "epoch": 3.2643288084464555, + "grad_norm": 0.1507254023645922, + "learning_rate": 2.59438803630942e-05, + "loss": 0.3975, + "step": 8657 + }, + { + "epoch": 3.264705882352941, + "grad_norm": 0.1272639892091351, + "learning_rate": 2.5934022511450528e-05, + "loss": 0.4011, + "step": 8658 + }, + { + "epoch": 3.265082956259427, + "grad_norm": 0.13534116171548444, + "learning_rate": 2.5924165634566637e-05, + "loss": 0.3837, + "step": 8659 + }, + { + "epoch": 3.2654600301659125, + "grad_norm": 0.1632613332692209, + "learning_rate": 2.5914309733125578e-05, + "loss": 0.4004, + "step": 8660 + }, + { + "epoch": 3.265837104072398, + "grad_norm": 0.1495015606281828, + "learning_rate": 2.5904454807810348e-05, + "loss": 0.3963, + "step": 8661 + }, + { + "epoch": 3.266214177978884, + "grad_norm": 0.1407276343123092, + "learning_rate": 2.5894600859303915e-05, + "loss": 0.3968, + "step": 8662 + }, + { + "epoch": 3.2665912518853695, + "grad_norm": 0.1558661287829066, + "learning_rate": 2.5884747888289122e-05, + "loss": 0.3974, + "step": 8663 + }, + { + "epoch": 3.266968325791855, + "grad_norm": 0.1900893502891346, + "learning_rate": 2.5874895895448776e-05, + "loss": 0.3941, + "step": 8664 + }, + { + "epoch": 3.267345399698341, + "grad_norm": 0.1453414498115986, + "learning_rate": 2.5865044881465624e-05, + "loss": 0.3989, + "step": 8665 + }, + { + "epoch": 3.2677224736048265, + "grad_norm": 0.1727693332873207, + "learning_rate": 2.585519484702233e-05, + "loss": 0.3925, + "step": 8666 + }, + { + "epoch": 3.268099547511312, + "grad_norm": 0.19793767788316935, + "learning_rate": 2.5845345792801464e-05, + "loss": 0.4004, + "step": 8667 + }, + { + "epoch": 3.268476621417798, + "grad_norm": 0.12135928426207243, + "learning_rate": 2.5835497719485596e-05, + "loss": 0.3936, + "step": 8668 + }, + { + "epoch": 3.2688536953242835, + "grad_norm": 0.16349012381631067, + "learning_rate": 2.5825650627757173e-05, + "loss": 0.4079, + "step": 8669 + }, + { + "epoch": 3.269230769230769, + "grad_norm": 0.1645476365695181, + "learning_rate": 2.5815804518298575e-05, + "loss": 0.3924, + "step": 8670 + }, + { + "epoch": 3.269607843137255, + "grad_norm": 0.1268201232328754, + "learning_rate": 2.5805959391792157e-05, + "loss": 0.4052, + "step": 8671 + }, + { + "epoch": 3.2699849170437405, + "grad_norm": 0.15083769337136488, + "learning_rate": 2.579611524892016e-05, + "loss": 0.3941, + "step": 8672 + }, + { + "epoch": 3.270361990950226, + "grad_norm": 0.15814920789403347, + "learning_rate": 2.578627209036477e-05, + "loss": 0.3993, + "step": 8673 + }, + { + "epoch": 3.270739064856712, + "grad_norm": 0.14962777868079055, + "learning_rate": 2.5776429916808122e-05, + "loss": 0.3992, + "step": 8674 + }, + { + "epoch": 3.2711161387631975, + "grad_norm": 0.12824675128583468, + "learning_rate": 2.5766588728932274e-05, + "loss": 0.4007, + "step": 8675 + }, + { + "epoch": 3.271493212669683, + "grad_norm": 0.14039932866380095, + "learning_rate": 2.5756748527419185e-05, + "loss": 0.4026, + "step": 8676 + }, + { + "epoch": 3.271870286576169, + "grad_norm": 0.14166103642402417, + "learning_rate": 2.5746909312950797e-05, + "loss": 0.3952, + "step": 8677 + }, + { + "epoch": 3.2722473604826545, + "grad_norm": 0.11632877095300623, + "learning_rate": 2.573707108620896e-05, + "loss": 0.3876, + "step": 8678 + }, + { + "epoch": 3.27262443438914, + "grad_norm": 0.14419953958363693, + "learning_rate": 2.5727233847875423e-05, + "loss": 0.3938, + "step": 8679 + }, + { + "epoch": 3.273001508295626, + "grad_norm": 0.1410464773839592, + "learning_rate": 2.571739759863194e-05, + "loss": 0.3994, + "step": 8680 + }, + { + "epoch": 3.2733785822021115, + "grad_norm": 0.12393647543268253, + "learning_rate": 2.5707562339160134e-05, + "loss": 0.3906, + "step": 8681 + }, + { + "epoch": 3.273755656108597, + "grad_norm": 0.15144073222395554, + "learning_rate": 2.5697728070141565e-05, + "loss": 0.4012, + "step": 8682 + }, + { + "epoch": 3.274132730015083, + "grad_norm": 0.14070665649412686, + "learning_rate": 2.568789479225777e-05, + "loss": 0.387, + "step": 8683 + }, + { + "epoch": 3.2745098039215685, + "grad_norm": 0.14665169836030129, + "learning_rate": 2.5678062506190172e-05, + "loss": 0.3956, + "step": 8684 + }, + { + "epoch": 3.274886877828054, + "grad_norm": 0.12133216312468816, + "learning_rate": 2.566823121262012e-05, + "loss": 0.3948, + "step": 8685 + }, + { + "epoch": 3.27526395173454, + "grad_norm": 0.15716278665097358, + "learning_rate": 2.565840091222895e-05, + "loss": 0.3877, + "step": 8686 + }, + { + "epoch": 3.2756410256410255, + "grad_norm": 0.15280178314962609, + "learning_rate": 2.564857160569788e-05, + "loss": 0.3831, + "step": 8687 + }, + { + "epoch": 3.276018099547511, + "grad_norm": 0.1214725555101781, + "learning_rate": 2.5638743293708048e-05, + "loss": 0.3956, + "step": 8688 + }, + { + "epoch": 3.276395173453997, + "grad_norm": 0.18796844502454996, + "learning_rate": 2.5628915976940583e-05, + "loss": 0.3981, + "step": 8689 + }, + { + "epoch": 3.2767722473604826, + "grad_norm": 0.17640473583266073, + "learning_rate": 2.5619089656076495e-05, + "loss": 0.3945, + "step": 8690 + }, + { + "epoch": 3.2771493212669682, + "grad_norm": 0.12485844627576867, + "learning_rate": 2.560926433179672e-05, + "loss": 0.4048, + "step": 8691 + }, + { + "epoch": 3.277526395173454, + "grad_norm": 0.1373255600301807, + "learning_rate": 2.5599440004782182e-05, + "loss": 0.3981, + "step": 8692 + }, + { + "epoch": 3.2779034690799396, + "grad_norm": 0.12907634737668167, + "learning_rate": 2.5589616675713673e-05, + "loss": 0.3984, + "step": 8693 + }, + { + "epoch": 3.2782805429864252, + "grad_norm": 0.125024999210252, + "learning_rate": 2.5579794345271936e-05, + "loss": 0.3912, + "step": 8694 + }, + { + "epoch": 3.278657616892911, + "grad_norm": 0.11390683795771023, + "learning_rate": 2.556997301413767e-05, + "loss": 0.3901, + "step": 8695 + }, + { + "epoch": 3.2790346907993966, + "grad_norm": 0.1100824094867607, + "learning_rate": 2.556015268299148e-05, + "loss": 0.3909, + "step": 8696 + }, + { + "epoch": 3.2794117647058822, + "grad_norm": 0.12327212254097095, + "learning_rate": 2.5550333352513885e-05, + "loss": 0.3943, + "step": 8697 + }, + { + "epoch": 3.279788838612368, + "grad_norm": 0.12331872741331142, + "learning_rate": 2.5540515023385385e-05, + "loss": 0.4038, + "step": 8698 + }, + { + "epoch": 3.2801659125188536, + "grad_norm": 0.1266671511188262, + "learning_rate": 2.5530697696286373e-05, + "loss": 0.3898, + "step": 8699 + }, + { + "epoch": 3.2805429864253393, + "grad_norm": 0.16261015576318635, + "learning_rate": 2.5520881371897154e-05, + "loss": 0.4066, + "step": 8700 + }, + { + "epoch": 3.280920060331825, + "grad_norm": 0.11255361294467449, + "learning_rate": 2.551106605089803e-05, + "loss": 0.392, + "step": 8701 + }, + { + "epoch": 3.2812971342383106, + "grad_norm": 0.13426372309601392, + "learning_rate": 2.550125173396918e-05, + "loss": 0.3978, + "step": 8702 + }, + { + "epoch": 3.2816742081447963, + "grad_norm": 0.14928400094828603, + "learning_rate": 2.5491438421790707e-05, + "loss": 0.406, + "step": 8703 + }, + { + "epoch": 3.282051282051282, + "grad_norm": 0.12176217315393553, + "learning_rate": 2.54816261150427e-05, + "loss": 0.4106, + "step": 8704 + }, + { + "epoch": 3.2824283559577676, + "grad_norm": 0.13029984185643984, + "learning_rate": 2.5471814814405127e-05, + "loss": 0.3981, + "step": 8705 + }, + { + "epoch": 3.2828054298642533, + "grad_norm": 0.1491969187361493, + "learning_rate": 2.546200452055788e-05, + "loss": 0.3999, + "step": 8706 + }, + { + "epoch": 3.283182503770739, + "grad_norm": 0.13965007568702298, + "learning_rate": 2.5452195234180842e-05, + "loss": 0.4044, + "step": 8707 + }, + { + "epoch": 3.2835595776772246, + "grad_norm": 0.13606648763412324, + "learning_rate": 2.5442386955953772e-05, + "loss": 0.3991, + "step": 8708 + }, + { + "epoch": 3.2839366515837103, + "grad_norm": 0.14190852158967518, + "learning_rate": 2.5432579686556355e-05, + "loss": 0.3935, + "step": 8709 + }, + { + "epoch": 3.284313725490196, + "grad_norm": 0.14539555825615724, + "learning_rate": 2.5422773426668262e-05, + "loss": 0.3955, + "step": 8710 + }, + { + "epoch": 3.2846907993966816, + "grad_norm": 0.13088804639496876, + "learning_rate": 2.5412968176969043e-05, + "loss": 0.3874, + "step": 8711 + }, + { + "epoch": 3.2850678733031673, + "grad_norm": 0.1411278328950296, + "learning_rate": 2.540316393813817e-05, + "loss": 0.3817, + "step": 8712 + }, + { + "epoch": 3.285444947209653, + "grad_norm": 0.14332930857023188, + "learning_rate": 2.5393360710855105e-05, + "loss": 0.3943, + "step": 8713 + }, + { + "epoch": 3.2858220211161386, + "grad_norm": 0.16607588884460484, + "learning_rate": 2.5383558495799186e-05, + "loss": 0.4037, + "step": 8714 + }, + { + "epoch": 3.2861990950226243, + "grad_norm": 0.11043992462501692, + "learning_rate": 2.5373757293649687e-05, + "loss": 0.4053, + "step": 8715 + }, + { + "epoch": 3.28657616892911, + "grad_norm": 0.16214319626807042, + "learning_rate": 2.5363957105085847e-05, + "loss": 0.4005, + "step": 8716 + }, + { + "epoch": 3.2869532428355956, + "grad_norm": 0.159187573228353, + "learning_rate": 2.535415793078679e-05, + "loss": 0.3905, + "step": 8717 + }, + { + "epoch": 3.2873303167420813, + "grad_norm": 0.11762372196587473, + "learning_rate": 2.5344359771431596e-05, + "loss": 0.3959, + "step": 8718 + }, + { + "epoch": 3.287707390648567, + "grad_norm": 0.14924312398215153, + "learning_rate": 2.533456262769928e-05, + "loss": 0.3844, + "step": 8719 + }, + { + "epoch": 3.2880844645550527, + "grad_norm": 0.1321707479638726, + "learning_rate": 2.532476650026876e-05, + "loss": 0.3999, + "step": 8720 + }, + { + "epoch": 3.2884615384615383, + "grad_norm": 0.11954678967905819, + "learning_rate": 2.5314971389818902e-05, + "loss": 0.395, + "step": 8721 + }, + { + "epoch": 3.288838612368024, + "grad_norm": 0.13584514829212452, + "learning_rate": 2.5305177297028508e-05, + "loss": 0.4044, + "step": 8722 + }, + { + "epoch": 3.2892156862745097, + "grad_norm": 0.13950959762698018, + "learning_rate": 2.529538422257628e-05, + "loss": 0.3943, + "step": 8723 + }, + { + "epoch": 3.2895927601809953, + "grad_norm": 0.12720910039450883, + "learning_rate": 2.52855921671409e-05, + "loss": 0.3921, + "step": 8724 + }, + { + "epoch": 3.289969834087481, + "grad_norm": 0.13340918156250672, + "learning_rate": 2.5275801131400925e-05, + "loss": 0.3991, + "step": 8725 + }, + { + "epoch": 3.2903469079939667, + "grad_norm": 0.1434154561195071, + "learning_rate": 2.5266011116034862e-05, + "loss": 0.3922, + "step": 8726 + }, + { + "epoch": 3.2907239819004523, + "grad_norm": 0.1455884272958826, + "learning_rate": 2.5256222121721166e-05, + "loss": 0.3946, + "step": 8727 + }, + { + "epoch": 3.291101055806938, + "grad_norm": 0.11994674456623297, + "learning_rate": 2.524643414913821e-05, + "loss": 0.4044, + "step": 8728 + }, + { + "epoch": 3.2914781297134237, + "grad_norm": 0.13128424292668353, + "learning_rate": 2.5236647198964255e-05, + "loss": 0.3915, + "step": 8729 + }, + { + "epoch": 3.2918552036199094, + "grad_norm": 0.13334072113182496, + "learning_rate": 2.5226861271877577e-05, + "loss": 0.4046, + "step": 8730 + }, + { + "epoch": 3.292232277526395, + "grad_norm": 0.12188920411572272, + "learning_rate": 2.5217076368556308e-05, + "loss": 0.3894, + "step": 8731 + }, + { + "epoch": 3.2926093514328807, + "grad_norm": 0.1291579530360641, + "learning_rate": 2.5207292489678522e-05, + "loss": 0.3893, + "step": 8732 + }, + { + "epoch": 3.2929864253393664, + "grad_norm": 0.12827397842055366, + "learning_rate": 2.5197509635922252e-05, + "loss": 0.4027, + "step": 8733 + }, + { + "epoch": 3.293363499245852, + "grad_norm": 0.13056729481017026, + "learning_rate": 2.5187727807965433e-05, + "loss": 0.4026, + "step": 8734 + }, + { + "epoch": 3.2937405731523377, + "grad_norm": 0.11970459132120224, + "learning_rate": 2.5177947006485938e-05, + "loss": 0.3986, + "step": 8735 + }, + { + "epoch": 3.2941176470588234, + "grad_norm": 0.13151065246004232, + "learning_rate": 2.5168167232161574e-05, + "loss": 0.3932, + "step": 8736 + }, + { + "epoch": 3.294494720965309, + "grad_norm": 0.14488948222856485, + "learning_rate": 2.5158388485670057e-05, + "loss": 0.4044, + "step": 8737 + }, + { + "epoch": 3.2948717948717947, + "grad_norm": 0.10838944132595464, + "learning_rate": 2.5148610767689056e-05, + "loss": 0.3934, + "step": 8738 + }, + { + "epoch": 3.2952488687782804, + "grad_norm": 0.15088809583579282, + "learning_rate": 2.5138834078896162e-05, + "loss": 0.3978, + "step": 8739 + }, + { + "epoch": 3.295625942684766, + "grad_norm": 0.13201675087572487, + "learning_rate": 2.5129058419968874e-05, + "loss": 0.4001, + "step": 8740 + }, + { + "epoch": 3.2960030165912517, + "grad_norm": 0.11039220847222199, + "learning_rate": 2.511928379158465e-05, + "loss": 0.3928, + "step": 8741 + }, + { + "epoch": 3.2963800904977374, + "grad_norm": 0.11995999754790128, + "learning_rate": 2.5109510194420866e-05, + "loss": 0.4008, + "step": 8742 + }, + { + "epoch": 3.296757164404223, + "grad_norm": 0.10779658742669483, + "learning_rate": 2.5099737629154817e-05, + "loss": 0.3914, + "step": 8743 + }, + { + "epoch": 3.2971342383107087, + "grad_norm": 0.14429397980990166, + "learning_rate": 2.5089966096463726e-05, + "loss": 0.3869, + "step": 8744 + }, + { + "epoch": 3.2975113122171944, + "grad_norm": 0.12822528453776344, + "learning_rate": 2.508019559702477e-05, + "loss": 0.3874, + "step": 8745 + }, + { + "epoch": 3.29788838612368, + "grad_norm": 0.13564136810749633, + "learning_rate": 2.5070426131515015e-05, + "loss": 0.4001, + "step": 8746 + }, + { + "epoch": 3.2982654600301657, + "grad_norm": 0.11781092085704815, + "learning_rate": 2.506065770061149e-05, + "loss": 0.3986, + "step": 8747 + }, + { + "epoch": 3.2986425339366514, + "grad_norm": 0.1459715818305913, + "learning_rate": 2.505089030499114e-05, + "loss": 0.402, + "step": 8748 + }, + { + "epoch": 3.299019607843137, + "grad_norm": 0.14302387062299324, + "learning_rate": 2.504112394533083e-05, + "loss": 0.3947, + "step": 8749 + }, + { + "epoch": 3.2993966817496228, + "grad_norm": 0.13944541067070138, + "learning_rate": 2.5031358622307355e-05, + "loss": 0.3991, + "step": 8750 + }, + { + "epoch": 3.2997737556561084, + "grad_norm": 0.1479675810497832, + "learning_rate": 2.5021594336597464e-05, + "loss": 0.3994, + "step": 8751 + }, + { + "epoch": 3.300150829562594, + "grad_norm": 0.12299196546660439, + "learning_rate": 2.5011831088877783e-05, + "loss": 0.3924, + "step": 8752 + }, + { + "epoch": 3.3005279034690798, + "grad_norm": 0.15025308030841142, + "learning_rate": 2.500206887982492e-05, + "loss": 0.4069, + "step": 8753 + }, + { + "epoch": 3.3009049773755654, + "grad_norm": 0.13645407476350574, + "learning_rate": 2.499230771011538e-05, + "loss": 0.4108, + "step": 8754 + }, + { + "epoch": 3.301282051282051, + "grad_norm": 0.1286206218952415, + "learning_rate": 2.4982547580425612e-05, + "loss": 0.3917, + "step": 8755 + }, + { + "epoch": 3.3016591251885368, + "grad_norm": 0.1308919056759636, + "learning_rate": 2.4972788491431965e-05, + "loss": 0.3929, + "step": 8756 + }, + { + "epoch": 3.3020361990950224, + "grad_norm": 0.11678792801642011, + "learning_rate": 2.4963030443810753e-05, + "loss": 0.3944, + "step": 8757 + }, + { + "epoch": 3.302413273001508, + "grad_norm": 0.1087427441213665, + "learning_rate": 2.495327343823819e-05, + "loss": 0.4033, + "step": 8758 + }, + { + "epoch": 3.302790346907994, + "grad_norm": 0.127955355954563, + "learning_rate": 2.4943517475390434e-05, + "loss": 0.3948, + "step": 8759 + }, + { + "epoch": 3.3031674208144794, + "grad_norm": 0.11840534599303533, + "learning_rate": 2.4933762555943557e-05, + "loss": 0.3982, + "step": 8760 + }, + { + "epoch": 3.303544494720965, + "grad_norm": 0.12102951914706656, + "learning_rate": 2.492400868057358e-05, + "loss": 0.4024, + "step": 8761 + }, + { + "epoch": 3.303921568627451, + "grad_norm": 0.11993397684854652, + "learning_rate": 2.4914255849956424e-05, + "loss": 0.389, + "step": 8762 + }, + { + "epoch": 3.3042986425339365, + "grad_norm": 0.11257647113472401, + "learning_rate": 2.4904504064767957e-05, + "loss": 0.382, + "step": 8763 + }, + { + "epoch": 3.304675716440422, + "grad_norm": 0.12858654631605324, + "learning_rate": 2.4894753325683967e-05, + "loss": 0.3972, + "step": 8764 + }, + { + "epoch": 3.305052790346908, + "grad_norm": 0.12566457726057362, + "learning_rate": 2.4885003633380173e-05, + "loss": 0.391, + "step": 8765 + }, + { + "epoch": 3.3054298642533935, + "grad_norm": 0.13743265936673155, + "learning_rate": 2.4875254988532216e-05, + "loss": 0.4047, + "step": 8766 + }, + { + "epoch": 3.305806938159879, + "grad_norm": 0.14443735988075337, + "learning_rate": 2.486550739181569e-05, + "loss": 0.394, + "step": 8767 + }, + { + "epoch": 3.306184012066365, + "grad_norm": 0.11799430433798283, + "learning_rate": 2.4855760843906058e-05, + "loss": 0.394, + "step": 8768 + }, + { + "epoch": 3.3065610859728505, + "grad_norm": 0.1254756671637708, + "learning_rate": 2.484601534547878e-05, + "loss": 0.4015, + "step": 8769 + }, + { + "epoch": 3.306938159879336, + "grad_norm": 0.12472133327818412, + "learning_rate": 2.4836270897209193e-05, + "loss": 0.3946, + "step": 8770 + }, + { + "epoch": 3.307315233785822, + "grad_norm": 0.13357600276251058, + "learning_rate": 2.4826527499772567e-05, + "loss": 0.3947, + "step": 8771 + }, + { + "epoch": 3.3076923076923075, + "grad_norm": 0.12549028540253623, + "learning_rate": 2.481678515384415e-05, + "loss": 0.394, + "step": 8772 + }, + { + "epoch": 3.308069381598793, + "grad_norm": 0.12299155126957081, + "learning_rate": 2.4807043860099046e-05, + "loss": 0.3964, + "step": 8773 + }, + { + "epoch": 3.308446455505279, + "grad_norm": 0.15830514764713857, + "learning_rate": 2.479730361921232e-05, + "loss": 0.4093, + "step": 8774 + }, + { + "epoch": 3.3088235294117645, + "grad_norm": 0.12438570516542817, + "learning_rate": 2.4787564431858977e-05, + "loss": 0.3851, + "step": 8775 + }, + { + "epoch": 3.30920060331825, + "grad_norm": 0.14172639129286566, + "learning_rate": 2.4777826298713924e-05, + "loss": 0.3897, + "step": 8776 + }, + { + "epoch": 3.309577677224736, + "grad_norm": 0.12509153932656777, + "learning_rate": 2.4768089220451997e-05, + "loss": 0.3988, + "step": 8777 + }, + { + "epoch": 3.3099547511312215, + "grad_norm": 0.1779834778749211, + "learning_rate": 2.4758353197747988e-05, + "loss": 0.3971, + "step": 8778 + }, + { + "epoch": 3.310331825037707, + "grad_norm": 0.17580927310003686, + "learning_rate": 2.4748618231276584e-05, + "loss": 0.3983, + "step": 8779 + }, + { + "epoch": 3.310708898944193, + "grad_norm": 0.12845665212729782, + "learning_rate": 2.473888432171239e-05, + "loss": 0.3818, + "step": 8780 + }, + { + "epoch": 3.3110859728506785, + "grad_norm": 0.11635071776327802, + "learning_rate": 2.4729151469729998e-05, + "loss": 0.3945, + "step": 8781 + }, + { + "epoch": 3.311463046757164, + "grad_norm": 0.1389917056999923, + "learning_rate": 2.471941967600386e-05, + "loss": 0.4053, + "step": 8782 + }, + { + "epoch": 3.31184012066365, + "grad_norm": 0.1315240623965747, + "learning_rate": 2.4709688941208373e-05, + "loss": 0.3959, + "step": 8783 + }, + { + "epoch": 3.3122171945701355, + "grad_norm": 0.13246647538185982, + "learning_rate": 2.469995926601789e-05, + "loss": 0.4026, + "step": 8784 + }, + { + "epoch": 3.312594268476621, + "grad_norm": 0.12811246037568247, + "learning_rate": 2.469023065110666e-05, + "loss": 0.3937, + "step": 8785 + }, + { + "epoch": 3.312971342383107, + "grad_norm": 0.15578877766187296, + "learning_rate": 2.4680503097148848e-05, + "loss": 0.3938, + "step": 8786 + }, + { + "epoch": 3.3133484162895925, + "grad_norm": 0.1350400124136766, + "learning_rate": 2.4670776604818603e-05, + "loss": 0.398, + "step": 8787 + }, + { + "epoch": 3.313725490196078, + "grad_norm": 0.11682174262575908, + "learning_rate": 2.4661051174789944e-05, + "loss": 0.4007, + "step": 8788 + }, + { + "epoch": 3.314102564102564, + "grad_norm": 0.11871103381461211, + "learning_rate": 2.4651326807736812e-05, + "loss": 0.3884, + "step": 8789 + }, + { + "epoch": 3.3144796380090495, + "grad_norm": 0.13462479045461107, + "learning_rate": 2.4641603504333134e-05, + "loss": 0.4009, + "step": 8790 + }, + { + "epoch": 3.314856711915535, + "grad_norm": 0.11314272672338356, + "learning_rate": 2.4631881265252716e-05, + "loss": 0.3943, + "step": 8791 + }, + { + "epoch": 3.315233785822021, + "grad_norm": 0.13995000828646595, + "learning_rate": 2.4622160091169277e-05, + "loss": 0.388, + "step": 8792 + }, + { + "epoch": 3.3156108597285066, + "grad_norm": 0.11750914970373065, + "learning_rate": 2.4612439982756514e-05, + "loss": 0.3906, + "step": 8793 + }, + { + "epoch": 3.3159879336349922, + "grad_norm": 0.14141819194189487, + "learning_rate": 2.4602720940688017e-05, + "loss": 0.392, + "step": 8794 + }, + { + "epoch": 3.316365007541478, + "grad_norm": 0.14060707367237774, + "learning_rate": 2.4593002965637284e-05, + "loss": 0.4074, + "step": 8795 + }, + { + "epoch": 3.3167420814479636, + "grad_norm": 0.145680812475643, + "learning_rate": 2.4583286058277795e-05, + "loss": 0.392, + "step": 8796 + }, + { + "epoch": 3.3171191553544492, + "grad_norm": 0.1418210997190883, + "learning_rate": 2.4573570219282904e-05, + "loss": 0.3939, + "step": 8797 + }, + { + "epoch": 3.317496229260935, + "grad_norm": 0.12339945496139859, + "learning_rate": 2.4563855449325904e-05, + "loss": 0.4117, + "step": 8798 + }, + { + "epoch": 3.3178733031674206, + "grad_norm": 0.15803916351000372, + "learning_rate": 2.455414174908004e-05, + "loss": 0.3932, + "step": 8799 + }, + { + "epoch": 3.3182503770739062, + "grad_norm": 0.15392587359074406, + "learning_rate": 2.4544429119218462e-05, + "loss": 0.3978, + "step": 8800 + }, + { + "epoch": 3.318627450980392, + "grad_norm": 0.11518557783639817, + "learning_rate": 2.453471756041421e-05, + "loss": 0.3857, + "step": 8801 + }, + { + "epoch": 3.3190045248868776, + "grad_norm": 0.17817744919950063, + "learning_rate": 2.4525007073340333e-05, + "loss": 0.3873, + "step": 8802 + }, + { + "epoch": 3.3193815987933633, + "grad_norm": 0.16615786619973816, + "learning_rate": 2.4515297658669744e-05, + "loss": 0.3955, + "step": 8803 + }, + { + "epoch": 3.319758672699849, + "grad_norm": 0.11843580174951668, + "learning_rate": 2.450558931707527e-05, + "loss": 0.3923, + "step": 8804 + }, + { + "epoch": 3.3201357466063346, + "grad_norm": 0.12503618472582606, + "learning_rate": 2.4495882049229732e-05, + "loss": 0.3951, + "step": 8805 + }, + { + "epoch": 3.3205128205128207, + "grad_norm": 0.13826333048511355, + "learning_rate": 2.448617585580581e-05, + "loss": 0.4017, + "step": 8806 + }, + { + "epoch": 3.3208898944193064, + "grad_norm": 0.12463013625217706, + "learning_rate": 2.4476470737476134e-05, + "loss": 0.403, + "step": 8807 + }, + { + "epoch": 3.321266968325792, + "grad_norm": 0.1490875748229196, + "learning_rate": 2.4466766694913275e-05, + "loss": 0.3959, + "step": 8808 + }, + { + "epoch": 3.3216440422322777, + "grad_norm": 0.15329201739106826, + "learning_rate": 2.4457063728789708e-05, + "loss": 0.3969, + "step": 8809 + }, + { + "epoch": 3.3220211161387634, + "grad_norm": 0.12577542706828923, + "learning_rate": 2.4447361839777825e-05, + "loss": 0.3989, + "step": 8810 + }, + { + "epoch": 3.322398190045249, + "grad_norm": 0.16647631499464946, + "learning_rate": 2.4437661028549986e-05, + "loss": 0.3976, + "step": 8811 + }, + { + "epoch": 3.3227752639517347, + "grad_norm": 0.18945621439629487, + "learning_rate": 2.4427961295778434e-05, + "loss": 0.3992, + "step": 8812 + }, + { + "epoch": 3.3231523378582204, + "grad_norm": 0.1232642389903967, + "learning_rate": 2.4418262642135338e-05, + "loss": 0.3998, + "step": 8813 + }, + { + "epoch": 3.323529411764706, + "grad_norm": 0.14533704511975365, + "learning_rate": 2.4408565068292827e-05, + "loss": 0.3911, + "step": 8814 + }, + { + "epoch": 3.3239064856711917, + "grad_norm": 0.14363453025764839, + "learning_rate": 2.439886857492294e-05, + "loss": 0.3977, + "step": 8815 + }, + { + "epoch": 3.3242835595776774, + "grad_norm": 0.12802693582710475, + "learning_rate": 2.43891731626976e-05, + "loss": 0.385, + "step": 8816 + }, + { + "epoch": 3.324660633484163, + "grad_norm": 0.14829495213386343, + "learning_rate": 2.437947883228873e-05, + "loss": 0.3975, + "step": 8817 + }, + { + "epoch": 3.3250377073906487, + "grad_norm": 0.11679543218809023, + "learning_rate": 2.4369785584368116e-05, + "loss": 0.3924, + "step": 8818 + }, + { + "epoch": 3.3254147812971344, + "grad_norm": 0.1451983377320573, + "learning_rate": 2.4360093419607486e-05, + "loss": 0.4093, + "step": 8819 + }, + { + "epoch": 3.32579185520362, + "grad_norm": 0.12738994655866834, + "learning_rate": 2.4350402338678523e-05, + "loss": 0.406, + "step": 8820 + }, + { + "epoch": 3.3261689291101058, + "grad_norm": 0.12355115355214036, + "learning_rate": 2.4340712342252802e-05, + "loss": 0.3922, + "step": 8821 + }, + { + "epoch": 3.3265460030165914, + "grad_norm": 0.14528640851581928, + "learning_rate": 2.4331023431001803e-05, + "loss": 0.4042, + "step": 8822 + }, + { + "epoch": 3.326923076923077, + "grad_norm": 0.11452962766249306, + "learning_rate": 2.4321335605597006e-05, + "loss": 0.3915, + "step": 8823 + }, + { + "epoch": 3.3273001508295628, + "grad_norm": 0.13172289890656536, + "learning_rate": 2.431164886670974e-05, + "loss": 0.3927, + "step": 8824 + }, + { + "epoch": 3.3276772247360484, + "grad_norm": 0.13555707940953157, + "learning_rate": 2.430196321501127e-05, + "loss": 0.3961, + "step": 8825 + }, + { + "epoch": 3.328054298642534, + "grad_norm": 0.12053779242006023, + "learning_rate": 2.4292278651172852e-05, + "loss": 0.402, + "step": 8826 + }, + { + "epoch": 3.3284313725490198, + "grad_norm": 0.13036759951609037, + "learning_rate": 2.4282595175865585e-05, + "loss": 0.3948, + "step": 8827 + }, + { + "epoch": 3.3288084464555054, + "grad_norm": 0.159929299195437, + "learning_rate": 2.4272912789760515e-05, + "loss": 0.3966, + "step": 8828 + }, + { + "epoch": 3.329185520361991, + "grad_norm": 0.11281834784498176, + "learning_rate": 2.4263231493528653e-05, + "loss": 0.3993, + "step": 8829 + }, + { + "epoch": 3.329562594268477, + "grad_norm": 0.13438054139546354, + "learning_rate": 2.4253551287840895e-05, + "loss": 0.3859, + "step": 8830 + }, + { + "epoch": 3.3299396681749625, + "grad_norm": 0.1351625947338887, + "learning_rate": 2.424387217336805e-05, + "loss": 0.3895, + "step": 8831 + }, + { + "epoch": 3.330316742081448, + "grad_norm": 0.12974571561459114, + "learning_rate": 2.42341941507809e-05, + "loss": 0.3915, + "step": 8832 + }, + { + "epoch": 3.330693815987934, + "grad_norm": 0.1345480465528601, + "learning_rate": 2.4224517220750113e-05, + "loss": 0.3949, + "step": 8833 + }, + { + "epoch": 3.3310708898944195, + "grad_norm": 0.11918912232402212, + "learning_rate": 2.4214841383946272e-05, + "loss": 0.404, + "step": 8834 + }, + { + "epoch": 3.331447963800905, + "grad_norm": 0.1303824615294782, + "learning_rate": 2.4205166641039935e-05, + "loss": 0.3933, + "step": 8835 + }, + { + "epoch": 3.331825037707391, + "grad_norm": 0.12990943360060878, + "learning_rate": 2.4195492992701547e-05, + "loss": 0.3957, + "step": 8836 + }, + { + "epoch": 3.3322021116138765, + "grad_norm": 0.11785638798090013, + "learning_rate": 2.4185820439601456e-05, + "loss": 0.3941, + "step": 8837 + }, + { + "epoch": 3.332579185520362, + "grad_norm": 0.12869626012152943, + "learning_rate": 2.4176148982409993e-05, + "loss": 0.3825, + "step": 8838 + }, + { + "epoch": 3.332956259426848, + "grad_norm": 0.11315137297084235, + "learning_rate": 2.416647862179738e-05, + "loss": 0.4016, + "step": 8839 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 0.1204158073427281, + "learning_rate": 2.4156809358433728e-05, + "loss": 0.3807, + "step": 8840 + }, + { + "epoch": 3.333710407239819, + "grad_norm": 0.11387115575417964, + "learning_rate": 2.414714119298916e-05, + "loss": 0.3988, + "step": 8841 + }, + { + "epoch": 3.334087481146305, + "grad_norm": 0.1346528683931007, + "learning_rate": 2.413747412613364e-05, + "loss": 0.4033, + "step": 8842 + }, + { + "epoch": 3.3344645550527905, + "grad_norm": 0.11868318954327768, + "learning_rate": 2.4127808158537075e-05, + "loss": 0.4002, + "step": 8843 + }, + { + "epoch": 3.334841628959276, + "grad_norm": 0.12107279388571585, + "learning_rate": 2.4118143290869345e-05, + "loss": 0.4018, + "step": 8844 + }, + { + "epoch": 3.335218702865762, + "grad_norm": 0.1073375669625344, + "learning_rate": 2.4108479523800198e-05, + "loss": 0.385, + "step": 8845 + }, + { + "epoch": 3.3355957767722475, + "grad_norm": 0.13866962802978752, + "learning_rate": 2.409881685799931e-05, + "loss": 0.3898, + "step": 8846 + }, + { + "epoch": 3.335972850678733, + "grad_norm": 0.12240635351028421, + "learning_rate": 2.408915529413632e-05, + "loss": 0.3895, + "step": 8847 + }, + { + "epoch": 3.336349924585219, + "grad_norm": 0.1452869627097657, + "learning_rate": 2.4079494832880747e-05, + "loss": 0.4009, + "step": 8848 + }, + { + "epoch": 3.3367269984917045, + "grad_norm": 0.14930830427011474, + "learning_rate": 2.406983547490207e-05, + "loss": 0.4139, + "step": 8849 + }, + { + "epoch": 3.33710407239819, + "grad_norm": 0.11896640478550317, + "learning_rate": 2.4060177220869666e-05, + "loss": 0.4051, + "step": 8850 + }, + { + "epoch": 3.337481146304676, + "grad_norm": 0.12735098074114257, + "learning_rate": 2.4050520071452827e-05, + "loss": 0.4, + "step": 8851 + }, + { + "epoch": 3.3378582202111615, + "grad_norm": 0.13634684978823153, + "learning_rate": 2.4040864027320818e-05, + "loss": 0.4012, + "step": 8852 + }, + { + "epoch": 3.338235294117647, + "grad_norm": 0.13172407773302205, + "learning_rate": 2.4031209089142773e-05, + "loss": 0.3966, + "step": 8853 + }, + { + "epoch": 3.338612368024133, + "grad_norm": 0.11126095110478082, + "learning_rate": 2.4021555257587758e-05, + "loss": 0.3976, + "step": 8854 + }, + { + "epoch": 3.3389894419306185, + "grad_norm": 0.1230070349892988, + "learning_rate": 2.4011902533324807e-05, + "loss": 0.4053, + "step": 8855 + }, + { + "epoch": 3.339366515837104, + "grad_norm": 0.12227938605256705, + "learning_rate": 2.400225091702283e-05, + "loss": 0.3983, + "step": 8856 + }, + { + "epoch": 3.33974358974359, + "grad_norm": 0.12722025125977193, + "learning_rate": 2.399260040935066e-05, + "loss": 0.3931, + "step": 8857 + }, + { + "epoch": 3.3401206636500755, + "grad_norm": 0.11088294251103341, + "learning_rate": 2.3982951010977096e-05, + "loss": 0.3962, + "step": 8858 + }, + { + "epoch": 3.340497737556561, + "grad_norm": 0.12748665252176314, + "learning_rate": 2.3973302722570822e-05, + "loss": 0.3916, + "step": 8859 + }, + { + "epoch": 3.340874811463047, + "grad_norm": 0.12593194432541682, + "learning_rate": 2.3963655544800437e-05, + "loss": 0.3916, + "step": 8860 + }, + { + "epoch": 3.3412518853695325, + "grad_norm": 0.1198963989726914, + "learning_rate": 2.3954009478334514e-05, + "loss": 0.4023, + "step": 8861 + }, + { + "epoch": 3.341628959276018, + "grad_norm": 0.13099084682868392, + "learning_rate": 2.3944364523841498e-05, + "loss": 0.3854, + "step": 8862 + }, + { + "epoch": 3.342006033182504, + "grad_norm": 0.1642548685880207, + "learning_rate": 2.3934720681989764e-05, + "loss": 0.4086, + "step": 8863 + }, + { + "epoch": 3.3423831070889896, + "grad_norm": 0.1375688578966589, + "learning_rate": 2.392507795344766e-05, + "loss": 0.3912, + "step": 8864 + }, + { + "epoch": 3.3427601809954752, + "grad_norm": 0.13728827446026431, + "learning_rate": 2.3915436338883388e-05, + "loss": 0.4041, + "step": 8865 + }, + { + "epoch": 3.343137254901961, + "grad_norm": 0.15848722263151444, + "learning_rate": 2.3905795838965098e-05, + "loss": 0.3834, + "step": 8866 + }, + { + "epoch": 3.3435143288084466, + "grad_norm": 0.15242806931573058, + "learning_rate": 2.3896156454360897e-05, + "loss": 0.3971, + "step": 8867 + }, + { + "epoch": 3.3438914027149322, + "grad_norm": 0.1316176449116502, + "learning_rate": 2.388651818573877e-05, + "loss": 0.4, + "step": 8868 + }, + { + "epoch": 3.344268476621418, + "grad_norm": 0.1739565601799625, + "learning_rate": 2.3876881033766623e-05, + "loss": 0.3982, + "step": 8869 + }, + { + "epoch": 3.3446455505279036, + "grad_norm": 0.14890664816608548, + "learning_rate": 2.3867244999112337e-05, + "loss": 0.4017, + "step": 8870 + }, + { + "epoch": 3.3450226244343892, + "grad_norm": 0.1576122849743633, + "learning_rate": 2.385761008244366e-05, + "loss": 0.4013, + "step": 8871 + }, + { + "epoch": 3.345399698340875, + "grad_norm": 0.1341112507940579, + "learning_rate": 2.3847976284428267e-05, + "loss": 0.3975, + "step": 8872 + }, + { + "epoch": 3.3457767722473606, + "grad_norm": 0.16404172237780212, + "learning_rate": 2.3838343605733812e-05, + "loss": 0.3892, + "step": 8873 + }, + { + "epoch": 3.3461538461538463, + "grad_norm": 0.1365370590315432, + "learning_rate": 2.382871204702781e-05, + "loss": 0.4054, + "step": 8874 + }, + { + "epoch": 3.346530920060332, + "grad_norm": 0.14810621215172814, + "learning_rate": 2.38190816089777e-05, + "loss": 0.405, + "step": 8875 + }, + { + "epoch": 3.3469079939668176, + "grad_norm": 0.14217655923983633, + "learning_rate": 2.3809452292250902e-05, + "loss": 0.3793, + "step": 8876 + }, + { + "epoch": 3.3472850678733033, + "grad_norm": 0.12900312185538115, + "learning_rate": 2.3799824097514693e-05, + "loss": 0.4053, + "step": 8877 + }, + { + "epoch": 3.347662141779789, + "grad_norm": 0.13518619588725259, + "learning_rate": 2.3790197025436298e-05, + "loss": 0.4001, + "step": 8878 + }, + { + "epoch": 3.3480392156862746, + "grad_norm": 0.17974018179680587, + "learning_rate": 2.378057107668288e-05, + "loss": 0.3934, + "step": 8879 + }, + { + "epoch": 3.3484162895927603, + "grad_norm": 0.1712618382012229, + "learning_rate": 2.3770946251921502e-05, + "loss": 0.3956, + "step": 8880 + }, + { + "epoch": 3.348793363499246, + "grad_norm": 0.11100485786006031, + "learning_rate": 2.376132255181913e-05, + "loss": 0.3996, + "step": 8881 + }, + { + "epoch": 3.3491704374057316, + "grad_norm": 0.2105182106523246, + "learning_rate": 2.375169997704273e-05, + "loss": 0.4023, + "step": 8882 + }, + { + "epoch": 3.3495475113122173, + "grad_norm": 0.19964641099781627, + "learning_rate": 2.3742078528259098e-05, + "loss": 0.3977, + "step": 8883 + }, + { + "epoch": 3.349924585218703, + "grad_norm": 0.12851598028745242, + "learning_rate": 2.3732458206134997e-05, + "loss": 0.4031, + "step": 8884 + }, + { + "epoch": 3.3503016591251886, + "grad_norm": 0.1721032729497728, + "learning_rate": 2.3722839011337125e-05, + "loss": 0.3909, + "step": 8885 + }, + { + "epoch": 3.3506787330316743, + "grad_norm": 0.17643029443030114, + "learning_rate": 2.3713220944532064e-05, + "loss": 0.3997, + "step": 8886 + }, + { + "epoch": 3.35105580693816, + "grad_norm": 0.1647016743872002, + "learning_rate": 2.3703604006386343e-05, + "loss": 0.4008, + "step": 8887 + }, + { + "epoch": 3.3514328808446456, + "grad_norm": 0.1686439562647831, + "learning_rate": 2.369398819756642e-05, + "loss": 0.4037, + "step": 8888 + }, + { + "epoch": 3.3518099547511313, + "grad_norm": 0.19018999913244145, + "learning_rate": 2.3684373518738646e-05, + "loss": 0.391, + "step": 8889 + }, + { + "epoch": 3.352187028657617, + "grad_norm": 0.17607033135354316, + "learning_rate": 2.367475997056931e-05, + "loss": 0.3939, + "step": 8890 + }, + { + "epoch": 3.3525641025641026, + "grad_norm": 0.1705956498449579, + "learning_rate": 2.3665147553724638e-05, + "loss": 0.4016, + "step": 8891 + }, + { + "epoch": 3.3529411764705883, + "grad_norm": 0.17300960129288923, + "learning_rate": 2.3655536268870744e-05, + "loss": 0.4041, + "step": 8892 + }, + { + "epoch": 3.353318250377074, + "grad_norm": 0.17817706061023675, + "learning_rate": 2.3645926116673684e-05, + "loss": 0.3999, + "step": 8893 + }, + { + "epoch": 3.3536953242835597, + "grad_norm": 0.15401633928748168, + "learning_rate": 2.363631709779945e-05, + "loss": 0.3937, + "step": 8894 + }, + { + "epoch": 3.3540723981900453, + "grad_norm": 0.13617450907667314, + "learning_rate": 2.3626709212913922e-05, + "loss": 0.4058, + "step": 8895 + }, + { + "epoch": 3.354449472096531, + "grad_norm": 0.1876706892499079, + "learning_rate": 2.3617102462682918e-05, + "loss": 0.3986, + "step": 8896 + }, + { + "epoch": 3.3548265460030167, + "grad_norm": 0.14432467344124855, + "learning_rate": 2.3607496847772193e-05, + "loss": 0.4081, + "step": 8897 + }, + { + "epoch": 3.3552036199095023, + "grad_norm": 0.1358253501940006, + "learning_rate": 2.359789236884739e-05, + "loss": 0.3855, + "step": 8898 + }, + { + "epoch": 3.355580693815988, + "grad_norm": 0.13383423370915956, + "learning_rate": 2.358828902657409e-05, + "loss": 0.3974, + "step": 8899 + }, + { + "epoch": 3.3559577677224737, + "grad_norm": 0.13867500665517507, + "learning_rate": 2.3578686821617812e-05, + "loss": 0.3911, + "step": 8900 + }, + { + "epoch": 3.3563348416289593, + "grad_norm": 0.1166094521182452, + "learning_rate": 2.356908575464397e-05, + "loss": 0.4018, + "step": 8901 + }, + { + "epoch": 3.356711915535445, + "grad_norm": 0.13528904405028241, + "learning_rate": 2.355948582631791e-05, + "loss": 0.3996, + "step": 8902 + }, + { + "epoch": 3.3570889894419307, + "grad_norm": 0.11367269225406536, + "learning_rate": 2.3549887037304894e-05, + "loss": 0.3826, + "step": 8903 + }, + { + "epoch": 3.3574660633484164, + "grad_norm": 0.1265200609859795, + "learning_rate": 2.3540289388270124e-05, + "loss": 0.3971, + "step": 8904 + }, + { + "epoch": 3.357843137254902, + "grad_norm": 0.12220533885081888, + "learning_rate": 2.3530692879878692e-05, + "loss": 0.4058, + "step": 8905 + }, + { + "epoch": 3.3582202111613877, + "grad_norm": 0.12942469602753107, + "learning_rate": 2.3521097512795636e-05, + "loss": 0.4003, + "step": 8906 + }, + { + "epoch": 3.3585972850678734, + "grad_norm": 0.13100012302437997, + "learning_rate": 2.3511503287685916e-05, + "loss": 0.3934, + "step": 8907 + }, + { + "epoch": 3.358974358974359, + "grad_norm": 0.12932347409558365, + "learning_rate": 2.3501910205214386e-05, + "loss": 0.4002, + "step": 8908 + }, + { + "epoch": 3.3593514328808447, + "grad_norm": 0.14214051189977794, + "learning_rate": 2.3492318266045837e-05, + "loss": 0.3859, + "step": 8909 + }, + { + "epoch": 3.3597285067873304, + "grad_norm": 0.13123900574840353, + "learning_rate": 2.3482727470845007e-05, + "loss": 0.3984, + "step": 8910 + }, + { + "epoch": 3.360105580693816, + "grad_norm": 0.14672960859316195, + "learning_rate": 2.3473137820276496e-05, + "loss": 0.3869, + "step": 8911 + }, + { + "epoch": 3.3604826546003017, + "grad_norm": 0.14871857932878502, + "learning_rate": 2.3463549315004882e-05, + "loss": 0.397, + "step": 8912 + }, + { + "epoch": 3.3608597285067874, + "grad_norm": 0.11443071748623901, + "learning_rate": 2.345396195569464e-05, + "loss": 0.3929, + "step": 8913 + }, + { + "epoch": 3.361236802413273, + "grad_norm": 0.14466696145826521, + "learning_rate": 2.3444375743010146e-05, + "loss": 0.3951, + "step": 8914 + }, + { + "epoch": 3.3616138763197587, + "grad_norm": 0.11663428189452217, + "learning_rate": 2.343479067761573e-05, + "loss": 0.394, + "step": 8915 + }, + { + "epoch": 3.3619909502262444, + "grad_norm": 0.13907575660968224, + "learning_rate": 2.3425206760175635e-05, + "loss": 0.3927, + "step": 8916 + }, + { + "epoch": 3.36236802413273, + "grad_norm": 0.16847658516844372, + "learning_rate": 2.3415623991354e-05, + "loss": 0.4012, + "step": 8917 + }, + { + "epoch": 3.3627450980392157, + "grad_norm": 0.142583988468848, + "learning_rate": 2.3406042371814915e-05, + "loss": 0.3939, + "step": 8918 + }, + { + "epoch": 3.3631221719457014, + "grad_norm": 0.12465751763191024, + "learning_rate": 2.3396461902222384e-05, + "loss": 0.4054, + "step": 8919 + }, + { + "epoch": 3.363499245852187, + "grad_norm": 0.13865475244406275, + "learning_rate": 2.3386882583240293e-05, + "loss": 0.3883, + "step": 8920 + }, + { + "epoch": 3.3638763197586727, + "grad_norm": 0.15459878827202186, + "learning_rate": 2.3377304415532526e-05, + "loss": 0.3921, + "step": 8921 + }, + { + "epoch": 3.3642533936651584, + "grad_norm": 0.11786335626813553, + "learning_rate": 2.3367727399762817e-05, + "loss": 0.4043, + "step": 8922 + }, + { + "epoch": 3.364630467571644, + "grad_norm": 0.13678914090893993, + "learning_rate": 2.335815153659483e-05, + "loss": 0.3952, + "step": 8923 + }, + { + "epoch": 3.3650075414781297, + "grad_norm": 0.15056731899073564, + "learning_rate": 2.3348576826692196e-05, + "loss": 0.4069, + "step": 8924 + }, + { + "epoch": 3.3653846153846154, + "grad_norm": 0.11098295876777348, + "learning_rate": 2.3339003270718424e-05, + "loss": 0.3924, + "step": 8925 + }, + { + "epoch": 3.365761689291101, + "grad_norm": 0.13247458464218542, + "learning_rate": 2.3329430869336917e-05, + "loss": 0.3972, + "step": 8926 + }, + { + "epoch": 3.3661387631975868, + "grad_norm": 0.13970467538806478, + "learning_rate": 2.3319859623211096e-05, + "loss": 0.3961, + "step": 8927 + }, + { + "epoch": 3.3665158371040724, + "grad_norm": 0.12206481781217315, + "learning_rate": 2.3310289533004192e-05, + "loss": 0.3937, + "step": 8928 + }, + { + "epoch": 3.366892911010558, + "grad_norm": 0.12419589382995094, + "learning_rate": 2.3300720599379423e-05, + "loss": 0.396, + "step": 8929 + }, + { + "epoch": 3.3672699849170438, + "grad_norm": 0.16555552228873635, + "learning_rate": 2.3291152822999895e-05, + "loss": 0.3953, + "step": 8930 + }, + { + "epoch": 3.3676470588235294, + "grad_norm": 0.1347396539126926, + "learning_rate": 2.328158620452868e-05, + "loss": 0.4041, + "step": 8931 + }, + { + "epoch": 3.368024132730015, + "grad_norm": 0.11949442138784454, + "learning_rate": 2.3272020744628678e-05, + "loss": 0.3996, + "step": 8932 + }, + { + "epoch": 3.3684012066365008, + "grad_norm": 0.1627911033904635, + "learning_rate": 2.3262456443962832e-05, + "loss": 0.3987, + "step": 8933 + }, + { + "epoch": 3.3687782805429864, + "grad_norm": 0.15387811419105382, + "learning_rate": 2.325289330319389e-05, + "loss": 0.4122, + "step": 8934 + }, + { + "epoch": 3.369155354449472, + "grad_norm": 0.11588517967347073, + "learning_rate": 2.324333132298459e-05, + "loss": 0.3877, + "step": 8935 + }, + { + "epoch": 3.369532428355958, + "grad_norm": 0.1633235059202232, + "learning_rate": 2.3233770503997562e-05, + "loss": 0.4007, + "step": 8936 + }, + { + "epoch": 3.3699095022624435, + "grad_norm": 0.1616736947390285, + "learning_rate": 2.322421084689539e-05, + "loss": 0.3919, + "step": 8937 + }, + { + "epoch": 3.370286576168929, + "grad_norm": 0.12931030484941847, + "learning_rate": 2.3214652352340493e-05, + "loss": 0.39, + "step": 8938 + }, + { + "epoch": 3.370663650075415, + "grad_norm": 0.1394722771760827, + "learning_rate": 2.3205095020995323e-05, + "loss": 0.4042, + "step": 8939 + }, + { + "epoch": 3.3710407239819005, + "grad_norm": 0.12172278047544924, + "learning_rate": 2.3195538853522167e-05, + "loss": 0.3941, + "step": 8940 + }, + { + "epoch": 3.371417797888386, + "grad_norm": 0.12331979489579749, + "learning_rate": 2.3185983850583257e-05, + "loss": 0.3927, + "step": 8941 + }, + { + "epoch": 3.371794871794872, + "grad_norm": 0.12884770404161125, + "learning_rate": 2.3176430012840748e-05, + "loss": 0.3919, + "step": 8942 + }, + { + "epoch": 3.3721719457013575, + "grad_norm": 0.12913205799723376, + "learning_rate": 2.316687734095674e-05, + "loss": 0.392, + "step": 8943 + }, + { + "epoch": 3.372549019607843, + "grad_norm": 0.11722787456596498, + "learning_rate": 2.3157325835593162e-05, + "loss": 0.4054, + "step": 8944 + }, + { + "epoch": 3.372926093514329, + "grad_norm": 0.1217638050780869, + "learning_rate": 2.3147775497411998e-05, + "loss": 0.3908, + "step": 8945 + }, + { + "epoch": 3.3733031674208145, + "grad_norm": 0.1262629326482507, + "learning_rate": 2.313822632707503e-05, + "loss": 0.4018, + "step": 8946 + }, + { + "epoch": 3.3736802413273, + "grad_norm": 0.10983761334980666, + "learning_rate": 2.3128678325244024e-05, + "loss": 0.3928, + "step": 8947 + }, + { + "epoch": 3.374057315233786, + "grad_norm": 0.11937662358176668, + "learning_rate": 2.3119131492580638e-05, + "loss": 0.3932, + "step": 8948 + }, + { + "epoch": 3.3744343891402715, + "grad_norm": 0.12922637268273668, + "learning_rate": 2.3109585829746474e-05, + "loss": 0.3973, + "step": 8949 + }, + { + "epoch": 3.374811463046757, + "grad_norm": 0.11127950698763045, + "learning_rate": 2.3100041337403026e-05, + "loss": 0.3956, + "step": 8950 + }, + { + "epoch": 3.375188536953243, + "grad_norm": 0.1211940369484829, + "learning_rate": 2.309049801621174e-05, + "loss": 0.4014, + "step": 8951 + }, + { + "epoch": 3.3755656108597285, + "grad_norm": 0.11089561833540497, + "learning_rate": 2.3080955866833926e-05, + "loss": 0.4013, + "step": 8952 + }, + { + "epoch": 3.375942684766214, + "grad_norm": 0.11514163830945115, + "learning_rate": 2.307141488993087e-05, + "loss": 0.4, + "step": 8953 + }, + { + "epoch": 3.3763197586727, + "grad_norm": 0.12129113040722592, + "learning_rate": 2.3061875086163737e-05, + "loss": 0.3968, + "step": 8954 + }, + { + "epoch": 3.3766968325791855, + "grad_norm": 0.1134127672800099, + "learning_rate": 2.3052336456193637e-05, + "loss": 0.4008, + "step": 8955 + }, + { + "epoch": 3.377073906485671, + "grad_norm": 0.11524832601153645, + "learning_rate": 2.304279900068159e-05, + "loss": 0.3927, + "step": 8956 + }, + { + "epoch": 3.377450980392157, + "grad_norm": 0.14808974644998274, + "learning_rate": 2.303326272028855e-05, + "loss": 0.4013, + "step": 8957 + }, + { + "epoch": 3.3778280542986425, + "grad_norm": 0.11798542759274219, + "learning_rate": 2.302372761567534e-05, + "loss": 0.4011, + "step": 8958 + }, + { + "epoch": 3.378205128205128, + "grad_norm": 0.14189210504874783, + "learning_rate": 2.3014193687502743e-05, + "loss": 0.3915, + "step": 8959 + }, + { + "epoch": 3.378582202111614, + "grad_norm": 0.13914215762091708, + "learning_rate": 2.3004660936431455e-05, + "loss": 0.3814, + "step": 8960 + }, + { + "epoch": 3.3789592760180995, + "grad_norm": 0.12351688192203715, + "learning_rate": 2.299512936312209e-05, + "loss": 0.3986, + "step": 8961 + }, + { + "epoch": 3.379336349924585, + "grad_norm": 0.13896526412018087, + "learning_rate": 2.2985598968235183e-05, + "loss": 0.3793, + "step": 8962 + }, + { + "epoch": 3.379713423831071, + "grad_norm": 0.15239540975815238, + "learning_rate": 2.2976069752431185e-05, + "loss": 0.3847, + "step": 8963 + }, + { + "epoch": 3.3800904977375565, + "grad_norm": 0.10672541806066382, + "learning_rate": 2.2966541716370444e-05, + "loss": 0.3923, + "step": 8964 + }, + { + "epoch": 3.380467571644042, + "grad_norm": 0.15986040585111386, + "learning_rate": 2.295701486071325e-05, + "loss": 0.397, + "step": 8965 + }, + { + "epoch": 3.380844645550528, + "grad_norm": 0.14428769442700642, + "learning_rate": 2.2947489186119807e-05, + "loss": 0.4042, + "step": 8966 + }, + { + "epoch": 3.3812217194570136, + "grad_norm": 0.1319384239511992, + "learning_rate": 2.2937964693250237e-05, + "loss": 0.4065, + "step": 8967 + }, + { + "epoch": 3.3815987933634992, + "grad_norm": 0.12652740834316262, + "learning_rate": 2.2928441382764585e-05, + "loss": 0.3931, + "step": 8968 + }, + { + "epoch": 3.381975867269985, + "grad_norm": 0.13130856110008246, + "learning_rate": 2.29189192553228e-05, + "loss": 0.3977, + "step": 8969 + }, + { + "epoch": 3.3823529411764706, + "grad_norm": 0.1160670339928223, + "learning_rate": 2.2909398311584775e-05, + "loss": 0.4008, + "step": 8970 + }, + { + "epoch": 3.3827300150829562, + "grad_norm": 0.12953379903365328, + "learning_rate": 2.2899878552210276e-05, + "loss": 0.3901, + "step": 8971 + }, + { + "epoch": 3.383107088989442, + "grad_norm": 0.11310062275112644, + "learning_rate": 2.2890359977859017e-05, + "loss": 0.3987, + "step": 8972 + }, + { + "epoch": 3.3834841628959276, + "grad_norm": 0.1231355449541035, + "learning_rate": 2.288084258919064e-05, + "loss": 0.3902, + "step": 8973 + }, + { + "epoch": 3.3838612368024132, + "grad_norm": 0.12222320057573305, + "learning_rate": 2.287132638686469e-05, + "loss": 0.3975, + "step": 8974 + }, + { + "epoch": 3.384238310708899, + "grad_norm": 0.12660724020018593, + "learning_rate": 2.2861811371540626e-05, + "loss": 0.3954, + "step": 8975 + }, + { + "epoch": 3.3846153846153846, + "grad_norm": 0.10708617800477142, + "learning_rate": 2.285229754387783e-05, + "loss": 0.4046, + "step": 8976 + }, + { + "epoch": 3.3849924585218703, + "grad_norm": 0.13738700974434706, + "learning_rate": 2.284278490453562e-05, + "loss": 0.4032, + "step": 8977 + }, + { + "epoch": 3.385369532428356, + "grad_norm": 0.12867809189669632, + "learning_rate": 2.2833273454173188e-05, + "loss": 0.3881, + "step": 8978 + }, + { + "epoch": 3.3857466063348416, + "grad_norm": 0.1109411826738322, + "learning_rate": 2.282376319344967e-05, + "loss": 0.3907, + "step": 8979 + }, + { + "epoch": 3.3861236802413273, + "grad_norm": 0.12009632326821322, + "learning_rate": 2.2814254123024133e-05, + "loss": 0.3942, + "step": 8980 + }, + { + "epoch": 3.386500754147813, + "grad_norm": 0.1114514904333011, + "learning_rate": 2.280474624355554e-05, + "loss": 0.4013, + "step": 8981 + }, + { + "epoch": 3.3868778280542986, + "grad_norm": 0.10755323864809826, + "learning_rate": 2.2795239555702784e-05, + "loss": 0.3867, + "step": 8982 + }, + { + "epoch": 3.3872549019607843, + "grad_norm": 0.11394011078349148, + "learning_rate": 2.2785734060124676e-05, + "loss": 0.3949, + "step": 8983 + }, + { + "epoch": 3.38763197586727, + "grad_norm": 0.11840634638980471, + "learning_rate": 2.2776229757479914e-05, + "loss": 0.3915, + "step": 8984 + }, + { + "epoch": 3.3880090497737556, + "grad_norm": 0.1191327634043139, + "learning_rate": 2.2766726648427158e-05, + "loss": 0.3824, + "step": 8985 + }, + { + "epoch": 3.3883861236802413, + "grad_norm": 0.1279744917993156, + "learning_rate": 2.275722473362495e-05, + "loss": 0.3958, + "step": 8986 + }, + { + "epoch": 3.388763197586727, + "grad_norm": 0.11990787204624129, + "learning_rate": 2.274772401373178e-05, + "loss": 0.3932, + "step": 8987 + }, + { + "epoch": 3.3891402714932126, + "grad_norm": 0.1089029915050049, + "learning_rate": 2.2738224489406032e-05, + "loss": 0.3892, + "step": 8988 + }, + { + "epoch": 3.3895173453996983, + "grad_norm": 0.13257559669029823, + "learning_rate": 2.2728726161306013e-05, + "loss": 0.4011, + "step": 8989 + }, + { + "epoch": 3.389894419306184, + "grad_norm": 0.13305424064890636, + "learning_rate": 2.271922903008997e-05, + "loss": 0.3965, + "step": 8990 + }, + { + "epoch": 3.3902714932126696, + "grad_norm": 0.12756275086477029, + "learning_rate": 2.2709733096416013e-05, + "loss": 0.3993, + "step": 8991 + }, + { + "epoch": 3.3906485671191553, + "grad_norm": 0.1337978097812285, + "learning_rate": 2.2700238360942213e-05, + "loss": 0.4023, + "step": 8992 + }, + { + "epoch": 3.391025641025641, + "grad_norm": 0.14931569911012124, + "learning_rate": 2.2690744824326552e-05, + "loss": 0.3956, + "step": 8993 + }, + { + "epoch": 3.3914027149321266, + "grad_norm": 0.13689533130509643, + "learning_rate": 2.268125248722692e-05, + "loss": 0.4027, + "step": 8994 + }, + { + "epoch": 3.3917797888386123, + "grad_norm": 0.1379773056627328, + "learning_rate": 2.2671761350301126e-05, + "loss": 0.4046, + "step": 8995 + }, + { + "epoch": 3.392156862745098, + "grad_norm": 0.13231156886217402, + "learning_rate": 2.266227141420692e-05, + "loss": 0.3959, + "step": 8996 + }, + { + "epoch": 3.3925339366515836, + "grad_norm": 0.13934961545189914, + "learning_rate": 2.2652782679601915e-05, + "loss": 0.3988, + "step": 8997 + }, + { + "epoch": 3.3929110105580693, + "grad_norm": 0.12344484242773111, + "learning_rate": 2.2643295147143674e-05, + "loss": 0.3999, + "step": 8998 + }, + { + "epoch": 3.393288084464555, + "grad_norm": 0.12653826676544752, + "learning_rate": 2.263380881748969e-05, + "loss": 0.3867, + "step": 8999 + }, + { + "epoch": 3.3936651583710407, + "grad_norm": 0.1346821134316575, + "learning_rate": 2.2624323691297347e-05, + "loss": 0.4013, + "step": 9000 + }, + { + "epoch": 3.3940422322775263, + "grad_norm": 0.13819019238433689, + "learning_rate": 2.261483976922396e-05, + "loss": 0.4067, + "step": 9001 + }, + { + "epoch": 3.394419306184012, + "grad_norm": 0.14148353563540467, + "learning_rate": 2.2605357051926777e-05, + "loss": 0.3988, + "step": 9002 + }, + { + "epoch": 3.3947963800904977, + "grad_norm": 0.13997347558026432, + "learning_rate": 2.2595875540062905e-05, + "loss": 0.3902, + "step": 9003 + }, + { + "epoch": 3.3951734539969833, + "grad_norm": 0.19971908384000286, + "learning_rate": 2.258639523428942e-05, + "loss": 0.3976, + "step": 9004 + }, + { + "epoch": 3.395550527903469, + "grad_norm": 0.12514946259141843, + "learning_rate": 2.2576916135263292e-05, + "loss": 0.3901, + "step": 9005 + }, + { + "epoch": 3.3959276018099547, + "grad_norm": 0.16845356225621183, + "learning_rate": 2.2567438243641425e-05, + "loss": 0.3919, + "step": 9006 + }, + { + "epoch": 3.3963046757164403, + "grad_norm": 0.1522836736409811, + "learning_rate": 2.255796156008063e-05, + "loss": 0.4063, + "step": 9007 + }, + { + "epoch": 3.396681749622926, + "grad_norm": 0.11929901583571974, + "learning_rate": 2.2548486085237645e-05, + "loss": 0.3978, + "step": 9008 + }, + { + "epoch": 3.3970588235294117, + "grad_norm": 0.14780552299093747, + "learning_rate": 2.2539011819769056e-05, + "loss": 0.3909, + "step": 9009 + }, + { + "epoch": 3.3974358974358974, + "grad_norm": 0.12503636245088512, + "learning_rate": 2.2529538764331495e-05, + "loss": 0.3922, + "step": 9010 + }, + { + "epoch": 3.397812971342383, + "grad_norm": 0.14966002310496462, + "learning_rate": 2.2520066919581385e-05, + "loss": 0.397, + "step": 9011 + }, + { + "epoch": 3.3981900452488687, + "grad_norm": 0.13249549189316115, + "learning_rate": 2.2510596286175136e-05, + "loss": 0.3906, + "step": 9012 + }, + { + "epoch": 3.3985671191553544, + "grad_norm": 0.12113595326256502, + "learning_rate": 2.2501126864769046e-05, + "loss": 0.3967, + "step": 9013 + }, + { + "epoch": 3.39894419306184, + "grad_norm": 0.16516378700034837, + "learning_rate": 2.2491658656019366e-05, + "loss": 0.3895, + "step": 9014 + }, + { + "epoch": 3.3993212669683257, + "grad_norm": 0.10571356603964893, + "learning_rate": 2.248219166058217e-05, + "loss": 0.386, + "step": 9015 + }, + { + "epoch": 3.3996983408748114, + "grad_norm": 0.15032359282589777, + "learning_rate": 2.2472725879113596e-05, + "loss": 0.395, + "step": 9016 + }, + { + "epoch": 3.400075414781297, + "grad_norm": 0.139581071121035, + "learning_rate": 2.2463261312269547e-05, + "loss": 0.3951, + "step": 9017 + }, + { + "epoch": 3.4004524886877827, + "grad_norm": 0.14162133758933806, + "learning_rate": 2.2453797960705934e-05, + "loss": 0.408, + "step": 9018 + }, + { + "epoch": 3.4008295625942684, + "grad_norm": 0.15446039214624296, + "learning_rate": 2.2444335825078562e-05, + "loss": 0.4013, + "step": 9019 + }, + { + "epoch": 3.401206636500754, + "grad_norm": 0.12636280832748592, + "learning_rate": 2.2434874906043164e-05, + "loss": 0.3927, + "step": 9020 + }, + { + "epoch": 3.4015837104072397, + "grad_norm": 0.15653869194321324, + "learning_rate": 2.242541520425532e-05, + "loss": 0.392, + "step": 9021 + }, + { + "epoch": 3.4019607843137254, + "grad_norm": 0.12672894076332727, + "learning_rate": 2.2415956720370645e-05, + "loss": 0.3925, + "step": 9022 + }, + { + "epoch": 3.402337858220211, + "grad_norm": 0.13236813614791834, + "learning_rate": 2.2406499455044558e-05, + "loss": 0.4, + "step": 9023 + }, + { + "epoch": 3.4027149321266967, + "grad_norm": 0.1561057916429159, + "learning_rate": 2.239704340893245e-05, + "loss": 0.3964, + "step": 9024 + }, + { + "epoch": 3.4030920060331824, + "grad_norm": 0.1441393201947853, + "learning_rate": 2.2387588582689627e-05, + "loss": 0.412, + "step": 9025 + }, + { + "epoch": 3.403469079939668, + "grad_norm": 0.1257680302040572, + "learning_rate": 2.23781349769713e-05, + "loss": 0.3966, + "step": 9026 + }, + { + "epoch": 3.4038461538461537, + "grad_norm": 0.1444950532710061, + "learning_rate": 2.2368682592432562e-05, + "loss": 0.403, + "step": 9027 + }, + { + "epoch": 3.4042232277526394, + "grad_norm": 0.12034030814040544, + "learning_rate": 2.235923142972851e-05, + "loss": 0.4048, + "step": 9028 + }, + { + "epoch": 3.404600301659125, + "grad_norm": 0.12510205779983027, + "learning_rate": 2.234978148951406e-05, + "loss": 0.3872, + "step": 9029 + }, + { + "epoch": 3.4049773755656108, + "grad_norm": 0.10997652390413039, + "learning_rate": 2.23403327724441e-05, + "loss": 0.3861, + "step": 9030 + }, + { + "epoch": 3.4053544494720964, + "grad_norm": 0.12720362121705028, + "learning_rate": 2.2330885279173405e-05, + "loss": 0.3872, + "step": 9031 + }, + { + "epoch": 3.405731523378582, + "grad_norm": 0.13053509837982502, + "learning_rate": 2.2321439010356712e-05, + "loss": 0.3976, + "step": 9032 + }, + { + "epoch": 3.4061085972850678, + "grad_norm": 0.10885720736284664, + "learning_rate": 2.2311993966648582e-05, + "loss": 0.394, + "step": 9033 + }, + { + "epoch": 3.4064856711915534, + "grad_norm": 0.10978587707320311, + "learning_rate": 2.230255014870361e-05, + "loss": 0.3922, + "step": 9034 + }, + { + "epoch": 3.406862745098039, + "grad_norm": 0.12925866190839608, + "learning_rate": 2.22931075571762e-05, + "loss": 0.4038, + "step": 9035 + }, + { + "epoch": 3.4072398190045248, + "grad_norm": 0.11694224803214161, + "learning_rate": 2.2283666192720733e-05, + "loss": 0.3875, + "step": 9036 + }, + { + "epoch": 3.4076168929110104, + "grad_norm": 0.14687650732265262, + "learning_rate": 2.2274226055991475e-05, + "loss": 0.3982, + "step": 9037 + }, + { + "epoch": 3.407993966817496, + "grad_norm": 0.10406822649746508, + "learning_rate": 2.226478714764266e-05, + "loss": 0.4022, + "step": 9038 + }, + { + "epoch": 3.408371040723982, + "grad_norm": 0.12175070876570575, + "learning_rate": 2.225534946832832e-05, + "loss": 0.3965, + "step": 9039 + }, + { + "epoch": 3.4087481146304675, + "grad_norm": 0.1290641546971016, + "learning_rate": 2.2245913018702557e-05, + "loss": 0.4012, + "step": 9040 + }, + { + "epoch": 3.409125188536953, + "grad_norm": 0.11450493286773494, + "learning_rate": 2.223647779941926e-05, + "loss": 0.3972, + "step": 9041 + }, + { + "epoch": 3.409502262443439, + "grad_norm": 0.12404927650728395, + "learning_rate": 2.2227043811132293e-05, + "loss": 0.3856, + "step": 9042 + }, + { + "epoch": 3.4098793363499245, + "grad_norm": 0.11344935586970556, + "learning_rate": 2.221761105449543e-05, + "loss": 0.3888, + "step": 9043 + }, + { + "epoch": 3.41025641025641, + "grad_norm": 0.11263931773653987, + "learning_rate": 2.220817953016236e-05, + "loss": 0.3943, + "step": 9044 + }, + { + "epoch": 3.410633484162896, + "grad_norm": 0.12819051708917828, + "learning_rate": 2.2198749238786633e-05, + "loss": 0.4031, + "step": 9045 + }, + { + "epoch": 3.4110105580693815, + "grad_norm": 0.14481559664307367, + "learning_rate": 2.2189320181021833e-05, + "loss": 0.3945, + "step": 9046 + }, + { + "epoch": 3.411387631975867, + "grad_norm": 0.12300745181667294, + "learning_rate": 2.217989235752133e-05, + "loss": 0.3797, + "step": 9047 + }, + { + "epoch": 3.411764705882353, + "grad_norm": 0.16421541422277808, + "learning_rate": 2.2170465768938473e-05, + "loss": 0.3934, + "step": 9048 + }, + { + "epoch": 3.4121417797888385, + "grad_norm": 0.15361086293563658, + "learning_rate": 2.2161040415926526e-05, + "loss": 0.3977, + "step": 9049 + }, + { + "epoch": 3.412518853695324, + "grad_norm": 0.13364289655568168, + "learning_rate": 2.215161629913867e-05, + "loss": 0.3915, + "step": 9050 + }, + { + "epoch": 3.41289592760181, + "grad_norm": 0.11688650216315716, + "learning_rate": 2.2142193419227946e-05, + "loss": 0.3837, + "step": 9051 + }, + { + "epoch": 3.4132730015082955, + "grad_norm": 0.1385293492578538, + "learning_rate": 2.2132771776847408e-05, + "loss": 0.3951, + "step": 9052 + }, + { + "epoch": 3.413650075414781, + "grad_norm": 0.11801537481083561, + "learning_rate": 2.212335137264992e-05, + "loss": 0.4091, + "step": 9053 + }, + { + "epoch": 3.414027149321267, + "grad_norm": 0.13168971270579857, + "learning_rate": 2.2113932207288328e-05, + "loss": 0.3931, + "step": 9054 + }, + { + "epoch": 3.4144042232277525, + "grad_norm": 0.14309603783147093, + "learning_rate": 2.2104514281415367e-05, + "loss": 0.4043, + "step": 9055 + }, + { + "epoch": 3.414781297134238, + "grad_norm": 0.11795838560692912, + "learning_rate": 2.209509759568371e-05, + "loss": 0.3847, + "step": 9056 + }, + { + "epoch": 3.415158371040724, + "grad_norm": 0.1287493661927277, + "learning_rate": 2.2085682150745873e-05, + "loss": 0.3949, + "step": 9057 + }, + { + "epoch": 3.4155354449472095, + "grad_norm": 0.13689339072708695, + "learning_rate": 2.2076267947254408e-05, + "loss": 0.4061, + "step": 9058 + }, + { + "epoch": 3.415912518853695, + "grad_norm": 0.12384691094122344, + "learning_rate": 2.2066854985861662e-05, + "loss": 0.3859, + "step": 9059 + }, + { + "epoch": 3.416289592760181, + "grad_norm": 0.10460778230844052, + "learning_rate": 2.2057443267219956e-05, + "loss": 0.4027, + "step": 9060 + }, + { + "epoch": 3.4166666666666665, + "grad_norm": 0.11334415197426642, + "learning_rate": 2.2048032791981515e-05, + "loss": 0.4001, + "step": 9061 + }, + { + "epoch": 3.417043740573152, + "grad_norm": 0.10691829309675874, + "learning_rate": 2.2038623560798494e-05, + "loss": 0.4068, + "step": 9062 + }, + { + "epoch": 3.417420814479638, + "grad_norm": 0.12358764576075928, + "learning_rate": 2.2029215574322897e-05, + "loss": 0.3969, + "step": 9063 + }, + { + "epoch": 3.4177978883861235, + "grad_norm": 0.11629242050998174, + "learning_rate": 2.2019808833206758e-05, + "loss": 0.3975, + "step": 9064 + }, + { + "epoch": 3.418174962292609, + "grad_norm": 0.11225989799322858, + "learning_rate": 2.20104033381019e-05, + "loss": 0.394, + "step": 9065 + }, + { + "epoch": 3.418552036199095, + "grad_norm": 0.10990144740620282, + "learning_rate": 2.200099908966013e-05, + "loss": 0.3847, + "step": 9066 + }, + { + "epoch": 3.4189291101055805, + "grad_norm": 0.11914255185597426, + "learning_rate": 2.199159608853316e-05, + "loss": 0.3988, + "step": 9067 + }, + { + "epoch": 3.419306184012066, + "grad_norm": 0.12384463114887705, + "learning_rate": 2.198219433537262e-05, + "loss": 0.4019, + "step": 9068 + }, + { + "epoch": 3.419683257918552, + "grad_norm": 0.12993459965053208, + "learning_rate": 2.197279383083001e-05, + "loss": 0.3954, + "step": 9069 + }, + { + "epoch": 3.4200603318250375, + "grad_norm": 0.13712008038633489, + "learning_rate": 2.196339457555682e-05, + "loss": 0.4003, + "step": 9070 + }, + { + "epoch": 3.420437405731523, + "grad_norm": 0.11596292030081581, + "learning_rate": 2.1953996570204375e-05, + "loss": 0.3974, + "step": 9071 + }, + { + "epoch": 3.420814479638009, + "grad_norm": 0.15293519278740716, + "learning_rate": 2.194459981542396e-05, + "loss": 0.4005, + "step": 9072 + }, + { + "epoch": 3.4211915535444946, + "grad_norm": 0.11105482349050355, + "learning_rate": 2.193520431186676e-05, + "loss": 0.4006, + "step": 9073 + }, + { + "epoch": 3.4215686274509802, + "grad_norm": 0.1401411952784895, + "learning_rate": 2.1925810060183897e-05, + "loss": 0.3985, + "step": 9074 + }, + { + "epoch": 3.421945701357466, + "grad_norm": 0.1416327058821298, + "learning_rate": 2.191641706102633e-05, + "loss": 0.3932, + "step": 9075 + }, + { + "epoch": 3.4223227752639516, + "grad_norm": 0.13551810705170392, + "learning_rate": 2.1907025315045055e-05, + "loss": 0.3872, + "step": 9076 + }, + { + "epoch": 3.4226998491704372, + "grad_norm": 0.13904588616709368, + "learning_rate": 2.1897634822890863e-05, + "loss": 0.3924, + "step": 9077 + }, + { + "epoch": 3.423076923076923, + "grad_norm": 0.14846642805919608, + "learning_rate": 2.188824558521452e-05, + "loss": 0.3981, + "step": 9078 + }, + { + "epoch": 3.4234539969834086, + "grad_norm": 0.13679090400609706, + "learning_rate": 2.187885760266669e-05, + "loss": 0.4079, + "step": 9079 + }, + { + "epoch": 3.4238310708898942, + "grad_norm": 0.13732789367786738, + "learning_rate": 2.186947087589797e-05, + "loss": 0.41, + "step": 9080 + }, + { + "epoch": 3.42420814479638, + "grad_norm": 0.12474134084489114, + "learning_rate": 2.186008540555881e-05, + "loss": 0.3914, + "step": 9081 + }, + { + "epoch": 3.4245852187028656, + "grad_norm": 0.14007820140004387, + "learning_rate": 2.185070119229967e-05, + "loss": 0.383, + "step": 9082 + }, + { + "epoch": 3.4249622926093513, + "grad_norm": 0.12838298849393506, + "learning_rate": 2.184131823677082e-05, + "loss": 0.3934, + "step": 9083 + }, + { + "epoch": 3.425339366515837, + "grad_norm": 0.12978491098704997, + "learning_rate": 2.1831936539622514e-05, + "loss": 0.3891, + "step": 9084 + }, + { + "epoch": 3.4257164404223226, + "grad_norm": 0.1250373123616409, + "learning_rate": 2.182255610150489e-05, + "loss": 0.3927, + "step": 9085 + }, + { + "epoch": 3.4260935143288083, + "grad_norm": 0.1305609337033009, + "learning_rate": 2.181317692306802e-05, + "loss": 0.385, + "step": 9086 + }, + { + "epoch": 3.426470588235294, + "grad_norm": 0.11360797337358404, + "learning_rate": 2.1803799004961824e-05, + "loss": 0.3921, + "step": 9087 + }, + { + "epoch": 3.4268476621417796, + "grad_norm": 0.1426232959098214, + "learning_rate": 2.179442234783625e-05, + "loss": 0.4007, + "step": 9088 + }, + { + "epoch": 3.4272247360482653, + "grad_norm": 0.10854715530765699, + "learning_rate": 2.1785046952341045e-05, + "loss": 0.3924, + "step": 9089 + }, + { + "epoch": 3.427601809954751, + "grad_norm": 0.12671548670179142, + "learning_rate": 2.1775672819125928e-05, + "loss": 0.3888, + "step": 9090 + }, + { + "epoch": 3.4279788838612366, + "grad_norm": 0.11875396800210211, + "learning_rate": 2.176629994884052e-05, + "loss": 0.4055, + "step": 9091 + }, + { + "epoch": 3.4283559577677223, + "grad_norm": 0.14039681852289804, + "learning_rate": 2.175692834213437e-05, + "loss": 0.4001, + "step": 9092 + }, + { + "epoch": 3.428733031674208, + "grad_norm": 0.13466165546059153, + "learning_rate": 2.174755799965687e-05, + "loss": 0.3983, + "step": 9093 + }, + { + "epoch": 3.4291101055806936, + "grad_norm": 0.11645783834068801, + "learning_rate": 2.173818892205744e-05, + "loss": 0.3953, + "step": 9094 + }, + { + "epoch": 3.4294871794871793, + "grad_norm": 0.12835591177007724, + "learning_rate": 2.172882110998531e-05, + "loss": 0.3971, + "step": 9095 + }, + { + "epoch": 3.4298642533936654, + "grad_norm": 0.11440356098440459, + "learning_rate": 2.1719454564089664e-05, + "loss": 0.3923, + "step": 9096 + }, + { + "epoch": 3.430241327300151, + "grad_norm": 0.11230106734351965, + "learning_rate": 2.1710089285019604e-05, + "loss": 0.3977, + "step": 9097 + }, + { + "epoch": 3.4306184012066367, + "grad_norm": 0.12338487662367882, + "learning_rate": 2.170072527342414e-05, + "loss": 0.4033, + "step": 9098 + }, + { + "epoch": 3.4309954751131224, + "grad_norm": 0.11699228449287274, + "learning_rate": 2.1691362529952174e-05, + "loss": 0.399, + "step": 9099 + }, + { + "epoch": 3.431372549019608, + "grad_norm": 0.12328603270193302, + "learning_rate": 2.1682001055252565e-05, + "loss": 0.4001, + "step": 9100 + }, + { + "epoch": 3.4317496229260938, + "grad_norm": 0.1509043851852186, + "learning_rate": 2.1672640849974004e-05, + "loss": 0.3793, + "step": 9101 + }, + { + "epoch": 3.4321266968325794, + "grad_norm": 0.1285078576634741, + "learning_rate": 2.1663281914765203e-05, + "loss": 0.4044, + "step": 9102 + }, + { + "epoch": 3.432503770739065, + "grad_norm": 0.1455144322808683, + "learning_rate": 2.1653924250274687e-05, + "loss": 0.3912, + "step": 9103 + }, + { + "epoch": 3.4328808446455508, + "grad_norm": 0.15930799743751586, + "learning_rate": 2.1644567857150948e-05, + "loss": 0.3991, + "step": 9104 + }, + { + "epoch": 3.4332579185520364, + "grad_norm": 0.13706807342335464, + "learning_rate": 2.1635212736042373e-05, + "loss": 0.4021, + "step": 9105 + }, + { + "epoch": 3.433634992458522, + "grad_norm": 0.17864323095156046, + "learning_rate": 2.162585888759729e-05, + "loss": 0.4001, + "step": 9106 + }, + { + "epoch": 3.4340120663650078, + "grad_norm": 0.1301660832543313, + "learning_rate": 2.161650631246385e-05, + "loss": 0.3902, + "step": 9107 + }, + { + "epoch": 3.4343891402714934, + "grad_norm": 0.17453450091075945, + "learning_rate": 2.1607155011290247e-05, + "loss": 0.394, + "step": 9108 + }, + { + "epoch": 3.434766214177979, + "grad_norm": 0.19254419461890762, + "learning_rate": 2.159780498472447e-05, + "loss": 0.3995, + "step": 9109 + }, + { + "epoch": 3.435143288084465, + "grad_norm": 0.1241219455352949, + "learning_rate": 2.1588456233414492e-05, + "loss": 0.392, + "step": 9110 + }, + { + "epoch": 3.4355203619909505, + "grad_norm": 0.17535194901633322, + "learning_rate": 2.157910875800816e-05, + "loss": 0.3966, + "step": 9111 + }, + { + "epoch": 3.435897435897436, + "grad_norm": 0.12567686461411964, + "learning_rate": 2.156976255915327e-05, + "loss": 0.3908, + "step": 9112 + }, + { + "epoch": 3.436274509803922, + "grad_norm": 0.1647017355415061, + "learning_rate": 2.1560417637497455e-05, + "loss": 0.4, + "step": 9113 + }, + { + "epoch": 3.4366515837104075, + "grad_norm": 0.14068025208314697, + "learning_rate": 2.1551073993688378e-05, + "loss": 0.3889, + "step": 9114 + }, + { + "epoch": 3.437028657616893, + "grad_norm": 0.12210070957201054, + "learning_rate": 2.1541731628373485e-05, + "loss": 0.4081, + "step": 9115 + }, + { + "epoch": 3.437405731523379, + "grad_norm": 0.12363871652546081, + "learning_rate": 2.1532390542200226e-05, + "loss": 0.3924, + "step": 9116 + }, + { + "epoch": 3.4377828054298645, + "grad_norm": 0.13017031439276505, + "learning_rate": 2.1523050735815912e-05, + "loss": 0.3993, + "step": 9117 + }, + { + "epoch": 3.43815987933635, + "grad_norm": 0.12422103880256687, + "learning_rate": 2.15137122098678e-05, + "loss": 0.3905, + "step": 9118 + }, + { + "epoch": 3.438536953242836, + "grad_norm": 0.15438129022258557, + "learning_rate": 2.1504374965003033e-05, + "loss": 0.3788, + "step": 9119 + }, + { + "epoch": 3.4389140271493215, + "grad_norm": 0.12368406798417579, + "learning_rate": 2.1495039001868694e-05, + "loss": 0.3925, + "step": 9120 + }, + { + "epoch": 3.439291101055807, + "grad_norm": 0.12653351689420086, + "learning_rate": 2.1485704321111712e-05, + "loss": 0.4008, + "step": 9121 + }, + { + "epoch": 3.439668174962293, + "grad_norm": 0.1454305524688539, + "learning_rate": 2.147637092337901e-05, + "loss": 0.3956, + "step": 9122 + }, + { + "epoch": 3.4400452488687785, + "grad_norm": 0.11673257969950172, + "learning_rate": 2.1467038809317372e-05, + "loss": 0.4005, + "step": 9123 + }, + { + "epoch": 3.440422322775264, + "grad_norm": 0.1340588904528114, + "learning_rate": 2.14577079795735e-05, + "loss": 0.3963, + "step": 9124 + }, + { + "epoch": 3.44079939668175, + "grad_norm": 0.13436738192856287, + "learning_rate": 2.144837843479402e-05, + "loss": 0.4022, + "step": 9125 + }, + { + "epoch": 3.4411764705882355, + "grad_norm": 0.11968918477272521, + "learning_rate": 2.1439050175625474e-05, + "loss": 0.3919, + "step": 9126 + }, + { + "epoch": 3.441553544494721, + "grad_norm": 0.12451699668316368, + "learning_rate": 2.142972320271427e-05, + "loss": 0.3928, + "step": 9127 + }, + { + "epoch": 3.441930618401207, + "grad_norm": 0.12717505320563224, + "learning_rate": 2.1420397516706782e-05, + "loss": 0.3937, + "step": 9128 + }, + { + "epoch": 3.4423076923076925, + "grad_norm": 0.13316362381718458, + "learning_rate": 2.141107311824926e-05, + "loss": 0.3897, + "step": 9129 + }, + { + "epoch": 3.442684766214178, + "grad_norm": 0.11957709062606023, + "learning_rate": 2.140175000798789e-05, + "loss": 0.3991, + "step": 9130 + }, + { + "epoch": 3.443061840120664, + "grad_norm": 0.12845599380835074, + "learning_rate": 2.1392428186568743e-05, + "loss": 0.387, + "step": 9131 + }, + { + "epoch": 3.4434389140271495, + "grad_norm": 0.1581302855017624, + "learning_rate": 2.138310765463783e-05, + "loss": 0.3966, + "step": 9132 + }, + { + "epoch": 3.443815987933635, + "grad_norm": 0.1471829040273968, + "learning_rate": 2.137378841284103e-05, + "loss": 0.4124, + "step": 9133 + }, + { + "epoch": 3.444193061840121, + "grad_norm": 0.1366336775835417, + "learning_rate": 2.1364470461824175e-05, + "loss": 0.4034, + "step": 9134 + }, + { + "epoch": 3.4445701357466065, + "grad_norm": 0.15134353495885397, + "learning_rate": 2.1355153802232985e-05, + "loss": 0.3923, + "step": 9135 + }, + { + "epoch": 3.444947209653092, + "grad_norm": 0.12408920648574107, + "learning_rate": 2.13458384347131e-05, + "loss": 0.3963, + "step": 9136 + }, + { + "epoch": 3.445324283559578, + "grad_norm": 0.13871273327110814, + "learning_rate": 2.1336524359910063e-05, + "loss": 0.4022, + "step": 9137 + }, + { + "epoch": 3.4457013574660635, + "grad_norm": 0.10929605323668683, + "learning_rate": 2.132721157846933e-05, + "loss": 0.3996, + "step": 9138 + }, + { + "epoch": 3.446078431372549, + "grad_norm": 0.13092229517324747, + "learning_rate": 2.1317900091036297e-05, + "loss": 0.3967, + "step": 9139 + }, + { + "epoch": 3.446455505279035, + "grad_norm": 0.13026792344834617, + "learning_rate": 2.13085898982562e-05, + "loss": 0.3985, + "step": 9140 + }, + { + "epoch": 3.4468325791855206, + "grad_norm": 0.11436152711391365, + "learning_rate": 2.1299281000774235e-05, + "loss": 0.4054, + "step": 9141 + }, + { + "epoch": 3.447209653092006, + "grad_norm": 0.12023618827037573, + "learning_rate": 2.1289973399235512e-05, + "loss": 0.392, + "step": 9142 + }, + { + "epoch": 3.447586726998492, + "grad_norm": 0.12082315065020773, + "learning_rate": 2.128066709428504e-05, + "loss": 0.3924, + "step": 9143 + }, + { + "epoch": 3.4479638009049776, + "grad_norm": 0.10874112806818936, + "learning_rate": 2.127136208656773e-05, + "loss": 0.3859, + "step": 9144 + }, + { + "epoch": 3.4483408748114632, + "grad_norm": 0.11434574782719058, + "learning_rate": 2.126205837672844e-05, + "loss": 0.3837, + "step": 9145 + }, + { + "epoch": 3.448717948717949, + "grad_norm": 0.10978638899030856, + "learning_rate": 2.1252755965411857e-05, + "loss": 0.3967, + "step": 9146 + }, + { + "epoch": 3.4490950226244346, + "grad_norm": 0.10226059446413184, + "learning_rate": 2.1243454853262663e-05, + "loss": 0.3946, + "step": 9147 + }, + { + "epoch": 3.4494720965309202, + "grad_norm": 0.11379132815856353, + "learning_rate": 2.1234155040925408e-05, + "loss": 0.3761, + "step": 9148 + }, + { + "epoch": 3.449849170437406, + "grad_norm": 0.11617928167613034, + "learning_rate": 2.1224856529044565e-05, + "loss": 0.3874, + "step": 9149 + }, + { + "epoch": 3.4502262443438916, + "grad_norm": 0.12336213145617979, + "learning_rate": 2.121555931826451e-05, + "loss": 0.3962, + "step": 9150 + }, + { + "epoch": 3.4506033182503772, + "grad_norm": 0.10789699779991908, + "learning_rate": 2.120626340922955e-05, + "loss": 0.3981, + "step": 9151 + }, + { + "epoch": 3.450980392156863, + "grad_norm": 0.11940693345630805, + "learning_rate": 2.1196968802583836e-05, + "loss": 0.3872, + "step": 9152 + }, + { + "epoch": 3.4513574660633486, + "grad_norm": 0.11183349040492122, + "learning_rate": 2.118767549897153e-05, + "loss": 0.3887, + "step": 9153 + }, + { + "epoch": 3.4517345399698343, + "grad_norm": 0.1380830656872108, + "learning_rate": 2.1178383499036617e-05, + "loss": 0.3835, + "step": 9154 + }, + { + "epoch": 3.45211161387632, + "grad_norm": 0.10622693577052571, + "learning_rate": 2.116909280342303e-05, + "loss": 0.4108, + "step": 9155 + }, + { + "epoch": 3.4524886877828056, + "grad_norm": 0.149732562175128, + "learning_rate": 2.115980341277461e-05, + "loss": 0.3901, + "step": 9156 + }, + { + "epoch": 3.4528657616892913, + "grad_norm": 0.12073988575213081, + "learning_rate": 2.1150515327735114e-05, + "loss": 0.3896, + "step": 9157 + }, + { + "epoch": 3.453242835595777, + "grad_norm": 0.13007747400739805, + "learning_rate": 2.1141228548948155e-05, + "loss": 0.389, + "step": 9158 + }, + { + "epoch": 3.4536199095022626, + "grad_norm": 0.13285056341607265, + "learning_rate": 2.1131943077057367e-05, + "loss": 0.3839, + "step": 9159 + }, + { + "epoch": 3.4539969834087483, + "grad_norm": 0.11989747555083805, + "learning_rate": 2.1122658912706168e-05, + "loss": 0.3953, + "step": 9160 + }, + { + "epoch": 3.454374057315234, + "grad_norm": 0.13670378785064616, + "learning_rate": 2.1113376056537963e-05, + "loss": 0.3916, + "step": 9161 + }, + { + "epoch": 3.4547511312217196, + "grad_norm": 0.12337785868211143, + "learning_rate": 2.1104094509196045e-05, + "loss": 0.4006, + "step": 9162 + }, + { + "epoch": 3.4551282051282053, + "grad_norm": 0.12262012428346847, + "learning_rate": 2.109481427132363e-05, + "loss": 0.3891, + "step": 9163 + }, + { + "epoch": 3.455505279034691, + "grad_norm": 0.11695283889755805, + "learning_rate": 2.108553534356379e-05, + "loss": 0.3962, + "step": 9164 + }, + { + "epoch": 3.4558823529411766, + "grad_norm": 0.12351933531275465, + "learning_rate": 2.1076257726559603e-05, + "loss": 0.4035, + "step": 9165 + }, + { + "epoch": 3.4562594268476623, + "grad_norm": 0.12092806181115805, + "learning_rate": 2.1066981420953956e-05, + "loss": 0.3893, + "step": 9166 + }, + { + "epoch": 3.456636500754148, + "grad_norm": 0.11042182675488973, + "learning_rate": 2.1057706427389704e-05, + "loss": 0.3952, + "step": 9167 + }, + { + "epoch": 3.4570135746606336, + "grad_norm": 0.12475409694792666, + "learning_rate": 2.104843274650959e-05, + "loss": 0.3969, + "step": 9168 + }, + { + "epoch": 3.4573906485671193, + "grad_norm": 0.09785492289437346, + "learning_rate": 2.1039160378956294e-05, + "loss": 0.3955, + "step": 9169 + }, + { + "epoch": 3.457767722473605, + "grad_norm": 0.12421992911430993, + "learning_rate": 2.102988932537233e-05, + "loss": 0.3966, + "step": 9170 + }, + { + "epoch": 3.4581447963800906, + "grad_norm": 0.10964727652027552, + "learning_rate": 2.1020619586400254e-05, + "loss": 0.4014, + "step": 9171 + }, + { + "epoch": 3.4585218702865763, + "grad_norm": 0.11015826910011269, + "learning_rate": 2.101135116268238e-05, + "loss": 0.3897, + "step": 9172 + }, + { + "epoch": 3.458898944193062, + "grad_norm": 0.11504471432984857, + "learning_rate": 2.100208405486103e-05, + "loss": 0.3971, + "step": 9173 + }, + { + "epoch": 3.4592760180995477, + "grad_norm": 0.1187201528910951, + "learning_rate": 2.09928182635784e-05, + "loss": 0.3958, + "step": 9174 + }, + { + "epoch": 3.4596530920060333, + "grad_norm": 0.1346161264806991, + "learning_rate": 2.0983553789476626e-05, + "loss": 0.3967, + "step": 9175 + }, + { + "epoch": 3.460030165912519, + "grad_norm": 0.11632706882468594, + "learning_rate": 2.0974290633197673e-05, + "loss": 0.4005, + "step": 9176 + }, + { + "epoch": 3.4604072398190047, + "grad_norm": 0.13191945723694656, + "learning_rate": 2.0965028795383534e-05, + "loss": 0.3894, + "step": 9177 + }, + { + "epoch": 3.4607843137254903, + "grad_norm": 0.11682986004639771, + "learning_rate": 2.0955768276676e-05, + "loss": 0.4012, + "step": 9178 + }, + { + "epoch": 3.461161387631976, + "grad_norm": 0.15695145433594965, + "learning_rate": 2.0946509077716833e-05, + "loss": 0.4054, + "step": 9179 + }, + { + "epoch": 3.4615384615384617, + "grad_norm": 0.12958485751272344, + "learning_rate": 2.0937251199147684e-05, + "loss": 0.3937, + "step": 9180 + }, + { + "epoch": 3.4619155354449473, + "grad_norm": 0.1528932745946515, + "learning_rate": 2.092799464161014e-05, + "loss": 0.4054, + "step": 9181 + }, + { + "epoch": 3.462292609351433, + "grad_norm": 0.13351144600602735, + "learning_rate": 2.091873940574561e-05, + "loss": 0.3976, + "step": 9182 + }, + { + "epoch": 3.4626696832579187, + "grad_norm": 0.11970972074030897, + "learning_rate": 2.0909485492195554e-05, + "loss": 0.3902, + "step": 9183 + }, + { + "epoch": 3.4630467571644044, + "grad_norm": 0.1235607091041802, + "learning_rate": 2.0900232901601206e-05, + "loss": 0.4005, + "step": 9184 + }, + { + "epoch": 3.46342383107089, + "grad_norm": 0.10793553959938565, + "learning_rate": 2.0890981634603767e-05, + "loss": 0.3981, + "step": 9185 + }, + { + "epoch": 3.4638009049773757, + "grad_norm": 0.12608160290030157, + "learning_rate": 2.0881731691844357e-05, + "loss": 0.4004, + "step": 9186 + }, + { + "epoch": 3.4641779788838614, + "grad_norm": 0.12053567146868709, + "learning_rate": 2.0872483073964e-05, + "loss": 0.4002, + "step": 9187 + }, + { + "epoch": 3.464555052790347, + "grad_norm": 0.11638644988871723, + "learning_rate": 2.0863235781603565e-05, + "loss": 0.4067, + "step": 9188 + }, + { + "epoch": 3.4649321266968327, + "grad_norm": 0.11915424713614024, + "learning_rate": 2.0853989815403945e-05, + "loss": 0.394, + "step": 9189 + }, + { + "epoch": 3.4653092006033184, + "grad_norm": 0.12258880859296847, + "learning_rate": 2.084474517600584e-05, + "loss": 0.4065, + "step": 9190 + }, + { + "epoch": 3.465686274509804, + "grad_norm": 0.1174261840453406, + "learning_rate": 2.08355018640499e-05, + "loss": 0.3904, + "step": 9191 + }, + { + "epoch": 3.4660633484162897, + "grad_norm": 0.11744432375277089, + "learning_rate": 2.082625988017668e-05, + "loss": 0.3971, + "step": 9192 + }, + { + "epoch": 3.4664404223227754, + "grad_norm": 0.10817783244126539, + "learning_rate": 2.0817019225026662e-05, + "loss": 0.396, + "step": 9193 + }, + { + "epoch": 3.466817496229261, + "grad_norm": 0.12388531897202602, + "learning_rate": 2.0807779899240162e-05, + "loss": 0.4005, + "step": 9194 + }, + { + "epoch": 3.4671945701357467, + "grad_norm": 0.11980728287235928, + "learning_rate": 2.079854190345752e-05, + "loss": 0.4012, + "step": 9195 + }, + { + "epoch": 3.4675716440422324, + "grad_norm": 0.13601815793925387, + "learning_rate": 2.0789305238318884e-05, + "loss": 0.3886, + "step": 9196 + }, + { + "epoch": 3.467948717948718, + "grad_norm": 0.12185573122649708, + "learning_rate": 2.0780069904464356e-05, + "loss": 0.3965, + "step": 9197 + }, + { + "epoch": 3.4683257918552037, + "grad_norm": 0.14305333914453378, + "learning_rate": 2.077083590253393e-05, + "loss": 0.3956, + "step": 9198 + }, + { + "epoch": 3.4687028657616894, + "grad_norm": 0.14364099329613, + "learning_rate": 2.0761603233167536e-05, + "loss": 0.4024, + "step": 9199 + }, + { + "epoch": 3.469079939668175, + "grad_norm": 0.12131096450915083, + "learning_rate": 2.0752371897004942e-05, + "loss": 0.3942, + "step": 9200 + }, + { + "epoch": 3.4694570135746607, + "grad_norm": 0.1377530860203112, + "learning_rate": 2.0743141894685937e-05, + "loss": 0.3818, + "step": 9201 + }, + { + "epoch": 3.4698340874811464, + "grad_norm": 0.10787793901036688, + "learning_rate": 2.07339132268501e-05, + "loss": 0.3975, + "step": 9202 + }, + { + "epoch": 3.470211161387632, + "grad_norm": 0.14278158891684023, + "learning_rate": 2.0724685894136992e-05, + "loss": 0.3936, + "step": 9203 + }, + { + "epoch": 3.4705882352941178, + "grad_norm": 0.12362168063239418, + "learning_rate": 2.0715459897186046e-05, + "loss": 0.3979, + "step": 9204 + }, + { + "epoch": 3.4709653092006034, + "grad_norm": 0.12322637303727016, + "learning_rate": 2.0706235236636646e-05, + "loss": 0.4025, + "step": 9205 + }, + { + "epoch": 3.471342383107089, + "grad_norm": 0.11383309450076981, + "learning_rate": 2.0697011913127997e-05, + "loss": 0.3908, + "step": 9206 + }, + { + "epoch": 3.4717194570135748, + "grad_norm": 0.12851369511165595, + "learning_rate": 2.068778992729933e-05, + "loss": 0.3987, + "step": 9207 + }, + { + "epoch": 3.4720965309200604, + "grad_norm": 0.12142404215254143, + "learning_rate": 2.067856927978968e-05, + "loss": 0.3919, + "step": 9208 + }, + { + "epoch": 3.472473604826546, + "grad_norm": 0.11889107110203323, + "learning_rate": 2.066934997123804e-05, + "loss": 0.4067, + "step": 9209 + }, + { + "epoch": 3.4728506787330318, + "grad_norm": 0.11916406889593606, + "learning_rate": 2.06601320022833e-05, + "loss": 0.4087, + "step": 9210 + }, + { + "epoch": 3.4732277526395174, + "grad_norm": 0.11273962740781224, + "learning_rate": 2.0650915373564268e-05, + "loss": 0.3961, + "step": 9211 + }, + { + "epoch": 3.473604826546003, + "grad_norm": 0.1259635999548672, + "learning_rate": 2.0641700085719615e-05, + "loss": 0.3864, + "step": 9212 + }, + { + "epoch": 3.473981900452489, + "grad_norm": 0.11387843558290545, + "learning_rate": 2.0632486139388e-05, + "loss": 0.388, + "step": 9213 + }, + { + "epoch": 3.4743589743589745, + "grad_norm": 0.12916578106953036, + "learning_rate": 2.0623273535207904e-05, + "loss": 0.3938, + "step": 9214 + }, + { + "epoch": 3.47473604826546, + "grad_norm": 0.11937446584051223, + "learning_rate": 2.0614062273817765e-05, + "loss": 0.3867, + "step": 9215 + }, + { + "epoch": 3.475113122171946, + "grad_norm": 0.11500565145870598, + "learning_rate": 2.0604852355855914e-05, + "loss": 0.4014, + "step": 9216 + }, + { + "epoch": 3.4754901960784315, + "grad_norm": 0.11127765233441629, + "learning_rate": 2.0595643781960607e-05, + "loss": 0.3984, + "step": 9217 + }, + { + "epoch": 3.475867269984917, + "grad_norm": 0.12617362938086263, + "learning_rate": 2.0586436552769946e-05, + "loss": 0.3974, + "step": 9218 + }, + { + "epoch": 3.476244343891403, + "grad_norm": 0.11791935189912099, + "learning_rate": 2.0577230668922043e-05, + "loss": 0.3955, + "step": 9219 + }, + { + "epoch": 3.4766214177978885, + "grad_norm": 0.17017121149996312, + "learning_rate": 2.0568026131054792e-05, + "loss": 0.4024, + "step": 9220 + }, + { + "epoch": 3.476998491704374, + "grad_norm": 0.14089272567437383, + "learning_rate": 2.055882293980613e-05, + "loss": 0.4056, + "step": 9221 + }, + { + "epoch": 3.47737556561086, + "grad_norm": 0.11797115977805445, + "learning_rate": 2.0549621095813773e-05, + "loss": 0.4033, + "step": 9222 + }, + { + "epoch": 3.4777526395173455, + "grad_norm": 0.1175624738747666, + "learning_rate": 2.0540420599715426e-05, + "loss": 0.3958, + "step": 9223 + }, + { + "epoch": 3.478129713423831, + "grad_norm": 0.12115501160094247, + "learning_rate": 2.0531221452148662e-05, + "loss": 0.3975, + "step": 9224 + }, + { + "epoch": 3.478506787330317, + "grad_norm": 0.12437434291507017, + "learning_rate": 2.0522023653751003e-05, + "loss": 0.3923, + "step": 9225 + }, + { + "epoch": 3.4788838612368025, + "grad_norm": 0.1326098280068343, + "learning_rate": 2.05128272051598e-05, + "loss": 0.4017, + "step": 9226 + }, + { + "epoch": 3.479260935143288, + "grad_norm": 0.11872293218171681, + "learning_rate": 2.050363210701242e-05, + "loss": 0.3953, + "step": 9227 + }, + { + "epoch": 3.479638009049774, + "grad_norm": 0.12593914494076525, + "learning_rate": 2.049443835994602e-05, + "loss": 0.384, + "step": 9228 + }, + { + "epoch": 3.4800150829562595, + "grad_norm": 0.13269173013356436, + "learning_rate": 2.0485245964597747e-05, + "loss": 0.4048, + "step": 9229 + }, + { + "epoch": 3.480392156862745, + "grad_norm": 0.11330282271205694, + "learning_rate": 2.047605492160461e-05, + "loss": 0.4039, + "step": 9230 + }, + { + "epoch": 3.480769230769231, + "grad_norm": 0.12504380905109316, + "learning_rate": 2.046686523160357e-05, + "loss": 0.3842, + "step": 9231 + }, + { + "epoch": 3.4811463046757165, + "grad_norm": 0.12245487578651174, + "learning_rate": 2.0457676895231416e-05, + "loss": 0.3926, + "step": 9232 + }, + { + "epoch": 3.481523378582202, + "grad_norm": 0.11337058715855043, + "learning_rate": 2.0448489913124945e-05, + "loss": 0.399, + "step": 9233 + }, + { + "epoch": 3.481900452488688, + "grad_norm": 0.13241144492706544, + "learning_rate": 2.0439304285920762e-05, + "loss": 0.4049, + "step": 9234 + }, + { + "epoch": 3.4822775263951735, + "grad_norm": 0.11491698726109824, + "learning_rate": 2.0430120014255447e-05, + "loss": 0.4051, + "step": 9235 + }, + { + "epoch": 3.482654600301659, + "grad_norm": 0.11342199480370695, + "learning_rate": 2.0420937098765455e-05, + "loss": 0.3918, + "step": 9236 + }, + { + "epoch": 3.483031674208145, + "grad_norm": 0.11171169773463321, + "learning_rate": 2.0411755540087174e-05, + "loss": 0.3928, + "step": 9237 + }, + { + "epoch": 3.4834087481146305, + "grad_norm": 0.11506278487972392, + "learning_rate": 2.0402575338856823e-05, + "loss": 0.3865, + "step": 9238 + }, + { + "epoch": 3.483785822021116, + "grad_norm": 0.14290682509991584, + "learning_rate": 2.0393396495710656e-05, + "loss": 0.3952, + "step": 9239 + }, + { + "epoch": 3.484162895927602, + "grad_norm": 0.13575919236280926, + "learning_rate": 2.0384219011284697e-05, + "loss": 0.392, + "step": 9240 + }, + { + "epoch": 3.4845399698340875, + "grad_norm": 0.129585165624803, + "learning_rate": 2.037504288621497e-05, + "loss": 0.3958, + "step": 9241 + }, + { + "epoch": 3.484917043740573, + "grad_norm": 0.1624438374012628, + "learning_rate": 2.0365868121137357e-05, + "loss": 0.398, + "step": 9242 + }, + { + "epoch": 3.485294117647059, + "grad_norm": 0.12280963478420412, + "learning_rate": 2.0356694716687687e-05, + "loss": 0.3966, + "step": 9243 + }, + { + "epoch": 3.4856711915535445, + "grad_norm": 0.15466691895337842, + "learning_rate": 2.0347522673501617e-05, + "loss": 0.3923, + "step": 9244 + }, + { + "epoch": 3.48604826546003, + "grad_norm": 0.14572094683528172, + "learning_rate": 2.0338351992214826e-05, + "loss": 0.3919, + "step": 9245 + }, + { + "epoch": 3.486425339366516, + "grad_norm": 0.1451159552840621, + "learning_rate": 2.032918267346278e-05, + "loss": 0.3976, + "step": 9246 + }, + { + "epoch": 3.4868024132730016, + "grad_norm": 0.10884458468433046, + "learning_rate": 2.0320014717880935e-05, + "loss": 0.411, + "step": 9247 + }, + { + "epoch": 3.4871794871794872, + "grad_norm": 0.11851504352818062, + "learning_rate": 2.0310848126104612e-05, + "loss": 0.3965, + "step": 9248 + }, + { + "epoch": 3.487556561085973, + "grad_norm": 0.11545749376350244, + "learning_rate": 2.0301682898769066e-05, + "loss": 0.4092, + "step": 9249 + }, + { + "epoch": 3.4879336349924586, + "grad_norm": 0.10505904500999794, + "learning_rate": 2.0292519036509385e-05, + "loss": 0.3815, + "step": 9250 + }, + { + "epoch": 3.4883107088989442, + "grad_norm": 0.11111573037071856, + "learning_rate": 2.028335653996069e-05, + "loss": 0.3879, + "step": 9251 + }, + { + "epoch": 3.48868778280543, + "grad_norm": 0.11205137957999173, + "learning_rate": 2.027419540975789e-05, + "loss": 0.3948, + "step": 9252 + }, + { + "epoch": 3.4890648567119156, + "grad_norm": 0.1116892008189292, + "learning_rate": 2.026503564653584e-05, + "loss": 0.3926, + "step": 9253 + }, + { + "epoch": 3.4894419306184012, + "grad_norm": 0.11275062404551414, + "learning_rate": 2.0255877250929324e-05, + "loss": 0.3995, + "step": 9254 + }, + { + "epoch": 3.489819004524887, + "grad_norm": 0.11217157316853228, + "learning_rate": 2.024672022357301e-05, + "loss": 0.3893, + "step": 9255 + }, + { + "epoch": 3.4901960784313726, + "grad_norm": 0.10922121604102857, + "learning_rate": 2.023756456510143e-05, + "loss": 0.4079, + "step": 9256 + }, + { + "epoch": 3.4905731523378583, + "grad_norm": 0.11525522106826687, + "learning_rate": 2.0228410276149138e-05, + "loss": 0.393, + "step": 9257 + }, + { + "epoch": 3.490950226244344, + "grad_norm": 0.10221091082296394, + "learning_rate": 2.021925735735045e-05, + "loss": 0.4034, + "step": 9258 + }, + { + "epoch": 3.4913273001508296, + "grad_norm": 0.12158444231085787, + "learning_rate": 2.021010580933969e-05, + "loss": 0.3796, + "step": 9259 + }, + { + "epoch": 3.4917043740573153, + "grad_norm": 0.10994479476574794, + "learning_rate": 2.0200955632751036e-05, + "loss": 0.4001, + "step": 9260 + }, + { + "epoch": 3.492081447963801, + "grad_norm": 0.11741229629769749, + "learning_rate": 2.019180682821862e-05, + "loss": 0.3967, + "step": 9261 + }, + { + "epoch": 3.4924585218702866, + "grad_norm": 0.10411198333115569, + "learning_rate": 2.0182659396376386e-05, + "loss": 0.3806, + "step": 9262 + }, + { + "epoch": 3.4928355957767723, + "grad_norm": 0.11712835965354566, + "learning_rate": 2.0173513337858303e-05, + "loss": 0.3987, + "step": 9263 + }, + { + "epoch": 3.493212669683258, + "grad_norm": 0.11487810203080752, + "learning_rate": 2.0164368653298147e-05, + "loss": 0.4051, + "step": 9264 + }, + { + "epoch": 3.4935897435897436, + "grad_norm": 0.11295594717509515, + "learning_rate": 2.0155225343329655e-05, + "loss": 0.4054, + "step": 9265 + }, + { + "epoch": 3.4939668174962293, + "grad_norm": 0.11082403539358143, + "learning_rate": 2.0146083408586437e-05, + "loss": 0.4012, + "step": 9266 + }, + { + "epoch": 3.494343891402715, + "grad_norm": 0.10873803894519066, + "learning_rate": 2.013694284970203e-05, + "loss": 0.3983, + "step": 9267 + }, + { + "epoch": 3.4947209653092006, + "grad_norm": 0.11640174441584829, + "learning_rate": 2.0127803667309868e-05, + "loss": 0.4019, + "step": 9268 + }, + { + "epoch": 3.4950980392156863, + "grad_norm": 0.10570118203856232, + "learning_rate": 2.01186658620433e-05, + "loss": 0.3999, + "step": 9269 + }, + { + "epoch": 3.495475113122172, + "grad_norm": 0.10863626580083062, + "learning_rate": 2.0109529434535536e-05, + "loss": 0.3948, + "step": 9270 + }, + { + "epoch": 3.4958521870286576, + "grad_norm": 0.12945682029353714, + "learning_rate": 2.0100394385419742e-05, + "loss": 0.4025, + "step": 9271 + }, + { + "epoch": 3.4962292609351433, + "grad_norm": 0.10729081520620368, + "learning_rate": 2.0091260715328966e-05, + "loss": 0.3997, + "step": 9272 + }, + { + "epoch": 3.496606334841629, + "grad_norm": 0.11685660101416659, + "learning_rate": 2.0082128424896164e-05, + "loss": 0.4006, + "step": 9273 + }, + { + "epoch": 3.4969834087481146, + "grad_norm": 0.13067164504408363, + "learning_rate": 2.0072997514754188e-05, + "loss": 0.3938, + "step": 9274 + }, + { + "epoch": 3.4973604826546003, + "grad_norm": 0.12838620250026284, + "learning_rate": 2.006386798553583e-05, + "loss": 0.391, + "step": 9275 + }, + { + "epoch": 3.497737556561086, + "grad_norm": 0.1462004714194169, + "learning_rate": 2.005473983787372e-05, + "loss": 0.3955, + "step": 9276 + }, + { + "epoch": 3.4981146304675717, + "grad_norm": 0.09807509695539858, + "learning_rate": 2.004561307240045e-05, + "loss": 0.3941, + "step": 9277 + }, + { + "epoch": 3.4984917043740573, + "grad_norm": 0.14778242274422004, + "learning_rate": 2.0036487689748495e-05, + "loss": 0.4012, + "step": 9278 + }, + { + "epoch": 3.498868778280543, + "grad_norm": 0.1158878957634528, + "learning_rate": 2.0027363690550237e-05, + "loss": 0.4045, + "step": 9279 + }, + { + "epoch": 3.4992458521870287, + "grad_norm": 0.12125647268666025, + "learning_rate": 2.0018241075437953e-05, + "loss": 0.3963, + "step": 9280 + }, + { + "epoch": 3.4996229260935143, + "grad_norm": 0.15022035611370751, + "learning_rate": 2.000911984504386e-05, + "loss": 0.3808, + "step": 9281 + }, + { + "epoch": 3.5, + "grad_norm": 0.1032964645383513, + "learning_rate": 2.0000000000000012e-05, + "loss": 0.3894, + "step": 9282 + }, + { + "epoch": 3.5003770739064857, + "grad_norm": 0.13836732980681174, + "learning_rate": 1.9990881540938422e-05, + "loss": 0.3979, + "step": 9283 + }, + { + "epoch": 3.5007541478129713, + "grad_norm": 0.11833620306303667, + "learning_rate": 1.9981764468490996e-05, + "loss": 0.4038, + "step": 9284 + }, + { + "epoch": 3.501131221719457, + "grad_norm": 0.12095374359731359, + "learning_rate": 1.997264878328953e-05, + "loss": 0.3926, + "step": 9285 + }, + { + "epoch": 3.5015082956259427, + "grad_norm": 0.11473455049552543, + "learning_rate": 1.996353448596575e-05, + "loss": 0.3964, + "step": 9286 + }, + { + "epoch": 3.5018853695324283, + "grad_norm": 0.12088242061002, + "learning_rate": 1.9954421577151253e-05, + "loss": 0.3918, + "step": 9287 + }, + { + "epoch": 3.502262443438914, + "grad_norm": 0.12278534475501095, + "learning_rate": 1.9945310057477573e-05, + "loss": 0.4081, + "step": 9288 + }, + { + "epoch": 3.5026395173453997, + "grad_norm": 0.12051997785202617, + "learning_rate": 1.9936199927576102e-05, + "loss": 0.4029, + "step": 9289 + }, + { + "epoch": 3.5030165912518854, + "grad_norm": 0.1122961843655721, + "learning_rate": 1.9927091188078184e-05, + "loss": 0.3948, + "step": 9290 + }, + { + "epoch": 3.503393665158371, + "grad_norm": 0.10188273029502092, + "learning_rate": 1.9917983839615037e-05, + "loss": 0.3944, + "step": 9291 + }, + { + "epoch": 3.5037707390648567, + "grad_norm": 0.12872749177637963, + "learning_rate": 1.9908877882817804e-05, + "loss": 0.4009, + "step": 9292 + }, + { + "epoch": 3.5041478129713424, + "grad_norm": 0.11839036054851759, + "learning_rate": 1.9899773318317507e-05, + "loss": 0.3884, + "step": 9293 + }, + { + "epoch": 3.504524886877828, + "grad_norm": 0.12803550951378354, + "learning_rate": 1.9890670146745104e-05, + "loss": 0.3895, + "step": 9294 + }, + { + "epoch": 3.5049019607843137, + "grad_norm": 0.13265844503791374, + "learning_rate": 1.988156836873141e-05, + "loss": 0.3948, + "step": 9295 + }, + { + "epoch": 3.5052790346907994, + "grad_norm": 0.13111088215486233, + "learning_rate": 1.987246798490719e-05, + "loss": 0.3897, + "step": 9296 + }, + { + "epoch": 3.505656108597285, + "grad_norm": 0.16820584835808924, + "learning_rate": 1.986336899590308e-05, + "loss": 0.3982, + "step": 9297 + }, + { + "epoch": 3.5060331825037707, + "grad_norm": 0.14551915853442068, + "learning_rate": 1.9854271402349633e-05, + "loss": 0.3991, + "step": 9298 + }, + { + "epoch": 3.5064102564102564, + "grad_norm": 0.13688825081234693, + "learning_rate": 1.9845175204877313e-05, + "loss": 0.3991, + "step": 9299 + }, + { + "epoch": 3.506787330316742, + "grad_norm": 0.1470584202683723, + "learning_rate": 1.983608040411649e-05, + "loss": 0.3965, + "step": 9300 + }, + { + "epoch": 3.5071644042232277, + "grad_norm": 0.1531747443435663, + "learning_rate": 1.9826987000697375e-05, + "loss": 0.3818, + "step": 9301 + }, + { + "epoch": 3.5075414781297134, + "grad_norm": 0.12566194805009093, + "learning_rate": 1.9817894995250204e-05, + "loss": 0.3971, + "step": 9302 + }, + { + "epoch": 3.507918552036199, + "grad_norm": 0.15490891786188332, + "learning_rate": 1.9808804388404986e-05, + "loss": 0.3981, + "step": 9303 + }, + { + "epoch": 3.5082956259426847, + "grad_norm": 0.12148088269174462, + "learning_rate": 1.9799715180791722e-05, + "loss": 0.3881, + "step": 9304 + }, + { + "epoch": 3.5086726998491704, + "grad_norm": 0.1395272548615298, + "learning_rate": 1.9790627373040273e-05, + "loss": 0.4018, + "step": 9305 + }, + { + "epoch": 3.509049773755656, + "grad_norm": 0.13617234500370282, + "learning_rate": 1.9781540965780434e-05, + "loss": 0.4011, + "step": 9306 + }, + { + "epoch": 3.5094268476621417, + "grad_norm": 0.11291478194429706, + "learning_rate": 1.977245595964185e-05, + "loss": 0.3808, + "step": 9307 + }, + { + "epoch": 3.5098039215686274, + "grad_norm": 0.1374755609505516, + "learning_rate": 1.9763372355254147e-05, + "loss": 0.408, + "step": 9308 + }, + { + "epoch": 3.510180995475113, + "grad_norm": 0.12414337875996677, + "learning_rate": 1.9754290153246784e-05, + "loss": 0.4053, + "step": 9309 + }, + { + "epoch": 3.5105580693815988, + "grad_norm": 0.11631509540558684, + "learning_rate": 1.974520935424915e-05, + "loss": 0.3954, + "step": 9310 + }, + { + "epoch": 3.5109351432880844, + "grad_norm": 0.14125664212975278, + "learning_rate": 1.9736129958890548e-05, + "loss": 0.393, + "step": 9311 + }, + { + "epoch": 3.51131221719457, + "grad_norm": 0.11636120163349813, + "learning_rate": 1.9727051967800173e-05, + "loss": 0.3882, + "step": 9312 + }, + { + "epoch": 3.5116892911010558, + "grad_norm": 0.13519950959108615, + "learning_rate": 1.9717975381607087e-05, + "loss": 0.3921, + "step": 9313 + }, + { + "epoch": 3.5120663650075414, + "grad_norm": 0.13375848546073388, + "learning_rate": 1.970890020094035e-05, + "loss": 0.4027, + "step": 9314 + }, + { + "epoch": 3.512443438914027, + "grad_norm": 0.12618774088578688, + "learning_rate": 1.9699826426428812e-05, + "loss": 0.3891, + "step": 9315 + }, + { + "epoch": 3.5128205128205128, + "grad_norm": 0.12277306245459138, + "learning_rate": 1.96907540587013e-05, + "loss": 0.3935, + "step": 9316 + }, + { + "epoch": 3.5131975867269984, + "grad_norm": 0.11172459410963696, + "learning_rate": 1.9681683098386514e-05, + "loss": 0.3929, + "step": 9317 + }, + { + "epoch": 3.513574660633484, + "grad_norm": 0.1293859184682345, + "learning_rate": 1.9672613546113082e-05, + "loss": 0.3951, + "step": 9318 + }, + { + "epoch": 3.51395173453997, + "grad_norm": 0.11059120292271843, + "learning_rate": 1.9663545402509473e-05, + "loss": 0.3925, + "step": 9319 + }, + { + "epoch": 3.5143288084464555, + "grad_norm": 0.1314194078761392, + "learning_rate": 1.965447866820416e-05, + "loss": 0.4043, + "step": 9320 + }, + { + "epoch": 3.514705882352941, + "grad_norm": 0.11176264185476989, + "learning_rate": 1.964541334382541e-05, + "loss": 0.4005, + "step": 9321 + }, + { + "epoch": 3.515082956259427, + "grad_norm": 0.122819527419827, + "learning_rate": 1.9636349430001464e-05, + "loss": 0.3881, + "step": 9322 + }, + { + "epoch": 3.5154600301659125, + "grad_norm": 0.11308907773959352, + "learning_rate": 1.9627286927360435e-05, + "loss": 0.4012, + "step": 9323 + }, + { + "epoch": 3.515837104072398, + "grad_norm": 0.12944271768334864, + "learning_rate": 1.9618225836530372e-05, + "loss": 0.3961, + "step": 9324 + }, + { + "epoch": 3.516214177978884, + "grad_norm": 0.1080634182740302, + "learning_rate": 1.960916615813915e-05, + "loss": 0.4023, + "step": 9325 + }, + { + "epoch": 3.5165912518853695, + "grad_norm": 0.13092811431204704, + "learning_rate": 1.9600107892814653e-05, + "loss": 0.3908, + "step": 9326 + }, + { + "epoch": 3.516968325791855, + "grad_norm": 0.10929928768914962, + "learning_rate": 1.9591051041184574e-05, + "loss": 0.3968, + "step": 9327 + }, + { + "epoch": 3.517345399698341, + "grad_norm": 0.13896551197304371, + "learning_rate": 1.958199560387655e-05, + "loss": 0.3852, + "step": 9328 + }, + { + "epoch": 3.5177224736048265, + "grad_norm": 0.10323280113055011, + "learning_rate": 1.9572941581518123e-05, + "loss": 0.4004, + "step": 9329 + }, + { + "epoch": 3.518099547511312, + "grad_norm": 0.12887258235903365, + "learning_rate": 1.956388897473674e-05, + "loss": 0.4021, + "step": 9330 + }, + { + "epoch": 3.518476621417798, + "grad_norm": 0.11877170096169123, + "learning_rate": 1.9554837784159693e-05, + "loss": 0.3945, + "step": 9331 + }, + { + "epoch": 3.5188536953242835, + "grad_norm": 0.12228118300253123, + "learning_rate": 1.9545788010414286e-05, + "loss": 0.4032, + "step": 9332 + }, + { + "epoch": 3.519230769230769, + "grad_norm": 0.13064548057978623, + "learning_rate": 1.9536739654127606e-05, + "loss": 0.3953, + "step": 9333 + }, + { + "epoch": 3.519607843137255, + "grad_norm": 0.12078184727866847, + "learning_rate": 1.9527692715926725e-05, + "loss": 0.4061, + "step": 9334 + }, + { + "epoch": 3.5199849170437405, + "grad_norm": 0.12742079919483498, + "learning_rate": 1.9518647196438577e-05, + "loss": 0.3935, + "step": 9335 + }, + { + "epoch": 3.520361990950226, + "grad_norm": 0.12099183893933664, + "learning_rate": 1.9509603096290023e-05, + "loss": 0.3978, + "step": 9336 + }, + { + "epoch": 3.520739064856712, + "grad_norm": 0.1254405318660002, + "learning_rate": 1.9500560416107773e-05, + "loss": 0.3911, + "step": 9337 + }, + { + "epoch": 3.5211161387631975, + "grad_norm": 0.11878873879607432, + "learning_rate": 1.949151915651853e-05, + "loss": 0.3966, + "step": 9338 + }, + { + "epoch": 3.521493212669683, + "grad_norm": 0.1200758269661856, + "learning_rate": 1.9482479318148805e-05, + "loss": 0.3937, + "step": 9339 + }, + { + "epoch": 3.521870286576169, + "grad_norm": 0.1175038312684731, + "learning_rate": 1.9473440901625066e-05, + "loss": 0.4052, + "step": 9340 + }, + { + "epoch": 3.5222473604826545, + "grad_norm": 0.10843360225958608, + "learning_rate": 1.9464403907573663e-05, + "loss": 0.4031, + "step": 9341 + }, + { + "epoch": 3.52262443438914, + "grad_norm": 0.10790816125602254, + "learning_rate": 1.9455368336620867e-05, + "loss": 0.4031, + "step": 9342 + }, + { + "epoch": 3.523001508295626, + "grad_norm": 0.11745198248951495, + "learning_rate": 1.94463341893928e-05, + "loss": 0.4014, + "step": 9343 + }, + { + "epoch": 3.5233785822021115, + "grad_norm": 0.10744832239526357, + "learning_rate": 1.9437301466515566e-05, + "loss": 0.3967, + "step": 9344 + }, + { + "epoch": 3.523755656108597, + "grad_norm": 0.10475485683324995, + "learning_rate": 1.9428270168615074e-05, + "loss": 0.3968, + "step": 9345 + }, + { + "epoch": 3.524132730015083, + "grad_norm": 0.11076097868244637, + "learning_rate": 1.9419240296317243e-05, + "loss": 0.3962, + "step": 9346 + }, + { + "epoch": 3.5245098039215685, + "grad_norm": 0.10310696057696414, + "learning_rate": 1.9410211850247786e-05, + "loss": 0.3901, + "step": 9347 + }, + { + "epoch": 3.524886877828054, + "grad_norm": 0.11774899200094043, + "learning_rate": 1.9401184831032386e-05, + "loss": 0.3988, + "step": 9348 + }, + { + "epoch": 3.52526395173454, + "grad_norm": 0.11847392875978059, + "learning_rate": 1.9392159239296604e-05, + "loss": 0.3881, + "step": 9349 + }, + { + "epoch": 3.5256410256410255, + "grad_norm": 0.11406828079627994, + "learning_rate": 1.9383135075665927e-05, + "loss": 0.3934, + "step": 9350 + }, + { + "epoch": 3.526018099547511, + "grad_norm": 0.11136270480307975, + "learning_rate": 1.9374112340765672e-05, + "loss": 0.3966, + "step": 9351 + }, + { + "epoch": 3.526395173453997, + "grad_norm": 0.11681252009205448, + "learning_rate": 1.9365091035221165e-05, + "loss": 0.3952, + "step": 9352 + }, + { + "epoch": 3.5267722473604826, + "grad_norm": 0.1112319611518189, + "learning_rate": 1.9356071159657528e-05, + "loss": 0.3892, + "step": 9353 + }, + { + "epoch": 3.5271493212669682, + "grad_norm": 0.11744530236821281, + "learning_rate": 1.9347052714699853e-05, + "loss": 0.4022, + "step": 9354 + }, + { + "epoch": 3.527526395173454, + "grad_norm": 0.09804234802281463, + "learning_rate": 1.9338035700973106e-05, + "loss": 0.3917, + "step": 9355 + }, + { + "epoch": 3.5279034690799396, + "grad_norm": 0.1237438523255016, + "learning_rate": 1.932902011910217e-05, + "loss": 0.4074, + "step": 9356 + }, + { + "epoch": 3.5282805429864252, + "grad_norm": 0.09434906611560183, + "learning_rate": 1.9320005969711773e-05, + "loss": 0.3943, + "step": 9357 + }, + { + "epoch": 3.528657616892911, + "grad_norm": 0.1260377221221759, + "learning_rate": 1.9310993253426657e-05, + "loss": 0.3879, + "step": 9358 + }, + { + "epoch": 3.5290346907993966, + "grad_norm": 0.1165541174235149, + "learning_rate": 1.9301981970871333e-05, + "loss": 0.4046, + "step": 9359 + }, + { + "epoch": 3.5294117647058822, + "grad_norm": 0.11398032249729614, + "learning_rate": 1.9292972122670303e-05, + "loss": 0.398, + "step": 9360 + }, + { + "epoch": 3.529788838612368, + "grad_norm": 0.13134712540686833, + "learning_rate": 1.9283963709447938e-05, + "loss": 0.3916, + "step": 9361 + }, + { + "epoch": 3.5301659125188536, + "grad_norm": 0.11271377703016308, + "learning_rate": 1.9274956731828525e-05, + "loss": 0.4092, + "step": 9362 + }, + { + "epoch": 3.5305429864253393, + "grad_norm": 0.11704661226903156, + "learning_rate": 1.9265951190436203e-05, + "loss": 0.3961, + "step": 9363 + }, + { + "epoch": 3.530920060331825, + "grad_norm": 0.11472151376281145, + "learning_rate": 1.9256947085895098e-05, + "loss": 0.39, + "step": 9364 + }, + { + "epoch": 3.5312971342383106, + "grad_norm": 0.1197835881825647, + "learning_rate": 1.9247944418829152e-05, + "loss": 0.406, + "step": 9365 + }, + { + "epoch": 3.5316742081447963, + "grad_norm": 0.1126339291585866, + "learning_rate": 1.9238943189862246e-05, + "loss": 0.3785, + "step": 9366 + }, + { + "epoch": 3.532051282051282, + "grad_norm": 0.11743042427315308, + "learning_rate": 1.9229943399618162e-05, + "loss": 0.3936, + "step": 9367 + }, + { + "epoch": 3.5324283559577676, + "grad_norm": 0.12750276663453697, + "learning_rate": 1.9220945048720594e-05, + "loss": 0.3991, + "step": 9368 + }, + { + "epoch": 3.5328054298642533, + "grad_norm": 0.11638895094479591, + "learning_rate": 1.921194813779307e-05, + "loss": 0.3971, + "step": 9369 + }, + { + "epoch": 3.533182503770739, + "grad_norm": 0.1248040249729486, + "learning_rate": 1.9202952667459134e-05, + "loss": 0.4004, + "step": 9370 + }, + { + "epoch": 3.5335595776772246, + "grad_norm": 0.11800793649624303, + "learning_rate": 1.9193958638342116e-05, + "loss": 0.3937, + "step": 9371 + }, + { + "epoch": 3.5339366515837103, + "grad_norm": 0.11647065753615732, + "learning_rate": 1.9184966051065306e-05, + "loss": 0.3904, + "step": 9372 + }, + { + "epoch": 3.534313725490196, + "grad_norm": 0.12480432696029793, + "learning_rate": 1.917597490625189e-05, + "loss": 0.398, + "step": 9373 + }, + { + "epoch": 3.5346907993966816, + "grad_norm": 0.12684895152380654, + "learning_rate": 1.916698520452496e-05, + "loss": 0.394, + "step": 9374 + }, + { + "epoch": 3.5350678733031673, + "grad_norm": 0.11179311462996733, + "learning_rate": 1.9157996946507437e-05, + "loss": 0.3949, + "step": 9375 + }, + { + "epoch": 3.535444947209653, + "grad_norm": 0.12963839597291735, + "learning_rate": 1.9149010132822274e-05, + "loss": 0.4075, + "step": 9376 + }, + { + "epoch": 3.5358220211161386, + "grad_norm": 0.09675119526777569, + "learning_rate": 1.9140024764092198e-05, + "loss": 0.3968, + "step": 9377 + }, + { + "epoch": 3.5361990950226243, + "grad_norm": 0.12774447937804279, + "learning_rate": 1.9131040840939904e-05, + "loss": 0.3888, + "step": 9378 + }, + { + "epoch": 3.53657616892911, + "grad_norm": 0.12063999535490855, + "learning_rate": 1.9122058363987967e-05, + "loss": 0.3934, + "step": 9379 + }, + { + "epoch": 3.5369532428355956, + "grad_norm": 0.11886604589858941, + "learning_rate": 1.9113077333858885e-05, + "loss": 0.3888, + "step": 9380 + }, + { + "epoch": 3.5373303167420813, + "grad_norm": 0.11697559481010887, + "learning_rate": 1.9104097751174985e-05, + "loss": 0.3939, + "step": 9381 + }, + { + "epoch": 3.537707390648567, + "grad_norm": 0.1157232134954378, + "learning_rate": 1.9095119616558608e-05, + "loss": 0.3979, + "step": 9382 + }, + { + "epoch": 3.5380844645550527, + "grad_norm": 0.12183433017742584, + "learning_rate": 1.9086142930631882e-05, + "loss": 0.4034, + "step": 9383 + }, + { + "epoch": 3.5384615384615383, + "grad_norm": 0.11858427641333219, + "learning_rate": 1.9077167694016903e-05, + "loss": 0.3864, + "step": 9384 + }, + { + "epoch": 3.538838612368024, + "grad_norm": 0.1467340685652513, + "learning_rate": 1.906819390733564e-05, + "loss": 0.3919, + "step": 9385 + }, + { + "epoch": 3.5392156862745097, + "grad_norm": 0.11158679822588836, + "learning_rate": 1.905922157120999e-05, + "loss": 0.4094, + "step": 9386 + }, + { + "epoch": 3.5395927601809953, + "grad_norm": 0.12138988789990765, + "learning_rate": 1.9050250686261682e-05, + "loss": 0.3968, + "step": 9387 + }, + { + "epoch": 3.539969834087481, + "grad_norm": 0.12164573370909089, + "learning_rate": 1.9041281253112447e-05, + "loss": 0.4037, + "step": 9388 + }, + { + "epoch": 3.5403469079939667, + "grad_norm": 0.10656244778666106, + "learning_rate": 1.9032313272383813e-05, + "loss": 0.3921, + "step": 9389 + }, + { + "epoch": 3.5407239819004523, + "grad_norm": 0.1245044668409161, + "learning_rate": 1.9023346744697276e-05, + "loss": 0.3908, + "step": 9390 + }, + { + "epoch": 3.541101055806938, + "grad_norm": 0.11682796511217643, + "learning_rate": 1.9014381670674203e-05, + "loss": 0.4014, + "step": 9391 + }, + { + "epoch": 3.5414781297134237, + "grad_norm": 0.13174750920376863, + "learning_rate": 1.900541805093588e-05, + "loss": 0.3923, + "step": 9392 + }, + { + "epoch": 3.5418552036199094, + "grad_norm": 0.12284103849034407, + "learning_rate": 1.8996455886103433e-05, + "loss": 0.4058, + "step": 9393 + }, + { + "epoch": 3.542232277526395, + "grad_norm": 0.10939991172667062, + "learning_rate": 1.8987495176797992e-05, + "loss": 0.4009, + "step": 9394 + }, + { + "epoch": 3.5426093514328807, + "grad_norm": 0.13938724522728158, + "learning_rate": 1.8978535923640488e-05, + "loss": 0.3866, + "step": 9395 + }, + { + "epoch": 3.5429864253393664, + "grad_norm": 0.11993592906920005, + "learning_rate": 1.8969578127251795e-05, + "loss": 0.3977, + "step": 9396 + }, + { + "epoch": 3.543363499245852, + "grad_norm": 0.13505409326056986, + "learning_rate": 1.896062178825268e-05, + "loss": 0.4005, + "step": 9397 + }, + { + "epoch": 3.5437405731523377, + "grad_norm": 0.12483953566497671, + "learning_rate": 1.895166690726383e-05, + "loss": 0.3811, + "step": 9398 + }, + { + "epoch": 3.5441176470588234, + "grad_norm": 0.14800353295501756, + "learning_rate": 1.8942713484905762e-05, + "loss": 0.3985, + "step": 9399 + }, + { + "epoch": 3.544494720965309, + "grad_norm": 0.10413207318997059, + "learning_rate": 1.893376152179901e-05, + "loss": 0.3927, + "step": 9400 + }, + { + "epoch": 3.5448717948717947, + "grad_norm": 0.1283897750175932, + "learning_rate": 1.8924811018563874e-05, + "loss": 0.3994, + "step": 9401 + }, + { + "epoch": 3.5452488687782804, + "grad_norm": 0.12388671348139454, + "learning_rate": 1.8915861975820647e-05, + "loss": 0.39, + "step": 9402 + }, + { + "epoch": 3.545625942684766, + "grad_norm": 0.1320293227151082, + "learning_rate": 1.890691439418948e-05, + "loss": 0.3983, + "step": 9403 + }, + { + "epoch": 3.5460030165912517, + "grad_norm": 0.139367173052056, + "learning_rate": 1.8897968274290454e-05, + "loss": 0.3995, + "step": 9404 + }, + { + "epoch": 3.5463800904977374, + "grad_norm": 0.12179910088944738, + "learning_rate": 1.8889023616743477e-05, + "loss": 0.4003, + "step": 9405 + }, + { + "epoch": 3.546757164404223, + "grad_norm": 0.11112853786471745, + "learning_rate": 1.888008042216847e-05, + "loss": 0.3941, + "step": 9406 + }, + { + "epoch": 3.5471342383107087, + "grad_norm": 0.13754965687608176, + "learning_rate": 1.887113869118514e-05, + "loss": 0.405, + "step": 9407 + }, + { + "epoch": 3.5475113122171944, + "grad_norm": 0.1264277008766783, + "learning_rate": 1.8862198424413155e-05, + "loss": 0.392, + "step": 9408 + }, + { + "epoch": 3.54788838612368, + "grad_norm": 0.1217388514335498, + "learning_rate": 1.8853259622472074e-05, + "loss": 0.3979, + "step": 9409 + }, + { + "epoch": 3.5482654600301657, + "grad_norm": 0.14789521665143898, + "learning_rate": 1.8844322285981354e-05, + "loss": 0.3919, + "step": 9410 + }, + { + "epoch": 3.5486425339366514, + "grad_norm": 0.1029407290142451, + "learning_rate": 1.8835386415560308e-05, + "loss": 0.393, + "step": 9411 + }, + { + "epoch": 3.549019607843137, + "grad_norm": 0.13540616290399948, + "learning_rate": 1.882645201182824e-05, + "loss": 0.3826, + "step": 9412 + }, + { + "epoch": 3.5493966817496228, + "grad_norm": 0.11681764290701584, + "learning_rate": 1.881751907540425e-05, + "loss": 0.4061, + "step": 9413 + }, + { + "epoch": 3.5497737556561084, + "grad_norm": 0.12369788250501385, + "learning_rate": 1.8808587606907403e-05, + "loss": 0.4004, + "step": 9414 + }, + { + "epoch": 3.550150829562594, + "grad_norm": 0.12772250510942218, + "learning_rate": 1.879965760695664e-05, + "loss": 0.3989, + "step": 9415 + }, + { + "epoch": 3.5505279034690798, + "grad_norm": 0.11627484150171158, + "learning_rate": 1.8790729076170792e-05, + "loss": 0.3869, + "step": 9416 + }, + { + "epoch": 3.5509049773755654, + "grad_norm": 0.15711112573736757, + "learning_rate": 1.878180201516861e-05, + "loss": 0.4073, + "step": 9417 + }, + { + "epoch": 3.551282051282051, + "grad_norm": 0.12900416724207944, + "learning_rate": 1.8772876424568745e-05, + "loss": 0.3936, + "step": 9418 + }, + { + "epoch": 3.5516591251885368, + "grad_norm": 0.1255337557537473, + "learning_rate": 1.87639523049897e-05, + "loss": 0.4147, + "step": 9419 + }, + { + "epoch": 3.5520361990950224, + "grad_norm": 0.12439064922922825, + "learning_rate": 1.875502965704992e-05, + "loss": 0.3938, + "step": 9420 + }, + { + "epoch": 3.552413273001508, + "grad_norm": 0.13312429950744692, + "learning_rate": 1.8746108481367745e-05, + "loss": 0.3861, + "step": 9421 + }, + { + "epoch": 3.552790346907994, + "grad_norm": 0.12540947933177313, + "learning_rate": 1.87371887785614e-05, + "loss": 0.4066, + "step": 9422 + }, + { + "epoch": 3.5531674208144794, + "grad_norm": 0.1241626970884063, + "learning_rate": 1.8728270549249014e-05, + "loss": 0.4, + "step": 9423 + }, + { + "epoch": 3.553544494720965, + "grad_norm": 0.123689400109073, + "learning_rate": 1.871935379404862e-05, + "loss": 0.4027, + "step": 9424 + }, + { + "epoch": 3.553921568627451, + "grad_norm": 0.12448059617769978, + "learning_rate": 1.8710438513578116e-05, + "loss": 0.4058, + "step": 9425 + }, + { + "epoch": 3.5542986425339365, + "grad_norm": 0.10549418820366038, + "learning_rate": 1.8701524708455348e-05, + "loss": 0.39, + "step": 9426 + }, + { + "epoch": 3.554675716440422, + "grad_norm": 0.13023235377280787, + "learning_rate": 1.869261237929802e-05, + "loss": 0.3932, + "step": 9427 + }, + { + "epoch": 3.555052790346908, + "grad_norm": 0.11993384391637729, + "learning_rate": 1.8683701526723753e-05, + "loss": 0.4055, + "step": 9428 + }, + { + "epoch": 3.5554298642533935, + "grad_norm": 0.12953312769705733, + "learning_rate": 1.8674792151350067e-05, + "loss": 0.3892, + "step": 9429 + }, + { + "epoch": 3.555806938159879, + "grad_norm": 0.1298744205207571, + "learning_rate": 1.866588425379438e-05, + "loss": 0.3932, + "step": 9430 + }, + { + "epoch": 3.556184012066365, + "grad_norm": 0.11976806301453098, + "learning_rate": 1.8656977834673983e-05, + "loss": 0.3959, + "step": 9431 + }, + { + "epoch": 3.5565610859728505, + "grad_norm": 0.15280898523895792, + "learning_rate": 1.8648072894606084e-05, + "loss": 0.3978, + "step": 9432 + }, + { + "epoch": 3.556938159879336, + "grad_norm": 0.11009437894898372, + "learning_rate": 1.86391694342078e-05, + "loss": 0.3875, + "step": 9433 + }, + { + "epoch": 3.557315233785822, + "grad_norm": 0.15914225938253965, + "learning_rate": 1.863026745409612e-05, + "loss": 0.3956, + "step": 9434 + }, + { + "epoch": 3.5576923076923075, + "grad_norm": 0.1181821015534817, + "learning_rate": 1.8621366954887958e-05, + "loss": 0.4013, + "step": 9435 + }, + { + "epoch": 3.558069381598793, + "grad_norm": 0.14946404575792738, + "learning_rate": 1.8612467937200097e-05, + "loss": 0.4008, + "step": 9436 + }, + { + "epoch": 3.558446455505279, + "grad_norm": 0.13255138158869895, + "learning_rate": 1.8603570401649258e-05, + "loss": 0.398, + "step": 9437 + }, + { + "epoch": 3.5588235294117645, + "grad_norm": 0.11919840953736172, + "learning_rate": 1.8594674348851992e-05, + "loss": 0.3913, + "step": 9438 + }, + { + "epoch": 3.55920060331825, + "grad_norm": 0.14137589840458467, + "learning_rate": 1.8585779779424807e-05, + "loss": 0.3911, + "step": 9439 + }, + { + "epoch": 3.559577677224736, + "grad_norm": 0.11639661011637288, + "learning_rate": 1.8576886693984092e-05, + "loss": 0.3979, + "step": 9440 + }, + { + "epoch": 3.5599547511312215, + "grad_norm": 0.1429587074968547, + "learning_rate": 1.8567995093146125e-05, + "loss": 0.3894, + "step": 9441 + }, + { + "epoch": 3.560331825037707, + "grad_norm": 0.13424677808905547, + "learning_rate": 1.8559104977527088e-05, + "loss": 0.4035, + "step": 9442 + }, + { + "epoch": 3.560708898944193, + "grad_norm": 0.11841843437341162, + "learning_rate": 1.8550216347743072e-05, + "loss": 0.4095, + "step": 9443 + }, + { + "epoch": 3.5610859728506785, + "grad_norm": 0.13283521975084472, + "learning_rate": 1.8541329204410027e-05, + "loss": 0.3889, + "step": 9444 + }, + { + "epoch": 3.561463046757164, + "grad_norm": 0.10886570656822123, + "learning_rate": 1.853244354814383e-05, + "loss": 0.395, + "step": 9445 + }, + { + "epoch": 3.56184012066365, + "grad_norm": 0.14808075013174712, + "learning_rate": 1.8523559379560252e-05, + "loss": 0.3991, + "step": 9446 + }, + { + "epoch": 3.5622171945701355, + "grad_norm": 0.1366345633384518, + "learning_rate": 1.8514676699274963e-05, + "loss": 0.4026, + "step": 9447 + }, + { + "epoch": 3.562594268476621, + "grad_norm": 0.13156792876552414, + "learning_rate": 1.8505795507903518e-05, + "loss": 0.4095, + "step": 9448 + }, + { + "epoch": 3.562971342383107, + "grad_norm": 0.11737767135743525, + "learning_rate": 1.8496915806061398e-05, + "loss": 0.3859, + "step": 9449 + }, + { + "epoch": 3.5633484162895925, + "grad_norm": 0.12512090550111307, + "learning_rate": 1.8488037594363904e-05, + "loss": 0.3901, + "step": 9450 + }, + { + "epoch": 3.563725490196078, + "grad_norm": 0.0974587963527636, + "learning_rate": 1.8479160873426356e-05, + "loss": 0.3865, + "step": 9451 + }, + { + "epoch": 3.564102564102564, + "grad_norm": 0.11055729887900705, + "learning_rate": 1.8470285643863856e-05, + "loss": 0.4029, + "step": 9452 + }, + { + "epoch": 3.5644796380090495, + "grad_norm": 0.10673579491830816, + "learning_rate": 1.8461411906291466e-05, + "loss": 0.3998, + "step": 9453 + }, + { + "epoch": 3.564856711915535, + "grad_norm": 0.10381871901938516, + "learning_rate": 1.8452539661324124e-05, + "loss": 0.4077, + "step": 9454 + }, + { + "epoch": 3.565233785822021, + "grad_norm": 0.10937519437158044, + "learning_rate": 1.8443668909576688e-05, + "loss": 0.3993, + "step": 9455 + }, + { + "epoch": 3.5656108597285066, + "grad_norm": 0.10890779102112583, + "learning_rate": 1.8434799651663845e-05, + "loss": 0.3973, + "step": 9456 + }, + { + "epoch": 3.5659879336349922, + "grad_norm": 0.11718229758002399, + "learning_rate": 1.8425931888200287e-05, + "loss": 0.392, + "step": 9457 + }, + { + "epoch": 3.566365007541478, + "grad_norm": 0.11210175813308743, + "learning_rate": 1.8417065619800497e-05, + "loss": 0.3995, + "step": 9458 + }, + { + "epoch": 3.5667420814479636, + "grad_norm": 0.11380479414997957, + "learning_rate": 1.840820084707892e-05, + "loss": 0.3816, + "step": 9459 + }, + { + "epoch": 3.5671191553544492, + "grad_norm": 0.10612372555436618, + "learning_rate": 1.8399337570649867e-05, + "loss": 0.4056, + "step": 9460 + }, + { + "epoch": 3.567496229260935, + "grad_norm": 0.11520819800183954, + "learning_rate": 1.8390475791127578e-05, + "loss": 0.403, + "step": 9461 + }, + { + "epoch": 3.5678733031674206, + "grad_norm": 0.10627934831626756, + "learning_rate": 1.8381615509126122e-05, + "loss": 0.3936, + "step": 9462 + }, + { + "epoch": 3.5682503770739062, + "grad_norm": 0.11520688340576377, + "learning_rate": 1.8372756725259558e-05, + "loss": 0.3867, + "step": 9463 + }, + { + "epoch": 3.568627450980392, + "grad_norm": 0.1094378116540573, + "learning_rate": 1.8363899440141758e-05, + "loss": 0.387, + "step": 9464 + }, + { + "epoch": 3.5690045248868776, + "grad_norm": 0.10718323364109755, + "learning_rate": 1.835504365438654e-05, + "loss": 0.3911, + "step": 9465 + }, + { + "epoch": 3.5693815987933633, + "grad_norm": 0.1270030620774799, + "learning_rate": 1.834618936860759e-05, + "loss": 0.4061, + "step": 9466 + }, + { + "epoch": 3.569758672699849, + "grad_norm": 0.11132098231514485, + "learning_rate": 1.8337336583418533e-05, + "loss": 0.3951, + "step": 9467 + }, + { + "epoch": 3.5701357466063346, + "grad_norm": 0.10838280914750636, + "learning_rate": 1.8328485299432803e-05, + "loss": 0.4048, + "step": 9468 + }, + { + "epoch": 3.5705128205128203, + "grad_norm": 0.11654799367693941, + "learning_rate": 1.8319635517263846e-05, + "loss": 0.3914, + "step": 9469 + }, + { + "epoch": 3.570889894419306, + "grad_norm": 0.11096792319318206, + "learning_rate": 1.831078723752489e-05, + "loss": 0.388, + "step": 9470 + }, + { + "epoch": 3.5712669683257916, + "grad_norm": 0.1090611134948627, + "learning_rate": 1.8301940460829174e-05, + "loss": 0.3903, + "step": 9471 + }, + { + "epoch": 3.5716440422322773, + "grad_norm": 0.11483816499219719, + "learning_rate": 1.8293095187789717e-05, + "loss": 0.4039, + "step": 9472 + }, + { + "epoch": 3.572021116138763, + "grad_norm": 0.10626452956215925, + "learning_rate": 1.828425141901951e-05, + "loss": 0.3939, + "step": 9473 + }, + { + "epoch": 3.5723981900452486, + "grad_norm": 0.10959939531776215, + "learning_rate": 1.8275409155131417e-05, + "loss": 0.3975, + "step": 9474 + }, + { + "epoch": 3.5727752639517343, + "grad_norm": 0.12953178422109615, + "learning_rate": 1.826656839673822e-05, + "loss": 0.4038, + "step": 9475 + }, + { + "epoch": 3.57315233785822, + "grad_norm": 0.11796400515537782, + "learning_rate": 1.8257729144452524e-05, + "loss": 0.3823, + "step": 9476 + }, + { + "epoch": 3.5735294117647056, + "grad_norm": 0.10331037991187925, + "learning_rate": 1.824889139888694e-05, + "loss": 0.386, + "step": 9477 + }, + { + "epoch": 3.5739064856711913, + "grad_norm": 0.12476703618857515, + "learning_rate": 1.8240055160653875e-05, + "loss": 0.3972, + "step": 9478 + }, + { + "epoch": 3.574283559577677, + "grad_norm": 0.12833456149275174, + "learning_rate": 1.8231220430365684e-05, + "loss": 0.3926, + "step": 9479 + }, + { + "epoch": 3.5746606334841626, + "grad_norm": 0.11683268933268456, + "learning_rate": 1.822238720863461e-05, + "loss": 0.4014, + "step": 9480 + }, + { + "epoch": 3.5750377073906483, + "grad_norm": 0.11316935692687273, + "learning_rate": 1.8213555496072802e-05, + "loss": 0.3922, + "step": 9481 + }, + { + "epoch": 3.575414781297134, + "grad_norm": 0.11683961185416572, + "learning_rate": 1.8204725293292246e-05, + "loss": 0.3916, + "step": 9482 + }, + { + "epoch": 3.5757918552036196, + "grad_norm": 0.10526294069331629, + "learning_rate": 1.8195896600904924e-05, + "loss": 0.3928, + "step": 9483 + }, + { + "epoch": 3.5761689291101053, + "grad_norm": 0.11942483339255744, + "learning_rate": 1.818706941952261e-05, + "loss": 0.3896, + "step": 9484 + }, + { + "epoch": 3.576546003016591, + "grad_norm": 0.10063334765390602, + "learning_rate": 1.8178243749757033e-05, + "loss": 0.3965, + "step": 9485 + }, + { + "epoch": 3.5769230769230766, + "grad_norm": 0.11906670667009439, + "learning_rate": 1.8169419592219813e-05, + "loss": 0.3946, + "step": 9486 + }, + { + "epoch": 3.5773001508295623, + "grad_norm": 0.11866572319670679, + "learning_rate": 1.8160596947522465e-05, + "loss": 0.4089, + "step": 9487 + }, + { + "epoch": 3.577677224736048, + "grad_norm": 0.10168044277069532, + "learning_rate": 1.815177581627635e-05, + "loss": 0.3963, + "step": 9488 + }, + { + "epoch": 3.5780542986425337, + "grad_norm": 0.12668135157947683, + "learning_rate": 1.8142956199092823e-05, + "loss": 0.3978, + "step": 9489 + }, + { + "epoch": 3.5784313725490198, + "grad_norm": 0.10452530492726272, + "learning_rate": 1.8134138096583024e-05, + "loss": 0.3939, + "step": 9490 + }, + { + "epoch": 3.5788084464555054, + "grad_norm": 0.11872346115207599, + "learning_rate": 1.8125321509358064e-05, + "loss": 0.3897, + "step": 9491 + }, + { + "epoch": 3.579185520361991, + "grad_norm": 0.11704937067138065, + "learning_rate": 1.8116506438028918e-05, + "loss": 0.3902, + "step": 9492 + }, + { + "epoch": 3.579562594268477, + "grad_norm": 0.12815523442639615, + "learning_rate": 1.810769288320648e-05, + "loss": 0.3874, + "step": 9493 + }, + { + "epoch": 3.5799396681749625, + "grad_norm": 0.1236498980823928, + "learning_rate": 1.809888084550148e-05, + "loss": 0.3866, + "step": 9494 + }, + { + "epoch": 3.580316742081448, + "grad_norm": 0.1389072966324836, + "learning_rate": 1.8090070325524634e-05, + "loss": 0.4073, + "step": 9495 + }, + { + "epoch": 3.580693815987934, + "grad_norm": 0.113994873320034, + "learning_rate": 1.808126132388647e-05, + "loss": 0.3888, + "step": 9496 + }, + { + "epoch": 3.5810708898944195, + "grad_norm": 0.14059760117199485, + "learning_rate": 1.8072453841197454e-05, + "loss": 0.3876, + "step": 9497 + }, + { + "epoch": 3.581447963800905, + "grad_norm": 0.11242633329258941, + "learning_rate": 1.806364787806794e-05, + "loss": 0.3798, + "step": 9498 + }, + { + "epoch": 3.581825037707391, + "grad_norm": 0.1411112970012984, + "learning_rate": 1.805484343510818e-05, + "loss": 0.3892, + "step": 9499 + }, + { + "epoch": 3.5822021116138765, + "grad_norm": 0.10152178785792092, + "learning_rate": 1.8046040512928278e-05, + "loss": 0.3948, + "step": 9500 + }, + { + "epoch": 3.582579185520362, + "grad_norm": 0.12423176065317001, + "learning_rate": 1.8037239112138327e-05, + "loss": 0.3932, + "step": 9501 + }, + { + "epoch": 3.582956259426848, + "grad_norm": 0.11754424675280498, + "learning_rate": 1.8028439233348204e-05, + "loss": 0.3935, + "step": 9502 + }, + { + "epoch": 3.5833333333333335, + "grad_norm": 0.11090557163576721, + "learning_rate": 1.8019640877167763e-05, + "loss": 0.3921, + "step": 9503 + }, + { + "epoch": 3.583710407239819, + "grad_norm": 0.11481562860943782, + "learning_rate": 1.8010844044206705e-05, + "loss": 0.3997, + "step": 9504 + }, + { + "epoch": 3.584087481146305, + "grad_norm": 0.1024383965085693, + "learning_rate": 1.8002048735074663e-05, + "loss": 0.3933, + "step": 9505 + }, + { + "epoch": 3.5844645550527905, + "grad_norm": 0.11699511861931848, + "learning_rate": 1.7993254950381108e-05, + "loss": 0.3937, + "step": 9506 + }, + { + "epoch": 3.584841628959276, + "grad_norm": 0.13383467791313358, + "learning_rate": 1.798446269073549e-05, + "loss": 0.4054, + "step": 9507 + }, + { + "epoch": 3.585218702865762, + "grad_norm": 0.12175617924102666, + "learning_rate": 1.797567195674707e-05, + "loss": 0.3976, + "step": 9508 + }, + { + "epoch": 3.5855957767722475, + "grad_norm": 0.11488599292337254, + "learning_rate": 1.7966882749025045e-05, + "loss": 0.3995, + "step": 9509 + }, + { + "epoch": 3.585972850678733, + "grad_norm": 0.13481412186748606, + "learning_rate": 1.79580950681785e-05, + "loss": 0.3854, + "step": 9510 + }, + { + "epoch": 3.586349924585219, + "grad_norm": 0.10922402863035155, + "learning_rate": 1.7949308914816433e-05, + "loss": 0.3821, + "step": 9511 + }, + { + "epoch": 3.5867269984917045, + "grad_norm": 0.132897142687829, + "learning_rate": 1.794052428954767e-05, + "loss": 0.3887, + "step": 9512 + }, + { + "epoch": 3.58710407239819, + "grad_norm": 0.12214509692473748, + "learning_rate": 1.7931741192981036e-05, + "loss": 0.3944, + "step": 9513 + }, + { + "epoch": 3.587481146304676, + "grad_norm": 0.11430950603368414, + "learning_rate": 1.7922959625725148e-05, + "loss": 0.3943, + "step": 9514 + }, + { + "epoch": 3.5878582202111615, + "grad_norm": 0.13296237908614802, + "learning_rate": 1.7914179588388575e-05, + "loss": 0.3937, + "step": 9515 + }, + { + "epoch": 3.588235294117647, + "grad_norm": 0.11130714354264062, + "learning_rate": 1.790540108157977e-05, + "loss": 0.3934, + "step": 9516 + }, + { + "epoch": 3.588612368024133, + "grad_norm": 0.11757064839491813, + "learning_rate": 1.789662410590708e-05, + "loss": 0.392, + "step": 9517 + }, + { + "epoch": 3.5889894419306185, + "grad_norm": 0.12275741480677055, + "learning_rate": 1.788784866197871e-05, + "loss": 0.3986, + "step": 9518 + }, + { + "epoch": 3.589366515837104, + "grad_norm": 0.11779118046659999, + "learning_rate": 1.7879074750402844e-05, + "loss": 0.4003, + "step": 9519 + }, + { + "epoch": 3.58974358974359, + "grad_norm": 0.11832583940595591, + "learning_rate": 1.787030237178746e-05, + "loss": 0.4035, + "step": 9520 + }, + { + "epoch": 3.5901206636500755, + "grad_norm": 0.12092789006778921, + "learning_rate": 1.786153152674049e-05, + "loss": 0.404, + "step": 9521 + }, + { + "epoch": 3.590497737556561, + "grad_norm": 0.13405139305074376, + "learning_rate": 1.785276221586975e-05, + "loss": 0.3892, + "step": 9522 + }, + { + "epoch": 3.590874811463047, + "grad_norm": 0.10801239356120704, + "learning_rate": 1.784399443978296e-05, + "loss": 0.3988, + "step": 9523 + }, + { + "epoch": 3.5912518853695325, + "grad_norm": 0.13468696645942493, + "learning_rate": 1.7835228199087672e-05, + "loss": 0.3998, + "step": 9524 + }, + { + "epoch": 3.591628959276018, + "grad_norm": 0.1019059068204522, + "learning_rate": 1.7826463494391442e-05, + "loss": 0.3993, + "step": 9525 + }, + { + "epoch": 3.592006033182504, + "grad_norm": 0.12346283151439492, + "learning_rate": 1.781770032630161e-05, + "loss": 0.3861, + "step": 9526 + }, + { + "epoch": 3.5923831070889896, + "grad_norm": 0.10633553593168965, + "learning_rate": 1.7808938695425463e-05, + "loss": 0.3912, + "step": 9527 + }, + { + "epoch": 3.5927601809954752, + "grad_norm": 0.13535217053042023, + "learning_rate": 1.7800178602370187e-05, + "loss": 0.3983, + "step": 9528 + }, + { + "epoch": 3.593137254901961, + "grad_norm": 0.11009739360229669, + "learning_rate": 1.7791420047742855e-05, + "loss": 0.4059, + "step": 9529 + }, + { + "epoch": 3.5935143288084466, + "grad_norm": 0.13285599452453464, + "learning_rate": 1.7782663032150386e-05, + "loss": 0.3926, + "step": 9530 + }, + { + "epoch": 3.5938914027149322, + "grad_norm": 0.10620327159804085, + "learning_rate": 1.777390755619969e-05, + "loss": 0.392, + "step": 9531 + }, + { + "epoch": 3.594268476621418, + "grad_norm": 0.12580272417958369, + "learning_rate": 1.776515362049747e-05, + "loss": 0.4068, + "step": 9532 + }, + { + "epoch": 3.5946455505279036, + "grad_norm": 0.12062765698273406, + "learning_rate": 1.7756401225650385e-05, + "loss": 0.399, + "step": 9533 + }, + { + "epoch": 3.5950226244343892, + "grad_norm": 0.11923943754586455, + "learning_rate": 1.7747650372264962e-05, + "loss": 0.4014, + "step": 9534 + }, + { + "epoch": 3.595399698340875, + "grad_norm": 0.13060263078691625, + "learning_rate": 1.773890106094765e-05, + "loss": 0.4002, + "step": 9535 + }, + { + "epoch": 3.5957767722473606, + "grad_norm": 0.11868605113860757, + "learning_rate": 1.7730153292304714e-05, + "loss": 0.4034, + "step": 9536 + }, + { + "epoch": 3.5961538461538463, + "grad_norm": 0.11030115002797526, + "learning_rate": 1.7721407066942435e-05, + "loss": 0.3953, + "step": 9537 + }, + { + "epoch": 3.596530920060332, + "grad_norm": 0.12365891375464759, + "learning_rate": 1.771266238546687e-05, + "loss": 0.3963, + "step": 9538 + }, + { + "epoch": 3.5969079939668176, + "grad_norm": 0.10499319849997399, + "learning_rate": 1.770391924848403e-05, + "loss": 0.3902, + "step": 9539 + }, + { + "epoch": 3.5972850678733033, + "grad_norm": 0.11481179875525185, + "learning_rate": 1.7695177656599812e-05, + "loss": 0.3966, + "step": 9540 + }, + { + "epoch": 3.597662141779789, + "grad_norm": 0.11605971085420033, + "learning_rate": 1.768643761042001e-05, + "loss": 0.3907, + "step": 9541 + }, + { + "epoch": 3.5980392156862746, + "grad_norm": 0.12278501711126498, + "learning_rate": 1.767769911055026e-05, + "loss": 0.39, + "step": 9542 + }, + { + "epoch": 3.5984162895927603, + "grad_norm": 0.10446505416955543, + "learning_rate": 1.7668962157596198e-05, + "loss": 0.3901, + "step": 9543 + }, + { + "epoch": 3.598793363499246, + "grad_norm": 0.12081294078872268, + "learning_rate": 1.766022675216323e-05, + "loss": 0.3899, + "step": 9544 + }, + { + "epoch": 3.5991704374057316, + "grad_norm": 0.10663714581320895, + "learning_rate": 1.7651492894856736e-05, + "loss": 0.3869, + "step": 9545 + }, + { + "epoch": 3.5995475113122173, + "grad_norm": 0.10312611517689561, + "learning_rate": 1.764276058628196e-05, + "loss": 0.394, + "step": 9546 + }, + { + "epoch": 3.599924585218703, + "grad_norm": 0.12695336860999046, + "learning_rate": 1.7634029827044058e-05, + "loss": 0.4117, + "step": 9547 + }, + { + "epoch": 3.6003016591251886, + "grad_norm": 0.10380191540189974, + "learning_rate": 1.7625300617748022e-05, + "loss": 0.3901, + "step": 9548 + }, + { + "epoch": 3.6006787330316743, + "grad_norm": 0.12498486239058447, + "learning_rate": 1.7616572958998834e-05, + "loss": 0.4039, + "step": 9549 + }, + { + "epoch": 3.60105580693816, + "grad_norm": 0.11948874865956761, + "learning_rate": 1.7607846851401273e-05, + "loss": 0.4013, + "step": 9550 + }, + { + "epoch": 3.6014328808446456, + "grad_norm": 0.1293322603082425, + "learning_rate": 1.7599122295560064e-05, + "loss": 0.3987, + "step": 9551 + }, + { + "epoch": 3.6018099547511313, + "grad_norm": 0.12425145726012089, + "learning_rate": 1.759039929207981e-05, + "loss": 0.4041, + "step": 9552 + }, + { + "epoch": 3.602187028657617, + "grad_norm": 0.1166341860269711, + "learning_rate": 1.7581677841565023e-05, + "loss": 0.4036, + "step": 9553 + }, + { + "epoch": 3.6025641025641026, + "grad_norm": 0.1341016508140899, + "learning_rate": 1.7572957944620045e-05, + "loss": 0.3935, + "step": 9554 + }, + { + "epoch": 3.6029411764705883, + "grad_norm": 0.10523115105417578, + "learning_rate": 1.756423960184922e-05, + "loss": 0.3943, + "step": 9555 + }, + { + "epoch": 3.603318250377074, + "grad_norm": 0.12246069053236244, + "learning_rate": 1.7555522813856674e-05, + "loss": 0.3923, + "step": 9556 + }, + { + "epoch": 3.6036953242835597, + "grad_norm": 0.11537582702374102, + "learning_rate": 1.754680758124649e-05, + "loss": 0.4015, + "step": 9557 + }, + { + "epoch": 3.6040723981900453, + "grad_norm": 0.10950253761705923, + "learning_rate": 1.7538093904622627e-05, + "loss": 0.4094, + "step": 9558 + }, + { + "epoch": 3.604449472096531, + "grad_norm": 0.11794002775748612, + "learning_rate": 1.7529381784588943e-05, + "loss": 0.3907, + "step": 9559 + }, + { + "epoch": 3.6048265460030167, + "grad_norm": 0.10836465207911646, + "learning_rate": 1.752067122174914e-05, + "loss": 0.3883, + "step": 9560 + }, + { + "epoch": 3.6052036199095023, + "grad_norm": 0.10548732249512269, + "learning_rate": 1.7511962216706915e-05, + "loss": 0.3967, + "step": 9561 + }, + { + "epoch": 3.605580693815988, + "grad_norm": 0.11734668022512011, + "learning_rate": 1.7503254770065747e-05, + "loss": 0.3962, + "step": 9562 + }, + { + "epoch": 3.6059577677224737, + "grad_norm": 0.10807107792574423, + "learning_rate": 1.7494548882429066e-05, + "loss": 0.3911, + "step": 9563 + }, + { + "epoch": 3.6063348416289593, + "grad_norm": 0.12566236716571935, + "learning_rate": 1.748584455440019e-05, + "loss": 0.4025, + "step": 9564 + }, + { + "epoch": 3.606711915535445, + "grad_norm": 0.10135264480008976, + "learning_rate": 1.747714178658231e-05, + "loss": 0.3884, + "step": 9565 + }, + { + "epoch": 3.6070889894419307, + "grad_norm": 0.1302653771851551, + "learning_rate": 1.7468440579578533e-05, + "loss": 0.4003, + "step": 9566 + }, + { + "epoch": 3.6074660633484164, + "grad_norm": 0.10713268886559074, + "learning_rate": 1.7459740933991848e-05, + "loss": 0.386, + "step": 9567 + }, + { + "epoch": 3.607843137254902, + "grad_norm": 0.11282178843632638, + "learning_rate": 1.7451042850425112e-05, + "loss": 0.3934, + "step": 9568 + }, + { + "epoch": 3.6082202111613877, + "grad_norm": 0.11770541308813116, + "learning_rate": 1.74423463294811e-05, + "loss": 0.3926, + "step": 9569 + }, + { + "epoch": 3.6085972850678734, + "grad_norm": 0.10620249964702332, + "learning_rate": 1.743365137176248e-05, + "loss": 0.3951, + "step": 9570 + }, + { + "epoch": 3.608974358974359, + "grad_norm": 0.11500935780164562, + "learning_rate": 1.7424957977871808e-05, + "loss": 0.3989, + "step": 9571 + }, + { + "epoch": 3.6093514328808447, + "grad_norm": 0.11320210986636002, + "learning_rate": 1.741626614841152e-05, + "loss": 0.3955, + "step": 9572 + }, + { + "epoch": 3.6097285067873304, + "grad_norm": 0.112391225192846, + "learning_rate": 1.7407575883983974e-05, + "loss": 0.3993, + "step": 9573 + }, + { + "epoch": 3.610105580693816, + "grad_norm": 0.109880427957602, + "learning_rate": 1.7398887185191365e-05, + "loss": 0.3911, + "step": 9574 + }, + { + "epoch": 3.6104826546003017, + "grad_norm": 0.10179833496918422, + "learning_rate": 1.7390200052635825e-05, + "loss": 0.3919, + "step": 9575 + }, + { + "epoch": 3.6108597285067874, + "grad_norm": 0.10374773271809314, + "learning_rate": 1.7381514486919368e-05, + "loss": 0.3887, + "step": 9576 + }, + { + "epoch": 3.611236802413273, + "grad_norm": 0.11226640727906283, + "learning_rate": 1.737283048864389e-05, + "loss": 0.3947, + "step": 9577 + }, + { + "epoch": 3.6116138763197587, + "grad_norm": 0.11513311825898188, + "learning_rate": 1.7364148058411195e-05, + "loss": 0.3999, + "step": 9578 + }, + { + "epoch": 3.6119909502262444, + "grad_norm": 0.11406098159711237, + "learning_rate": 1.7355467196822976e-05, + "loss": 0.3944, + "step": 9579 + }, + { + "epoch": 3.61236802413273, + "grad_norm": 0.09648727946798587, + "learning_rate": 1.7346787904480774e-05, + "loss": 0.3987, + "step": 9580 + }, + { + "epoch": 3.6127450980392157, + "grad_norm": 0.1061389921703705, + "learning_rate": 1.7338110181986082e-05, + "loss": 0.3859, + "step": 9581 + }, + { + "epoch": 3.6131221719457014, + "grad_norm": 0.10036690263869985, + "learning_rate": 1.7329434029940257e-05, + "loss": 0.3872, + "step": 9582 + }, + { + "epoch": 3.613499245852187, + "grad_norm": 0.10549519951839666, + "learning_rate": 1.7320759448944535e-05, + "loss": 0.3948, + "step": 9583 + }, + { + "epoch": 3.6138763197586727, + "grad_norm": 0.1132480162832139, + "learning_rate": 1.7312086439600072e-05, + "loss": 0.4073, + "step": 9584 + }, + { + "epoch": 3.6142533936651584, + "grad_norm": 0.1095902224360867, + "learning_rate": 1.7303415002507894e-05, + "loss": 0.395, + "step": 9585 + }, + { + "epoch": 3.614630467571644, + "grad_norm": 0.12441049597732368, + "learning_rate": 1.7294745138268942e-05, + "loss": 0.3973, + "step": 9586 + }, + { + "epoch": 3.6150075414781297, + "grad_norm": 0.11060090664625075, + "learning_rate": 1.7286076847483994e-05, + "loss": 0.3836, + "step": 9587 + }, + { + "epoch": 3.6153846153846154, + "grad_norm": 0.1116275883734547, + "learning_rate": 1.7277410130753775e-05, + "loss": 0.3889, + "step": 9588 + }, + { + "epoch": 3.615761689291101, + "grad_norm": 0.1120866712667075, + "learning_rate": 1.7268744988678875e-05, + "loss": 0.3933, + "step": 9589 + }, + { + "epoch": 3.6161387631975868, + "grad_norm": 0.10896869141975998, + "learning_rate": 1.7260081421859788e-05, + "loss": 0.3933, + "step": 9590 + }, + { + "epoch": 3.6165158371040724, + "grad_norm": 0.11321342911462025, + "learning_rate": 1.725141943089689e-05, + "loss": 0.401, + "step": 9591 + }, + { + "epoch": 3.616892911010558, + "grad_norm": 0.11470799709611844, + "learning_rate": 1.7242759016390462e-05, + "loss": 0.405, + "step": 9592 + }, + { + "epoch": 3.6172699849170438, + "grad_norm": 0.10911427396425226, + "learning_rate": 1.723410017894063e-05, + "loss": 0.3861, + "step": 9593 + }, + { + "epoch": 3.6176470588235294, + "grad_norm": 0.12035100027427041, + "learning_rate": 1.7225442919147467e-05, + "loss": 0.4037, + "step": 9594 + }, + { + "epoch": 3.618024132730015, + "grad_norm": 0.11109480388929949, + "learning_rate": 1.7216787237610904e-05, + "loss": 0.3878, + "step": 9595 + }, + { + "epoch": 3.6184012066365008, + "grad_norm": 0.1355361121440108, + "learning_rate": 1.720813313493078e-05, + "loss": 0.395, + "step": 9596 + }, + { + "epoch": 3.6187782805429864, + "grad_norm": 0.11097569040799941, + "learning_rate": 1.719948061170681e-05, + "loss": 0.4026, + "step": 9597 + }, + { + "epoch": 3.619155354449472, + "grad_norm": 0.1273331810641056, + "learning_rate": 1.7190829668538613e-05, + "loss": 0.3984, + "step": 9598 + }, + { + "epoch": 3.619532428355958, + "grad_norm": 0.11047688805649228, + "learning_rate": 1.7182180306025687e-05, + "loss": 0.391, + "step": 9599 + }, + { + "epoch": 3.6199095022624435, + "grad_norm": 0.11381813099968763, + "learning_rate": 1.7173532524767443e-05, + "loss": 0.3943, + "step": 9600 + }, + { + "epoch": 3.620286576168929, + "grad_norm": 0.1019116872458892, + "learning_rate": 1.7164886325363133e-05, + "loss": 0.3914, + "step": 9601 + }, + { + "epoch": 3.620663650075415, + "grad_norm": 0.1212860194278063, + "learning_rate": 1.7156241708411942e-05, + "loss": 0.3924, + "step": 9602 + }, + { + "epoch": 3.6210407239819005, + "grad_norm": 0.09747535035458943, + "learning_rate": 1.714759867451294e-05, + "loss": 0.3962, + "step": 9603 + }, + { + "epoch": 3.621417797888386, + "grad_norm": 0.12149887982555085, + "learning_rate": 1.713895722426508e-05, + "loss": 0.4059, + "step": 9604 + }, + { + "epoch": 3.621794871794872, + "grad_norm": 0.10679469399749894, + "learning_rate": 1.713031735826721e-05, + "loss": 0.3963, + "step": 9605 + }, + { + "epoch": 3.6221719457013575, + "grad_norm": 0.11971504250372428, + "learning_rate": 1.7121679077118072e-05, + "loss": 0.4, + "step": 9606 + }, + { + "epoch": 3.622549019607843, + "grad_norm": 0.10545888671216126, + "learning_rate": 1.7113042381416268e-05, + "loss": 0.3846, + "step": 9607 + }, + { + "epoch": 3.622926093514329, + "grad_norm": 0.11336566470584372, + "learning_rate": 1.7104407271760327e-05, + "loss": 0.3821, + "step": 9608 + }, + { + "epoch": 3.6233031674208145, + "grad_norm": 0.11406779889029066, + "learning_rate": 1.709577374874865e-05, + "loss": 0.3978, + "step": 9609 + }, + { + "epoch": 3.6236802413273, + "grad_norm": 0.10701212165480731, + "learning_rate": 1.7087141812979543e-05, + "loss": 0.3808, + "step": 9610 + }, + { + "epoch": 3.624057315233786, + "grad_norm": 0.10514976561029764, + "learning_rate": 1.7078511465051177e-05, + "loss": 0.3929, + "step": 9611 + }, + { + "epoch": 3.6244343891402715, + "grad_norm": 0.10354335068327819, + "learning_rate": 1.7069882705561655e-05, + "loss": 0.3942, + "step": 9612 + }, + { + "epoch": 3.624811463046757, + "grad_norm": 0.10193558417048591, + "learning_rate": 1.7061255535108906e-05, + "loss": 0.3922, + "step": 9613 + }, + { + "epoch": 3.625188536953243, + "grad_norm": 0.11189790355934183, + "learning_rate": 1.70526299542908e-05, + "loss": 0.3888, + "step": 9614 + }, + { + "epoch": 3.6255656108597285, + "grad_norm": 0.10789574357474216, + "learning_rate": 1.7044005963705087e-05, + "loss": 0.3922, + "step": 9615 + }, + { + "epoch": 3.625942684766214, + "grad_norm": 0.10246297442325622, + "learning_rate": 1.70353835639494e-05, + "loss": 0.3889, + "step": 9616 + }, + { + "epoch": 3.6263197586727, + "grad_norm": 0.10158784732905404, + "learning_rate": 1.702676275562126e-05, + "loss": 0.4096, + "step": 9617 + }, + { + "epoch": 3.6266968325791855, + "grad_norm": 0.09780938048155206, + "learning_rate": 1.7018143539318095e-05, + "loss": 0.3937, + "step": 9618 + }, + { + "epoch": 3.627073906485671, + "grad_norm": 0.1159724400658872, + "learning_rate": 1.7009525915637173e-05, + "loss": 0.4051, + "step": 9619 + }, + { + "epoch": 3.627450980392157, + "grad_norm": 0.09916876342899804, + "learning_rate": 1.7000909885175742e-05, + "loss": 0.3993, + "step": 9620 + }, + { + "epoch": 3.6278280542986425, + "grad_norm": 0.11601513980977887, + "learning_rate": 1.699229544853084e-05, + "loss": 0.3967, + "step": 9621 + }, + { + "epoch": 3.628205128205128, + "grad_norm": 0.10817787594268534, + "learning_rate": 1.6983682606299457e-05, + "loss": 0.397, + "step": 9622 + }, + { + "epoch": 3.628582202111614, + "grad_norm": 0.15008579081947188, + "learning_rate": 1.697507135907845e-05, + "loss": 0.394, + "step": 9623 + }, + { + "epoch": 3.6289592760180995, + "grad_norm": 0.12052412616071215, + "learning_rate": 1.6966461707464592e-05, + "loss": 0.3848, + "step": 9624 + }, + { + "epoch": 3.629336349924585, + "grad_norm": 0.13156362340878944, + "learning_rate": 1.6957853652054478e-05, + "loss": 0.39, + "step": 9625 + }, + { + "epoch": 3.629713423831071, + "grad_norm": 0.1168114045469867, + "learning_rate": 1.69492471934447e-05, + "loss": 0.3869, + "step": 9626 + }, + { + "epoch": 3.6300904977375565, + "grad_norm": 0.1380952618937795, + "learning_rate": 1.694064233223162e-05, + "loss": 0.3957, + "step": 9627 + }, + { + "epoch": 3.630467571644042, + "grad_norm": 0.09808841595692518, + "learning_rate": 1.6932039069011582e-05, + "loss": 0.3972, + "step": 9628 + }, + { + "epoch": 3.630844645550528, + "grad_norm": 0.11570425437721127, + "learning_rate": 1.6923437404380772e-05, + "loss": 0.4015, + "step": 9629 + }, + { + "epoch": 3.6312217194570136, + "grad_norm": 0.09724777360927153, + "learning_rate": 1.6914837338935298e-05, + "loss": 0.4003, + "step": 9630 + }, + { + "epoch": 3.6315987933634992, + "grad_norm": 0.12703482295397897, + "learning_rate": 1.6906238873271092e-05, + "loss": 0.4083, + "step": 9631 + }, + { + "epoch": 3.631975867269985, + "grad_norm": 0.12341778550792414, + "learning_rate": 1.6897642007984072e-05, + "loss": 0.3883, + "step": 9632 + }, + { + "epoch": 3.6323529411764706, + "grad_norm": 0.10900980958025625, + "learning_rate": 1.6889046743669957e-05, + "loss": 0.3957, + "step": 9633 + }, + { + "epoch": 3.6327300150829562, + "grad_norm": 0.11534795449483101, + "learning_rate": 1.6880453080924403e-05, + "loss": 0.383, + "step": 9634 + }, + { + "epoch": 3.633107088989442, + "grad_norm": 0.11550151967533044, + "learning_rate": 1.6871861020342947e-05, + "loss": 0.3968, + "step": 9635 + }, + { + "epoch": 3.6334841628959276, + "grad_norm": 0.1282886258538828, + "learning_rate": 1.6863270562521017e-05, + "loss": 0.3911, + "step": 9636 + }, + { + "epoch": 3.6338612368024132, + "grad_norm": 0.12796696670365118, + "learning_rate": 1.685468170805389e-05, + "loss": 0.3857, + "step": 9637 + }, + { + "epoch": 3.634238310708899, + "grad_norm": 0.1564975644119005, + "learning_rate": 1.6846094457536818e-05, + "loss": 0.3915, + "step": 9638 + }, + { + "epoch": 3.6346153846153846, + "grad_norm": 0.1026143193139107, + "learning_rate": 1.6837508811564846e-05, + "loss": 0.3923, + "step": 9639 + }, + { + "epoch": 3.6349924585218703, + "grad_norm": 0.15016381333407486, + "learning_rate": 1.682892477073297e-05, + "loss": 0.394, + "step": 9640 + }, + { + "epoch": 3.635369532428356, + "grad_norm": 0.1110238298428765, + "learning_rate": 1.6820342335636057e-05, + "loss": 0.3992, + "step": 9641 + }, + { + "epoch": 3.6357466063348416, + "grad_norm": 0.12751434054021218, + "learning_rate": 1.681176150686887e-05, + "loss": 0.3964, + "step": 9642 + }, + { + "epoch": 3.6361236802413273, + "grad_norm": 0.11659004985172322, + "learning_rate": 1.6803182285026017e-05, + "loss": 0.4048, + "step": 9643 + }, + { + "epoch": 3.636500754147813, + "grad_norm": 0.11796990965698853, + "learning_rate": 1.679460467070208e-05, + "loss": 0.4022, + "step": 9644 + }, + { + "epoch": 3.6368778280542986, + "grad_norm": 0.12792608686425658, + "learning_rate": 1.6786028664491446e-05, + "loss": 0.4001, + "step": 9645 + }, + { + "epoch": 3.6372549019607843, + "grad_norm": 0.1126774728516033, + "learning_rate": 1.6777454266988438e-05, + "loss": 0.3999, + "step": 9646 + }, + { + "epoch": 3.63763197586727, + "grad_norm": 0.11985821855498903, + "learning_rate": 1.6768881478787243e-05, + "loss": 0.391, + "step": 9647 + }, + { + "epoch": 3.6380090497737556, + "grad_norm": 0.12933154626965052, + "learning_rate": 1.6760310300481974e-05, + "loss": 0.3971, + "step": 9648 + }, + { + "epoch": 3.6383861236802413, + "grad_norm": 0.10661590263233395, + "learning_rate": 1.6751740732666562e-05, + "loss": 0.3945, + "step": 9649 + }, + { + "epoch": 3.638763197586727, + "grad_norm": 0.12847916230649628, + "learning_rate": 1.6743172775934922e-05, + "loss": 0.3928, + "step": 9650 + }, + { + "epoch": 3.6391402714932126, + "grad_norm": 0.12135690888078554, + "learning_rate": 1.673460643088076e-05, + "loss": 0.403, + "step": 9651 + }, + { + "epoch": 3.6395173453996983, + "grad_norm": 0.1358012816212816, + "learning_rate": 1.6726041698097746e-05, + "loss": 0.3952, + "step": 9652 + }, + { + "epoch": 3.639894419306184, + "grad_norm": 0.12208852137385619, + "learning_rate": 1.6717478578179394e-05, + "loss": 0.3897, + "step": 9653 + }, + { + "epoch": 3.6402714932126696, + "grad_norm": 0.1303065254735782, + "learning_rate": 1.670891707171914e-05, + "loss": 0.4095, + "step": 9654 + }, + { + "epoch": 3.6406485671191553, + "grad_norm": 0.12479693235564294, + "learning_rate": 1.670035717931024e-05, + "loss": 0.3985, + "step": 9655 + }, + { + "epoch": 3.641025641025641, + "grad_norm": 0.11941649141490418, + "learning_rate": 1.669179890154595e-05, + "loss": 0.3953, + "step": 9656 + }, + { + "epoch": 3.6414027149321266, + "grad_norm": 0.11933683457555495, + "learning_rate": 1.6683242239019316e-05, + "loss": 0.3907, + "step": 9657 + }, + { + "epoch": 3.6417797888386123, + "grad_norm": 0.13038864538931197, + "learning_rate": 1.6674687192323306e-05, + "loss": 0.4005, + "step": 9658 + }, + { + "epoch": 3.642156862745098, + "grad_norm": 0.11776219378278074, + "learning_rate": 1.6666133762050785e-05, + "loss": 0.3949, + "step": 9659 + }, + { + "epoch": 3.6425339366515836, + "grad_norm": 0.1308641098878258, + "learning_rate": 1.6657581948794512e-05, + "loss": 0.3933, + "step": 9660 + }, + { + "epoch": 3.6429110105580693, + "grad_norm": 0.11341786572291421, + "learning_rate": 1.6649031753147074e-05, + "loss": 0.3883, + "step": 9661 + }, + { + "epoch": 3.643288084464555, + "grad_norm": 0.12027280880692492, + "learning_rate": 1.6640483175701055e-05, + "loss": 0.3945, + "step": 9662 + }, + { + "epoch": 3.6436651583710407, + "grad_norm": 0.13121848574139947, + "learning_rate": 1.663193621704882e-05, + "loss": 0.3828, + "step": 9663 + }, + { + "epoch": 3.6440422322775263, + "grad_norm": 0.11832204651449714, + "learning_rate": 1.6623390877782676e-05, + "loss": 0.4088, + "step": 9664 + }, + { + "epoch": 3.644419306184012, + "grad_norm": 0.11814990396818717, + "learning_rate": 1.661484715849481e-05, + "loss": 0.4119, + "step": 9665 + }, + { + "epoch": 3.6447963800904977, + "grad_norm": 0.12361338485173838, + "learning_rate": 1.6606305059777305e-05, + "loss": 0.3961, + "step": 9666 + }, + { + "epoch": 3.6451734539969833, + "grad_norm": 0.12442317308990314, + "learning_rate": 1.659776458222208e-05, + "loss": 0.4107, + "step": 9667 + }, + { + "epoch": 3.645550527903469, + "grad_norm": 0.13231676466441866, + "learning_rate": 1.658922572642104e-05, + "loss": 0.3891, + "step": 9668 + }, + { + "epoch": 3.6459276018099547, + "grad_norm": 0.12353326962465394, + "learning_rate": 1.658068849296588e-05, + "loss": 0.3941, + "step": 9669 + }, + { + "epoch": 3.6463046757164403, + "grad_norm": 0.116997386149964, + "learning_rate": 1.657215288244823e-05, + "loss": 0.3832, + "step": 9670 + }, + { + "epoch": 3.646681749622926, + "grad_norm": 0.11207358611677229, + "learning_rate": 1.656361889545961e-05, + "loss": 0.3984, + "step": 9671 + }, + { + "epoch": 3.6470588235294117, + "grad_norm": 0.10321814300349683, + "learning_rate": 1.6555086532591425e-05, + "loss": 0.4002, + "step": 9672 + }, + { + "epoch": 3.6474358974358974, + "grad_norm": 0.12138169384288482, + "learning_rate": 1.654655579443492e-05, + "loss": 0.3891, + "step": 9673 + }, + { + "epoch": 3.647812971342383, + "grad_norm": 0.10280579433414765, + "learning_rate": 1.653802668158132e-05, + "loss": 0.3846, + "step": 9674 + }, + { + "epoch": 3.6481900452488687, + "grad_norm": 0.11728128941127137, + "learning_rate": 1.652949919462165e-05, + "loss": 0.3974, + "step": 9675 + }, + { + "epoch": 3.6485671191553544, + "grad_norm": 0.11411002628840607, + "learning_rate": 1.652097333414687e-05, + "loss": 0.4065, + "step": 9676 + }, + { + "epoch": 3.64894419306184, + "grad_norm": 0.10614712681853739, + "learning_rate": 1.651244910074781e-05, + "loss": 0.3881, + "step": 9677 + }, + { + "epoch": 3.6493212669683257, + "grad_norm": 0.11207568798251329, + "learning_rate": 1.6503926495015203e-05, + "loss": 0.394, + "step": 9678 + }, + { + "epoch": 3.6496983408748114, + "grad_norm": 0.09616382895971101, + "learning_rate": 1.6495405517539626e-05, + "loss": 0.3956, + "step": 9679 + }, + { + "epoch": 3.650075414781297, + "grad_norm": 0.10183900922190176, + "learning_rate": 1.6486886168911625e-05, + "loss": 0.4022, + "step": 9680 + }, + { + "epoch": 3.6504524886877827, + "grad_norm": 0.10389813734306995, + "learning_rate": 1.6478368449721546e-05, + "loss": 0.3923, + "step": 9681 + }, + { + "epoch": 3.6508295625942684, + "grad_norm": 0.10040751293454514, + "learning_rate": 1.6469852360559664e-05, + "loss": 0.4023, + "step": 9682 + }, + { + "epoch": 3.651206636500754, + "grad_norm": 0.10547956917166225, + "learning_rate": 1.6461337902016142e-05, + "loss": 0.39, + "step": 9683 + }, + { + "epoch": 3.6515837104072397, + "grad_norm": 0.11003624852185583, + "learning_rate": 1.645282507468104e-05, + "loss": 0.397, + "step": 9684 + }, + { + "epoch": 3.6519607843137254, + "grad_norm": 0.11172470469633954, + "learning_rate": 1.6444313879144247e-05, + "loss": 0.4063, + "step": 9685 + }, + { + "epoch": 3.652337858220211, + "grad_norm": 0.10789468817801974, + "learning_rate": 1.643580431599564e-05, + "loss": 0.3911, + "step": 9686 + }, + { + "epoch": 3.6527149321266967, + "grad_norm": 0.12176749770207489, + "learning_rate": 1.642729638582487e-05, + "loss": 0.4007, + "step": 9687 + }, + { + "epoch": 3.6530920060331824, + "grad_norm": 0.10108044105716153, + "learning_rate": 1.6418790089221554e-05, + "loss": 0.401, + "step": 9688 + }, + { + "epoch": 3.653469079939668, + "grad_norm": 0.1179007137872483, + "learning_rate": 1.6410285426775172e-05, + "loss": 0.4082, + "step": 9689 + }, + { + "epoch": 3.6538461538461537, + "grad_norm": 0.09590799421638238, + "learning_rate": 1.6401782399075098e-05, + "loss": 0.4009, + "step": 9690 + }, + { + "epoch": 3.6542232277526394, + "grad_norm": 0.14823512096390856, + "learning_rate": 1.6393281006710546e-05, + "loss": 0.3898, + "step": 9691 + }, + { + "epoch": 3.654600301659125, + "grad_norm": 0.11896576233794412, + "learning_rate": 1.6384781250270708e-05, + "loss": 0.3886, + "step": 9692 + }, + { + "epoch": 3.6549773755656108, + "grad_norm": 0.11188949810580431, + "learning_rate": 1.6376283130344565e-05, + "loss": 0.4034, + "step": 9693 + }, + { + "epoch": 3.6553544494720964, + "grad_norm": 0.10893911280269059, + "learning_rate": 1.636778664752105e-05, + "loss": 0.4017, + "step": 9694 + }, + { + "epoch": 3.655731523378582, + "grad_norm": 0.10301000482187748, + "learning_rate": 1.6359291802388957e-05, + "loss": 0.402, + "step": 9695 + }, + { + "epoch": 3.6561085972850678, + "grad_norm": 0.10151721298646611, + "learning_rate": 1.635079859553699e-05, + "loss": 0.401, + "step": 9696 + }, + { + "epoch": 3.6564856711915534, + "grad_norm": 0.11822032357693404, + "learning_rate": 1.6342307027553674e-05, + "loss": 0.3966, + "step": 9697 + }, + { + "epoch": 3.656862745098039, + "grad_norm": 0.10403891550965047, + "learning_rate": 1.6333817099027522e-05, + "loss": 0.3987, + "step": 9698 + }, + { + "epoch": 3.6572398190045248, + "grad_norm": 0.10934166206654852, + "learning_rate": 1.6325328810546845e-05, + "loss": 0.4039, + "step": 9699 + }, + { + "epoch": 3.6576168929110104, + "grad_norm": 0.10859960635405252, + "learning_rate": 1.631684216269988e-05, + "loss": 0.4029, + "step": 9700 + }, + { + "epoch": 3.657993966817496, + "grad_norm": 0.11210863038735669, + "learning_rate": 1.6308357156074742e-05, + "loss": 0.3963, + "step": 9701 + }, + { + "epoch": 3.658371040723982, + "grad_norm": 0.11007709641109666, + "learning_rate": 1.629987379125946e-05, + "loss": 0.4022, + "step": 9702 + }, + { + "epoch": 3.6587481146304675, + "grad_norm": 0.11263940541818544, + "learning_rate": 1.6291392068841867e-05, + "loss": 0.3894, + "step": 9703 + }, + { + "epoch": 3.659125188536953, + "grad_norm": 0.09459666416601129, + "learning_rate": 1.6282911989409804e-05, + "loss": 0.3992, + "step": 9704 + }, + { + "epoch": 3.659502262443439, + "grad_norm": 0.11417613044482769, + "learning_rate": 1.627443355355089e-05, + "loss": 0.3961, + "step": 9705 + }, + { + "epoch": 3.6598793363499245, + "grad_norm": 0.10116083481661368, + "learning_rate": 1.6265956761852684e-05, + "loss": 0.4002, + "step": 9706 + }, + { + "epoch": 3.66025641025641, + "grad_norm": 0.10670479988877532, + "learning_rate": 1.6257481614902622e-05, + "loss": 0.3896, + "step": 9707 + }, + { + "epoch": 3.660633484162896, + "grad_norm": 0.11728093282679655, + "learning_rate": 1.6249008113288046e-05, + "loss": 0.3924, + "step": 9708 + }, + { + "epoch": 3.6610105580693815, + "grad_norm": 0.10635152197964143, + "learning_rate": 1.6240536257596106e-05, + "loss": 0.4025, + "step": 9709 + }, + { + "epoch": 3.661387631975867, + "grad_norm": 0.11488101946344569, + "learning_rate": 1.623206604841396e-05, + "loss": 0.3921, + "step": 9710 + }, + { + "epoch": 3.661764705882353, + "grad_norm": 0.10766263047590387, + "learning_rate": 1.6223597486328534e-05, + "loss": 0.3938, + "step": 9711 + }, + { + "epoch": 3.6621417797888385, + "grad_norm": 0.11375207121130095, + "learning_rate": 1.6215130571926718e-05, + "loss": 0.3962, + "step": 9712 + }, + { + "epoch": 3.662518853695324, + "grad_norm": 0.1256601379125045, + "learning_rate": 1.6206665305795258e-05, + "loss": 0.3958, + "step": 9713 + }, + { + "epoch": 3.66289592760181, + "grad_norm": 0.11101120942554418, + "learning_rate": 1.6198201688520783e-05, + "loss": 0.402, + "step": 9714 + }, + { + "epoch": 3.6632730015082955, + "grad_norm": 0.1486121069663394, + "learning_rate": 1.6189739720689823e-05, + "loss": 0.3922, + "step": 9715 + }, + { + "epoch": 3.663650075414781, + "grad_norm": 0.1044776344110843, + "learning_rate": 1.618127940288879e-05, + "loss": 0.3889, + "step": 9716 + }, + { + "epoch": 3.664027149321267, + "grad_norm": 0.1305847135600323, + "learning_rate": 1.617282073570396e-05, + "loss": 0.3919, + "step": 9717 + }, + { + "epoch": 3.6644042232277525, + "grad_norm": 0.12098253625114211, + "learning_rate": 1.6164363719721513e-05, + "loss": 0.3989, + "step": 9718 + }, + { + "epoch": 3.664781297134238, + "grad_norm": 0.1061071096070873, + "learning_rate": 1.615590835552752e-05, + "loss": 0.3988, + "step": 9719 + }, + { + "epoch": 3.665158371040724, + "grad_norm": 0.12558171263109902, + "learning_rate": 1.6147454643707928e-05, + "loss": 0.3887, + "step": 9720 + }, + { + "epoch": 3.6655354449472095, + "grad_norm": 0.10351997183506768, + "learning_rate": 1.6139002584848574e-05, + "loss": 0.3947, + "step": 9721 + }, + { + "epoch": 3.665912518853695, + "grad_norm": 0.12819214169589135, + "learning_rate": 1.6130552179535186e-05, + "loss": 0.4007, + "step": 9722 + }, + { + "epoch": 3.666289592760181, + "grad_norm": 0.10224194620180854, + "learning_rate": 1.6122103428353334e-05, + "loss": 0.3947, + "step": 9723 + }, + { + "epoch": 3.6666666666666665, + "grad_norm": 0.1314128188156237, + "learning_rate": 1.6113656331888563e-05, + "loss": 0.3872, + "step": 9724 + }, + { + "epoch": 3.667043740573152, + "grad_norm": 0.11435699121288126, + "learning_rate": 1.6105210890726204e-05, + "loss": 0.381, + "step": 9725 + }, + { + "epoch": 3.667420814479638, + "grad_norm": 0.1252945945190665, + "learning_rate": 1.6096767105451537e-05, + "loss": 0.4032, + "step": 9726 + }, + { + "epoch": 3.6677978883861235, + "grad_norm": 0.124609971006986, + "learning_rate": 1.6088324976649712e-05, + "loss": 0.3943, + "step": 9727 + }, + { + "epoch": 3.668174962292609, + "grad_norm": 0.12199284764987, + "learning_rate": 1.6079884504905767e-05, + "loss": 0.3912, + "step": 9728 + }, + { + "epoch": 3.668552036199095, + "grad_norm": 0.12676121035614682, + "learning_rate": 1.6071445690804577e-05, + "loss": 0.3887, + "step": 9729 + }, + { + "epoch": 3.6689291101055805, + "grad_norm": 0.13410915780167257, + "learning_rate": 1.6063008534931007e-05, + "loss": 0.4111, + "step": 9730 + }, + { + "epoch": 3.669306184012066, + "grad_norm": 0.11337234747252561, + "learning_rate": 1.6054573037869705e-05, + "loss": 0.3997, + "step": 9731 + }, + { + "epoch": 3.669683257918552, + "grad_norm": 0.1509237943452893, + "learning_rate": 1.6046139200205245e-05, + "loss": 0.3984, + "step": 9732 + }, + { + "epoch": 3.6700603318250375, + "grad_norm": 0.11335847922956954, + "learning_rate": 1.6037707022522094e-05, + "loss": 0.3932, + "step": 9733 + }, + { + "epoch": 3.670437405731523, + "grad_norm": 0.11475566339873296, + "learning_rate": 1.602927650540459e-05, + "loss": 0.4006, + "step": 9734 + }, + { + "epoch": 3.670814479638009, + "grad_norm": 0.1167870872537062, + "learning_rate": 1.6020847649436965e-05, + "loss": 0.4112, + "step": 9735 + }, + { + "epoch": 3.6711915535444946, + "grad_norm": 0.12092659987690371, + "learning_rate": 1.601242045520334e-05, + "loss": 0.3958, + "step": 9736 + }, + { + "epoch": 3.6715686274509802, + "grad_norm": 0.1315910505480237, + "learning_rate": 1.600399492328769e-05, + "loss": 0.3991, + "step": 9737 + }, + { + "epoch": 3.6719457013574663, + "grad_norm": 0.12633399013780455, + "learning_rate": 1.5995571054273904e-05, + "loss": 0.4064, + "step": 9738 + }, + { + "epoch": 3.672322775263952, + "grad_norm": 0.1321998678217812, + "learning_rate": 1.5987148848745752e-05, + "loss": 0.3975, + "step": 9739 + }, + { + "epoch": 3.6726998491704377, + "grad_norm": 0.11648067943457253, + "learning_rate": 1.5978728307286884e-05, + "loss": 0.3887, + "step": 9740 + }, + { + "epoch": 3.6730769230769234, + "grad_norm": 0.12866258066301542, + "learning_rate": 1.597030943048084e-05, + "loss": 0.3924, + "step": 9741 + }, + { + "epoch": 3.673453996983409, + "grad_norm": 0.11060155421698037, + "learning_rate": 1.5961892218911056e-05, + "loss": 0.4017, + "step": 9742 + }, + { + "epoch": 3.6738310708898947, + "grad_norm": 0.11763581530379318, + "learning_rate": 1.5953476673160797e-05, + "loss": 0.3999, + "step": 9743 + }, + { + "epoch": 3.6742081447963804, + "grad_norm": 0.098862084781099, + "learning_rate": 1.594506279381328e-05, + "loss": 0.3934, + "step": 9744 + }, + { + "epoch": 3.674585218702866, + "grad_norm": 0.10786288745765048, + "learning_rate": 1.5936650581451573e-05, + "loss": 0.3757, + "step": 9745 + }, + { + "epoch": 3.6749622926093517, + "grad_norm": 0.11209273511293746, + "learning_rate": 1.592824003665864e-05, + "loss": 0.3883, + "step": 9746 + }, + { + "epoch": 3.6753393665158374, + "grad_norm": 0.09959841263609671, + "learning_rate": 1.591983116001732e-05, + "loss": 0.3905, + "step": 9747 + }, + { + "epoch": 3.675716440422323, + "grad_norm": 0.1109874192303009, + "learning_rate": 1.591142395211034e-05, + "loss": 0.3961, + "step": 9748 + }, + { + "epoch": 3.6760935143288087, + "grad_norm": 0.09053141660297796, + "learning_rate": 1.590301841352033e-05, + "loss": 0.3931, + "step": 9749 + }, + { + "epoch": 3.6764705882352944, + "grad_norm": 0.10225964550491651, + "learning_rate": 1.589461454482975e-05, + "loss": 0.3963, + "step": 9750 + }, + { + "epoch": 3.67684766214178, + "grad_norm": 0.09595112553538257, + "learning_rate": 1.5886212346621003e-05, + "loss": 0.3921, + "step": 9751 + }, + { + "epoch": 3.6772247360482657, + "grad_norm": 0.1104152825304261, + "learning_rate": 1.5877811819476356e-05, + "loss": 0.3982, + "step": 9752 + }, + { + "epoch": 3.6776018099547514, + "grad_norm": 0.09440650111465368, + "learning_rate": 1.586941296397795e-05, + "loss": 0.3858, + "step": 9753 + }, + { + "epoch": 3.677978883861237, + "grad_norm": 0.10570339534249601, + "learning_rate": 1.5861015780707822e-05, + "loss": 0.3953, + "step": 9754 + }, + { + "epoch": 3.6783559577677227, + "grad_norm": 0.09466963920982792, + "learning_rate": 1.5852620270247904e-05, + "loss": 0.3961, + "step": 9755 + }, + { + "epoch": 3.6787330316742084, + "grad_norm": 0.11127198514406146, + "learning_rate": 1.584422643317997e-05, + "loss": 0.3984, + "step": 9756 + }, + { + "epoch": 3.679110105580694, + "grad_norm": 0.09275252559862944, + "learning_rate": 1.5835834270085713e-05, + "loss": 0.3995, + "step": 9757 + }, + { + "epoch": 3.6794871794871797, + "grad_norm": 0.11236458870971389, + "learning_rate": 1.5827443781546707e-05, + "loss": 0.4031, + "step": 9758 + }, + { + "epoch": 3.6798642533936654, + "grad_norm": 0.10841008733286686, + "learning_rate": 1.581905496814441e-05, + "loss": 0.4011, + "step": 9759 + }, + { + "epoch": 3.680241327300151, + "grad_norm": 0.12916276195179527, + "learning_rate": 1.5810667830460155e-05, + "loss": 0.407, + "step": 9760 + }, + { + "epoch": 3.6806184012066367, + "grad_norm": 0.11960029471027096, + "learning_rate": 1.5802282369075176e-05, + "loss": 0.4094, + "step": 9761 + }, + { + "epoch": 3.6809954751131224, + "grad_norm": 0.12760512466371368, + "learning_rate": 1.5793898584570552e-05, + "loss": 0.3998, + "step": 9762 + }, + { + "epoch": 3.681372549019608, + "grad_norm": 0.11635827457501972, + "learning_rate": 1.578551647752729e-05, + "loss": 0.3923, + "step": 9763 + }, + { + "epoch": 3.6817496229260938, + "grad_norm": 0.10200306671206735, + "learning_rate": 1.577713604852625e-05, + "loss": 0.3948, + "step": 9764 + }, + { + "epoch": 3.6821266968325794, + "grad_norm": 0.12299243497769173, + "learning_rate": 1.57687572981482e-05, + "loss": 0.4011, + "step": 9765 + }, + { + "epoch": 3.682503770739065, + "grad_norm": 0.0914065161639184, + "learning_rate": 1.5760380226973775e-05, + "loss": 0.396, + "step": 9766 + }, + { + "epoch": 3.6828808446455508, + "grad_norm": 0.11636493410445937, + "learning_rate": 1.5752004835583513e-05, + "loss": 0.4014, + "step": 9767 + }, + { + "epoch": 3.6832579185520364, + "grad_norm": 0.10622129761982639, + "learning_rate": 1.5743631124557785e-05, + "loss": 0.3842, + "step": 9768 + }, + { + "epoch": 3.683634992458522, + "grad_norm": 0.10922358388089956, + "learning_rate": 1.573525909447693e-05, + "loss": 0.4031, + "step": 9769 + }, + { + "epoch": 3.6840120663650078, + "grad_norm": 0.09959422905857515, + "learning_rate": 1.572688874592108e-05, + "loss": 0.3924, + "step": 9770 + }, + { + "epoch": 3.6843891402714934, + "grad_norm": 0.11392617748285822, + "learning_rate": 1.5718520079470315e-05, + "loss": 0.3971, + "step": 9771 + }, + { + "epoch": 3.684766214177979, + "grad_norm": 0.10669993000667755, + "learning_rate": 1.5710153095704564e-05, + "loss": 0.3975, + "step": 9772 + }, + { + "epoch": 3.685143288084465, + "grad_norm": 0.1093290833848072, + "learning_rate": 1.570178779520368e-05, + "loss": 0.3946, + "step": 9773 + }, + { + "epoch": 3.6855203619909505, + "grad_norm": 0.10868392247008392, + "learning_rate": 1.5693424178547315e-05, + "loss": 0.3965, + "step": 9774 + }, + { + "epoch": 3.685897435897436, + "grad_norm": 0.11551593350904371, + "learning_rate": 1.5685062246315124e-05, + "loss": 0.3915, + "step": 9775 + }, + { + "epoch": 3.686274509803922, + "grad_norm": 0.0969452647097194, + "learning_rate": 1.567670199908654e-05, + "loss": 0.386, + "step": 9776 + }, + { + "epoch": 3.6866515837104075, + "grad_norm": 0.1267332610108101, + "learning_rate": 1.5668343437440934e-05, + "loss": 0.3883, + "step": 9777 + }, + { + "epoch": 3.687028657616893, + "grad_norm": 0.10862567314521875, + "learning_rate": 1.5659986561957542e-05, + "loss": 0.3933, + "step": 9778 + }, + { + "epoch": 3.687405731523379, + "grad_norm": 0.1213342594051951, + "learning_rate": 1.565163137321551e-05, + "loss": 0.3916, + "step": 9779 + }, + { + "epoch": 3.6877828054298645, + "grad_norm": 0.12506997707664388, + "learning_rate": 1.564327787179379e-05, + "loss": 0.3903, + "step": 9780 + }, + { + "epoch": 3.68815987933635, + "grad_norm": 0.11363445218521616, + "learning_rate": 1.5634926058271343e-05, + "loss": 0.3853, + "step": 9781 + }, + { + "epoch": 3.688536953242836, + "grad_norm": 0.1243310282800101, + "learning_rate": 1.5626575933226897e-05, + "loss": 0.4069, + "step": 9782 + }, + { + "epoch": 3.6889140271493215, + "grad_norm": 0.10633365809167908, + "learning_rate": 1.5618227497239117e-05, + "loss": 0.3931, + "step": 9783 + }, + { + "epoch": 3.689291101055807, + "grad_norm": 0.11564238872270559, + "learning_rate": 1.5609880750886545e-05, + "loss": 0.377, + "step": 9784 + }, + { + "epoch": 3.689668174962293, + "grad_norm": 0.10018847457949677, + "learning_rate": 1.560153569474762e-05, + "loss": 0.3887, + "step": 9785 + }, + { + "epoch": 3.6900452488687785, + "grad_norm": 0.11840919263171548, + "learning_rate": 1.5593192329400596e-05, + "loss": 0.3872, + "step": 9786 + }, + { + "epoch": 3.690422322775264, + "grad_norm": 0.10773471764405415, + "learning_rate": 1.5584850655423727e-05, + "loss": 0.4034, + "step": 9787 + }, + { + "epoch": 3.69079939668175, + "grad_norm": 0.10708158072643705, + "learning_rate": 1.5576510673395035e-05, + "loss": 0.3984, + "step": 9788 + }, + { + "epoch": 3.6911764705882355, + "grad_norm": 0.10989530439902817, + "learning_rate": 1.556817238389249e-05, + "loss": 0.4094, + "step": 9789 + }, + { + "epoch": 3.691553544494721, + "grad_norm": 0.10411931504119457, + "learning_rate": 1.555983578749393e-05, + "loss": 0.3934, + "step": 9790 + }, + { + "epoch": 3.691930618401207, + "grad_norm": 0.1100526364678541, + "learning_rate": 1.5551500884777083e-05, + "loss": 0.4011, + "step": 9791 + }, + { + "epoch": 3.6923076923076925, + "grad_norm": 0.117024461612295, + "learning_rate": 1.554316767631951e-05, + "loss": 0.3985, + "step": 9792 + }, + { + "epoch": 3.692684766214178, + "grad_norm": 0.11906481593185257, + "learning_rate": 1.5534836162698753e-05, + "loss": 0.4031, + "step": 9793 + }, + { + "epoch": 3.693061840120664, + "grad_norm": 0.1163547721597758, + "learning_rate": 1.5526506344492136e-05, + "loss": 0.3979, + "step": 9794 + }, + { + "epoch": 3.6934389140271495, + "grad_norm": 0.11033454388520318, + "learning_rate": 1.5518178222276922e-05, + "loss": 0.3814, + "step": 9795 + }, + { + "epoch": 3.693815987933635, + "grad_norm": 0.12809815902208543, + "learning_rate": 1.550985179663024e-05, + "loss": 0.3859, + "step": 9796 + }, + { + "epoch": 3.694193061840121, + "grad_norm": 0.11583935367067529, + "learning_rate": 1.550152706812912e-05, + "loss": 0.3886, + "step": 9797 + }, + { + "epoch": 3.6945701357466065, + "grad_norm": 0.12304434619874313, + "learning_rate": 1.5493204037350416e-05, + "loss": 0.3812, + "step": 9798 + }, + { + "epoch": 3.694947209653092, + "grad_norm": 0.12426226546115744, + "learning_rate": 1.5484882704870968e-05, + "loss": 0.3965, + "step": 9799 + }, + { + "epoch": 3.695324283559578, + "grad_norm": 0.12923373626079634, + "learning_rate": 1.5476563071267383e-05, + "loss": 0.3819, + "step": 9800 + }, + { + "epoch": 3.6957013574660635, + "grad_norm": 0.11790474882523774, + "learning_rate": 1.546824513711623e-05, + "loss": 0.3986, + "step": 9801 + }, + { + "epoch": 3.696078431372549, + "grad_norm": 0.12899733383853035, + "learning_rate": 1.5459928902993925e-05, + "loss": 0.403, + "step": 9802 + }, + { + "epoch": 3.696455505279035, + "grad_norm": 0.1135308342160018, + "learning_rate": 1.5451614369476796e-05, + "loss": 0.4021, + "step": 9803 + }, + { + "epoch": 3.6968325791855206, + "grad_norm": 0.12907838687666118, + "learning_rate": 1.5443301537140988e-05, + "loss": 0.3946, + "step": 9804 + }, + { + "epoch": 3.697209653092006, + "grad_norm": 0.12723210647613786, + "learning_rate": 1.543499040656263e-05, + "loss": 0.3848, + "step": 9805 + }, + { + "epoch": 3.697586726998492, + "grad_norm": 0.11921001022998849, + "learning_rate": 1.5426680978317638e-05, + "loss": 0.4136, + "step": 9806 + }, + { + "epoch": 3.6979638009049776, + "grad_norm": 0.12456239709321162, + "learning_rate": 1.541837325298185e-05, + "loss": 0.3861, + "step": 9807 + }, + { + "epoch": 3.6983408748114632, + "grad_norm": 0.12129309131174891, + "learning_rate": 1.5410067231130995e-05, + "loss": 0.3914, + "step": 9808 + }, + { + "epoch": 3.698717948717949, + "grad_norm": 0.11133211500373212, + "learning_rate": 1.5401762913340684e-05, + "loss": 0.3838, + "step": 9809 + }, + { + "epoch": 3.6990950226244346, + "grad_norm": 0.12779126668061322, + "learning_rate": 1.5393460300186357e-05, + "loss": 0.4042, + "step": 9810 + }, + { + "epoch": 3.6994720965309202, + "grad_norm": 0.11835278101311697, + "learning_rate": 1.5385159392243432e-05, + "loss": 0.3931, + "step": 9811 + }, + { + "epoch": 3.699849170437406, + "grad_norm": 0.11844490349528854, + "learning_rate": 1.5376860190087115e-05, + "loss": 0.3968, + "step": 9812 + }, + { + "epoch": 3.7002262443438916, + "grad_norm": 0.12152867839774577, + "learning_rate": 1.5368562694292544e-05, + "loss": 0.3937, + "step": 9813 + }, + { + "epoch": 3.7006033182503772, + "grad_norm": 0.10466131034048966, + "learning_rate": 1.5360266905434734e-05, + "loss": 0.3911, + "step": 9814 + }, + { + "epoch": 3.700980392156863, + "grad_norm": 0.12274871315087676, + "learning_rate": 1.5351972824088585e-05, + "loss": 0.4038, + "step": 9815 + }, + { + "epoch": 3.7013574660633486, + "grad_norm": 0.10112085626944892, + "learning_rate": 1.5343680450828826e-05, + "loss": 0.3984, + "step": 9816 + }, + { + "epoch": 3.7017345399698343, + "grad_norm": 0.10393392462283499, + "learning_rate": 1.533538978623017e-05, + "loss": 0.3803, + "step": 9817 + }, + { + "epoch": 3.70211161387632, + "grad_norm": 0.11681289741527845, + "learning_rate": 1.532710083086711e-05, + "loss": 0.389, + "step": 9818 + }, + { + "epoch": 3.7024886877828056, + "grad_norm": 0.09818031458279541, + "learning_rate": 1.531881358531408e-05, + "loss": 0.3865, + "step": 9819 + }, + { + "epoch": 3.7028657616892913, + "grad_norm": 0.11819084503774041, + "learning_rate": 1.5310528050145376e-05, + "loss": 0.3938, + "step": 9820 + }, + { + "epoch": 3.703242835595777, + "grad_norm": 0.09405920586565653, + "learning_rate": 1.530224422593519e-05, + "loss": 0.386, + "step": 9821 + }, + { + "epoch": 3.7036199095022626, + "grad_norm": 0.11110838888522426, + "learning_rate": 1.5293962113257547e-05, + "loss": 0.4039, + "step": 9822 + }, + { + "epoch": 3.7039969834087483, + "grad_norm": 0.10569126862918694, + "learning_rate": 1.5285681712686444e-05, + "loss": 0.4112, + "step": 9823 + }, + { + "epoch": 3.704374057315234, + "grad_norm": 0.1046971314190711, + "learning_rate": 1.527740302479566e-05, + "loss": 0.404, + "step": 9824 + }, + { + "epoch": 3.7047511312217196, + "grad_norm": 0.11719997233288934, + "learning_rate": 1.526912605015892e-05, + "loss": 0.3789, + "step": 9825 + }, + { + "epoch": 3.7051282051282053, + "grad_norm": 0.09698883673187889, + "learning_rate": 1.5260850789349805e-05, + "loss": 0.3968, + "step": 9826 + }, + { + "epoch": 3.705505279034691, + "grad_norm": 0.11781299596311472, + "learning_rate": 1.5252577242941801e-05, + "loss": 0.3976, + "step": 9827 + }, + { + "epoch": 3.7058823529411766, + "grad_norm": 0.0988437599779828, + "learning_rate": 1.5244305411508217e-05, + "loss": 0.3923, + "step": 9828 + }, + { + "epoch": 3.7062594268476623, + "grad_norm": 0.09524348084975247, + "learning_rate": 1.5236035295622337e-05, + "loss": 0.3964, + "step": 9829 + }, + { + "epoch": 3.706636500754148, + "grad_norm": 0.10278407013689053, + "learning_rate": 1.5227766895857233e-05, + "loss": 0.4065, + "step": 9830 + }, + { + "epoch": 3.7070135746606336, + "grad_norm": 0.10000966075311647, + "learning_rate": 1.5219500212785909e-05, + "loss": 0.3915, + "step": 9831 + }, + { + "epoch": 3.7073906485671193, + "grad_norm": 0.10259931429432755, + "learning_rate": 1.521123524698124e-05, + "loss": 0.4035, + "step": 9832 + }, + { + "epoch": 3.707767722473605, + "grad_norm": 0.10150498181280977, + "learning_rate": 1.5202971999015997e-05, + "loss": 0.3864, + "step": 9833 + }, + { + "epoch": 3.7081447963800906, + "grad_norm": 0.10238023580068821, + "learning_rate": 1.519471046946277e-05, + "loss": 0.4046, + "step": 9834 + }, + { + "epoch": 3.7085218702865763, + "grad_norm": 0.10045336359694126, + "learning_rate": 1.518645065889413e-05, + "loss": 0.3902, + "step": 9835 + }, + { + "epoch": 3.708898944193062, + "grad_norm": 0.11353572493864351, + "learning_rate": 1.517819256788244e-05, + "loss": 0.3938, + "step": 9836 + }, + { + "epoch": 3.7092760180995477, + "grad_norm": 0.09098277885495812, + "learning_rate": 1.5169936196999984e-05, + "loss": 0.3843, + "step": 9837 + }, + { + "epoch": 3.7096530920060333, + "grad_norm": 0.11278538610614768, + "learning_rate": 1.5161681546818932e-05, + "loss": 0.3945, + "step": 9838 + }, + { + "epoch": 3.710030165912519, + "grad_norm": 0.1059063438913782, + "learning_rate": 1.5153428617911324e-05, + "loss": 0.3866, + "step": 9839 + }, + { + "epoch": 3.7104072398190047, + "grad_norm": 0.10891546964792542, + "learning_rate": 1.5145177410849047e-05, + "loss": 0.3924, + "step": 9840 + }, + { + "epoch": 3.7107843137254903, + "grad_norm": 0.11088904477292146, + "learning_rate": 1.5136927926203964e-05, + "loss": 0.4006, + "step": 9841 + }, + { + "epoch": 3.711161387631976, + "grad_norm": 0.12507376961545535, + "learning_rate": 1.5128680164547688e-05, + "loss": 0.3949, + "step": 9842 + }, + { + "epoch": 3.7115384615384617, + "grad_norm": 0.11282235161192387, + "learning_rate": 1.5120434126451846e-05, + "loss": 0.3928, + "step": 9843 + }, + { + "epoch": 3.7119155354449473, + "grad_norm": 0.11359230541934637, + "learning_rate": 1.5112189812487832e-05, + "loss": 0.3893, + "step": 9844 + }, + { + "epoch": 3.712292609351433, + "grad_norm": 0.11865204646192334, + "learning_rate": 1.5103947223226993e-05, + "loss": 0.4047, + "step": 9845 + }, + { + "epoch": 3.7126696832579187, + "grad_norm": 0.13953895743024056, + "learning_rate": 1.5095706359240532e-05, + "loss": 0.3969, + "step": 9846 + }, + { + "epoch": 3.7130467571644044, + "grad_norm": 0.1112525900015829, + "learning_rate": 1.5087467221099537e-05, + "loss": 0.3883, + "step": 9847 + }, + { + "epoch": 3.71342383107089, + "grad_norm": 0.11951074945942787, + "learning_rate": 1.5079229809374942e-05, + "loss": 0.3832, + "step": 9848 + }, + { + "epoch": 3.7138009049773757, + "grad_norm": 0.11305650181019616, + "learning_rate": 1.5070994124637643e-05, + "loss": 0.3814, + "step": 9849 + }, + { + "epoch": 3.7141779788838614, + "grad_norm": 0.10810040142054383, + "learning_rate": 1.506276016745833e-05, + "loss": 0.3864, + "step": 9850 + }, + { + "epoch": 3.714555052790347, + "grad_norm": 0.10573212522972612, + "learning_rate": 1.5054527938407613e-05, + "loss": 0.3923, + "step": 9851 + }, + { + "epoch": 3.7149321266968327, + "grad_norm": 0.12256275124594851, + "learning_rate": 1.5046297438055986e-05, + "loss": 0.3958, + "step": 9852 + }, + { + "epoch": 3.7153092006033184, + "grad_norm": 0.11338723284374648, + "learning_rate": 1.5038068666973824e-05, + "loss": 0.3899, + "step": 9853 + }, + { + "epoch": 3.715686274509804, + "grad_norm": 0.11439958014136709, + "learning_rate": 1.5029841625731334e-05, + "loss": 0.4009, + "step": 9854 + }, + { + "epoch": 3.7160633484162897, + "grad_norm": 0.10966792241356561, + "learning_rate": 1.5021616314898699e-05, + "loss": 0.3837, + "step": 9855 + }, + { + "epoch": 3.7164404223227754, + "grad_norm": 0.11987972904035912, + "learning_rate": 1.5013392735045887e-05, + "loss": 0.3877, + "step": 9856 + }, + { + "epoch": 3.716817496229261, + "grad_norm": 0.11737988324545406, + "learning_rate": 1.5005170886742795e-05, + "loss": 0.3976, + "step": 9857 + }, + { + "epoch": 3.7171945701357467, + "grad_norm": 0.10277960424038182, + "learning_rate": 1.4996950770559186e-05, + "loss": 0.3956, + "step": 9858 + }, + { + "epoch": 3.7175716440422324, + "grad_norm": 0.11792884069239355, + "learning_rate": 1.4988732387064726e-05, + "loss": 0.3953, + "step": 9859 + }, + { + "epoch": 3.717948717948718, + "grad_norm": 0.09593873322057085, + "learning_rate": 1.49805157368289e-05, + "loss": 0.3907, + "step": 9860 + }, + { + "epoch": 3.7183257918552037, + "grad_norm": 0.12522878619928413, + "learning_rate": 1.4972300820421168e-05, + "loss": 0.4014, + "step": 9861 + }, + { + "epoch": 3.7187028657616894, + "grad_norm": 0.10111349891720428, + "learning_rate": 1.4964087638410778e-05, + "loss": 0.3948, + "step": 9862 + }, + { + "epoch": 3.719079939668175, + "grad_norm": 0.11240311245489401, + "learning_rate": 1.495587619136691e-05, + "loss": 0.4012, + "step": 9863 + }, + { + "epoch": 3.7194570135746607, + "grad_norm": 0.12674133788968606, + "learning_rate": 1.4947666479858601e-05, + "loss": 0.3999, + "step": 9864 + }, + { + "epoch": 3.7198340874811464, + "grad_norm": 0.09612656253117108, + "learning_rate": 1.49394585044548e-05, + "loss": 0.3937, + "step": 9865 + }, + { + "epoch": 3.720211161387632, + "grad_norm": 0.12012161114677572, + "learning_rate": 1.4931252265724272e-05, + "loss": 0.3962, + "step": 9866 + }, + { + "epoch": 3.7205882352941178, + "grad_norm": 0.10515466859997015, + "learning_rate": 1.4923047764235752e-05, + "loss": 0.4022, + "step": 9867 + }, + { + "epoch": 3.7209653092006034, + "grad_norm": 0.1015548810212938, + "learning_rate": 1.4914845000557762e-05, + "loss": 0.3925, + "step": 9868 + }, + { + "epoch": 3.721342383107089, + "grad_norm": 0.11055365973871248, + "learning_rate": 1.4906643975258765e-05, + "loss": 0.3947, + "step": 9869 + }, + { + "epoch": 3.7217194570135748, + "grad_norm": 0.09374630930963213, + "learning_rate": 1.4898444688907083e-05, + "loss": 0.4002, + "step": 9870 + }, + { + "epoch": 3.7220965309200604, + "grad_norm": 0.10188319187333347, + "learning_rate": 1.4890247142070927e-05, + "loss": 0.3821, + "step": 9871 + }, + { + "epoch": 3.722473604826546, + "grad_norm": 0.10694519896448344, + "learning_rate": 1.4882051335318353e-05, + "loss": 0.3836, + "step": 9872 + }, + { + "epoch": 3.7228506787330318, + "grad_norm": 0.11377332699695472, + "learning_rate": 1.4873857269217365e-05, + "loss": 0.398, + "step": 9873 + }, + { + "epoch": 3.7232277526395174, + "grad_norm": 0.09826717400880038, + "learning_rate": 1.4865664944335768e-05, + "loss": 0.3947, + "step": 9874 + }, + { + "epoch": 3.723604826546003, + "grad_norm": 0.1051167757490706, + "learning_rate": 1.4857474361241297e-05, + "loss": 0.3881, + "step": 9875 + }, + { + "epoch": 3.723981900452489, + "grad_norm": 0.08657406040506772, + "learning_rate": 1.484928552050155e-05, + "loss": 0.3989, + "step": 9876 + }, + { + "epoch": 3.7243589743589745, + "grad_norm": 0.11542316586819132, + "learning_rate": 1.484109842268401e-05, + "loss": 0.3746, + "step": 9877 + }, + { + "epoch": 3.72473604826546, + "grad_norm": 0.09954647952647494, + "learning_rate": 1.4832913068356032e-05, + "loss": 0.3868, + "step": 9878 + }, + { + "epoch": 3.725113122171946, + "grad_norm": 0.11422915348054001, + "learning_rate": 1.4824729458084867e-05, + "loss": 0.3967, + "step": 9879 + }, + { + "epoch": 3.7254901960784315, + "grad_norm": 0.11655614511812608, + "learning_rate": 1.4816547592437611e-05, + "loss": 0.3874, + "step": 9880 + }, + { + "epoch": 3.725867269984917, + "grad_norm": 0.09720428631677824, + "learning_rate": 1.4808367471981266e-05, + "loss": 0.4024, + "step": 9881 + }, + { + "epoch": 3.726244343891403, + "grad_norm": 0.11539823779439605, + "learning_rate": 1.4800189097282709e-05, + "loss": 0.3925, + "step": 9882 + }, + { + "epoch": 3.7266214177978885, + "grad_norm": 0.11642154966345046, + "learning_rate": 1.47920124689087e-05, + "loss": 0.388, + "step": 9883 + }, + { + "epoch": 3.726998491704374, + "grad_norm": 0.09917328474300606, + "learning_rate": 1.4783837587425862e-05, + "loss": 0.3886, + "step": 9884 + }, + { + "epoch": 3.72737556561086, + "grad_norm": 0.12982705045342804, + "learning_rate": 1.4775664453400729e-05, + "loss": 0.402, + "step": 9885 + }, + { + "epoch": 3.7277526395173455, + "grad_norm": 0.10022795879543854, + "learning_rate": 1.4767493067399663e-05, + "loss": 0.3872, + "step": 9886 + }, + { + "epoch": 3.728129713423831, + "grad_norm": 0.11894662806680355, + "learning_rate": 1.4759323429988946e-05, + "loss": 0.3958, + "step": 9887 + }, + { + "epoch": 3.728506787330317, + "grad_norm": 0.11344109287192468, + "learning_rate": 1.4751155541734723e-05, + "loss": 0.3955, + "step": 9888 + }, + { + "epoch": 3.7288838612368025, + "grad_norm": 0.10008946339846196, + "learning_rate": 1.4742989403203027e-05, + "loss": 0.3948, + "step": 9889 + }, + { + "epoch": 3.729260935143288, + "grad_norm": 0.1288333315964717, + "learning_rate": 1.4734825014959766e-05, + "loss": 0.3959, + "step": 9890 + }, + { + "epoch": 3.729638009049774, + "grad_norm": 0.09381567470147352, + "learning_rate": 1.472666237757073e-05, + "loss": 0.3945, + "step": 9891 + }, + { + "epoch": 3.7300150829562595, + "grad_norm": 0.1212277059930914, + "learning_rate": 1.4718501491601562e-05, + "loss": 0.3854, + "step": 9892 + }, + { + "epoch": 3.730392156862745, + "grad_norm": 0.11047259407828655, + "learning_rate": 1.4710342357617817e-05, + "loss": 0.3961, + "step": 9893 + }, + { + "epoch": 3.730769230769231, + "grad_norm": 0.10469171615609994, + "learning_rate": 1.4702184976184915e-05, + "loss": 0.4034, + "step": 9894 + }, + { + "epoch": 3.7311463046757165, + "grad_norm": 0.10401779866430391, + "learning_rate": 1.4694029347868152e-05, + "loss": 0.3831, + "step": 9895 + }, + { + "epoch": 3.731523378582202, + "grad_norm": 0.10264172903386572, + "learning_rate": 1.4685875473232711e-05, + "loss": 0.4029, + "step": 9896 + }, + { + "epoch": 3.731900452488688, + "grad_norm": 0.10731473912835364, + "learning_rate": 1.4677723352843644e-05, + "loss": 0.4015, + "step": 9897 + }, + { + "epoch": 3.7322775263951735, + "grad_norm": 0.10852053059750738, + "learning_rate": 1.4669572987265905e-05, + "loss": 0.3949, + "step": 9898 + }, + { + "epoch": 3.732654600301659, + "grad_norm": 0.11001366698712839, + "learning_rate": 1.4661424377064278e-05, + "loss": 0.3997, + "step": 9899 + }, + { + "epoch": 3.733031674208145, + "grad_norm": 0.10442855582697363, + "learning_rate": 1.465327752280346e-05, + "loss": 0.3977, + "step": 9900 + }, + { + "epoch": 3.7334087481146305, + "grad_norm": 0.11630418211380864, + "learning_rate": 1.4645132425048032e-05, + "loss": 0.395, + "step": 9901 + }, + { + "epoch": 3.733785822021116, + "grad_norm": 0.1125082589094774, + "learning_rate": 1.4636989084362436e-05, + "loss": 0.3901, + "step": 9902 + }, + { + "epoch": 3.734162895927602, + "grad_norm": 0.11011658190204612, + "learning_rate": 1.4628847501310998e-05, + "loss": 0.3858, + "step": 9903 + }, + { + "epoch": 3.7345399698340875, + "grad_norm": 0.1113138299004518, + "learning_rate": 1.462070767645794e-05, + "loss": 0.3969, + "step": 9904 + }, + { + "epoch": 3.734917043740573, + "grad_norm": 0.10375184067289904, + "learning_rate": 1.4612569610367313e-05, + "loss": 0.3991, + "step": 9905 + }, + { + "epoch": 3.735294117647059, + "grad_norm": 0.11716305866165042, + "learning_rate": 1.4604433303603092e-05, + "loss": 0.4019, + "step": 9906 + }, + { + "epoch": 3.7356711915535445, + "grad_norm": 0.09745678494237343, + "learning_rate": 1.4596298756729122e-05, + "loss": 0.4068, + "step": 9907 + }, + { + "epoch": 3.73604826546003, + "grad_norm": 0.11890815592541805, + "learning_rate": 1.4588165970309112e-05, + "loss": 0.3915, + "step": 9908 + }, + { + "epoch": 3.736425339366516, + "grad_norm": 0.1131851330917234, + "learning_rate": 1.4580034944906656e-05, + "loss": 0.3981, + "step": 9909 + }, + { + "epoch": 3.7368024132730016, + "grad_norm": 0.11641214173904678, + "learning_rate": 1.457190568108525e-05, + "loss": 0.3917, + "step": 9910 + }, + { + "epoch": 3.7371794871794872, + "grad_norm": 0.11913182296682157, + "learning_rate": 1.4563778179408213e-05, + "loss": 0.3947, + "step": 9911 + }, + { + "epoch": 3.737556561085973, + "grad_norm": 0.09593376085196907, + "learning_rate": 1.4555652440438781e-05, + "loss": 0.3999, + "step": 9912 + }, + { + "epoch": 3.7379336349924586, + "grad_norm": 0.11415006092447756, + "learning_rate": 1.454752846474007e-05, + "loss": 0.3863, + "step": 9913 + }, + { + "epoch": 3.7383107088989442, + "grad_norm": 0.0989223243043803, + "learning_rate": 1.4539406252875057e-05, + "loss": 0.4022, + "step": 9914 + }, + { + "epoch": 3.73868778280543, + "grad_norm": 0.11901222716265533, + "learning_rate": 1.4531285805406609e-05, + "loss": 0.3962, + "step": 9915 + }, + { + "epoch": 3.7390648567119156, + "grad_norm": 0.09391256812795615, + "learning_rate": 1.4523167122897479e-05, + "loss": 0.401, + "step": 9916 + }, + { + "epoch": 3.7394419306184012, + "grad_norm": 0.11175232073030288, + "learning_rate": 1.4515050205910241e-05, + "loss": 0.3971, + "step": 9917 + }, + { + "epoch": 3.739819004524887, + "grad_norm": 0.09970098609606305, + "learning_rate": 1.4506935055007443e-05, + "loss": 0.3907, + "step": 9918 + }, + { + "epoch": 3.7401960784313726, + "grad_norm": 0.09485040276420004, + "learning_rate": 1.4498821670751423e-05, + "loss": 0.3902, + "step": 9919 + }, + { + "epoch": 3.7405731523378583, + "grad_norm": 0.09928798288740792, + "learning_rate": 1.4490710053704443e-05, + "loss": 0.3944, + "step": 9920 + }, + { + "epoch": 3.740950226244344, + "grad_norm": 0.09720996867345498, + "learning_rate": 1.4482600204428624e-05, + "loss": 0.3921, + "step": 9921 + }, + { + "epoch": 3.7413273001508296, + "grad_norm": 0.09638999414533457, + "learning_rate": 1.4474492123485994e-05, + "loss": 0.4001, + "step": 9922 + }, + { + "epoch": 3.7417043740573153, + "grad_norm": 0.100922705530285, + "learning_rate": 1.4466385811438395e-05, + "loss": 0.4166, + "step": 9923 + }, + { + "epoch": 3.742081447963801, + "grad_norm": 0.09483711189854237, + "learning_rate": 1.4458281268847634e-05, + "loss": 0.3954, + "step": 9924 + }, + { + "epoch": 3.7424585218702866, + "grad_norm": 0.09844165505196241, + "learning_rate": 1.445017849627531e-05, + "loss": 0.3875, + "step": 9925 + }, + { + "epoch": 3.7428355957767723, + "grad_norm": 0.10230919142678542, + "learning_rate": 1.444207749428296e-05, + "loss": 0.3961, + "step": 9926 + }, + { + "epoch": 3.743212669683258, + "grad_norm": 0.09857863575296248, + "learning_rate": 1.4433978263431967e-05, + "loss": 0.3892, + "step": 9927 + }, + { + "epoch": 3.7435897435897436, + "grad_norm": 0.08991468800616223, + "learning_rate": 1.4425880804283617e-05, + "loss": 0.3883, + "step": 9928 + }, + { + "epoch": 3.7439668174962293, + "grad_norm": 0.09848096044806048, + "learning_rate": 1.441778511739901e-05, + "loss": 0.3897, + "step": 9929 + }, + { + "epoch": 3.744343891402715, + "grad_norm": 0.09977475278614013, + "learning_rate": 1.4409691203339238e-05, + "loss": 0.4013, + "step": 9930 + }, + { + "epoch": 3.7447209653092006, + "grad_norm": 0.10310041193947002, + "learning_rate": 1.4401599062665148e-05, + "loss": 0.3956, + "step": 9931 + }, + { + "epoch": 3.7450980392156863, + "grad_norm": 0.10214118655233989, + "learning_rate": 1.4393508695937542e-05, + "loss": 0.395, + "step": 9932 + }, + { + "epoch": 3.745475113122172, + "grad_norm": 0.10142503084821276, + "learning_rate": 1.4385420103717068e-05, + "loss": 0.3996, + "step": 9933 + }, + { + "epoch": 3.7458521870286576, + "grad_norm": 0.09616950083247439, + "learning_rate": 1.4377333286564272e-05, + "loss": 0.3959, + "step": 9934 + }, + { + "epoch": 3.7462292609351433, + "grad_norm": 0.09770075646080528, + "learning_rate": 1.4369248245039531e-05, + "loss": 0.3815, + "step": 9935 + }, + { + "epoch": 3.746606334841629, + "grad_norm": 0.0992109598949085, + "learning_rate": 1.4361164979703173e-05, + "loss": 0.3948, + "step": 9936 + }, + { + "epoch": 3.7469834087481146, + "grad_norm": 0.11293223387313683, + "learning_rate": 1.4353083491115332e-05, + "loss": 0.3957, + "step": 9937 + }, + { + "epoch": 3.7473604826546003, + "grad_norm": 0.09643134399905566, + "learning_rate": 1.4345003779836054e-05, + "loss": 0.3996, + "step": 9938 + }, + { + "epoch": 3.747737556561086, + "grad_norm": 0.10712299867055176, + "learning_rate": 1.4336925846425254e-05, + "loss": 0.3916, + "step": 9939 + }, + { + "epoch": 3.7481146304675717, + "grad_norm": 0.09309910668586224, + "learning_rate": 1.4328849691442747e-05, + "loss": 0.3823, + "step": 9940 + }, + { + "epoch": 3.7484917043740573, + "grad_norm": 0.10320775389146797, + "learning_rate": 1.4320775315448158e-05, + "loss": 0.4019, + "step": 9941 + }, + { + "epoch": 3.748868778280543, + "grad_norm": 0.09409568358952332, + "learning_rate": 1.4312702719001093e-05, + "loss": 0.3939, + "step": 9942 + }, + { + "epoch": 3.7492458521870287, + "grad_norm": 0.10418660551365147, + "learning_rate": 1.4304631902660928e-05, + "loss": 0.3901, + "step": 9943 + }, + { + "epoch": 3.7496229260935143, + "grad_norm": 0.09962272621409192, + "learning_rate": 1.4296562866986977e-05, + "loss": 0.3958, + "step": 9944 + }, + { + "epoch": 3.75, + "grad_norm": 0.10359635980557623, + "learning_rate": 1.4288495612538427e-05, + "loss": 0.3965, + "step": 9945 + }, + { + "epoch": 3.7503770739064857, + "grad_norm": 0.0988272954517405, + "learning_rate": 1.4280430139874337e-05, + "loss": 0.3874, + "step": 9946 + }, + { + "epoch": 3.7507541478129713, + "grad_norm": 0.10829134140549763, + "learning_rate": 1.4272366449553597e-05, + "loss": 0.4009, + "step": 9947 + }, + { + "epoch": 3.751131221719457, + "grad_norm": 0.09842947190551642, + "learning_rate": 1.4264304542135068e-05, + "loss": 0.3845, + "step": 9948 + }, + { + "epoch": 3.7515082956259427, + "grad_norm": 0.10026833984656602, + "learning_rate": 1.4256244418177394e-05, + "loss": 0.3872, + "step": 9949 + }, + { + "epoch": 3.7518853695324283, + "grad_norm": 0.11342759274701338, + "learning_rate": 1.4248186078239146e-05, + "loss": 0.4001, + "step": 9950 + }, + { + "epoch": 3.752262443438914, + "grad_norm": 0.10491981325669006, + "learning_rate": 1.4240129522878765e-05, + "loss": 0.3852, + "step": 9951 + }, + { + "epoch": 3.7526395173453997, + "grad_norm": 0.1147803932311978, + "learning_rate": 1.4232074752654565e-05, + "loss": 0.3978, + "step": 9952 + }, + { + "epoch": 3.7530165912518854, + "grad_norm": 0.11428912037556596, + "learning_rate": 1.422402176812471e-05, + "loss": 0.3965, + "step": 9953 + }, + { + "epoch": 3.753393665158371, + "grad_norm": 0.10965180847843428, + "learning_rate": 1.4215970569847306e-05, + "loss": 0.3945, + "step": 9954 + }, + { + "epoch": 3.7537707390648567, + "grad_norm": 0.12032509917636348, + "learning_rate": 1.4207921158380256e-05, + "loss": 0.391, + "step": 9955 + }, + { + "epoch": 3.7541478129713424, + "grad_norm": 0.10462121793143922, + "learning_rate": 1.4199873534281397e-05, + "loss": 0.3932, + "step": 9956 + }, + { + "epoch": 3.754524886877828, + "grad_norm": 0.12233543085626064, + "learning_rate": 1.4191827698108411e-05, + "loss": 0.3864, + "step": 9957 + }, + { + "epoch": 3.7549019607843137, + "grad_norm": 0.09663564304217663, + "learning_rate": 1.418378365041889e-05, + "loss": 0.4055, + "step": 9958 + }, + { + "epoch": 3.7552790346907994, + "grad_norm": 0.11447470875421206, + "learning_rate": 1.4175741391770239e-05, + "loss": 0.3945, + "step": 9959 + }, + { + "epoch": 3.755656108597285, + "grad_norm": 0.09880854437745955, + "learning_rate": 1.4167700922719827e-05, + "loss": 0.3919, + "step": 9960 + }, + { + "epoch": 3.7560331825037707, + "grad_norm": 0.10458140231931273, + "learning_rate": 1.4159662243824816e-05, + "loss": 0.3881, + "step": 9961 + }, + { + "epoch": 3.7564102564102564, + "grad_norm": 0.10046314658652307, + "learning_rate": 1.415162535564229e-05, + "loss": 0.399, + "step": 9962 + }, + { + "epoch": 3.756787330316742, + "grad_norm": 0.09650751517557891, + "learning_rate": 1.4143590258729201e-05, + "loss": 0.3855, + "step": 9963 + }, + { + "epoch": 3.7571644042232277, + "grad_norm": 0.10155972511600934, + "learning_rate": 1.4135556953642385e-05, + "loss": 0.3993, + "step": 9964 + }, + { + "epoch": 3.7575414781297134, + "grad_norm": 0.09576344984460433, + "learning_rate": 1.4127525440938503e-05, + "loss": 0.3876, + "step": 9965 + }, + { + "epoch": 3.757918552036199, + "grad_norm": 0.09990601195846875, + "learning_rate": 1.4119495721174184e-05, + "loss": 0.3943, + "step": 9966 + }, + { + "epoch": 3.7582956259426847, + "grad_norm": 0.09043689236523068, + "learning_rate": 1.4111467794905833e-05, + "loss": 0.3879, + "step": 9967 + }, + { + "epoch": 3.7586726998491704, + "grad_norm": 0.0992563349165453, + "learning_rate": 1.4103441662689825e-05, + "loss": 0.3824, + "step": 9968 + }, + { + "epoch": 3.759049773755656, + "grad_norm": 0.11075298444564152, + "learning_rate": 1.4095417325082328e-05, + "loss": 0.3995, + "step": 9969 + }, + { + "epoch": 3.7594268476621417, + "grad_norm": 0.09669936552553751, + "learning_rate": 1.408739478263943e-05, + "loss": 0.3816, + "step": 9970 + }, + { + "epoch": 3.7598039215686274, + "grad_norm": 0.11050938099893887, + "learning_rate": 1.407937403591709e-05, + "loss": 0.3936, + "step": 9971 + }, + { + "epoch": 3.760180995475113, + "grad_norm": 0.109844005989154, + "learning_rate": 1.4071355085471155e-05, + "loss": 0.4034, + "step": 9972 + }, + { + "epoch": 3.7605580693815988, + "grad_norm": 0.11748558263023001, + "learning_rate": 1.4063337931857284e-05, + "loss": 0.4041, + "step": 9973 + }, + { + "epoch": 3.7609351432880844, + "grad_norm": 0.12350680081970493, + "learning_rate": 1.4055322575631113e-05, + "loss": 0.4133, + "step": 9974 + }, + { + "epoch": 3.76131221719457, + "grad_norm": 0.10393611632735722, + "learning_rate": 1.4047309017348063e-05, + "loss": 0.3964, + "step": 9975 + }, + { + "epoch": 3.7616892911010558, + "grad_norm": 0.1361190960684288, + "learning_rate": 1.4039297257563478e-05, + "loss": 0.3933, + "step": 9976 + }, + { + "epoch": 3.7620663650075414, + "grad_norm": 0.09764237505284283, + "learning_rate": 1.4031287296832568e-05, + "loss": 0.3927, + "step": 9977 + }, + { + "epoch": 3.762443438914027, + "grad_norm": 0.1317621719923297, + "learning_rate": 1.4023279135710425e-05, + "loss": 0.393, + "step": 9978 + }, + { + "epoch": 3.7628205128205128, + "grad_norm": 0.08976326255238767, + "learning_rate": 1.4015272774751973e-05, + "loss": 0.3896, + "step": 9979 + }, + { + "epoch": 3.7631975867269984, + "grad_norm": 0.127264476075956, + "learning_rate": 1.4007268214512095e-05, + "loss": 0.3899, + "step": 9980 + }, + { + "epoch": 3.763574660633484, + "grad_norm": 0.09382586349190511, + "learning_rate": 1.3999265455545458e-05, + "loss": 0.3964, + "step": 9981 + }, + { + "epoch": 3.76395173453997, + "grad_norm": 0.11818846112348856, + "learning_rate": 1.3991264498406666e-05, + "loss": 0.3902, + "step": 9982 + }, + { + "epoch": 3.7643288084464555, + "grad_norm": 0.09548072608425928, + "learning_rate": 1.3983265343650172e-05, + "loss": 0.3844, + "step": 9983 + }, + { + "epoch": 3.764705882352941, + "grad_norm": 0.11578130737733247, + "learning_rate": 1.3975267991830327e-05, + "loss": 0.3821, + "step": 9984 + }, + { + "epoch": 3.765082956259427, + "grad_norm": 0.1067665098704012, + "learning_rate": 1.3967272443501299e-05, + "loss": 0.3969, + "step": 9985 + }, + { + "epoch": 3.7654600301659125, + "grad_norm": 0.09975562551905566, + "learning_rate": 1.3959278699217226e-05, + "loss": 0.3945, + "step": 9986 + }, + { + "epoch": 3.765837104072398, + "grad_norm": 0.11518303043441833, + "learning_rate": 1.395128675953203e-05, + "loss": 0.3979, + "step": 9987 + }, + { + "epoch": 3.766214177978884, + "grad_norm": 0.11009160839800597, + "learning_rate": 1.3943296624999554e-05, + "loss": 0.3993, + "step": 9988 + }, + { + "epoch": 3.7665912518853695, + "grad_norm": 0.10688897696281209, + "learning_rate": 1.393530829617351e-05, + "loss": 0.3975, + "step": 9989 + }, + { + "epoch": 3.766968325791855, + "grad_norm": 0.10403386542971431, + "learning_rate": 1.3927321773607493e-05, + "loss": 0.3853, + "step": 9990 + }, + { + "epoch": 3.767345399698341, + "grad_norm": 0.10710234379975185, + "learning_rate": 1.391933705785493e-05, + "loss": 0.4008, + "step": 9991 + }, + { + "epoch": 3.7677224736048265, + "grad_norm": 0.10578666858109335, + "learning_rate": 1.3911354149469199e-05, + "loss": 0.4, + "step": 9992 + }, + { + "epoch": 3.768099547511312, + "grad_norm": 0.12099851259342202, + "learning_rate": 1.390337304900347e-05, + "loss": 0.4038, + "step": 9993 + }, + { + "epoch": 3.768476621417798, + "grad_norm": 0.10065418850301676, + "learning_rate": 1.3895393757010843e-05, + "loss": 0.3926, + "step": 9994 + }, + { + "epoch": 3.7688536953242835, + "grad_norm": 0.11525613039887303, + "learning_rate": 1.3887416274044272e-05, + "loss": 0.3999, + "step": 9995 + }, + { + "epoch": 3.769230769230769, + "grad_norm": 0.0991024028989141, + "learning_rate": 1.3879440600656607e-05, + "loss": 0.3972, + "step": 9996 + }, + { + "epoch": 3.769607843137255, + "grad_norm": 0.09479484030912741, + "learning_rate": 1.387146673740051e-05, + "loss": 0.3903, + "step": 9997 + }, + { + "epoch": 3.7699849170437405, + "grad_norm": 0.10405434335340727, + "learning_rate": 1.386349468482862e-05, + "loss": 0.3903, + "step": 9998 + }, + { + "epoch": 3.770361990950226, + "grad_norm": 0.0914772394110681, + "learning_rate": 1.3855524443493353e-05, + "loss": 0.3769, + "step": 9999 + }, + { + "epoch": 3.770739064856712, + "grad_norm": 0.10613207966232484, + "learning_rate": 1.384755601394705e-05, + "loss": 0.4104, + "step": 10000 + }, + { + "epoch": 3.7711161387631975, + "grad_norm": 0.08918195049758998, + "learning_rate": 1.3839589396741922e-05, + "loss": 0.3836, + "step": 10001 + }, + { + "epoch": 3.771493212669683, + "grad_norm": 0.12054106111865176, + "learning_rate": 1.3831624592430055e-05, + "loss": 0.3915, + "step": 10002 + }, + { + "epoch": 3.771870286576169, + "grad_norm": 0.1017574216493148, + "learning_rate": 1.3823661601563364e-05, + "loss": 0.3963, + "step": 10003 + }, + { + "epoch": 3.7722473604826545, + "grad_norm": 0.12731124746992026, + "learning_rate": 1.3815700424693735e-05, + "loss": 0.3901, + "step": 10004 + }, + { + "epoch": 3.77262443438914, + "grad_norm": 0.11375651190494554, + "learning_rate": 1.3807741062372828e-05, + "loss": 0.4013, + "step": 10005 + }, + { + "epoch": 3.773001508295626, + "grad_norm": 0.10733168132248389, + "learning_rate": 1.3799783515152231e-05, + "loss": 0.3974, + "step": 10006 + }, + { + "epoch": 3.7733785822021115, + "grad_norm": 0.10409875227425182, + "learning_rate": 1.3791827783583398e-05, + "loss": 0.3945, + "step": 10007 + }, + { + "epoch": 3.773755656108597, + "grad_norm": 0.09756630174376639, + "learning_rate": 1.3783873868217663e-05, + "loss": 0.3945, + "step": 10008 + }, + { + "epoch": 3.774132730015083, + "grad_norm": 0.10754004792834725, + "learning_rate": 1.3775921769606195e-05, + "loss": 0.3943, + "step": 10009 + }, + { + "epoch": 3.7745098039215685, + "grad_norm": 0.09450822681658445, + "learning_rate": 1.376797148830011e-05, + "loss": 0.414, + "step": 10010 + }, + { + "epoch": 3.774886877828054, + "grad_norm": 0.10197548525744708, + "learning_rate": 1.376002302485032e-05, + "loss": 0.3877, + "step": 10011 + }, + { + "epoch": 3.77526395173454, + "grad_norm": 0.1100087959422512, + "learning_rate": 1.3752076379807662e-05, + "loss": 0.3977, + "step": 10012 + }, + { + "epoch": 3.7756410256410255, + "grad_norm": 0.1241000533427918, + "learning_rate": 1.3744131553722823e-05, + "loss": 0.4039, + "step": 10013 + }, + { + "epoch": 3.776018099547511, + "grad_norm": 0.11598596581100017, + "learning_rate": 1.3736188547146397e-05, + "loss": 0.3925, + "step": 10014 + }, + { + "epoch": 3.776395173453997, + "grad_norm": 0.11051475428593553, + "learning_rate": 1.3728247360628784e-05, + "loss": 0.3916, + "step": 10015 + }, + { + "epoch": 3.7767722473604826, + "grad_norm": 0.11848547200606245, + "learning_rate": 1.3720307994720345e-05, + "loss": 0.3959, + "step": 10016 + }, + { + "epoch": 3.7771493212669682, + "grad_norm": 0.10229526106813504, + "learning_rate": 1.3712370449971246e-05, + "loss": 0.4001, + "step": 10017 + }, + { + "epoch": 3.777526395173454, + "grad_norm": 0.12464669456905232, + "learning_rate": 1.3704434726931553e-05, + "loss": 0.4009, + "step": 10018 + }, + { + "epoch": 3.7779034690799396, + "grad_norm": 0.09595588131337093, + "learning_rate": 1.3696500826151212e-05, + "loss": 0.387, + "step": 10019 + }, + { + "epoch": 3.7782805429864252, + "grad_norm": 0.12431865563648517, + "learning_rate": 1.3688568748180044e-05, + "loss": 0.3952, + "step": 10020 + }, + { + "epoch": 3.778657616892911, + "grad_norm": 0.10624195245258178, + "learning_rate": 1.3680638493567697e-05, + "loss": 0.396, + "step": 10021 + }, + { + "epoch": 3.7790346907993966, + "grad_norm": 0.10568208961730653, + "learning_rate": 1.3672710062863788e-05, + "loss": 0.3999, + "step": 10022 + }, + { + "epoch": 3.7794117647058822, + "grad_norm": 0.09286925366093421, + "learning_rate": 1.3664783456617703e-05, + "loss": 0.3938, + "step": 10023 + }, + { + "epoch": 3.779788838612368, + "grad_norm": 0.10820332803307796, + "learning_rate": 1.3656858675378764e-05, + "loss": 0.3784, + "step": 10024 + }, + { + "epoch": 3.7801659125188536, + "grad_norm": 0.09879402424452044, + "learning_rate": 1.3648935719696153e-05, + "loss": 0.3846, + "step": 10025 + }, + { + "epoch": 3.7805429864253393, + "grad_norm": 0.0970306076661584, + "learning_rate": 1.3641014590118924e-05, + "loss": 0.3797, + "step": 10026 + }, + { + "epoch": 3.780920060331825, + "grad_norm": 0.10341897538953841, + "learning_rate": 1.3633095287196004e-05, + "loss": 0.3938, + "step": 10027 + }, + { + "epoch": 3.7812971342383106, + "grad_norm": 0.11273161620006263, + "learning_rate": 1.3625177811476205e-05, + "loss": 0.3891, + "step": 10028 + }, + { + "epoch": 3.7816742081447963, + "grad_norm": 0.10818914460638722, + "learning_rate": 1.3617262163508174e-05, + "loss": 0.3955, + "step": 10029 + }, + { + "epoch": 3.782051282051282, + "grad_norm": 0.12412133042517022, + "learning_rate": 1.3609348343840477e-05, + "loss": 0.4029, + "step": 10030 + }, + { + "epoch": 3.7824283559577676, + "grad_norm": 0.09824837533340024, + "learning_rate": 1.3601436353021532e-05, + "loss": 0.4015, + "step": 10031 + }, + { + "epoch": 3.7828054298642533, + "grad_norm": 0.11303639403954155, + "learning_rate": 1.359352619159963e-05, + "loss": 0.393, + "step": 10032 + }, + { + "epoch": 3.783182503770739, + "grad_norm": 0.12510330377582957, + "learning_rate": 1.358561786012294e-05, + "loss": 0.3958, + "step": 10033 + }, + { + "epoch": 3.7835595776772246, + "grad_norm": 0.10156490317370012, + "learning_rate": 1.3577711359139519e-05, + "loss": 0.3923, + "step": 10034 + }, + { + "epoch": 3.7839366515837103, + "grad_norm": 0.10286982478466397, + "learning_rate": 1.3569806689197246e-05, + "loss": 0.3953, + "step": 10035 + }, + { + "epoch": 3.784313725490196, + "grad_norm": 0.10468546534796874, + "learning_rate": 1.3561903850843931e-05, + "loss": 0.39, + "step": 10036 + }, + { + "epoch": 3.7846907993966816, + "grad_norm": 0.12110705249466013, + "learning_rate": 1.3554002844627228e-05, + "loss": 0.3962, + "step": 10037 + }, + { + "epoch": 3.7850678733031673, + "grad_norm": 0.09976817933983707, + "learning_rate": 1.3546103671094665e-05, + "loss": 0.41, + "step": 10038 + }, + { + "epoch": 3.785444947209653, + "grad_norm": 0.1281749344052191, + "learning_rate": 1.3538206330793657e-05, + "loss": 0.3981, + "step": 10039 + }, + { + "epoch": 3.7858220211161386, + "grad_norm": 0.10107842378900186, + "learning_rate": 1.3530310824271494e-05, + "loss": 0.3986, + "step": 10040 + }, + { + "epoch": 3.7861990950226243, + "grad_norm": 0.1304945954165782, + "learning_rate": 1.3522417152075295e-05, + "loss": 0.3831, + "step": 10041 + }, + { + "epoch": 3.78657616892911, + "grad_norm": 0.10558732165785191, + "learning_rate": 1.3514525314752103e-05, + "loss": 0.3905, + "step": 10042 + }, + { + "epoch": 3.7869532428355956, + "grad_norm": 0.11984998398615683, + "learning_rate": 1.3506635312848815e-05, + "loss": 0.3863, + "step": 10043 + }, + { + "epoch": 3.7873303167420813, + "grad_norm": 0.11145653329730114, + "learning_rate": 1.3498747146912198e-05, + "loss": 0.4089, + "step": 10044 + }, + { + "epoch": 3.787707390648567, + "grad_norm": 0.12318403139238517, + "learning_rate": 1.3490860817488903e-05, + "loss": 0.4024, + "step": 10045 + }, + { + "epoch": 3.7880844645550527, + "grad_norm": 0.10758837126924571, + "learning_rate": 1.3482976325125434e-05, + "loss": 0.394, + "step": 10046 + }, + { + "epoch": 3.7884615384615383, + "grad_norm": 0.10887331336033236, + "learning_rate": 1.3475093670368202e-05, + "loss": 0.3985, + "step": 10047 + }, + { + "epoch": 3.788838612368024, + "grad_norm": 0.09874651948321682, + "learning_rate": 1.3467212853763435e-05, + "loss": 0.3987, + "step": 10048 + }, + { + "epoch": 3.7892156862745097, + "grad_norm": 0.11332022582707799, + "learning_rate": 1.3459333875857286e-05, + "loss": 0.3942, + "step": 10049 + }, + { + "epoch": 3.7895927601809953, + "grad_norm": 0.1133530374079175, + "learning_rate": 1.3451456737195754e-05, + "loss": 0.413, + "step": 10050 + }, + { + "epoch": 3.789969834087481, + "grad_norm": 0.10147909791759605, + "learning_rate": 1.344358143832472e-05, + "loss": 0.3982, + "step": 10051 + }, + { + "epoch": 3.7903469079939667, + "grad_norm": 0.11644737702723448, + "learning_rate": 1.3435707979789942e-05, + "loss": 0.3944, + "step": 10052 + }, + { + "epoch": 3.7907239819004523, + "grad_norm": 0.10757536325413694, + "learning_rate": 1.3427836362137048e-05, + "loss": 0.3902, + "step": 10053 + }, + { + "epoch": 3.791101055806938, + "grad_norm": 0.12310025400023866, + "learning_rate": 1.3419966585911515e-05, + "loss": 0.3968, + "step": 10054 + }, + { + "epoch": 3.7914781297134237, + "grad_norm": 0.11992385792091736, + "learning_rate": 1.341209865165872e-05, + "loss": 0.3976, + "step": 10055 + }, + { + "epoch": 3.7918552036199094, + "grad_norm": 0.11203941461512719, + "learning_rate": 1.3404232559923905e-05, + "loss": 0.396, + "step": 10056 + }, + { + "epoch": 3.792232277526395, + "grad_norm": 0.12917355520849672, + "learning_rate": 1.3396368311252182e-05, + "loss": 0.4036, + "step": 10057 + }, + { + "epoch": 3.7926093514328807, + "grad_norm": 0.09544799154822291, + "learning_rate": 1.3388505906188538e-05, + "loss": 0.4003, + "step": 10058 + }, + { + "epoch": 3.7929864253393664, + "grad_norm": 0.14380828298633586, + "learning_rate": 1.3380645345277841e-05, + "loss": 0.3954, + "step": 10059 + }, + { + "epoch": 3.793363499245852, + "grad_norm": 0.0994366590693513, + "learning_rate": 1.33727866290648e-05, + "loss": 0.3931, + "step": 10060 + }, + { + "epoch": 3.7937405731523377, + "grad_norm": 0.13055120715500024, + "learning_rate": 1.3364929758094024e-05, + "loss": 0.3964, + "step": 10061 + }, + { + "epoch": 3.7941176470588234, + "grad_norm": 0.11115446551250968, + "learning_rate": 1.3357074732909996e-05, + "loss": 0.3855, + "step": 10062 + }, + { + "epoch": 3.794494720965309, + "grad_norm": 0.11795811122649465, + "learning_rate": 1.3349221554057055e-05, + "loss": 0.3957, + "step": 10063 + }, + { + "epoch": 3.7948717948717947, + "grad_norm": 0.1103740870201617, + "learning_rate": 1.334137022207942e-05, + "loss": 0.3973, + "step": 10064 + }, + { + "epoch": 3.7952488687782804, + "grad_norm": 0.09213240210530305, + "learning_rate": 1.3333520737521198e-05, + "loss": 0.3897, + "step": 10065 + }, + { + "epoch": 3.795625942684766, + "grad_norm": 0.11748031075087534, + "learning_rate": 1.3325673100926305e-05, + "loss": 0.3963, + "step": 10066 + }, + { + "epoch": 3.7960030165912517, + "grad_norm": 0.09720827772177557, + "learning_rate": 1.3317827312838634e-05, + "loss": 0.4021, + "step": 10067 + }, + { + "epoch": 3.7963800904977374, + "grad_norm": 0.11167697152055171, + "learning_rate": 1.330998337380185e-05, + "loss": 0.3964, + "step": 10068 + }, + { + "epoch": 3.796757164404223, + "grad_norm": 0.09433003785098473, + "learning_rate": 1.3302141284359546e-05, + "loss": 0.3859, + "step": 10069 + }, + { + "epoch": 3.7971342383107087, + "grad_norm": 0.09817938808293379, + "learning_rate": 1.3294301045055167e-05, + "loss": 0.4009, + "step": 10070 + }, + { + "epoch": 3.7975113122171944, + "grad_norm": 0.09499941734947956, + "learning_rate": 1.3286462656432054e-05, + "loss": 0.3956, + "step": 10071 + }, + { + "epoch": 3.79788838612368, + "grad_norm": 0.0952147753170246, + "learning_rate": 1.3278626119033357e-05, + "loss": 0.3974, + "step": 10072 + }, + { + "epoch": 3.7982654600301657, + "grad_norm": 0.10449082822828942, + "learning_rate": 1.3270791433402192e-05, + "loss": 0.392, + "step": 10073 + }, + { + "epoch": 3.7986425339366514, + "grad_norm": 0.10004452917598818, + "learning_rate": 1.3262958600081462e-05, + "loss": 0.3997, + "step": 10074 + }, + { + "epoch": 3.799019607843137, + "grad_norm": 0.11097023269857233, + "learning_rate": 1.3255127619613983e-05, + "loss": 0.4004, + "step": 10075 + }, + { + "epoch": 3.7993966817496228, + "grad_norm": 0.1082202596993, + "learning_rate": 1.324729849254244e-05, + "loss": 0.3964, + "step": 10076 + }, + { + "epoch": 3.7997737556561084, + "grad_norm": 0.11159599359281007, + "learning_rate": 1.3239471219409392e-05, + "loss": 0.4059, + "step": 10077 + }, + { + "epoch": 3.800150829562594, + "grad_norm": 0.1035435639102926, + "learning_rate": 1.3231645800757225e-05, + "loss": 0.4107, + "step": 10078 + }, + { + "epoch": 3.8005279034690798, + "grad_norm": 0.09950256566634876, + "learning_rate": 1.322382223712829e-05, + "loss": 0.3833, + "step": 10079 + }, + { + "epoch": 3.8009049773755654, + "grad_norm": 0.12453921296197543, + "learning_rate": 1.3216000529064709e-05, + "loss": 0.3936, + "step": 10080 + }, + { + "epoch": 3.801282051282051, + "grad_norm": 0.12694759173587966, + "learning_rate": 1.3208180677108535e-05, + "loss": 0.3858, + "step": 10081 + }, + { + "epoch": 3.8016591251885368, + "grad_norm": 0.10812730033078041, + "learning_rate": 1.320036268180167e-05, + "loss": 0.3996, + "step": 10082 + }, + { + "epoch": 3.8020361990950224, + "grad_norm": 0.1142548360268956, + "learning_rate": 1.3192546543685922e-05, + "loss": 0.3918, + "step": 10083 + }, + { + "epoch": 3.802413273001508, + "grad_norm": 0.11305892347700805, + "learning_rate": 1.3184732263302889e-05, + "loss": 0.3952, + "step": 10084 + }, + { + "epoch": 3.802790346907994, + "grad_norm": 0.0967376101285273, + "learning_rate": 1.3176919841194149e-05, + "loss": 0.3795, + "step": 10085 + }, + { + "epoch": 3.8031674208144794, + "grad_norm": 0.142265571179302, + "learning_rate": 1.3169109277901061e-05, + "loss": 0.3941, + "step": 10086 + }, + { + "epoch": 3.803544494720965, + "grad_norm": 0.10429939401909634, + "learning_rate": 1.3161300573964906e-05, + "loss": 0.3993, + "step": 10087 + }, + { + "epoch": 3.803921568627451, + "grad_norm": 0.12961198202541135, + "learning_rate": 1.3153493729926812e-05, + "loss": 0.3963, + "step": 10088 + }, + { + "epoch": 3.8042986425339365, + "grad_norm": 0.09850284745602987, + "learning_rate": 1.31456887463278e-05, + "loss": 0.3909, + "step": 10089 + }, + { + "epoch": 3.804675716440422, + "grad_norm": 0.1237413757193921, + "learning_rate": 1.3137885623708715e-05, + "loss": 0.4042, + "step": 10090 + }, + { + "epoch": 3.805052790346908, + "grad_norm": 0.10696817813779544, + "learning_rate": 1.3130084362610354e-05, + "loss": 0.4036, + "step": 10091 + }, + { + "epoch": 3.8054298642533935, + "grad_norm": 0.13740968029633052, + "learning_rate": 1.312228496357329e-05, + "loss": 0.3965, + "step": 10092 + }, + { + "epoch": 3.805806938159879, + "grad_norm": 0.10028382163937165, + "learning_rate": 1.3114487427138057e-05, + "loss": 0.4001, + "step": 10093 + }, + { + "epoch": 3.806184012066365, + "grad_norm": 0.12922683912761188, + "learning_rate": 1.310669175384499e-05, + "loss": 0.3914, + "step": 10094 + }, + { + "epoch": 3.8065610859728505, + "grad_norm": 0.11008855774077425, + "learning_rate": 1.309889794423433e-05, + "loss": 0.3959, + "step": 10095 + }, + { + "epoch": 3.806938159879336, + "grad_norm": 0.11487718366930193, + "learning_rate": 1.309110599884618e-05, + "loss": 0.4011, + "step": 10096 + }, + { + "epoch": 3.807315233785822, + "grad_norm": 0.10590952639566856, + "learning_rate": 1.3083315918220532e-05, + "loss": 0.3957, + "step": 10097 + }, + { + "epoch": 3.8076923076923075, + "grad_norm": 0.11008351289499808, + "learning_rate": 1.3075527702897185e-05, + "loss": 0.4005, + "step": 10098 + }, + { + "epoch": 3.808069381598793, + "grad_norm": 0.09697323766235849, + "learning_rate": 1.3067741353415912e-05, + "loss": 0.4034, + "step": 10099 + }, + { + "epoch": 3.808446455505279, + "grad_norm": 0.11317737307709297, + "learning_rate": 1.3059956870316267e-05, + "loss": 0.3899, + "step": 10100 + }, + { + "epoch": 3.8088235294117645, + "grad_norm": 0.10068274668755725, + "learning_rate": 1.3052174254137713e-05, + "loss": 0.3874, + "step": 10101 + }, + { + "epoch": 3.80920060331825, + "grad_norm": 0.10480108815976055, + "learning_rate": 1.3044393505419577e-05, + "loss": 0.4038, + "step": 10102 + }, + { + "epoch": 3.809577677224736, + "grad_norm": 0.09543455638049343, + "learning_rate": 1.3036614624701076e-05, + "loss": 0.399, + "step": 10103 + }, + { + "epoch": 3.8099547511312215, + "grad_norm": 0.10853987783112821, + "learning_rate": 1.3028837612521241e-05, + "loss": 0.3904, + "step": 10104 + }, + { + "epoch": 3.810331825037707, + "grad_norm": 0.09666394940897287, + "learning_rate": 1.3021062469419063e-05, + "loss": 0.3959, + "step": 10105 + }, + { + "epoch": 3.810708898944193, + "grad_norm": 0.10934042862302634, + "learning_rate": 1.3013289195933308e-05, + "loss": 0.3877, + "step": 10106 + }, + { + "epoch": 3.8110859728506785, + "grad_norm": 0.10348191146023106, + "learning_rate": 1.300551779260268e-05, + "loss": 0.4002, + "step": 10107 + }, + { + "epoch": 3.811463046757164, + "grad_norm": 0.1073828094880813, + "learning_rate": 1.2997748259965718e-05, + "loss": 0.4005, + "step": 10108 + }, + { + "epoch": 3.81184012066365, + "grad_norm": 0.11146140103447438, + "learning_rate": 1.298998059856087e-05, + "loss": 0.3967, + "step": 10109 + }, + { + "epoch": 3.8122171945701355, + "grad_norm": 0.09782246727212539, + "learning_rate": 1.298221480892638e-05, + "loss": 0.3907, + "step": 10110 + }, + { + "epoch": 3.812594268476621, + "grad_norm": 0.12340569770001732, + "learning_rate": 1.2974450891600467e-05, + "loss": 0.3945, + "step": 10111 + }, + { + "epoch": 3.812971342383107, + "grad_norm": 0.11726809657946932, + "learning_rate": 1.2966688847121121e-05, + "loss": 0.3969, + "step": 10112 + }, + { + "epoch": 3.8133484162895925, + "grad_norm": 0.09632218939320268, + "learning_rate": 1.2958928676026257e-05, + "loss": 0.3898, + "step": 10113 + }, + { + "epoch": 3.813725490196078, + "grad_norm": 0.09329941202776253, + "learning_rate": 1.2951170378853655e-05, + "loss": 0.3863, + "step": 10114 + }, + { + "epoch": 3.814102564102564, + "grad_norm": 0.09886810969763017, + "learning_rate": 1.2943413956140965e-05, + "loss": 0.385, + "step": 10115 + }, + { + "epoch": 3.8144796380090495, + "grad_norm": 0.09950828136055168, + "learning_rate": 1.2935659408425658e-05, + "loss": 0.395, + "step": 10116 + }, + { + "epoch": 3.814856711915535, + "grad_norm": 0.09963250862569878, + "learning_rate": 1.2927906736245182e-05, + "loss": 0.3904, + "step": 10117 + }, + { + "epoch": 3.815233785822021, + "grad_norm": 0.1117306415701499, + "learning_rate": 1.2920155940136736e-05, + "loss": 0.3892, + "step": 10118 + }, + { + "epoch": 3.8156108597285066, + "grad_norm": 0.09223083513631154, + "learning_rate": 1.2912407020637466e-05, + "loss": 0.3827, + "step": 10119 + }, + { + "epoch": 3.8159879336349922, + "grad_norm": 0.1068858270035084, + "learning_rate": 1.2904659978284366e-05, + "loss": 0.3988, + "step": 10120 + }, + { + "epoch": 3.816365007541478, + "grad_norm": 0.09240362259881128, + "learning_rate": 1.2896914813614303e-05, + "loss": 0.3881, + "step": 10121 + }, + { + "epoch": 3.8167420814479636, + "grad_norm": 0.12870247597253967, + "learning_rate": 1.2889171527163985e-05, + "loss": 0.3968, + "step": 10122 + }, + { + "epoch": 3.8171191553544492, + "grad_norm": 0.10757660272427494, + "learning_rate": 1.288143011947005e-05, + "loss": 0.3887, + "step": 10123 + }, + { + "epoch": 3.817496229260935, + "grad_norm": 0.0948108602804053, + "learning_rate": 1.2873690591068946e-05, + "loss": 0.3825, + "step": 10124 + }, + { + "epoch": 3.8178733031674206, + "grad_norm": 0.11018103462593556, + "learning_rate": 1.2865952942497022e-05, + "loss": 0.4028, + "step": 10125 + }, + { + "epoch": 3.8182503770739062, + "grad_norm": 0.1053034449579393, + "learning_rate": 1.285821717429049e-05, + "loss": 0.3963, + "step": 10126 + }, + { + "epoch": 3.818627450980392, + "grad_norm": 0.13028608536003217, + "learning_rate": 1.2850483286985447e-05, + "loss": 0.3994, + "step": 10127 + }, + { + "epoch": 3.8190045248868776, + "grad_norm": 0.10316110460004459, + "learning_rate": 1.2842751281117805e-05, + "loss": 0.3951, + "step": 10128 + }, + { + "epoch": 3.8193815987933633, + "grad_norm": 0.14123730117090694, + "learning_rate": 1.2835021157223437e-05, + "loss": 0.3884, + "step": 10129 + }, + { + "epoch": 3.819758672699849, + "grad_norm": 0.09824767445899672, + "learning_rate": 1.2827292915837992e-05, + "loss": 0.3929, + "step": 10130 + }, + { + "epoch": 3.8201357466063346, + "grad_norm": 0.11337863422030649, + "learning_rate": 1.2819566557497049e-05, + "loss": 0.3942, + "step": 10131 + }, + { + "epoch": 3.8205128205128203, + "grad_norm": 0.11009631965045323, + "learning_rate": 1.2811842082736035e-05, + "loss": 0.3863, + "step": 10132 + }, + { + "epoch": 3.820889894419306, + "grad_norm": 0.12082822457963094, + "learning_rate": 1.280411949209026e-05, + "loss": 0.3989, + "step": 10133 + }, + { + "epoch": 3.8212669683257916, + "grad_norm": 0.10225552633393326, + "learning_rate": 1.2796398786094857e-05, + "loss": 0.3936, + "step": 10134 + }, + { + "epoch": 3.8216440422322773, + "grad_norm": 0.11906522622327143, + "learning_rate": 1.2788679965284909e-05, + "loss": 0.4059, + "step": 10135 + }, + { + "epoch": 3.822021116138763, + "grad_norm": 0.11287534004811911, + "learning_rate": 1.2780963030195297e-05, + "loss": 0.3934, + "step": 10136 + }, + { + "epoch": 3.8223981900452486, + "grad_norm": 0.11711958206029052, + "learning_rate": 1.27732479813608e-05, + "loss": 0.3936, + "step": 10137 + }, + { + "epoch": 3.8227752639517343, + "grad_norm": 0.10033230346098068, + "learning_rate": 1.276553481931607e-05, + "loss": 0.3911, + "step": 10138 + }, + { + "epoch": 3.82315233785822, + "grad_norm": 0.09862450140929509, + "learning_rate": 1.2757823544595632e-05, + "loss": 0.3995, + "step": 10139 + }, + { + "epoch": 3.8235294117647056, + "grad_norm": 0.10479320678195987, + "learning_rate": 1.275011415773383e-05, + "loss": 0.4008, + "step": 10140 + }, + { + "epoch": 3.8239064856711913, + "grad_norm": 0.10062674357600689, + "learning_rate": 1.2742406659264974e-05, + "loss": 0.3951, + "step": 10141 + }, + { + "epoch": 3.824283559577677, + "grad_norm": 0.10908136465020467, + "learning_rate": 1.2734701049723151e-05, + "loss": 0.3938, + "step": 10142 + }, + { + "epoch": 3.8246606334841626, + "grad_norm": 0.10651834439764908, + "learning_rate": 1.2726997329642354e-05, + "loss": 0.3794, + "step": 10143 + }, + { + "epoch": 3.8250377073906483, + "grad_norm": 0.10517827570772965, + "learning_rate": 1.2719295499556453e-05, + "loss": 0.4075, + "step": 10144 + }, + { + "epoch": 3.825414781297134, + "grad_norm": 0.11480490558862884, + "learning_rate": 1.271159555999919e-05, + "loss": 0.383, + "step": 10145 + }, + { + "epoch": 3.8257918552036196, + "grad_norm": 0.1054385239065863, + "learning_rate": 1.2703897511504124e-05, + "loss": 0.3905, + "step": 10146 + }, + { + "epoch": 3.8261689291101053, + "grad_norm": 0.1180452407027609, + "learning_rate": 1.2696201354604773e-05, + "loss": 0.4011, + "step": 10147 + }, + { + "epoch": 3.826546003016591, + "grad_norm": 0.10636045338973361, + "learning_rate": 1.268850708983444e-05, + "loss": 0.3928, + "step": 10148 + }, + { + "epoch": 3.8269230769230766, + "grad_norm": 0.09841337022546837, + "learning_rate": 1.2680814717726335e-05, + "loss": 0.3866, + "step": 10149 + }, + { + "epoch": 3.8273001508295623, + "grad_norm": 0.10756690241892534, + "learning_rate": 1.2673124238813541e-05, + "loss": 0.3995, + "step": 10150 + }, + { + "epoch": 3.827677224736048, + "grad_norm": 0.10061134494805583, + "learning_rate": 1.2665435653629015e-05, + "loss": 0.4037, + "step": 10151 + }, + { + "epoch": 3.8280542986425337, + "grad_norm": 0.0959627156329417, + "learning_rate": 1.265774896270552e-05, + "loss": 0.383, + "step": 10152 + }, + { + "epoch": 3.8284313725490198, + "grad_norm": 0.10346382490983398, + "learning_rate": 1.2650064166575801e-05, + "loss": 0.3869, + "step": 10153 + }, + { + "epoch": 3.8288084464555054, + "grad_norm": 0.08997337247510215, + "learning_rate": 1.2642381265772361e-05, + "loss": 0.3981, + "step": 10154 + }, + { + "epoch": 3.829185520361991, + "grad_norm": 0.09599412640142554, + "learning_rate": 1.263470026082763e-05, + "loss": 0.3912, + "step": 10155 + }, + { + "epoch": 3.829562594268477, + "grad_norm": 0.10191492855382221, + "learning_rate": 1.26270211522739e-05, + "loss": 0.3982, + "step": 10156 + }, + { + "epoch": 3.8299396681749625, + "grad_norm": 0.11248140454788139, + "learning_rate": 1.2619343940643338e-05, + "loss": 0.387, + "step": 10157 + }, + { + "epoch": 3.830316742081448, + "grad_norm": 0.10695092584693716, + "learning_rate": 1.261166862646793e-05, + "loss": 0.3983, + "step": 10158 + }, + { + "epoch": 3.830693815987934, + "grad_norm": 0.09833216774178449, + "learning_rate": 1.2603995210279618e-05, + "loss": 0.3997, + "step": 10159 + }, + { + "epoch": 3.8310708898944195, + "grad_norm": 0.10012849890690544, + "learning_rate": 1.259632369261012e-05, + "loss": 0.3983, + "step": 10160 + }, + { + "epoch": 3.831447963800905, + "grad_norm": 0.11676056182755656, + "learning_rate": 1.2588654073991089e-05, + "loss": 0.3897, + "step": 10161 + }, + { + "epoch": 3.831825037707391, + "grad_norm": 0.11549818479391868, + "learning_rate": 1.2580986354954012e-05, + "loss": 0.4022, + "step": 10162 + }, + { + "epoch": 3.8322021116138765, + "grad_norm": 0.11513471251837103, + "learning_rate": 1.2573320536030278e-05, + "loss": 0.4033, + "step": 10163 + }, + { + "epoch": 3.832579185520362, + "grad_norm": 0.11144218697085873, + "learning_rate": 1.2565656617751075e-05, + "loss": 0.3869, + "step": 10164 + }, + { + "epoch": 3.832956259426848, + "grad_norm": 0.11874958946393964, + "learning_rate": 1.2557994600647559e-05, + "loss": 0.3933, + "step": 10165 + }, + { + "epoch": 3.8333333333333335, + "grad_norm": 0.14560040983673633, + "learning_rate": 1.2550334485250661e-05, + "loss": 0.4013, + "step": 10166 + }, + { + "epoch": 3.833710407239819, + "grad_norm": 0.10647037908481798, + "learning_rate": 1.2542676272091244e-05, + "loss": 0.3909, + "step": 10167 + }, + { + "epoch": 3.834087481146305, + "grad_norm": 0.14423007061896978, + "learning_rate": 1.2535019961699998e-05, + "loss": 0.3887, + "step": 10168 + }, + { + "epoch": 3.8344645550527905, + "grad_norm": 0.10105621323950324, + "learning_rate": 1.2527365554607522e-05, + "loss": 0.3862, + "step": 10169 + }, + { + "epoch": 3.834841628959276, + "grad_norm": 0.1401906394990897, + "learning_rate": 1.251971305134422e-05, + "loss": 0.3935, + "step": 10170 + }, + { + "epoch": 3.835218702865762, + "grad_norm": 0.10610630106321516, + "learning_rate": 1.2512062452440454e-05, + "loss": 0.3848, + "step": 10171 + }, + { + "epoch": 3.8355957767722475, + "grad_norm": 0.12683446272844293, + "learning_rate": 1.250441375842637e-05, + "loss": 0.3873, + "step": 10172 + }, + { + "epoch": 3.835972850678733, + "grad_norm": 0.10847954754691067, + "learning_rate": 1.2496766969832019e-05, + "loss": 0.3936, + "step": 10173 + }, + { + "epoch": 3.836349924585219, + "grad_norm": 0.12460258111685248, + "learning_rate": 1.2489122087187324e-05, + "loss": 0.3861, + "step": 10174 + }, + { + "epoch": 3.8367269984917045, + "grad_norm": 0.11241485261598369, + "learning_rate": 1.2481479111022065e-05, + "loss": 0.3802, + "step": 10175 + }, + { + "epoch": 3.83710407239819, + "grad_norm": 0.13244340295106646, + "learning_rate": 1.247383804186589e-05, + "loss": 0.3877, + "step": 10176 + }, + { + "epoch": 3.837481146304676, + "grad_norm": 0.12116239903343418, + "learning_rate": 1.2466198880248345e-05, + "loss": 0.3995, + "step": 10177 + }, + { + "epoch": 3.8378582202111615, + "grad_norm": 0.16238765803083355, + "learning_rate": 1.2458561626698776e-05, + "loss": 0.3936, + "step": 10178 + }, + { + "epoch": 3.838235294117647, + "grad_norm": 0.10477935634103737, + "learning_rate": 1.2450926281746458e-05, + "loss": 0.3976, + "step": 10179 + }, + { + "epoch": 3.838612368024133, + "grad_norm": 0.15191082334323436, + "learning_rate": 1.2443292845920513e-05, + "loss": 0.387, + "step": 10180 + }, + { + "epoch": 3.8389894419306185, + "grad_norm": 0.1053551383967024, + "learning_rate": 1.2435661319749932e-05, + "loss": 0.3943, + "step": 10181 + }, + { + "epoch": 3.839366515837104, + "grad_norm": 0.16538205533840764, + "learning_rate": 1.2428031703763574e-05, + "loss": 0.401, + "step": 10182 + }, + { + "epoch": 3.83974358974359, + "grad_norm": 0.10982260968241779, + "learning_rate": 1.2420403998490174e-05, + "loss": 0.3843, + "step": 10183 + }, + { + "epoch": 3.8401206636500755, + "grad_norm": 0.1585272257804956, + "learning_rate": 1.2412778204458297e-05, + "loss": 0.3899, + "step": 10184 + }, + { + "epoch": 3.840497737556561, + "grad_norm": 0.11572845639686821, + "learning_rate": 1.2405154322196427e-05, + "loss": 0.3909, + "step": 10185 + }, + { + "epoch": 3.840874811463047, + "grad_norm": 0.14181567067284256, + "learning_rate": 1.239753235223288e-05, + "loss": 0.3898, + "step": 10186 + }, + { + "epoch": 3.8412518853695325, + "grad_norm": 0.12668947522523277, + "learning_rate": 1.2389912295095861e-05, + "loss": 0.3921, + "step": 10187 + }, + { + "epoch": 3.841628959276018, + "grad_norm": 0.13775771286651298, + "learning_rate": 1.2382294151313427e-05, + "loss": 0.399, + "step": 10188 + }, + { + "epoch": 3.842006033182504, + "grad_norm": 0.12229713895910768, + "learning_rate": 1.2374677921413527e-05, + "loss": 0.3905, + "step": 10189 + }, + { + "epoch": 3.8423831070889896, + "grad_norm": 0.13287670977176674, + "learning_rate": 1.2367063605923932e-05, + "loss": 0.4032, + "step": 10190 + }, + { + "epoch": 3.8427601809954752, + "grad_norm": 0.10792114066834707, + "learning_rate": 1.2359451205372315e-05, + "loss": 0.3881, + "step": 10191 + }, + { + "epoch": 3.843137254901961, + "grad_norm": 0.15512847380763767, + "learning_rate": 1.2351840720286212e-05, + "loss": 0.3948, + "step": 10192 + }, + { + "epoch": 3.8435143288084466, + "grad_norm": 0.10395507836772004, + "learning_rate": 1.2344232151193025e-05, + "loss": 0.4056, + "step": 10193 + }, + { + "epoch": 3.8438914027149322, + "grad_norm": 0.1496075720753241, + "learning_rate": 1.233662549862002e-05, + "loss": 0.387, + "step": 10194 + }, + { + "epoch": 3.844268476621418, + "grad_norm": 0.11159063153308958, + "learning_rate": 1.2329020763094333e-05, + "loss": 0.3995, + "step": 10195 + }, + { + "epoch": 3.8446455505279036, + "grad_norm": 0.1559596244466363, + "learning_rate": 1.2321417945142971e-05, + "loss": 0.4, + "step": 10196 + }, + { + "epoch": 3.8450226244343892, + "grad_norm": 0.10602855607777856, + "learning_rate": 1.2313817045292783e-05, + "loss": 0.3887, + "step": 10197 + }, + { + "epoch": 3.845399698340875, + "grad_norm": 0.1261566271164707, + "learning_rate": 1.2306218064070516e-05, + "loss": 0.3813, + "step": 10198 + }, + { + "epoch": 3.8457767722473606, + "grad_norm": 0.1058029257336973, + "learning_rate": 1.2298621002002772e-05, + "loss": 0.3863, + "step": 10199 + }, + { + "epoch": 3.8461538461538463, + "grad_norm": 0.12984599075878492, + "learning_rate": 1.2291025859616026e-05, + "loss": 0.3893, + "step": 10200 + }, + { + "epoch": 3.846530920060332, + "grad_norm": 0.13013616621822524, + "learning_rate": 1.2283432637436605e-05, + "loss": 0.398, + "step": 10201 + }, + { + "epoch": 3.8469079939668176, + "grad_norm": 0.15547991415572462, + "learning_rate": 1.2275841335990735e-05, + "loss": 0.4001, + "step": 10202 + }, + { + "epoch": 3.8472850678733033, + "grad_norm": 0.12396969835942552, + "learning_rate": 1.2268251955804451e-05, + "loss": 0.392, + "step": 10203 + }, + { + "epoch": 3.847662141779789, + "grad_norm": 0.1428874434072408, + "learning_rate": 1.226066449740371e-05, + "loss": 0.3868, + "step": 10204 + }, + { + "epoch": 3.8480392156862746, + "grad_norm": 0.1125356856015458, + "learning_rate": 1.2253078961314317e-05, + "loss": 0.3863, + "step": 10205 + }, + { + "epoch": 3.8484162895927603, + "grad_norm": 0.13799134174807298, + "learning_rate": 1.2245495348061941e-05, + "loss": 0.3966, + "step": 10206 + }, + { + "epoch": 3.848793363499246, + "grad_norm": 0.09919170898014842, + "learning_rate": 1.2237913658172116e-05, + "loss": 0.3839, + "step": 10207 + }, + { + "epoch": 3.8491704374057316, + "grad_norm": 0.13605642498686682, + "learning_rate": 1.2230333892170263e-05, + "loss": 0.4086, + "step": 10208 + }, + { + "epoch": 3.8495475113122173, + "grad_norm": 0.0925008184800931, + "learning_rate": 1.222275605058163e-05, + "loss": 0.3839, + "step": 10209 + }, + { + "epoch": 3.849924585218703, + "grad_norm": 0.10578295780757266, + "learning_rate": 1.2215180133931362e-05, + "loss": 0.3944, + "step": 10210 + }, + { + "epoch": 3.8503016591251886, + "grad_norm": 0.09585705368167426, + "learning_rate": 1.2207606142744463e-05, + "loss": 0.3972, + "step": 10211 + }, + { + "epoch": 3.8506787330316743, + "grad_norm": 0.11154215181748975, + "learning_rate": 1.220003407754581e-05, + "loss": 0.3862, + "step": 10212 + }, + { + "epoch": 3.85105580693816, + "grad_norm": 0.09182199672759477, + "learning_rate": 1.2192463938860137e-05, + "loss": 0.3853, + "step": 10213 + }, + { + "epoch": 3.8514328808446456, + "grad_norm": 0.11246678480376192, + "learning_rate": 1.218489572721206e-05, + "loss": 0.3948, + "step": 10214 + }, + { + "epoch": 3.8518099547511313, + "grad_norm": 0.10614970902165093, + "learning_rate": 1.2177329443126008e-05, + "loss": 0.3956, + "step": 10215 + }, + { + "epoch": 3.852187028657617, + "grad_norm": 0.10544926733316232, + "learning_rate": 1.2169765087126377e-05, + "loss": 0.4025, + "step": 10216 + }, + { + "epoch": 3.8525641025641026, + "grad_norm": 0.10697768758440745, + "learning_rate": 1.2162202659737328e-05, + "loss": 0.4099, + "step": 10217 + }, + { + "epoch": 3.8529411764705883, + "grad_norm": 0.10386761464746448, + "learning_rate": 1.2154642161482939e-05, + "loss": 0.3974, + "step": 10218 + }, + { + "epoch": 3.853318250377074, + "grad_norm": 0.11044232682216124, + "learning_rate": 1.2147083592887148e-05, + "loss": 0.3929, + "step": 10219 + }, + { + "epoch": 3.8536953242835597, + "grad_norm": 0.09305076683072679, + "learning_rate": 1.2139526954473761e-05, + "loss": 0.3885, + "step": 10220 + }, + { + "epoch": 3.8540723981900453, + "grad_norm": 0.0988755720651071, + "learning_rate": 1.2131972246766446e-05, + "loss": 0.3913, + "step": 10221 + }, + { + "epoch": 3.854449472096531, + "grad_norm": 0.10480616775937163, + "learning_rate": 1.2124419470288746e-05, + "loss": 0.4067, + "step": 10222 + }, + { + "epoch": 3.8548265460030167, + "grad_norm": 0.10375326812694759, + "learning_rate": 1.211686862556404e-05, + "loss": 0.3842, + "step": 10223 + }, + { + "epoch": 3.8552036199095023, + "grad_norm": 0.11723073054112984, + "learning_rate": 1.2109319713115601e-05, + "loss": 0.387, + "step": 10224 + }, + { + "epoch": 3.855580693815988, + "grad_norm": 0.09543150431523269, + "learning_rate": 1.2101772733466573e-05, + "loss": 0.3929, + "step": 10225 + }, + { + "epoch": 3.8559577677224737, + "grad_norm": 0.11493273666149594, + "learning_rate": 1.2094227687139948e-05, + "loss": 0.4009, + "step": 10226 + }, + { + "epoch": 3.8563348416289593, + "grad_norm": 0.08794334060890949, + "learning_rate": 1.2086684574658588e-05, + "loss": 0.3959, + "step": 10227 + }, + { + "epoch": 3.856711915535445, + "grad_norm": 0.1118215198924124, + "learning_rate": 1.2079143396545244e-05, + "loss": 0.3932, + "step": 10228 + }, + { + "epoch": 3.8570889894419307, + "grad_norm": 0.09916878617866207, + "learning_rate": 1.2071604153322483e-05, + "loss": 0.3986, + "step": 10229 + }, + { + "epoch": 3.8574660633484164, + "grad_norm": 0.10435557774331905, + "learning_rate": 1.2064066845512778e-05, + "loss": 0.3903, + "step": 10230 + }, + { + "epoch": 3.857843137254902, + "grad_norm": 0.09599421202888103, + "learning_rate": 1.2056531473638468e-05, + "loss": 0.3946, + "step": 10231 + }, + { + "epoch": 3.8582202111613877, + "grad_norm": 0.1111033121516502, + "learning_rate": 1.204899803822174e-05, + "loss": 0.3906, + "step": 10232 + }, + { + "epoch": 3.8585972850678734, + "grad_norm": 0.0969838220531993, + "learning_rate": 1.204146653978465e-05, + "loss": 0.3932, + "step": 10233 + }, + { + "epoch": 3.858974358974359, + "grad_norm": 0.11741444321220128, + "learning_rate": 1.2033936978849145e-05, + "loss": 0.3948, + "step": 10234 + }, + { + "epoch": 3.8593514328808447, + "grad_norm": 0.10174555951541804, + "learning_rate": 1.2026409355936978e-05, + "loss": 0.4043, + "step": 10235 + }, + { + "epoch": 3.8597285067873304, + "grad_norm": 0.10989585392656306, + "learning_rate": 1.2018883671569857e-05, + "loss": 0.3946, + "step": 10236 + }, + { + "epoch": 3.860105580693816, + "grad_norm": 0.10499408570117504, + "learning_rate": 1.2011359926269259e-05, + "loss": 0.3895, + "step": 10237 + }, + { + "epoch": 3.8604826546003017, + "grad_norm": 0.09434326263727698, + "learning_rate": 1.2003838120556598e-05, + "loss": 0.4047, + "step": 10238 + }, + { + "epoch": 3.8608597285067874, + "grad_norm": 0.0924094807418323, + "learning_rate": 1.1996318254953124e-05, + "loss": 0.3834, + "step": 10239 + }, + { + "epoch": 3.861236802413273, + "grad_norm": 0.10058073267533311, + "learning_rate": 1.1988800329979963e-05, + "loss": 0.4026, + "step": 10240 + }, + { + "epoch": 3.8616138763197587, + "grad_norm": 0.1097579260529056, + "learning_rate": 1.198128434615807e-05, + "loss": 0.4118, + "step": 10241 + }, + { + "epoch": 3.8619909502262444, + "grad_norm": 0.10215323406254714, + "learning_rate": 1.1973770304008343e-05, + "loss": 0.4016, + "step": 10242 + }, + { + "epoch": 3.86236802413273, + "grad_norm": 0.10718329239677347, + "learning_rate": 1.1966258204051466e-05, + "loss": 0.3796, + "step": 10243 + }, + { + "epoch": 3.8627450980392157, + "grad_norm": 0.10136701493403995, + "learning_rate": 1.195874804680803e-05, + "loss": 0.3953, + "step": 10244 + }, + { + "epoch": 3.8631221719457014, + "grad_norm": 0.09225271649793891, + "learning_rate": 1.1951239832798476e-05, + "loss": 0.3887, + "step": 10245 + }, + { + "epoch": 3.863499245852187, + "grad_norm": 0.10017602672111826, + "learning_rate": 1.1943733562543138e-05, + "loss": 0.3926, + "step": 10246 + }, + { + "epoch": 3.8638763197586727, + "grad_norm": 0.1029335861583472, + "learning_rate": 1.1936229236562155e-05, + "loss": 0.383, + "step": 10247 + }, + { + "epoch": 3.8642533936651584, + "grad_norm": 0.10321029736570099, + "learning_rate": 1.1928726855375615e-05, + "loss": 0.399, + "step": 10248 + }, + { + "epoch": 3.864630467571644, + "grad_norm": 0.10400913719816898, + "learning_rate": 1.1921226419503396e-05, + "loss": 0.405, + "step": 10249 + }, + { + "epoch": 3.8650075414781297, + "grad_norm": 0.11500958941169968, + "learning_rate": 1.1913727929465275e-05, + "loss": 0.4051, + "step": 10250 + }, + { + "epoch": 3.8653846153846154, + "grad_norm": 0.09655241001667297, + "learning_rate": 1.19062313857809e-05, + "loss": 0.4002, + "step": 10251 + }, + { + "epoch": 3.865761689291101, + "grad_norm": 0.10610568849412924, + "learning_rate": 1.1898736788969783e-05, + "loss": 0.4054, + "step": 10252 + }, + { + "epoch": 3.8661387631975868, + "grad_norm": 0.09762138778812951, + "learning_rate": 1.1891244139551254e-05, + "loss": 0.3993, + "step": 10253 + }, + { + "epoch": 3.8665158371040724, + "grad_norm": 0.0971438478667198, + "learning_rate": 1.1883753438044603e-05, + "loss": 0.3991, + "step": 10254 + }, + { + "epoch": 3.866892911010558, + "grad_norm": 0.09655384994626226, + "learning_rate": 1.1876264684968883e-05, + "loss": 0.4064, + "step": 10255 + }, + { + "epoch": 3.8672699849170438, + "grad_norm": 0.10442697297014775, + "learning_rate": 1.1868777880843076e-05, + "loss": 0.3954, + "step": 10256 + }, + { + "epoch": 3.8676470588235294, + "grad_norm": 0.08794986622479545, + "learning_rate": 1.1861293026186007e-05, + "loss": 0.3896, + "step": 10257 + }, + { + "epoch": 3.868024132730015, + "grad_norm": 0.0989927540499181, + "learning_rate": 1.1853810121516389e-05, + "loss": 0.3894, + "step": 10258 + }, + { + "epoch": 3.8684012066365008, + "grad_norm": 0.08875824968036203, + "learning_rate": 1.1846329167352738e-05, + "loss": 0.3957, + "step": 10259 + }, + { + "epoch": 3.8687782805429864, + "grad_norm": 0.09844180631937238, + "learning_rate": 1.1838850164213525e-05, + "loss": 0.396, + "step": 10260 + }, + { + "epoch": 3.869155354449472, + "grad_norm": 0.09321274406565715, + "learning_rate": 1.1831373112617009e-05, + "loss": 0.3974, + "step": 10261 + }, + { + "epoch": 3.869532428355958, + "grad_norm": 0.09169131213012661, + "learning_rate": 1.1823898013081348e-05, + "loss": 0.4052, + "step": 10262 + }, + { + "epoch": 3.8699095022624435, + "grad_norm": 0.09523742187201989, + "learning_rate": 1.1816424866124563e-05, + "loss": 0.3946, + "step": 10263 + }, + { + "epoch": 3.870286576168929, + "grad_norm": 0.09481273340121073, + "learning_rate": 1.1808953672264547e-05, + "loss": 0.3993, + "step": 10264 + }, + { + "epoch": 3.870663650075415, + "grad_norm": 0.09193752883368707, + "learning_rate": 1.1801484432019014e-05, + "loss": 0.3967, + "step": 10265 + }, + { + "epoch": 3.8710407239819005, + "grad_norm": 0.09139255532990545, + "learning_rate": 1.1794017145905619e-05, + "loss": 0.3966, + "step": 10266 + }, + { + "epoch": 3.871417797888386, + "grad_norm": 0.08968529776507958, + "learning_rate": 1.1786551814441807e-05, + "loss": 0.3872, + "step": 10267 + }, + { + "epoch": 3.871794871794872, + "grad_norm": 0.08993774117334649, + "learning_rate": 1.177908843814493e-05, + "loss": 0.3991, + "step": 10268 + }, + { + "epoch": 3.8721719457013575, + "grad_norm": 0.09794669563211462, + "learning_rate": 1.1771627017532193e-05, + "loss": 0.39, + "step": 10269 + }, + { + "epoch": 3.872549019607843, + "grad_norm": 0.09326438602467634, + "learning_rate": 1.1764167553120673e-05, + "loss": 0.3811, + "step": 10270 + }, + { + "epoch": 3.872926093514329, + "grad_norm": 0.09528229852743464, + "learning_rate": 1.1756710045427276e-05, + "loss": 0.3921, + "step": 10271 + }, + { + "epoch": 3.8733031674208145, + "grad_norm": 0.08446137868480257, + "learning_rate": 1.1749254494968843e-05, + "loss": 0.4003, + "step": 10272 + }, + { + "epoch": 3.8736802413273, + "grad_norm": 0.09154810987957873, + "learning_rate": 1.1741800902262001e-05, + "loss": 0.3782, + "step": 10273 + }, + { + "epoch": 3.874057315233786, + "grad_norm": 0.08935170942516713, + "learning_rate": 1.1734349267823295e-05, + "loss": 0.3879, + "step": 10274 + }, + { + "epoch": 3.8744343891402715, + "grad_norm": 0.09879490260166374, + "learning_rate": 1.1726899592169114e-05, + "loss": 0.3914, + "step": 10275 + }, + { + "epoch": 3.874811463046757, + "grad_norm": 0.09412618381377037, + "learning_rate": 1.1719451875815726e-05, + "loss": 0.3962, + "step": 10276 + }, + { + "epoch": 3.875188536953243, + "grad_norm": 0.12059340879559757, + "learning_rate": 1.171200611927921e-05, + "loss": 0.4002, + "step": 10277 + }, + { + "epoch": 3.8755656108597285, + "grad_norm": 0.10134823470805698, + "learning_rate": 1.1704562323075606e-05, + "loss": 0.406, + "step": 10278 + }, + { + "epoch": 3.875942684766214, + "grad_norm": 0.10436931579685732, + "learning_rate": 1.1697120487720719e-05, + "loss": 0.3977, + "step": 10279 + }, + { + "epoch": 3.8763197586727, + "grad_norm": 0.11804041226293037, + "learning_rate": 1.1689680613730281e-05, + "loss": 0.3869, + "step": 10280 + }, + { + "epoch": 3.8766968325791855, + "grad_norm": 0.10610333280297993, + "learning_rate": 1.1682242701619867e-05, + "loss": 0.3956, + "step": 10281 + }, + { + "epoch": 3.877073906485671, + "grad_norm": 0.1258214242319756, + "learning_rate": 1.1674806751904923e-05, + "loss": 0.39, + "step": 10282 + }, + { + "epoch": 3.877450980392157, + "grad_norm": 0.10924562559428576, + "learning_rate": 1.1667372765100722e-05, + "loss": 0.4029, + "step": 10283 + }, + { + "epoch": 3.8778280542986425, + "grad_norm": 0.13310888957502348, + "learning_rate": 1.1659940741722484e-05, + "loss": 0.3965, + "step": 10284 + }, + { + "epoch": 3.878205128205128, + "grad_norm": 0.09150583227920046, + "learning_rate": 1.1652510682285203e-05, + "loss": 0.3908, + "step": 10285 + }, + { + "epoch": 3.878582202111614, + "grad_norm": 0.10639841025894144, + "learning_rate": 1.1645082587303791e-05, + "loss": 0.3995, + "step": 10286 + }, + { + "epoch": 3.8789592760180995, + "grad_norm": 0.09405296008495971, + "learning_rate": 1.1637656457293e-05, + "loss": 0.3953, + "step": 10287 + }, + { + "epoch": 3.879336349924585, + "grad_norm": 0.0925963436624521, + "learning_rate": 1.1630232292767474e-05, + "loss": 0.4042, + "step": 10288 + }, + { + "epoch": 3.879713423831071, + "grad_norm": 0.10410606960352653, + "learning_rate": 1.162281009424166e-05, + "loss": 0.4021, + "step": 10289 + }, + { + "epoch": 3.8800904977375565, + "grad_norm": 0.0967509016262344, + "learning_rate": 1.161538986222996e-05, + "loss": 0.3965, + "step": 10290 + }, + { + "epoch": 3.880467571644042, + "grad_norm": 0.09624400451543759, + "learning_rate": 1.1607971597246555e-05, + "loss": 0.3938, + "step": 10291 + }, + { + "epoch": 3.880844645550528, + "grad_norm": 0.09345456675707926, + "learning_rate": 1.1600555299805532e-05, + "loss": 0.3838, + "step": 10292 + }, + { + "epoch": 3.8812217194570136, + "grad_norm": 0.10106663413600588, + "learning_rate": 1.159314097042084e-05, + "loss": 0.4023, + "step": 10293 + }, + { + "epoch": 3.8815987933634992, + "grad_norm": 0.10763124085080333, + "learning_rate": 1.158572860960629e-05, + "loss": 0.394, + "step": 10294 + }, + { + "epoch": 3.881975867269985, + "grad_norm": 0.10252917182974548, + "learning_rate": 1.1578318217875526e-05, + "loss": 0.3966, + "step": 10295 + }, + { + "epoch": 3.8823529411764706, + "grad_norm": 0.09256611851835334, + "learning_rate": 1.1570909795742118e-05, + "loss": 0.3914, + "step": 10296 + }, + { + "epoch": 3.8827300150829562, + "grad_norm": 0.13079180939193405, + "learning_rate": 1.1563503343719433e-05, + "loss": 0.3931, + "step": 10297 + }, + { + "epoch": 3.883107088989442, + "grad_norm": 0.1015798197763502, + "learning_rate": 1.1556098862320746e-05, + "loss": 0.3916, + "step": 10298 + }, + { + "epoch": 3.8834841628959276, + "grad_norm": 0.11442605767558037, + "learning_rate": 1.1548696352059174e-05, + "loss": 0.397, + "step": 10299 + }, + { + "epoch": 3.8838612368024132, + "grad_norm": 0.10507095791237665, + "learning_rate": 1.1541295813447725e-05, + "loss": 0.3965, + "step": 10300 + }, + { + "epoch": 3.884238310708899, + "grad_norm": 0.10190185721004238, + "learning_rate": 1.1533897246999209e-05, + "loss": 0.3818, + "step": 10301 + }, + { + "epoch": 3.8846153846153846, + "grad_norm": 0.10118967316948067, + "learning_rate": 1.1526500653226385e-05, + "loss": 0.403, + "step": 10302 + }, + { + "epoch": 3.8849924585218703, + "grad_norm": 0.09621785107931179, + "learning_rate": 1.15191060326418e-05, + "loss": 0.3752, + "step": 10303 + }, + { + "epoch": 3.885369532428356, + "grad_norm": 0.09981473099168645, + "learning_rate": 1.1511713385757904e-05, + "loss": 0.3985, + "step": 10304 + }, + { + "epoch": 3.8857466063348416, + "grad_norm": 0.1297281783776706, + "learning_rate": 1.1504322713087e-05, + "loss": 0.3811, + "step": 10305 + }, + { + "epoch": 3.8861236802413273, + "grad_norm": 0.12158258856937117, + "learning_rate": 1.149693401514127e-05, + "loss": 0.3906, + "step": 10306 + }, + { + "epoch": 3.886500754147813, + "grad_norm": 0.11215293539395366, + "learning_rate": 1.1489547292432705e-05, + "loss": 0.3869, + "step": 10307 + }, + { + "epoch": 3.8868778280542986, + "grad_norm": 0.09976163851381507, + "learning_rate": 1.1482162545473247e-05, + "loss": 0.3898, + "step": 10308 + }, + { + "epoch": 3.8872549019607843, + "grad_norm": 0.10111987628585302, + "learning_rate": 1.147477977477462e-05, + "loss": 0.3843, + "step": 10309 + }, + { + "epoch": 3.88763197586727, + "grad_norm": 0.09484782792289673, + "learning_rate": 1.1467398980848455e-05, + "loss": 0.3834, + "step": 10310 + }, + { + "epoch": 3.8880090497737556, + "grad_norm": 0.09394332513196649, + "learning_rate": 1.1460020164206229e-05, + "loss": 0.3884, + "step": 10311 + }, + { + "epoch": 3.8883861236802413, + "grad_norm": 0.09362434825598227, + "learning_rate": 1.1452643325359301e-05, + "loss": 0.3823, + "step": 10312 + }, + { + "epoch": 3.888763197586727, + "grad_norm": 0.0994608231383333, + "learning_rate": 1.1445268464818847e-05, + "loss": 0.3912, + "step": 10313 + }, + { + "epoch": 3.8891402714932126, + "grad_norm": 0.11089340678163646, + "learning_rate": 1.1437895583095986e-05, + "loss": 0.3959, + "step": 10314 + }, + { + "epoch": 3.8895173453996983, + "grad_norm": 0.12210105596281405, + "learning_rate": 1.1430524680701613e-05, + "loss": 0.3961, + "step": 10315 + }, + { + "epoch": 3.889894419306184, + "grad_norm": 0.11021686773400409, + "learning_rate": 1.1423155758146538e-05, + "loss": 0.4007, + "step": 10316 + }, + { + "epoch": 3.8902714932126696, + "grad_norm": 0.1171520530649113, + "learning_rate": 1.1415788815941423e-05, + "loss": 0.3946, + "step": 10317 + }, + { + "epoch": 3.8906485671191553, + "grad_norm": 0.11074251119880757, + "learning_rate": 1.14084238545968e-05, + "loss": 0.3933, + "step": 10318 + }, + { + "epoch": 3.891025641025641, + "grad_norm": 0.09261644872945017, + "learning_rate": 1.1401060874623018e-05, + "loss": 0.3939, + "step": 10319 + }, + { + "epoch": 3.8914027149321266, + "grad_norm": 0.10192026045749882, + "learning_rate": 1.1393699876530375e-05, + "loss": 0.3914, + "step": 10320 + }, + { + "epoch": 3.8917797888386123, + "grad_norm": 0.08270531870694875, + "learning_rate": 1.1386340860828945e-05, + "loss": 0.3998, + "step": 10321 + }, + { + "epoch": 3.892156862745098, + "grad_norm": 0.09399369080074754, + "learning_rate": 1.1378983828028712e-05, + "loss": 0.3851, + "step": 10322 + }, + { + "epoch": 3.8925339366515836, + "grad_norm": 0.09726868002360883, + "learning_rate": 1.1371628778639514e-05, + "loss": 0.3982, + "step": 10323 + }, + { + "epoch": 3.8929110105580693, + "grad_norm": 0.08932155864072623, + "learning_rate": 1.1364275713171047e-05, + "loss": 0.3862, + "step": 10324 + }, + { + "epoch": 3.893288084464555, + "grad_norm": 0.10055044620984127, + "learning_rate": 1.1356924632132876e-05, + "loss": 0.386, + "step": 10325 + }, + { + "epoch": 3.8936651583710407, + "grad_norm": 0.09976453280023806, + "learning_rate": 1.1349575536034432e-05, + "loss": 0.4023, + "step": 10326 + }, + { + "epoch": 3.8940422322775263, + "grad_norm": 0.09472917833674153, + "learning_rate": 1.1342228425384975e-05, + "loss": 0.3857, + "step": 10327 + }, + { + "epoch": 3.894419306184012, + "grad_norm": 0.10581683229770471, + "learning_rate": 1.1334883300693673e-05, + "loss": 0.3935, + "step": 10328 + }, + { + "epoch": 3.8947963800904977, + "grad_norm": 0.10132513097284135, + "learning_rate": 1.1327540162469531e-05, + "loss": 0.3978, + "step": 10329 + }, + { + "epoch": 3.8951734539969833, + "grad_norm": 0.0970447842571627, + "learning_rate": 1.1320199011221424e-05, + "loss": 0.3949, + "step": 10330 + }, + { + "epoch": 3.895550527903469, + "grad_norm": 0.10869186421656371, + "learning_rate": 1.1312859847458086e-05, + "loss": 0.4051, + "step": 10331 + }, + { + "epoch": 3.8959276018099547, + "grad_norm": 0.095817823349269, + "learning_rate": 1.1305522671688127e-05, + "loss": 0.4024, + "step": 10332 + }, + { + "epoch": 3.8963046757164403, + "grad_norm": 0.0958873205507633, + "learning_rate": 1.1298187484419985e-05, + "loss": 0.3908, + "step": 10333 + }, + { + "epoch": 3.896681749622926, + "grad_norm": 0.09649861836579775, + "learning_rate": 1.129085428616199e-05, + "loss": 0.399, + "step": 10334 + }, + { + "epoch": 3.8970588235294117, + "grad_norm": 0.09158373244506712, + "learning_rate": 1.1283523077422327e-05, + "loss": 0.3899, + "step": 10335 + }, + { + "epoch": 3.8974358974358974, + "grad_norm": 0.11304908213686064, + "learning_rate": 1.1276193858709044e-05, + "loss": 0.3959, + "step": 10336 + }, + { + "epoch": 3.897812971342383, + "grad_norm": 0.09769678492581432, + "learning_rate": 1.1268866630530049e-05, + "loss": 0.3962, + "step": 10337 + }, + { + "epoch": 3.8981900452488687, + "grad_norm": 0.09948665684260273, + "learning_rate": 1.1261541393393127e-05, + "loss": 0.3993, + "step": 10338 + }, + { + "epoch": 3.8985671191553544, + "grad_norm": 0.09361180291066878, + "learning_rate": 1.1254218147805864e-05, + "loss": 0.3869, + "step": 10339 + }, + { + "epoch": 3.89894419306184, + "grad_norm": 0.11694982034372373, + "learning_rate": 1.1246896894275818e-05, + "loss": 0.397, + "step": 10340 + }, + { + "epoch": 3.8993212669683257, + "grad_norm": 0.09734744307500078, + "learning_rate": 1.1239577633310294e-05, + "loss": 0.3903, + "step": 10341 + }, + { + "epoch": 3.8996983408748114, + "grad_norm": 0.09474745442157763, + "learning_rate": 1.1232260365416537e-05, + "loss": 0.3936, + "step": 10342 + }, + { + "epoch": 3.900075414781297, + "grad_norm": 0.08905401305482426, + "learning_rate": 1.1224945091101614e-05, + "loss": 0.3839, + "step": 10343 + }, + { + "epoch": 3.9004524886877827, + "grad_norm": 0.10591637801569999, + "learning_rate": 1.1217631810872476e-05, + "loss": 0.3956, + "step": 10344 + }, + { + "epoch": 3.9008295625942684, + "grad_norm": 0.09262754311739296, + "learning_rate": 1.1210320525235923e-05, + "loss": 0.3938, + "step": 10345 + }, + { + "epoch": 3.901206636500754, + "grad_norm": 0.10826713788331897, + "learning_rate": 1.120301123469863e-05, + "loss": 0.3873, + "step": 10346 + }, + { + "epoch": 3.9015837104072397, + "grad_norm": 0.10169186498250025, + "learning_rate": 1.1195703939767104e-05, + "loss": 0.3944, + "step": 10347 + }, + { + "epoch": 3.9019607843137254, + "grad_norm": 0.10892305106181605, + "learning_rate": 1.1188398640947748e-05, + "loss": 0.4005, + "step": 10348 + }, + { + "epoch": 3.902337858220211, + "grad_norm": 0.10277723470998447, + "learning_rate": 1.118109533874681e-05, + "loss": 0.392, + "step": 10349 + }, + { + "epoch": 3.9027149321266967, + "grad_norm": 0.09671988955054671, + "learning_rate": 1.1173794033670405e-05, + "loss": 0.3995, + "step": 10350 + }, + { + "epoch": 3.9030920060331824, + "grad_norm": 0.1064723332396815, + "learning_rate": 1.11664947262245e-05, + "loss": 0.3968, + "step": 10351 + }, + { + "epoch": 3.903469079939668, + "grad_norm": 0.09786186624608631, + "learning_rate": 1.1159197416914948e-05, + "loss": 0.3857, + "step": 10352 + }, + { + "epoch": 3.9038461538461537, + "grad_norm": 0.10861980904120667, + "learning_rate": 1.1151902106247423e-05, + "loss": 0.392, + "step": 10353 + }, + { + "epoch": 3.9042232277526394, + "grad_norm": 0.09703753805996046, + "learning_rate": 1.1144608794727496e-05, + "loss": 0.3874, + "step": 10354 + }, + { + "epoch": 3.904600301659125, + "grad_norm": 0.10272929892218415, + "learning_rate": 1.1137317482860582e-05, + "loss": 0.3962, + "step": 10355 + }, + { + "epoch": 3.9049773755656108, + "grad_norm": 0.10064846679274468, + "learning_rate": 1.1130028171151964e-05, + "loss": 0.3929, + "step": 10356 + }, + { + "epoch": 3.9053544494720964, + "grad_norm": 0.09165386948082456, + "learning_rate": 1.112274086010679e-05, + "loss": 0.3836, + "step": 10357 + }, + { + "epoch": 3.905731523378582, + "grad_norm": 0.09595761090008657, + "learning_rate": 1.1115455550230067e-05, + "loss": 0.3888, + "step": 10358 + }, + { + "epoch": 3.9061085972850678, + "grad_norm": 0.10836349476920858, + "learning_rate": 1.1108172242026644e-05, + "loss": 0.3858, + "step": 10359 + }, + { + "epoch": 3.9064856711915534, + "grad_norm": 0.10106820482321237, + "learning_rate": 1.1100890936001258e-05, + "loss": 0.3933, + "step": 10360 + }, + { + "epoch": 3.906862745098039, + "grad_norm": 0.09588762523754067, + "learning_rate": 1.10936116326585e-05, + "loss": 0.4028, + "step": 10361 + }, + { + "epoch": 3.9072398190045248, + "grad_norm": 0.1106041715688032, + "learning_rate": 1.1086334332502813e-05, + "loss": 0.3847, + "step": 10362 + }, + { + "epoch": 3.9076168929110104, + "grad_norm": 0.09672255428009398, + "learning_rate": 1.107905903603851e-05, + "loss": 0.3906, + "step": 10363 + }, + { + "epoch": 3.907993966817496, + "grad_norm": 0.09563516387024017, + "learning_rate": 1.107178574376976e-05, + "loss": 0.391, + "step": 10364 + }, + { + "epoch": 3.908371040723982, + "grad_norm": 0.10152614129069235, + "learning_rate": 1.1064514456200616e-05, + "loss": 0.4079, + "step": 10365 + }, + { + "epoch": 3.9087481146304675, + "grad_norm": 0.1204996738930815, + "learning_rate": 1.1057245173834938e-05, + "loss": 0.3953, + "step": 10366 + }, + { + "epoch": 3.909125188536953, + "grad_norm": 0.0999757851503586, + "learning_rate": 1.1049977897176501e-05, + "loss": 0.3973, + "step": 10367 + }, + { + "epoch": 3.909502262443439, + "grad_norm": 0.11592648331109899, + "learning_rate": 1.1042712626728917e-05, + "loss": 0.4021, + "step": 10368 + }, + { + "epoch": 3.9098793363499245, + "grad_norm": 0.09239247647560549, + "learning_rate": 1.1035449362995663e-05, + "loss": 0.3852, + "step": 10369 + }, + { + "epoch": 3.91025641025641, + "grad_norm": 0.09539937919985875, + "learning_rate": 1.102818810648007e-05, + "loss": 0.3955, + "step": 10370 + }, + { + "epoch": 3.910633484162896, + "grad_norm": 0.09448584590101987, + "learning_rate": 1.1020928857685362e-05, + "loss": 0.4045, + "step": 10371 + }, + { + "epoch": 3.9110105580693815, + "grad_norm": 0.09573398469590878, + "learning_rate": 1.1013671617114565e-05, + "loss": 0.3861, + "step": 10372 + }, + { + "epoch": 3.911387631975867, + "grad_norm": 0.08995148924359252, + "learning_rate": 1.100641638527062e-05, + "loss": 0.4053, + "step": 10373 + }, + { + "epoch": 3.911764705882353, + "grad_norm": 0.09434970841635713, + "learning_rate": 1.0999163162656296e-05, + "loss": 0.3834, + "step": 10374 + }, + { + "epoch": 3.9121417797888385, + "grad_norm": 0.09834566488767826, + "learning_rate": 1.0991911949774243e-05, + "loss": 0.3913, + "step": 10375 + }, + { + "epoch": 3.912518853695324, + "grad_norm": 0.09120344748797088, + "learning_rate": 1.0984662747126959e-05, + "loss": 0.3909, + "step": 10376 + }, + { + "epoch": 3.91289592760181, + "grad_norm": 0.09790089541806223, + "learning_rate": 1.0977415555216826e-05, + "loss": 0.3992, + "step": 10377 + }, + { + "epoch": 3.9132730015082955, + "grad_norm": 0.08952019403332963, + "learning_rate": 1.0970170374546036e-05, + "loss": 0.3923, + "step": 10378 + }, + { + "epoch": 3.913650075414781, + "grad_norm": 0.09097791523399362, + "learning_rate": 1.096292720561669e-05, + "loss": 0.402, + "step": 10379 + }, + { + "epoch": 3.914027149321267, + "grad_norm": 0.0949717957179875, + "learning_rate": 1.0955686048930736e-05, + "loss": 0.4056, + "step": 10380 + }, + { + "epoch": 3.9144042232277525, + "grad_norm": 0.09275214276909372, + "learning_rate": 1.0948446904989978e-05, + "loss": 0.3787, + "step": 10381 + }, + { + "epoch": 3.914781297134238, + "grad_norm": 0.09695774528305152, + "learning_rate": 1.0941209774296077e-05, + "loss": 0.3872, + "step": 10382 + }, + { + "epoch": 3.915158371040724, + "grad_norm": 0.07990895488387498, + "learning_rate": 1.0933974657350577e-05, + "loss": 0.3937, + "step": 10383 + }, + { + "epoch": 3.9155354449472095, + "grad_norm": 0.09489745691537556, + "learning_rate": 1.092674155465483e-05, + "loss": 0.3964, + "step": 10384 + }, + { + "epoch": 3.915912518853695, + "grad_norm": 0.10109310735086682, + "learning_rate": 1.0919510466710124e-05, + "loss": 0.3907, + "step": 10385 + }, + { + "epoch": 3.916289592760181, + "grad_norm": 0.08945525274611098, + "learning_rate": 1.0912281394017538e-05, + "loss": 0.3883, + "step": 10386 + }, + { + "epoch": 3.9166666666666665, + "grad_norm": 0.09163370355517829, + "learning_rate": 1.0905054337078051e-05, + "loss": 0.3901, + "step": 10387 + }, + { + "epoch": 3.917043740573152, + "grad_norm": 0.08971028776297386, + "learning_rate": 1.0897829296392492e-05, + "loss": 0.3948, + "step": 10388 + }, + { + "epoch": 3.917420814479638, + "grad_norm": 0.08886956044190189, + "learning_rate": 1.0890606272461563e-05, + "loss": 0.3951, + "step": 10389 + }, + { + "epoch": 3.9177978883861235, + "grad_norm": 0.0854491800792241, + "learning_rate": 1.0883385265785771e-05, + "loss": 0.3714, + "step": 10390 + }, + { + "epoch": 3.918174962292609, + "grad_norm": 0.08795786279873256, + "learning_rate": 1.0876166276865581e-05, + "loss": 0.3897, + "step": 10391 + }, + { + "epoch": 3.918552036199095, + "grad_norm": 0.09122011151504107, + "learning_rate": 1.0868949306201224e-05, + "loss": 0.3897, + "step": 10392 + }, + { + "epoch": 3.9189291101055805, + "grad_norm": 0.08854734307617981, + "learning_rate": 1.086173435429284e-05, + "loss": 0.3888, + "step": 10393 + }, + { + "epoch": 3.919306184012066, + "grad_norm": 0.09373965734372286, + "learning_rate": 1.0854521421640416e-05, + "loss": 0.3943, + "step": 10394 + }, + { + "epoch": 3.919683257918552, + "grad_norm": 0.09715583742783325, + "learning_rate": 1.0847310508743822e-05, + "loss": 0.3994, + "step": 10395 + }, + { + "epoch": 3.9200603318250375, + "grad_norm": 0.09327785553984386, + "learning_rate": 1.0840101616102725e-05, + "loss": 0.4, + "step": 10396 + }, + { + "epoch": 3.920437405731523, + "grad_norm": 0.1025559535115379, + "learning_rate": 1.0832894744216747e-05, + "loss": 0.381, + "step": 10397 + }, + { + "epoch": 3.920814479638009, + "grad_norm": 0.08887969964459268, + "learning_rate": 1.0825689893585279e-05, + "loss": 0.392, + "step": 10398 + }, + { + "epoch": 3.9211915535444946, + "grad_norm": 0.0940458114637271, + "learning_rate": 1.081848706470762e-05, + "loss": 0.3974, + "step": 10399 + }, + { + "epoch": 3.9215686274509802, + "grad_norm": 0.09679752022789957, + "learning_rate": 1.0811286258082929e-05, + "loss": 0.4054, + "step": 10400 + }, + { + "epoch": 3.9219457013574663, + "grad_norm": 0.09618502235301346, + "learning_rate": 1.0804087474210218e-05, + "loss": 0.3851, + "step": 10401 + }, + { + "epoch": 3.922322775263952, + "grad_norm": 0.10592839205179928, + "learning_rate": 1.0796890713588328e-05, + "loss": 0.398, + "step": 10402 + }, + { + "epoch": 3.9226998491704377, + "grad_norm": 0.09392037323132191, + "learning_rate": 1.0789695976716028e-05, + "loss": 0.3894, + "step": 10403 + }, + { + "epoch": 3.9230769230769234, + "grad_norm": 0.105700327059053, + "learning_rate": 1.078250326409188e-05, + "loss": 0.3905, + "step": 10404 + }, + { + "epoch": 3.923453996983409, + "grad_norm": 0.09007852999810843, + "learning_rate": 1.0775312576214336e-05, + "loss": 0.3958, + "step": 10405 + }, + { + "epoch": 3.9238310708898947, + "grad_norm": 0.09578759730491292, + "learning_rate": 1.0768123913581711e-05, + "loss": 0.3902, + "step": 10406 + }, + { + "epoch": 3.9242081447963804, + "grad_norm": 0.09130085251337237, + "learning_rate": 1.076093727669218e-05, + "loss": 0.3961, + "step": 10407 + }, + { + "epoch": 3.924585218702866, + "grad_norm": 0.10274503624894799, + "learning_rate": 1.075375266604374e-05, + "loss": 0.3944, + "step": 10408 + }, + { + "epoch": 3.9249622926093517, + "grad_norm": 0.08689868026418013, + "learning_rate": 1.074657008213432e-05, + "loss": 0.4034, + "step": 10409 + }, + { + "epoch": 3.9253393665158374, + "grad_norm": 0.09822900542092283, + "learning_rate": 1.0739389525461635e-05, + "loss": 0.3969, + "step": 10410 + }, + { + "epoch": 3.925716440422323, + "grad_norm": 0.10261112758440716, + "learning_rate": 1.0732210996523303e-05, + "loss": 0.4084, + "step": 10411 + }, + { + "epoch": 3.9260935143288087, + "grad_norm": 0.11717979700794309, + "learning_rate": 1.0725034495816785e-05, + "loss": 0.4088, + "step": 10412 + }, + { + "epoch": 3.9264705882352944, + "grad_norm": 0.10374441330778084, + "learning_rate": 1.0717860023839424e-05, + "loss": 0.3952, + "step": 10413 + }, + { + "epoch": 3.92684766214178, + "grad_norm": 0.11442029327289799, + "learning_rate": 1.0710687581088361e-05, + "loss": 0.3909, + "step": 10414 + }, + { + "epoch": 3.9272247360482657, + "grad_norm": 0.09867610820210534, + "learning_rate": 1.0703517168060698e-05, + "loss": 0.3971, + "step": 10415 + }, + { + "epoch": 3.9276018099547514, + "grad_norm": 0.09703175046418426, + "learning_rate": 1.0696348785253297e-05, + "loss": 0.3949, + "step": 10416 + }, + { + "epoch": 3.927978883861237, + "grad_norm": 0.09429909279120614, + "learning_rate": 1.0689182433162934e-05, + "loss": 0.3923, + "step": 10417 + }, + { + "epoch": 3.9283559577677227, + "grad_norm": 0.09819270187800264, + "learning_rate": 1.0682018112286228e-05, + "loss": 0.3927, + "step": 10418 + }, + { + "epoch": 3.9287330316742084, + "grad_norm": 0.10580233004463444, + "learning_rate": 1.0674855823119673e-05, + "loss": 0.3953, + "step": 10419 + }, + { + "epoch": 3.929110105580694, + "grad_norm": 0.09784695356870188, + "learning_rate": 1.0667695566159577e-05, + "loss": 0.3906, + "step": 10420 + }, + { + "epoch": 3.9294871794871797, + "grad_norm": 0.10529602055436549, + "learning_rate": 1.0660537341902181e-05, + "loss": 0.3999, + "step": 10421 + }, + { + "epoch": 3.9298642533936654, + "grad_norm": 0.10781409300758264, + "learning_rate": 1.0653381150843512e-05, + "loss": 0.3898, + "step": 10422 + }, + { + "epoch": 3.930241327300151, + "grad_norm": 0.11307769544467525, + "learning_rate": 1.0646226993479498e-05, + "loss": 0.3974, + "step": 10423 + }, + { + "epoch": 3.9306184012066367, + "grad_norm": 0.1077049213417317, + "learning_rate": 1.063907487030592e-05, + "loss": 0.4007, + "step": 10424 + }, + { + "epoch": 3.9309954751131224, + "grad_norm": 0.10444366648381351, + "learning_rate": 1.0631924781818422e-05, + "loss": 0.3901, + "step": 10425 + }, + { + "epoch": 3.931372549019608, + "grad_norm": 0.09636502499808365, + "learning_rate": 1.0624776728512458e-05, + "loss": 0.4001, + "step": 10426 + }, + { + "epoch": 3.9317496229260938, + "grad_norm": 0.09761744752331332, + "learning_rate": 1.0617630710883441e-05, + "loss": 0.3851, + "step": 10427 + }, + { + "epoch": 3.9321266968325794, + "grad_norm": 0.08818037300047121, + "learning_rate": 1.0610486729426537e-05, + "loss": 0.3885, + "step": 10428 + }, + { + "epoch": 3.932503770739065, + "grad_norm": 0.09475283855021831, + "learning_rate": 1.060334478463684e-05, + "loss": 0.3834, + "step": 10429 + }, + { + "epoch": 3.9328808446455508, + "grad_norm": 0.09487549034459218, + "learning_rate": 1.059620487700927e-05, + "loss": 0.3884, + "step": 10430 + }, + { + "epoch": 3.9332579185520364, + "grad_norm": 0.08804859092931704, + "learning_rate": 1.0589067007038633e-05, + "loss": 0.3942, + "step": 10431 + }, + { + "epoch": 3.933634992458522, + "grad_norm": 0.09760400304855027, + "learning_rate": 1.0581931175219542e-05, + "loss": 0.3861, + "step": 10432 + }, + { + "epoch": 3.9340120663650078, + "grad_norm": 0.08997596561335144, + "learning_rate": 1.0574797382046551e-05, + "loss": 0.3899, + "step": 10433 + }, + { + "epoch": 3.9343891402714934, + "grad_norm": 0.09052049431155383, + "learning_rate": 1.056766562801399e-05, + "loss": 0.3837, + "step": 10434 + }, + { + "epoch": 3.934766214177979, + "grad_norm": 0.08927307164367103, + "learning_rate": 1.056053591361609e-05, + "loss": 0.3851, + "step": 10435 + }, + { + "epoch": 3.935143288084465, + "grad_norm": 0.09440324414175183, + "learning_rate": 1.0553408239346945e-05, + "loss": 0.4058, + "step": 10436 + }, + { + "epoch": 3.9355203619909505, + "grad_norm": 0.08994330138604555, + "learning_rate": 1.0546282605700498e-05, + "loss": 0.3813, + "step": 10437 + }, + { + "epoch": 3.935897435897436, + "grad_norm": 0.10601178357506688, + "learning_rate": 1.0539159013170517e-05, + "loss": 0.3935, + "step": 10438 + }, + { + "epoch": 3.936274509803922, + "grad_norm": 0.08715879472064784, + "learning_rate": 1.053203746225071e-05, + "loss": 0.3869, + "step": 10439 + }, + { + "epoch": 3.9366515837104075, + "grad_norm": 0.10384428698751554, + "learning_rate": 1.0524917953434554e-05, + "loss": 0.4012, + "step": 10440 + }, + { + "epoch": 3.937028657616893, + "grad_norm": 0.09225715108557457, + "learning_rate": 1.0517800487215441e-05, + "loss": 0.3879, + "step": 10441 + }, + { + "epoch": 3.937405731523379, + "grad_norm": 0.09824190959613271, + "learning_rate": 1.0510685064086607e-05, + "loss": 0.3872, + "step": 10442 + }, + { + "epoch": 3.9377828054298645, + "grad_norm": 0.09589287531848537, + "learning_rate": 1.0503571684541151e-05, + "loss": 0.3876, + "step": 10443 + }, + { + "epoch": 3.93815987933635, + "grad_norm": 0.09691762727101855, + "learning_rate": 1.0496460349071987e-05, + "loss": 0.3805, + "step": 10444 + }, + { + "epoch": 3.938536953242836, + "grad_norm": 0.09585439054111074, + "learning_rate": 1.0489351058171979e-05, + "loss": 0.3977, + "step": 10445 + }, + { + "epoch": 3.9389140271493215, + "grad_norm": 0.0962384435780592, + "learning_rate": 1.0482243812333755e-05, + "loss": 0.3832, + "step": 10446 + }, + { + "epoch": 3.939291101055807, + "grad_norm": 0.0829934109316325, + "learning_rate": 1.0475138612049846e-05, + "loss": 0.3842, + "step": 10447 + }, + { + "epoch": 3.939668174962293, + "grad_norm": 0.09495546409318743, + "learning_rate": 1.0468035457812649e-05, + "loss": 0.3859, + "step": 10448 + }, + { + "epoch": 3.9400452488687785, + "grad_norm": 0.0961647940362943, + "learning_rate": 1.046093435011441e-05, + "loss": 0.4025, + "step": 10449 + }, + { + "epoch": 3.940422322775264, + "grad_norm": 0.10275620535915644, + "learning_rate": 1.0453835289447198e-05, + "loss": 0.3953, + "step": 10450 + }, + { + "epoch": 3.94079939668175, + "grad_norm": 0.0928358163189397, + "learning_rate": 1.0446738276303013e-05, + "loss": 0.3798, + "step": 10451 + }, + { + "epoch": 3.9411764705882355, + "grad_norm": 0.09760138872803742, + "learning_rate": 1.0439643311173642e-05, + "loss": 0.3984, + "step": 10452 + }, + { + "epoch": 3.941553544494721, + "grad_norm": 0.08328620934797787, + "learning_rate": 1.043255039455077e-05, + "loss": 0.3814, + "step": 10453 + }, + { + "epoch": 3.941930618401207, + "grad_norm": 0.09485733696358113, + "learning_rate": 1.042545952692593e-05, + "loss": 0.3878, + "step": 10454 + }, + { + "epoch": 3.9423076923076925, + "grad_norm": 0.0891265223130506, + "learning_rate": 1.0418370708790522e-05, + "loss": 0.4019, + "step": 10455 + }, + { + "epoch": 3.942684766214178, + "grad_norm": 0.0882589872402316, + "learning_rate": 1.0411283940635761e-05, + "loss": 0.4026, + "step": 10456 + }, + { + "epoch": 3.943061840120664, + "grad_norm": 0.07889061395578369, + "learning_rate": 1.0404199222952802e-05, + "loss": 0.3874, + "step": 10457 + }, + { + "epoch": 3.9434389140271495, + "grad_norm": 0.09119542147637977, + "learning_rate": 1.0397116556232572e-05, + "loss": 0.3818, + "step": 10458 + }, + { + "epoch": 3.943815987933635, + "grad_norm": 0.08558002975844235, + "learning_rate": 1.0390035940965907e-05, + "loss": 0.3963, + "step": 10459 + }, + { + "epoch": 3.944193061840121, + "grad_norm": 0.08979357530375341, + "learning_rate": 1.038295737764349e-05, + "loss": 0.3968, + "step": 10460 + }, + { + "epoch": 3.9445701357466065, + "grad_norm": 0.09192506014381538, + "learning_rate": 1.0375880866755868e-05, + "loss": 0.3999, + "step": 10461 + }, + { + "epoch": 3.944947209653092, + "grad_norm": 0.09270893401715331, + "learning_rate": 1.0368806408793404e-05, + "loss": 0.3917, + "step": 10462 + }, + { + "epoch": 3.945324283559578, + "grad_norm": 0.08994388608741877, + "learning_rate": 1.036173400424639e-05, + "loss": 0.3841, + "step": 10463 + }, + { + "epoch": 3.9457013574660635, + "grad_norm": 0.09584705505204182, + "learning_rate": 1.0354663653604905e-05, + "loss": 0.3942, + "step": 10464 + }, + { + "epoch": 3.946078431372549, + "grad_norm": 0.08827399027981955, + "learning_rate": 1.034759535735895e-05, + "loss": 0.399, + "step": 10465 + }, + { + "epoch": 3.946455505279035, + "grad_norm": 0.09273994977046562, + "learning_rate": 1.0340529115998326e-05, + "loss": 0.3814, + "step": 10466 + }, + { + "epoch": 3.9468325791855206, + "grad_norm": 0.09220360972417656, + "learning_rate": 1.0333464930012727e-05, + "loss": 0.3961, + "step": 10467 + }, + { + "epoch": 3.947209653092006, + "grad_norm": 0.09272779177829492, + "learning_rate": 1.032640279989169e-05, + "loss": 0.4022, + "step": 10468 + }, + { + "epoch": 3.947586726998492, + "grad_norm": 0.08749466159055218, + "learning_rate": 1.0319342726124635e-05, + "loss": 0.3953, + "step": 10469 + }, + { + "epoch": 3.9479638009049776, + "grad_norm": 0.08706610040809082, + "learning_rate": 1.0312284709200777e-05, + "loss": 0.396, + "step": 10470 + }, + { + "epoch": 3.9483408748114632, + "grad_norm": 0.09648207832671354, + "learning_rate": 1.0305228749609282e-05, + "loss": 0.399, + "step": 10471 + }, + { + "epoch": 3.948717948717949, + "grad_norm": 0.08605851814372953, + "learning_rate": 1.0298174847839082e-05, + "loss": 0.4002, + "step": 10472 + }, + { + "epoch": 3.9490950226244346, + "grad_norm": 0.09388533980290621, + "learning_rate": 1.0291123004379018e-05, + "loss": 0.3898, + "step": 10473 + }, + { + "epoch": 3.9494720965309202, + "grad_norm": 0.08895648869979103, + "learning_rate": 1.0284073219717778e-05, + "loss": 0.3892, + "step": 10474 + }, + { + "epoch": 3.949849170437406, + "grad_norm": 0.08690242580889207, + "learning_rate": 1.027702549434392e-05, + "loss": 0.3905, + "step": 10475 + }, + { + "epoch": 3.9502262443438916, + "grad_norm": 0.09001115796528443, + "learning_rate": 1.0269979828745807e-05, + "loss": 0.397, + "step": 10476 + }, + { + "epoch": 3.9506033182503772, + "grad_norm": 0.0881667425739455, + "learning_rate": 1.0262936223411741e-05, + "loss": 0.3969, + "step": 10477 + }, + { + "epoch": 3.950980392156863, + "grad_norm": 0.09533433331311839, + "learning_rate": 1.025589467882981e-05, + "loss": 0.3862, + "step": 10478 + }, + { + "epoch": 3.9513574660633486, + "grad_norm": 0.09113837933420105, + "learning_rate": 1.0248855195487994e-05, + "loss": 0.3891, + "step": 10479 + }, + { + "epoch": 3.9517345399698343, + "grad_norm": 0.12089360602118139, + "learning_rate": 1.0241817773874123e-05, + "loss": 0.4076, + "step": 10480 + }, + { + "epoch": 3.95211161387632, + "grad_norm": 0.09281832214653042, + "learning_rate": 1.02347824144759e-05, + "loss": 0.3965, + "step": 10481 + }, + { + "epoch": 3.9524886877828056, + "grad_norm": 0.11161584042449227, + "learning_rate": 1.0227749117780825e-05, + "loss": 0.3897, + "step": 10482 + }, + { + "epoch": 3.9528657616892913, + "grad_norm": 0.10952765752493238, + "learning_rate": 1.0220717884276356e-05, + "loss": 0.3965, + "step": 10483 + }, + { + "epoch": 3.953242835595777, + "grad_norm": 0.1080021099991216, + "learning_rate": 1.0213688714449708e-05, + "loss": 0.3918, + "step": 10484 + }, + { + "epoch": 3.9536199095022626, + "grad_norm": 0.10280222411817018, + "learning_rate": 1.0206661608788014e-05, + "loss": 0.3949, + "step": 10485 + }, + { + "epoch": 3.9539969834087483, + "grad_norm": 0.08989138418387323, + "learning_rate": 1.0199636567778244e-05, + "loss": 0.3887, + "step": 10486 + }, + { + "epoch": 3.954374057315234, + "grad_norm": 0.12078056964184612, + "learning_rate": 1.0192613591907237e-05, + "loss": 0.4004, + "step": 10487 + }, + { + "epoch": 3.9547511312217196, + "grad_norm": 0.08710695110638403, + "learning_rate": 1.0185592681661647e-05, + "loss": 0.4007, + "step": 10488 + }, + { + "epoch": 3.9551282051282053, + "grad_norm": 0.11365402083616281, + "learning_rate": 1.0178573837528062e-05, + "loss": 0.3982, + "step": 10489 + }, + { + "epoch": 3.955505279034691, + "grad_norm": 0.087323837974946, + "learning_rate": 1.0171557059992847e-05, + "loss": 0.3946, + "step": 10490 + }, + { + "epoch": 3.9558823529411766, + "grad_norm": 0.11430534413389792, + "learning_rate": 1.0164542349542273e-05, + "loss": 0.3974, + "step": 10491 + }, + { + "epoch": 3.9562594268476623, + "grad_norm": 0.09358725685655195, + "learning_rate": 1.0157529706662444e-05, + "loss": 0.3981, + "step": 10492 + }, + { + "epoch": 3.956636500754148, + "grad_norm": 0.11117797775598581, + "learning_rate": 1.015051913183934e-05, + "loss": 0.3927, + "step": 10493 + }, + { + "epoch": 3.9570135746606336, + "grad_norm": 0.0871621991634891, + "learning_rate": 1.0143510625558783e-05, + "loss": 0.3909, + "step": 10494 + }, + { + "epoch": 3.9573906485671193, + "grad_norm": 0.09546728252435972, + "learning_rate": 1.0136504188306469e-05, + "loss": 0.3869, + "step": 10495 + }, + { + "epoch": 3.957767722473605, + "grad_norm": 0.0954618792909686, + "learning_rate": 1.012949982056791e-05, + "loss": 0.3876, + "step": 10496 + }, + { + "epoch": 3.9581447963800906, + "grad_norm": 0.09216365560458666, + "learning_rate": 1.0122497522828522e-05, + "loss": 0.4059, + "step": 10497 + }, + { + "epoch": 3.9585218702865763, + "grad_norm": 0.11268662965376292, + "learning_rate": 1.0115497295573555e-05, + "loss": 0.3878, + "step": 10498 + }, + { + "epoch": 3.958898944193062, + "grad_norm": 0.08438922572775799, + "learning_rate": 1.0108499139288118e-05, + "loss": 0.3941, + "step": 10499 + }, + { + "epoch": 3.9592760180995477, + "grad_norm": 0.10306967372324435, + "learning_rate": 1.0101503054457176e-05, + "loss": 0.3875, + "step": 10500 + }, + { + "epoch": 3.9596530920060333, + "grad_norm": 0.09590808836764124, + "learning_rate": 1.0094509041565566e-05, + "loss": 0.4042, + "step": 10501 + }, + { + "epoch": 3.960030165912519, + "grad_norm": 0.09891078960888193, + "learning_rate": 1.0087517101097939e-05, + "loss": 0.3975, + "step": 10502 + }, + { + "epoch": 3.9604072398190047, + "grad_norm": 0.0861555551753582, + "learning_rate": 1.0080527233538842e-05, + "loss": 0.4026, + "step": 10503 + }, + { + "epoch": 3.9607843137254903, + "grad_norm": 0.09259060169849195, + "learning_rate": 1.0073539439372672e-05, + "loss": 0.3853, + "step": 10504 + }, + { + "epoch": 3.961161387631976, + "grad_norm": 0.0958037322353927, + "learning_rate": 1.0066553719083677e-05, + "loss": 0.3883, + "step": 10505 + }, + { + "epoch": 3.9615384615384617, + "grad_norm": 0.09289502149483199, + "learning_rate": 1.0059570073155953e-05, + "loss": 0.3895, + "step": 10506 + }, + { + "epoch": 3.9619155354449473, + "grad_norm": 0.09174120832910294, + "learning_rate": 1.0052588502073482e-05, + "loss": 0.3809, + "step": 10507 + }, + { + "epoch": 3.962292609351433, + "grad_norm": 0.08755952315558131, + "learning_rate": 1.0045609006320052e-05, + "loss": 0.3864, + "step": 10508 + }, + { + "epoch": 3.9626696832579187, + "grad_norm": 0.09247357474608836, + "learning_rate": 1.0038631586379348e-05, + "loss": 0.3985, + "step": 10509 + }, + { + "epoch": 3.9630467571644044, + "grad_norm": 0.09254973334387276, + "learning_rate": 1.00316562427349e-05, + "loss": 0.4014, + "step": 10510 + }, + { + "epoch": 3.96342383107089, + "grad_norm": 0.09462566438089724, + "learning_rate": 1.0024682975870089e-05, + "loss": 0.3964, + "step": 10511 + }, + { + "epoch": 3.9638009049773757, + "grad_norm": 0.09271831422678582, + "learning_rate": 1.0017711786268163e-05, + "loss": 0.3912, + "step": 10512 + }, + { + "epoch": 3.9641779788838614, + "grad_norm": 0.09394477863871482, + "learning_rate": 1.0010742674412212e-05, + "loss": 0.3893, + "step": 10513 + }, + { + "epoch": 3.964555052790347, + "grad_norm": 0.09680814786252964, + "learning_rate": 1.0003775640785211e-05, + "loss": 0.4019, + "step": 10514 + }, + { + "epoch": 3.9649321266968327, + "grad_norm": 0.09893059684628923, + "learning_rate": 9.996810685869933e-06, + "loss": 0.4111, + "step": 10515 + }, + { + "epoch": 3.9653092006033184, + "grad_norm": 0.09322906297331568, + "learning_rate": 9.989847810149068e-06, + "loss": 0.393, + "step": 10516 + }, + { + "epoch": 3.965686274509804, + "grad_norm": 0.10936556616604715, + "learning_rate": 9.98288701410513e-06, + "loss": 0.4038, + "step": 10517 + }, + { + "epoch": 3.9660633484162897, + "grad_norm": 0.09830119449537328, + "learning_rate": 9.975928298220494e-06, + "loss": 0.3891, + "step": 10518 + }, + { + "epoch": 3.9664404223227754, + "grad_norm": 0.11402689841560626, + "learning_rate": 9.968971662977394e-06, + "loss": 0.3906, + "step": 10519 + }, + { + "epoch": 3.966817496229261, + "grad_norm": 0.09542779549182014, + "learning_rate": 9.962017108857935e-06, + "loss": 0.3948, + "step": 10520 + }, + { + "epoch": 3.9671945701357467, + "grad_norm": 0.10228583239715268, + "learning_rate": 9.955064636344036e-06, + "loss": 0.3834, + "step": 10521 + }, + { + "epoch": 3.9675716440422324, + "grad_norm": 0.09824908996731292, + "learning_rate": 9.948114245917506e-06, + "loss": 0.3949, + "step": 10522 + }, + { + "epoch": 3.967948717948718, + "grad_norm": 0.09297590712040449, + "learning_rate": 9.94116593806e-06, + "loss": 0.3779, + "step": 10523 + }, + { + "epoch": 3.9683257918552037, + "grad_norm": 0.10278827887725447, + "learning_rate": 9.934219713253039e-06, + "loss": 0.386, + "step": 10524 + }, + { + "epoch": 3.9687028657616894, + "grad_norm": 0.08730348269194588, + "learning_rate": 9.927275571977977e-06, + "loss": 0.3942, + "step": 10525 + }, + { + "epoch": 3.969079939668175, + "grad_norm": 0.09571857116259208, + "learning_rate": 9.920333514716058e-06, + "loss": 0.3895, + "step": 10526 + }, + { + "epoch": 3.9694570135746607, + "grad_norm": 0.08994558342816947, + "learning_rate": 9.913393541948336e-06, + "loss": 0.3959, + "step": 10527 + }, + { + "epoch": 3.9698340874811464, + "grad_norm": 0.0973294518651194, + "learning_rate": 9.906455654155752e-06, + "loss": 0.398, + "step": 10528 + }, + { + "epoch": 3.970211161387632, + "grad_norm": 0.08388577868500253, + "learning_rate": 9.8995198518191e-06, + "loss": 0.3918, + "step": 10529 + }, + { + "epoch": 3.9705882352941178, + "grad_norm": 0.1006894512201068, + "learning_rate": 9.892586135419022e-06, + "loss": 0.4017, + "step": 10530 + }, + { + "epoch": 3.9709653092006034, + "grad_norm": 0.09084725399878736, + "learning_rate": 9.88565450543602e-06, + "loss": 0.3921, + "step": 10531 + }, + { + "epoch": 3.971342383107089, + "grad_norm": 0.09506040762760362, + "learning_rate": 9.878724962350459e-06, + "loss": 0.4015, + "step": 10532 + }, + { + "epoch": 3.9717194570135748, + "grad_norm": 0.09433505058174212, + "learning_rate": 9.87179750664252e-06, + "loss": 0.4003, + "step": 10533 + }, + { + "epoch": 3.9720965309200604, + "grad_norm": 0.08919405599602348, + "learning_rate": 9.864872138792313e-06, + "loss": 0.403, + "step": 10534 + }, + { + "epoch": 3.972473604826546, + "grad_norm": 0.09660335751771852, + "learning_rate": 9.857948859279726e-06, + "loss": 0.3947, + "step": 10535 + }, + { + "epoch": 3.9728506787330318, + "grad_norm": 0.08762495323429623, + "learning_rate": 9.851027668584545e-06, + "loss": 0.3834, + "step": 10536 + }, + { + "epoch": 3.9732277526395174, + "grad_norm": 0.10063947390072583, + "learning_rate": 9.844108567186405e-06, + "loss": 0.4139, + "step": 10537 + }, + { + "epoch": 3.973604826546003, + "grad_norm": 0.10927672618469916, + "learning_rate": 9.837191555564805e-06, + "loss": 0.3924, + "step": 10538 + }, + { + "epoch": 3.973981900452489, + "grad_norm": 0.08951562672016838, + "learning_rate": 9.830276634199052e-06, + "loss": 0.3906, + "step": 10539 + }, + { + "epoch": 3.9743589743589745, + "grad_norm": 0.1105778457583538, + "learning_rate": 9.823363803568395e-06, + "loss": 0.3859, + "step": 10540 + }, + { + "epoch": 3.97473604826546, + "grad_norm": 0.0961359348007651, + "learning_rate": 9.816453064151842e-06, + "loss": 0.3873, + "step": 10541 + }, + { + "epoch": 3.975113122171946, + "grad_norm": 0.08846935632198755, + "learning_rate": 9.809544416428323e-06, + "loss": 0.3964, + "step": 10542 + }, + { + "epoch": 3.9754901960784315, + "grad_norm": 0.10220834723358695, + "learning_rate": 9.802637860876594e-06, + "loss": 0.3989, + "step": 10543 + }, + { + "epoch": 3.975867269984917, + "grad_norm": 0.09623859964928408, + "learning_rate": 9.795733397975292e-06, + "loss": 0.3962, + "step": 10544 + }, + { + "epoch": 3.976244343891403, + "grad_norm": 0.094210994343403, + "learning_rate": 9.78883102820285e-06, + "loss": 0.3976, + "step": 10545 + }, + { + "epoch": 3.9766214177978885, + "grad_norm": 0.10460685466190531, + "learning_rate": 9.781930752037642e-06, + "loss": 0.3922, + "step": 10546 + }, + { + "epoch": 3.976998491704374, + "grad_norm": 0.08875273888082628, + "learning_rate": 9.775032569957816e-06, + "loss": 0.392, + "step": 10547 + }, + { + "epoch": 3.97737556561086, + "grad_norm": 0.10456717978183647, + "learning_rate": 9.768136482441428e-06, + "loss": 0.3905, + "step": 10548 + }, + { + "epoch": 3.9777526395173455, + "grad_norm": 0.08123418088603154, + "learning_rate": 9.761242489966363e-06, + "loss": 0.3827, + "step": 10549 + }, + { + "epoch": 3.978129713423831, + "grad_norm": 0.09051583331915791, + "learning_rate": 9.754350593010385e-06, + "loss": 0.3944, + "step": 10550 + }, + { + "epoch": 3.978506787330317, + "grad_norm": 0.09392110772079233, + "learning_rate": 9.747460792051058e-06, + "loss": 0.3992, + "step": 10551 + }, + { + "epoch": 3.9788838612368025, + "grad_norm": 0.08647659622034412, + "learning_rate": 9.74057308756589e-06, + "loss": 0.3972, + "step": 10552 + }, + { + "epoch": 3.979260935143288, + "grad_norm": 0.10591421606180629, + "learning_rate": 9.733687480032153e-06, + "loss": 0.398, + "step": 10553 + }, + { + "epoch": 3.979638009049774, + "grad_norm": 0.08924307714315363, + "learning_rate": 9.726803969927028e-06, + "loss": 0.3975, + "step": 10554 + }, + { + "epoch": 3.9800150829562595, + "grad_norm": 0.12301259889288048, + "learning_rate": 9.719922557727535e-06, + "loss": 0.4126, + "step": 10555 + }, + { + "epoch": 3.980392156862745, + "grad_norm": 0.09615387163387662, + "learning_rate": 9.713043243910563e-06, + "loss": 0.401, + "step": 10556 + }, + { + "epoch": 3.980769230769231, + "grad_norm": 0.10540474943715902, + "learning_rate": 9.706166028952807e-06, + "loss": 0.3992, + "step": 10557 + }, + { + "epoch": 3.9811463046757165, + "grad_norm": 0.09335872506545645, + "learning_rate": 9.699290913330897e-06, + "loss": 0.3915, + "step": 10558 + }, + { + "epoch": 3.981523378582202, + "grad_norm": 0.10056305151400331, + "learning_rate": 9.692417897521236e-06, + "loss": 0.4005, + "step": 10559 + }, + { + "epoch": 3.981900452488688, + "grad_norm": 0.10103692351337192, + "learning_rate": 9.685546982000136e-06, + "loss": 0.3988, + "step": 10560 + }, + { + "epoch": 3.9822775263951735, + "grad_norm": 0.10191477124654803, + "learning_rate": 9.678678167243736e-06, + "loss": 0.3824, + "step": 10561 + }, + { + "epoch": 3.982654600301659, + "grad_norm": 0.09165677742307086, + "learning_rate": 9.67181145372806e-06, + "loss": 0.4031, + "step": 10562 + }, + { + "epoch": 3.983031674208145, + "grad_norm": 0.10683719453982317, + "learning_rate": 9.664946841928925e-06, + "loss": 0.3882, + "step": 10563 + }, + { + "epoch": 3.9834087481146305, + "grad_norm": 0.09070162583608346, + "learning_rate": 9.658084332322092e-06, + "loss": 0.3935, + "step": 10564 + }, + { + "epoch": 3.983785822021116, + "grad_norm": 0.1022874673034916, + "learning_rate": 9.651223925383086e-06, + "loss": 0.395, + "step": 10565 + }, + { + "epoch": 3.984162895927602, + "grad_norm": 0.09636042557063161, + "learning_rate": 9.644365621587344e-06, + "loss": 0.3873, + "step": 10566 + }, + { + "epoch": 3.9845399698340875, + "grad_norm": 0.10187015507762993, + "learning_rate": 9.637509421410143e-06, + "loss": 0.407, + "step": 10567 + }, + { + "epoch": 3.984917043740573, + "grad_norm": 0.10132597949141346, + "learning_rate": 9.630655325326615e-06, + "loss": 0.4037, + "step": 10568 + }, + { + "epoch": 3.985294117647059, + "grad_norm": 0.09455290800054393, + "learning_rate": 9.623803333811713e-06, + "loss": 0.3944, + "step": 10569 + }, + { + "epoch": 3.9856711915535445, + "grad_norm": 0.08911528416008024, + "learning_rate": 9.616953447340318e-06, + "loss": 0.3907, + "step": 10570 + }, + { + "epoch": 3.98604826546003, + "grad_norm": 0.09062629543080818, + "learning_rate": 9.610105666387088e-06, + "loss": 0.4016, + "step": 10571 + }, + { + "epoch": 3.986425339366516, + "grad_norm": 0.08999602818163076, + "learning_rate": 9.60325999142658e-06, + "loss": 0.4117, + "step": 10572 + }, + { + "epoch": 3.9868024132730016, + "grad_norm": 0.10282193101506651, + "learning_rate": 9.596416422933195e-06, + "loss": 0.3972, + "step": 10573 + }, + { + "epoch": 3.9871794871794872, + "grad_norm": 0.09854394008085895, + "learning_rate": 9.589574961381193e-06, + "loss": 0.4008, + "step": 10574 + }, + { + "epoch": 3.987556561085973, + "grad_norm": 0.11152385625141623, + "learning_rate": 9.582735607244653e-06, + "loss": 0.3913, + "step": 10575 + }, + { + "epoch": 3.9879336349924586, + "grad_norm": 0.08991066855077581, + "learning_rate": 9.575898360997575e-06, + "loss": 0.3918, + "step": 10576 + }, + { + "epoch": 3.9883107088989442, + "grad_norm": 0.10965120386703665, + "learning_rate": 9.569063223113742e-06, + "loss": 0.3962, + "step": 10577 + }, + { + "epoch": 3.98868778280543, + "grad_norm": 0.10379681171054657, + "learning_rate": 9.562230194066839e-06, + "loss": 0.3885, + "step": 10578 + }, + { + "epoch": 3.9890648567119156, + "grad_norm": 0.10045102229829479, + "learning_rate": 9.555399274330384e-06, + "loss": 0.3853, + "step": 10579 + }, + { + "epoch": 3.9894419306184012, + "grad_norm": 0.11906246735310246, + "learning_rate": 9.548570464377765e-06, + "loss": 0.3967, + "step": 10580 + }, + { + "epoch": 3.989819004524887, + "grad_norm": 0.08293869464251538, + "learning_rate": 9.541743764682176e-06, + "loss": 0.3889, + "step": 10581 + }, + { + "epoch": 3.9901960784313726, + "grad_norm": 0.11177956813098772, + "learning_rate": 9.534919175716757e-06, + "loss": 0.4013, + "step": 10582 + }, + { + "epoch": 3.9905731523378583, + "grad_norm": 0.09440718352074173, + "learning_rate": 9.528096697954403e-06, + "loss": 0.3877, + "step": 10583 + }, + { + "epoch": 3.990950226244344, + "grad_norm": 0.08678021403155503, + "learning_rate": 9.52127633186792e-06, + "loss": 0.3923, + "step": 10584 + }, + { + "epoch": 3.9913273001508296, + "grad_norm": 0.09356429840471456, + "learning_rate": 9.514458077929949e-06, + "loss": 0.3902, + "step": 10585 + }, + { + "epoch": 3.9917043740573153, + "grad_norm": 0.09222955197848669, + "learning_rate": 9.507641936613008e-06, + "loss": 0.399, + "step": 10586 + }, + { + "epoch": 3.992081447963801, + "grad_norm": 0.08570477450975607, + "learning_rate": 9.500827908389411e-06, + "loss": 0.3896, + "step": 10587 + }, + { + "epoch": 3.9924585218702866, + "grad_norm": 0.08743795575695526, + "learning_rate": 9.494015993731409e-06, + "loss": 0.3944, + "step": 10588 + }, + { + "epoch": 3.9928355957767723, + "grad_norm": 0.09410768418475657, + "learning_rate": 9.487206193111014e-06, + "loss": 0.3911, + "step": 10589 + }, + { + "epoch": 3.993212669683258, + "grad_norm": 0.08534288843259151, + "learning_rate": 9.480398507000194e-06, + "loss": 0.4025, + "step": 10590 + }, + { + "epoch": 3.9935897435897436, + "grad_norm": 0.08529798171213468, + "learning_rate": 9.473592935870672e-06, + "loss": 0.3969, + "step": 10591 + }, + { + "epoch": 3.9939668174962293, + "grad_norm": 0.09507218479514375, + "learning_rate": 9.466789480194083e-06, + "loss": 0.39, + "step": 10592 + }, + { + "epoch": 3.994343891402715, + "grad_norm": 0.08777411535819325, + "learning_rate": 9.459988140441898e-06, + "loss": 0.3969, + "step": 10593 + }, + { + "epoch": 3.9947209653092006, + "grad_norm": 0.08372061144922247, + "learning_rate": 9.45318891708546e-06, + "loss": 0.397, + "step": 10594 + }, + { + "epoch": 3.9950980392156863, + "grad_norm": 0.09060105175834468, + "learning_rate": 9.44639181059591e-06, + "loss": 0.3906, + "step": 10595 + }, + { + "epoch": 3.995475113122172, + "grad_norm": 0.09037533935302547, + "learning_rate": 9.43959682144433e-06, + "loss": 0.3941, + "step": 10596 + }, + { + "epoch": 3.9958521870286576, + "grad_norm": 0.08898747518643763, + "learning_rate": 9.432803950101572e-06, + "loss": 0.4062, + "step": 10597 + }, + { + "epoch": 3.9962292609351433, + "grad_norm": 0.08417178583877394, + "learning_rate": 9.426013197038388e-06, + "loss": 0.3923, + "step": 10598 + }, + { + "epoch": 3.996606334841629, + "grad_norm": 0.08407505328996219, + "learning_rate": 9.419224562725371e-06, + "loss": 0.4001, + "step": 10599 + }, + { + "epoch": 3.9969834087481146, + "grad_norm": 0.0884363759492116, + "learning_rate": 9.41243804763298e-06, + "loss": 0.3989, + "step": 10600 + }, + { + "epoch": 3.9973604826546003, + "grad_norm": 0.07964907623603322, + "learning_rate": 9.405653652231472e-06, + "loss": 0.3824, + "step": 10601 + }, + { + "epoch": 3.997737556561086, + "grad_norm": 0.08752815741194396, + "learning_rate": 9.398871376991056e-06, + "loss": 0.3864, + "step": 10602 + }, + { + "epoch": 3.9981146304675717, + "grad_norm": 0.08957530753289616, + "learning_rate": 9.392091222381698e-06, + "loss": 0.3905, + "step": 10603 + }, + { + "epoch": 3.9984917043740573, + "grad_norm": 0.08618602667482593, + "learning_rate": 9.38531318887327e-06, + "loss": 0.3913, + "step": 10604 + }, + { + "epoch": 3.998868778280543, + "grad_norm": 0.09022292441194726, + "learning_rate": 9.378537276935482e-06, + "loss": 0.3987, + "step": 10605 + }, + { + "epoch": 3.9992458521870287, + "grad_norm": 0.08633300227502362, + "learning_rate": 9.371763487037908e-06, + "loss": 0.398, + "step": 10606 + }, + { + "epoch": 3.9996229260935143, + "grad_norm": 0.08949835609355714, + "learning_rate": 9.364991819649937e-06, + "loss": 0.401, + "step": 10607 + }, + { + "epoch": 4.0, + "grad_norm": 0.1006281680360939, + "learning_rate": 9.358222275240884e-06, + "loss": 0.3897, + "step": 10608 + }, + { + "epoch": 4.000377073906486, + "grad_norm": 0.10450602607208405, + "learning_rate": 9.351454854279835e-06, + "loss": 0.3817, + "step": 10609 + }, + { + "epoch": 4.000754147812971, + "grad_norm": 0.10497489712010172, + "learning_rate": 9.34468955723578e-06, + "loss": 0.3832, + "step": 10610 + }, + { + "epoch": 4.001131221719457, + "grad_norm": 0.10431016573159095, + "learning_rate": 9.337926384577548e-06, + "loss": 0.3781, + "step": 10611 + }, + { + "epoch": 4.001508295625943, + "grad_norm": 0.10296914968895107, + "learning_rate": 9.331165336773829e-06, + "loss": 0.3861, + "step": 10612 + }, + { + "epoch": 4.001885369532428, + "grad_norm": 0.10854930751488827, + "learning_rate": 9.32440641429313e-06, + "loss": 0.3828, + "step": 10613 + }, + { + "epoch": 4.002262443438914, + "grad_norm": 0.10681616350210608, + "learning_rate": 9.317649617603876e-06, + "loss": 0.3837, + "step": 10614 + }, + { + "epoch": 4.0026395173454, + "grad_norm": 0.10377877917636762, + "learning_rate": 9.310894947174281e-06, + "loss": 0.3828, + "step": 10615 + }, + { + "epoch": 4.003016591251885, + "grad_norm": 0.09671600602435976, + "learning_rate": 9.304142403472443e-06, + "loss": 0.3739, + "step": 10616 + }, + { + "epoch": 4.003393665158371, + "grad_norm": 0.11013936208966132, + "learning_rate": 9.29739198696631e-06, + "loss": 0.3706, + "step": 10617 + }, + { + "epoch": 4.003770739064857, + "grad_norm": 0.09214701396262638, + "learning_rate": 9.290643698123688e-06, + "loss": 0.3788, + "step": 10618 + }, + { + "epoch": 4.004147812971342, + "grad_norm": 0.10031729066703551, + "learning_rate": 9.2838975374122e-06, + "loss": 0.3805, + "step": 10619 + }, + { + "epoch": 4.004524886877828, + "grad_norm": 0.09603016498651684, + "learning_rate": 9.277153505299385e-06, + "loss": 0.3838, + "step": 10620 + }, + { + "epoch": 4.004901960784314, + "grad_norm": 0.10813550754943654, + "learning_rate": 9.270411602252572e-06, + "loss": 0.3838, + "step": 10621 + }, + { + "epoch": 4.005279034690799, + "grad_norm": 0.10561403509637331, + "learning_rate": 9.263671828738978e-06, + "loss": 0.3927, + "step": 10622 + }, + { + "epoch": 4.005656108597285, + "grad_norm": 0.09650378773568086, + "learning_rate": 9.256934185225664e-06, + "loss": 0.3809, + "step": 10623 + }, + { + "epoch": 4.006033182503771, + "grad_norm": 0.10432514346556584, + "learning_rate": 9.250198672179551e-06, + "loss": 0.3936, + "step": 10624 + }, + { + "epoch": 4.006410256410256, + "grad_norm": 0.10158888330198491, + "learning_rate": 9.243465290067374e-06, + "loss": 0.3743, + "step": 10625 + }, + { + "epoch": 4.006787330316742, + "grad_norm": 0.09697438853314039, + "learning_rate": 9.236734039355797e-06, + "loss": 0.3822, + "step": 10626 + }, + { + "epoch": 4.007164404223228, + "grad_norm": 0.09917549478651938, + "learning_rate": 9.230004920511249e-06, + "loss": 0.376, + "step": 10627 + }, + { + "epoch": 4.007541478129713, + "grad_norm": 0.09870775365104015, + "learning_rate": 9.223277934000067e-06, + "loss": 0.3796, + "step": 10628 + }, + { + "epoch": 4.007918552036199, + "grad_norm": 0.10439830215650125, + "learning_rate": 9.21655308028843e-06, + "loss": 0.3848, + "step": 10629 + }, + { + "epoch": 4.008295625942685, + "grad_norm": 0.10898842394898106, + "learning_rate": 9.209830359842362e-06, + "loss": 0.3803, + "step": 10630 + }, + { + "epoch": 4.00867269984917, + "grad_norm": 0.0909843969721473, + "learning_rate": 9.203109773127723e-06, + "loss": 0.3688, + "step": 10631 + }, + { + "epoch": 4.009049773755656, + "grad_norm": 0.10987803439694357, + "learning_rate": 9.196391320610281e-06, + "loss": 0.3727, + "step": 10632 + }, + { + "epoch": 4.009426847662142, + "grad_norm": 0.08955346029598306, + "learning_rate": 9.189675002755587e-06, + "loss": 0.3696, + "step": 10633 + }, + { + "epoch": 4.009803921568627, + "grad_norm": 0.11592378033879007, + "learning_rate": 9.182960820029087e-06, + "loss": 0.3721, + "step": 10634 + }, + { + "epoch": 4.010180995475113, + "grad_norm": 0.08638236476889037, + "learning_rate": 9.176248772896063e-06, + "loss": 0.3929, + "step": 10635 + }, + { + "epoch": 4.010558069381599, + "grad_norm": 0.11006334945030553, + "learning_rate": 9.169538861821676e-06, + "loss": 0.3839, + "step": 10636 + }, + { + "epoch": 4.010935143288084, + "grad_norm": 0.10424383413853659, + "learning_rate": 9.162831087270873e-06, + "loss": 0.3748, + "step": 10637 + }, + { + "epoch": 4.01131221719457, + "grad_norm": 0.09255562437147537, + "learning_rate": 9.156125449708545e-06, + "loss": 0.376, + "step": 10638 + }, + { + "epoch": 4.011689291101056, + "grad_norm": 0.10562595088890962, + "learning_rate": 9.149421949599357e-06, + "loss": 0.3837, + "step": 10639 + }, + { + "epoch": 4.012066365007541, + "grad_norm": 0.0922061529558753, + "learning_rate": 9.14272058740786e-06, + "loss": 0.3898, + "step": 10640 + }, + { + "epoch": 4.012443438914027, + "grad_norm": 0.09322436088204485, + "learning_rate": 9.136021363598462e-06, + "loss": 0.3894, + "step": 10641 + }, + { + "epoch": 4.012820512820513, + "grad_norm": 0.1046190086756837, + "learning_rate": 9.1293242786354e-06, + "loss": 0.384, + "step": 10642 + }, + { + "epoch": 4.0131975867269984, + "grad_norm": 0.09419426305903093, + "learning_rate": 9.12262933298279e-06, + "loss": 0.3752, + "step": 10643 + }, + { + "epoch": 4.013574660633484, + "grad_norm": 0.10153376144818746, + "learning_rate": 9.11593652710458e-06, + "loss": 0.3806, + "step": 10644 + }, + { + "epoch": 4.01395173453997, + "grad_norm": 0.0872122136497914, + "learning_rate": 9.109245861464573e-06, + "loss": 0.3768, + "step": 10645 + }, + { + "epoch": 4.0143288084464555, + "grad_norm": 0.09444340043658032, + "learning_rate": 9.102557336526426e-06, + "loss": 0.3787, + "step": 10646 + }, + { + "epoch": 4.014705882352941, + "grad_norm": 0.08373424095195524, + "learning_rate": 9.095870952753647e-06, + "loss": 0.3785, + "step": 10647 + }, + { + "epoch": 4.015082956259427, + "grad_norm": 0.09115464856047925, + "learning_rate": 9.089186710609596e-06, + "loss": 0.3666, + "step": 10648 + }, + { + "epoch": 4.0154600301659125, + "grad_norm": 0.0847886820130937, + "learning_rate": 9.082504610557494e-06, + "loss": 0.3806, + "step": 10649 + }, + { + "epoch": 4.015837104072398, + "grad_norm": 0.08789314441433226, + "learning_rate": 9.07582465306041e-06, + "loss": 0.3743, + "step": 10650 + }, + { + "epoch": 4.016214177978884, + "grad_norm": 0.0949679263871609, + "learning_rate": 9.069146838581236e-06, + "loss": 0.3794, + "step": 10651 + }, + { + "epoch": 4.0165912518853695, + "grad_norm": 0.09148954540530904, + "learning_rate": 9.062471167582747e-06, + "loss": 0.3845, + "step": 10652 + }, + { + "epoch": 4.016968325791855, + "grad_norm": 0.0835768732668276, + "learning_rate": 9.055797640527566e-06, + "loss": 0.3768, + "step": 10653 + }, + { + "epoch": 4.017345399698341, + "grad_norm": 0.09342994244946373, + "learning_rate": 9.049126257878163e-06, + "loss": 0.3755, + "step": 10654 + }, + { + "epoch": 4.0177224736048265, + "grad_norm": 0.0881539303088516, + "learning_rate": 9.042457020096855e-06, + "loss": 0.3894, + "step": 10655 + }, + { + "epoch": 4.018099547511312, + "grad_norm": 0.07968221793518772, + "learning_rate": 9.035789927645826e-06, + "loss": 0.3841, + "step": 10656 + }, + { + "epoch": 4.018476621417798, + "grad_norm": 0.0898227785203786, + "learning_rate": 9.029124980987083e-06, + "loss": 0.3739, + "step": 10657 + }, + { + "epoch": 4.0188536953242835, + "grad_norm": 0.08699802899892299, + "learning_rate": 9.022462180582505e-06, + "loss": 0.3773, + "step": 10658 + }, + { + "epoch": 4.019230769230769, + "grad_norm": 0.0877069501410759, + "learning_rate": 9.015801526893821e-06, + "loss": 0.376, + "step": 10659 + }, + { + "epoch": 4.019607843137255, + "grad_norm": 0.09440215344720462, + "learning_rate": 9.009143020382609e-06, + "loss": 0.3856, + "step": 10660 + }, + { + "epoch": 4.0199849170437405, + "grad_norm": 0.08602689985119648, + "learning_rate": 9.002486661510295e-06, + "loss": 0.3695, + "step": 10661 + }, + { + "epoch": 4.020361990950226, + "grad_norm": 0.09787637641746928, + "learning_rate": 8.995832450738163e-06, + "loss": 0.3922, + "step": 10662 + }, + { + "epoch": 4.020739064856712, + "grad_norm": 0.09734934353328743, + "learning_rate": 8.98918038852735e-06, + "loss": 0.3837, + "step": 10663 + }, + { + "epoch": 4.0211161387631975, + "grad_norm": 0.08229901438573509, + "learning_rate": 8.982530475338818e-06, + "loss": 0.3662, + "step": 10664 + }, + { + "epoch": 4.021493212669683, + "grad_norm": 0.10590849452905228, + "learning_rate": 8.975882711633415e-06, + "loss": 0.3819, + "step": 10665 + }, + { + "epoch": 4.021870286576169, + "grad_norm": 0.08434664764849999, + "learning_rate": 8.969237097871821e-06, + "loss": 0.3822, + "step": 10666 + }, + { + "epoch": 4.0222473604826545, + "grad_norm": 0.1031470198958896, + "learning_rate": 8.962593634514572e-06, + "loss": 0.3741, + "step": 10667 + }, + { + "epoch": 4.02262443438914, + "grad_norm": 0.09983996589746424, + "learning_rate": 8.955952322022052e-06, + "loss": 0.3823, + "step": 10668 + }, + { + "epoch": 4.023001508295626, + "grad_norm": 0.09726561508836262, + "learning_rate": 8.94931316085451e-06, + "loss": 0.3774, + "step": 10669 + }, + { + "epoch": 4.0233785822021115, + "grad_norm": 0.1066018305013229, + "learning_rate": 8.942676151472015e-06, + "loss": 0.3937, + "step": 10670 + }, + { + "epoch": 4.023755656108597, + "grad_norm": 0.08571610874565366, + "learning_rate": 8.936041294334514e-06, + "loss": 0.3703, + "step": 10671 + }, + { + "epoch": 4.024132730015083, + "grad_norm": 0.09812357087397491, + "learning_rate": 8.929408589901803e-06, + "loss": 0.3639, + "step": 10672 + }, + { + "epoch": 4.0245098039215685, + "grad_norm": 0.09368171979246759, + "learning_rate": 8.92277803863351e-06, + "loss": 0.3794, + "step": 10673 + }, + { + "epoch": 4.024886877828054, + "grad_norm": 0.09032693387360143, + "learning_rate": 8.916149640989139e-06, + "loss": 0.388, + "step": 10674 + }, + { + "epoch": 4.02526395173454, + "grad_norm": 0.08816331046956011, + "learning_rate": 8.909523397428033e-06, + "loss": 0.3775, + "step": 10675 + }, + { + "epoch": 4.0256410256410255, + "grad_norm": 0.08079220252185622, + "learning_rate": 8.90289930840937e-06, + "loss": 0.368, + "step": 10676 + }, + { + "epoch": 4.026018099547511, + "grad_norm": 0.085846390637627, + "learning_rate": 8.8962773743922e-06, + "loss": 0.3837, + "step": 10677 + }, + { + "epoch": 4.026395173453997, + "grad_norm": 0.08056114714569179, + "learning_rate": 8.889657595835418e-06, + "loss": 0.3801, + "step": 10678 + }, + { + "epoch": 4.026772247360483, + "grad_norm": 0.08622218246734094, + "learning_rate": 8.88303997319777e-06, + "loss": 0.3747, + "step": 10679 + }, + { + "epoch": 4.027149321266968, + "grad_norm": 0.0846677303579018, + "learning_rate": 8.876424506937854e-06, + "loss": 0.3832, + "step": 10680 + }, + { + "epoch": 4.027526395173454, + "grad_norm": 0.0871066179788682, + "learning_rate": 8.869811197514124e-06, + "loss": 0.3816, + "step": 10681 + }, + { + "epoch": 4.02790346907994, + "grad_norm": 0.09466941591726254, + "learning_rate": 8.863200045384844e-06, + "loss": 0.3778, + "step": 10682 + }, + { + "epoch": 4.028280542986425, + "grad_norm": 0.08414247364425631, + "learning_rate": 8.856591051008202e-06, + "loss": 0.3731, + "step": 10683 + }, + { + "epoch": 4.028657616892911, + "grad_norm": 0.09160729911104055, + "learning_rate": 8.849984214842165e-06, + "loss": 0.3785, + "step": 10684 + }, + { + "epoch": 4.029034690799397, + "grad_norm": 0.08185417665263557, + "learning_rate": 8.843379537344595e-06, + "loss": 0.3688, + "step": 10685 + }, + { + "epoch": 4.029411764705882, + "grad_norm": 0.10244961923779118, + "learning_rate": 8.83677701897318e-06, + "loss": 0.3718, + "step": 10686 + }, + { + "epoch": 4.029788838612368, + "grad_norm": 0.07981166283500886, + "learning_rate": 8.830176660185494e-06, + "loss": 0.3765, + "step": 10687 + }, + { + "epoch": 4.030165912518854, + "grad_norm": 0.09284941972502962, + "learning_rate": 8.823578461438895e-06, + "loss": 0.3895, + "step": 10688 + }, + { + "epoch": 4.030542986425339, + "grad_norm": 0.09849525501373865, + "learning_rate": 8.816982423190677e-06, + "loss": 0.3784, + "step": 10689 + }, + { + "epoch": 4.030920060331825, + "grad_norm": 0.08896172035636796, + "learning_rate": 8.810388545897903e-06, + "loss": 0.378, + "step": 10690 + }, + { + "epoch": 4.031297134238311, + "grad_norm": 0.09971096822146908, + "learning_rate": 8.803796830017539e-06, + "loss": 0.3716, + "step": 10691 + }, + { + "epoch": 4.031674208144796, + "grad_norm": 0.08231171326595861, + "learning_rate": 8.797207276006387e-06, + "loss": 0.38, + "step": 10692 + }, + { + "epoch": 4.032051282051282, + "grad_norm": 0.10307688431267788, + "learning_rate": 8.790619884321102e-06, + "loss": 0.3702, + "step": 10693 + }, + { + "epoch": 4.032428355957768, + "grad_norm": 0.082229466797875, + "learning_rate": 8.784034655418154e-06, + "loss": 0.3848, + "step": 10694 + }, + { + "epoch": 4.032805429864253, + "grad_norm": 0.09552164678377595, + "learning_rate": 8.777451589753943e-06, + "loss": 0.3814, + "step": 10695 + }, + { + "epoch": 4.033182503770739, + "grad_norm": 0.09477091357138868, + "learning_rate": 8.770870687784625e-06, + "loss": 0.3793, + "step": 10696 + }, + { + "epoch": 4.033559577677225, + "grad_norm": 0.09394756099268144, + "learning_rate": 8.764291949966269e-06, + "loss": 0.3685, + "step": 10697 + }, + { + "epoch": 4.03393665158371, + "grad_norm": 0.08755617018574431, + "learning_rate": 8.757715376754778e-06, + "loss": 0.375, + "step": 10698 + }, + { + "epoch": 4.034313725490196, + "grad_norm": 0.10120493296734143, + "learning_rate": 8.751140968605907e-06, + "loss": 0.3861, + "step": 10699 + }, + { + "epoch": 4.034690799396682, + "grad_norm": 0.08313343461609404, + "learning_rate": 8.74456872597523e-06, + "loss": 0.3783, + "step": 10700 + }, + { + "epoch": 4.035067873303167, + "grad_norm": 0.09667351555249287, + "learning_rate": 8.73799864931824e-06, + "loss": 0.3777, + "step": 10701 + }, + { + "epoch": 4.035444947209653, + "grad_norm": 0.10512249526178996, + "learning_rate": 8.7314307390902e-06, + "loss": 0.3686, + "step": 10702 + }, + { + "epoch": 4.035822021116139, + "grad_norm": 0.09095160968262532, + "learning_rate": 8.724864995746274e-06, + "loss": 0.382, + "step": 10703 + }, + { + "epoch": 4.036199095022624, + "grad_norm": 0.1080976194692548, + "learning_rate": 8.718301419741463e-06, + "loss": 0.3758, + "step": 10704 + }, + { + "epoch": 4.03657616892911, + "grad_norm": 0.0841744886413709, + "learning_rate": 8.711740011530633e-06, + "loss": 0.3752, + "step": 10705 + }, + { + "epoch": 4.036953242835596, + "grad_norm": 0.09917525080130628, + "learning_rate": 8.705180771568442e-06, + "loss": 0.3793, + "step": 10706 + }, + { + "epoch": 4.037330316742081, + "grad_norm": 0.10323282063258601, + "learning_rate": 8.698623700309485e-06, + "loss": 0.3727, + "step": 10707 + }, + { + "epoch": 4.037707390648567, + "grad_norm": 0.09018572086152725, + "learning_rate": 8.692068798208134e-06, + "loss": 0.3668, + "step": 10708 + }, + { + "epoch": 4.038084464555053, + "grad_norm": 0.10205085691735077, + "learning_rate": 8.685516065718644e-06, + "loss": 0.3739, + "step": 10709 + }, + { + "epoch": 4.038461538461538, + "grad_norm": 0.08919546630175704, + "learning_rate": 8.678965503295114e-06, + "loss": 0.3912, + "step": 10710 + }, + { + "epoch": 4.038838612368024, + "grad_norm": 0.0865575370459369, + "learning_rate": 8.672417111391502e-06, + "loss": 0.3821, + "step": 10711 + }, + { + "epoch": 4.03921568627451, + "grad_norm": 0.10686032690649311, + "learning_rate": 8.665870890461576e-06, + "loss": 0.3839, + "step": 10712 + }, + { + "epoch": 4.039592760180995, + "grad_norm": 0.091507511723392, + "learning_rate": 8.659326840959026e-06, + "loss": 0.3897, + "step": 10713 + }, + { + "epoch": 4.039969834087481, + "grad_norm": 0.09076930150770038, + "learning_rate": 8.652784963337306e-06, + "loss": 0.3813, + "step": 10714 + }, + { + "epoch": 4.040346907993967, + "grad_norm": 0.08711341124237215, + "learning_rate": 8.646245258049806e-06, + "loss": 0.3812, + "step": 10715 + }, + { + "epoch": 4.040723981900452, + "grad_norm": 0.08339542219594029, + "learning_rate": 8.639707725549687e-06, + "loss": 0.3812, + "step": 10716 + }, + { + "epoch": 4.041101055806938, + "grad_norm": 0.09088362870466654, + "learning_rate": 8.633172366290003e-06, + "loss": 0.3661, + "step": 10717 + }, + { + "epoch": 4.041478129713424, + "grad_norm": 0.08323539203780847, + "learning_rate": 8.626639180723652e-06, + "loss": 0.3886, + "step": 10718 + }, + { + "epoch": 4.041855203619909, + "grad_norm": 0.0944740838813542, + "learning_rate": 8.620108169303391e-06, + "loss": 0.3783, + "step": 10719 + }, + { + "epoch": 4.042232277526395, + "grad_norm": 0.08359338898952352, + "learning_rate": 8.613579332481778e-06, + "loss": 0.3779, + "step": 10720 + }, + { + "epoch": 4.042609351432881, + "grad_norm": 0.08439445267721923, + "learning_rate": 8.6070526707113e-06, + "loss": 0.3783, + "step": 10721 + }, + { + "epoch": 4.042986425339366, + "grad_norm": 0.096383257744843, + "learning_rate": 8.600528184444216e-06, + "loss": 0.3847, + "step": 10722 + }, + { + "epoch": 4.043363499245852, + "grad_norm": 0.09152138087456425, + "learning_rate": 8.594005874132673e-06, + "loss": 0.3813, + "step": 10723 + }, + { + "epoch": 4.043740573152338, + "grad_norm": 0.09854550633129842, + "learning_rate": 8.587485740228673e-06, + "loss": 0.3784, + "step": 10724 + }, + { + "epoch": 4.044117647058823, + "grad_norm": 0.09530861698931893, + "learning_rate": 8.580967783184055e-06, + "loss": 0.373, + "step": 10725 + }, + { + "epoch": 4.044494720965309, + "grad_norm": 0.10449884792794262, + "learning_rate": 8.57445200345048e-06, + "loss": 0.3807, + "step": 10726 + }, + { + "epoch": 4.044871794871795, + "grad_norm": 0.09573943171438146, + "learning_rate": 8.567938401479528e-06, + "loss": 0.373, + "step": 10727 + }, + { + "epoch": 4.04524886877828, + "grad_norm": 0.10687892723172403, + "learning_rate": 8.561426977722554e-06, + "loss": 0.3877, + "step": 10728 + }, + { + "epoch": 4.045625942684766, + "grad_norm": 0.09174638095377356, + "learning_rate": 8.5549177326308e-06, + "loss": 0.3913, + "step": 10729 + }, + { + "epoch": 4.046003016591252, + "grad_norm": 0.09922810536847694, + "learning_rate": 8.548410666655362e-06, + "loss": 0.378, + "step": 10730 + }, + { + "epoch": 4.046380090497737, + "grad_norm": 0.0833611159665225, + "learning_rate": 8.541905780247175e-06, + "loss": 0.3753, + "step": 10731 + }, + { + "epoch": 4.046757164404223, + "grad_norm": 0.08349427171227752, + "learning_rate": 8.535403073856993e-06, + "loss": 0.3756, + "step": 10732 + }, + { + "epoch": 4.047134238310709, + "grad_norm": 0.09189135015691477, + "learning_rate": 8.52890254793549e-06, + "loss": 0.3666, + "step": 10733 + }, + { + "epoch": 4.047511312217194, + "grad_norm": 0.08340782652623047, + "learning_rate": 8.522404202933115e-06, + "loss": 0.3878, + "step": 10734 + }, + { + "epoch": 4.04788838612368, + "grad_norm": 0.08987917065509668, + "learning_rate": 8.515908039300207e-06, + "loss": 0.3835, + "step": 10735 + }, + { + "epoch": 4.048265460030166, + "grad_norm": 0.10393571224467242, + "learning_rate": 8.509414057486949e-06, + "loss": 0.3928, + "step": 10736 + }, + { + "epoch": 4.048642533936651, + "grad_norm": 0.08429712002251556, + "learning_rate": 8.502922257943375e-06, + "loss": 0.3751, + "step": 10737 + }, + { + "epoch": 4.049019607843137, + "grad_norm": 0.09595167921765949, + "learning_rate": 8.496432641119327e-06, + "loss": 0.3881, + "step": 10738 + }, + { + "epoch": 4.049396681749623, + "grad_norm": 0.09608976286401427, + "learning_rate": 8.489945207464574e-06, + "loss": 0.3811, + "step": 10739 + }, + { + "epoch": 4.049773755656108, + "grad_norm": 0.0836343814948096, + "learning_rate": 8.483459957428661e-06, + "loss": 0.3738, + "step": 10740 + }, + { + "epoch": 4.050150829562594, + "grad_norm": 0.10568019794384387, + "learning_rate": 8.476976891461013e-06, + "loss": 0.38, + "step": 10741 + }, + { + "epoch": 4.05052790346908, + "grad_norm": 0.09486091280516647, + "learning_rate": 8.470496010010908e-06, + "loss": 0.3814, + "step": 10742 + }, + { + "epoch": 4.050904977375565, + "grad_norm": 0.08455361241690519, + "learning_rate": 8.464017313527479e-06, + "loss": 0.383, + "step": 10743 + }, + { + "epoch": 4.051282051282051, + "grad_norm": 0.08558337651081235, + "learning_rate": 8.457540802459649e-06, + "loss": 0.3626, + "step": 10744 + }, + { + "epoch": 4.051659125188537, + "grad_norm": 0.08677268001935923, + "learning_rate": 8.451066477256287e-06, + "loss": 0.3765, + "step": 10745 + }, + { + "epoch": 4.052036199095022, + "grad_norm": 0.08992045221484628, + "learning_rate": 8.444594338366024e-06, + "loss": 0.377, + "step": 10746 + }, + { + "epoch": 4.052413273001508, + "grad_norm": 0.08347429485715012, + "learning_rate": 8.438124386237385e-06, + "loss": 0.3693, + "step": 10747 + }, + { + "epoch": 4.052790346907994, + "grad_norm": 0.09576695779841067, + "learning_rate": 8.431656621318729e-06, + "loss": 0.3788, + "step": 10748 + }, + { + "epoch": 4.0531674208144794, + "grad_norm": 0.09451602476762097, + "learning_rate": 8.425191044058279e-06, + "loss": 0.3869, + "step": 10749 + }, + { + "epoch": 4.053544494720965, + "grad_norm": 0.0879385016182449, + "learning_rate": 8.418727654904062e-06, + "loss": 0.3695, + "step": 10750 + }, + { + "epoch": 4.053921568627451, + "grad_norm": 0.08719062373116193, + "learning_rate": 8.412266454304028e-06, + "loss": 0.374, + "step": 10751 + }, + { + "epoch": 4.0542986425339365, + "grad_norm": 0.08398501622641534, + "learning_rate": 8.405807442705898e-06, + "loss": 0.387, + "step": 10752 + }, + { + "epoch": 4.054675716440422, + "grad_norm": 0.09142580029569325, + "learning_rate": 8.399350620557287e-06, + "loss": 0.3742, + "step": 10753 + }, + { + "epoch": 4.055052790346908, + "grad_norm": 0.08077145887206863, + "learning_rate": 8.392895988305651e-06, + "loss": 0.3736, + "step": 10754 + }, + { + "epoch": 4.0554298642533935, + "grad_norm": 0.088608762458934, + "learning_rate": 8.3864435463983e-06, + "loss": 0.3867, + "step": 10755 + }, + { + "epoch": 4.055806938159879, + "grad_norm": 0.08493145565027171, + "learning_rate": 8.379993295282345e-06, + "loss": 0.3843, + "step": 10756 + }, + { + "epoch": 4.056184012066365, + "grad_norm": 0.08974028478098423, + "learning_rate": 8.37354523540483e-06, + "loss": 0.3853, + "step": 10757 + }, + { + "epoch": 4.0565610859728505, + "grad_norm": 0.09061623553112028, + "learning_rate": 8.367099367212566e-06, + "loss": 0.3835, + "step": 10758 + }, + { + "epoch": 4.056938159879336, + "grad_norm": 0.08245433794934696, + "learning_rate": 8.360655691152257e-06, + "loss": 0.3856, + "step": 10759 + }, + { + "epoch": 4.057315233785822, + "grad_norm": 0.09308817101870223, + "learning_rate": 8.354214207670446e-06, + "loss": 0.3958, + "step": 10760 + }, + { + "epoch": 4.0576923076923075, + "grad_norm": 0.08681784986770742, + "learning_rate": 8.347774917213534e-06, + "loss": 0.3719, + "step": 10761 + }, + { + "epoch": 4.058069381598793, + "grad_norm": 0.08027206189479624, + "learning_rate": 8.34133782022772e-06, + "loss": 0.3709, + "step": 10762 + }, + { + "epoch": 4.058446455505279, + "grad_norm": 0.10195703283669975, + "learning_rate": 8.334902917159136e-06, + "loss": 0.3703, + "step": 10763 + }, + { + "epoch": 4.0588235294117645, + "grad_norm": 0.08852888952207337, + "learning_rate": 8.328470208453683e-06, + "loss": 0.3808, + "step": 10764 + }, + { + "epoch": 4.05920060331825, + "grad_norm": 0.09341285613241698, + "learning_rate": 8.322039694557151e-06, + "loss": 0.3837, + "step": 10765 + }, + { + "epoch": 4.059577677224736, + "grad_norm": 0.09415773514235602, + "learning_rate": 8.315611375915167e-06, + "loss": 0.367, + "step": 10766 + }, + { + "epoch": 4.0599547511312215, + "grad_norm": 0.09683004473209507, + "learning_rate": 8.309185252973226e-06, + "loss": 0.3753, + "step": 10767 + }, + { + "epoch": 4.060331825037707, + "grad_norm": 0.08915457795420287, + "learning_rate": 8.302761326176614e-06, + "loss": 0.368, + "step": 10768 + }, + { + "epoch": 4.060708898944193, + "grad_norm": 0.08569738928305633, + "learning_rate": 8.296339595970546e-06, + "loss": 0.3835, + "step": 10769 + }, + { + "epoch": 4.0610859728506785, + "grad_norm": 0.09215904997604828, + "learning_rate": 8.289920062800013e-06, + "loss": 0.3738, + "step": 10770 + }, + { + "epoch": 4.061463046757164, + "grad_norm": 0.09055030188601114, + "learning_rate": 8.283502727109897e-06, + "loss": 0.3853, + "step": 10771 + }, + { + "epoch": 4.06184012066365, + "grad_norm": 0.09183354170115136, + "learning_rate": 8.277087589344908e-06, + "loss": 0.3745, + "step": 10772 + }, + { + "epoch": 4.0622171945701355, + "grad_norm": 0.080959755579577, + "learning_rate": 8.270674649949621e-06, + "loss": 0.3774, + "step": 10773 + }, + { + "epoch": 4.062594268476621, + "grad_norm": 0.10840433516316295, + "learning_rate": 8.264263909368413e-06, + "loss": 0.3794, + "step": 10774 + }, + { + "epoch": 4.062971342383107, + "grad_norm": 0.08325083203912016, + "learning_rate": 8.257855368045588e-06, + "loss": 0.3774, + "step": 10775 + }, + { + "epoch": 4.0633484162895925, + "grad_norm": 0.1071006778570037, + "learning_rate": 8.251449026425225e-06, + "loss": 0.381, + "step": 10776 + }, + { + "epoch": 4.063725490196078, + "grad_norm": 0.10109514694997931, + "learning_rate": 8.24504488495128e-06, + "loss": 0.3882, + "step": 10777 + }, + { + "epoch": 4.064102564102564, + "grad_norm": 0.09088455373347953, + "learning_rate": 8.238642944067559e-06, + "loss": 0.3822, + "step": 10778 + }, + { + "epoch": 4.0644796380090495, + "grad_norm": 0.10208909613582638, + "learning_rate": 8.232243204217716e-06, + "loss": 0.3727, + "step": 10779 + }, + { + "epoch": 4.064856711915535, + "grad_norm": 0.08853277172122827, + "learning_rate": 8.225845665845225e-06, + "loss": 0.3717, + "step": 10780 + }, + { + "epoch": 4.065233785822021, + "grad_norm": 0.10129424782018871, + "learning_rate": 8.219450329393464e-06, + "loss": 0.3788, + "step": 10781 + }, + { + "epoch": 4.0656108597285066, + "grad_norm": 0.09473078488011229, + "learning_rate": 8.213057195305598e-06, + "loss": 0.3817, + "step": 10782 + }, + { + "epoch": 4.065987933634992, + "grad_norm": 0.10156334753628679, + "learning_rate": 8.20666626402467e-06, + "loss": 0.3825, + "step": 10783 + }, + { + "epoch": 4.066365007541478, + "grad_norm": 0.10098104625562286, + "learning_rate": 8.200277535993573e-06, + "loss": 0.388, + "step": 10784 + }, + { + "epoch": 4.066742081447964, + "grad_norm": 0.09963404837790982, + "learning_rate": 8.193891011655046e-06, + "loss": 0.3847, + "step": 10785 + }, + { + "epoch": 4.067119155354449, + "grad_norm": 0.09781344045873601, + "learning_rate": 8.187506691451634e-06, + "loss": 0.3803, + "step": 10786 + }, + { + "epoch": 4.067496229260935, + "grad_norm": 0.0815218223571228, + "learning_rate": 8.18112457582582e-06, + "loss": 0.3807, + "step": 10787 + }, + { + "epoch": 4.067873303167421, + "grad_norm": 0.11197421993108546, + "learning_rate": 8.174744665219831e-06, + "loss": 0.3781, + "step": 10788 + }, + { + "epoch": 4.068250377073906, + "grad_norm": 0.0891452334242499, + "learning_rate": 8.168366960075813e-06, + "loss": 0.379, + "step": 10789 + }, + { + "epoch": 4.068627450980392, + "grad_norm": 0.09772601598490328, + "learning_rate": 8.16199146083573e-06, + "loss": 0.371, + "step": 10790 + }, + { + "epoch": 4.069004524886878, + "grad_norm": 0.10431937992775674, + "learning_rate": 8.155618167941393e-06, + "loss": 0.385, + "step": 10791 + }, + { + "epoch": 4.069381598793363, + "grad_norm": 0.08686375797569358, + "learning_rate": 8.149247081834476e-06, + "loss": 0.3782, + "step": 10792 + }, + { + "epoch": 4.069758672699849, + "grad_norm": 0.10233755379500692, + "learning_rate": 8.142878202956495e-06, + "loss": 0.3861, + "step": 10793 + }, + { + "epoch": 4.070135746606335, + "grad_norm": 0.0856755555454286, + "learning_rate": 8.136511531748788e-06, + "loss": 0.3774, + "step": 10794 + }, + { + "epoch": 4.07051282051282, + "grad_norm": 0.08781878633683964, + "learning_rate": 8.130147068652565e-06, + "loss": 0.3869, + "step": 10795 + }, + { + "epoch": 4.070889894419306, + "grad_norm": 0.09131760856374087, + "learning_rate": 8.123784814108883e-06, + "loss": 0.3702, + "step": 10796 + }, + { + "epoch": 4.071266968325792, + "grad_norm": 0.08334032128290678, + "learning_rate": 8.11742476855864e-06, + "loss": 0.3885, + "step": 10797 + }, + { + "epoch": 4.071644042232277, + "grad_norm": 0.08715287041001128, + "learning_rate": 8.111066932442582e-06, + "loss": 0.3817, + "step": 10798 + }, + { + "epoch": 4.072021116138763, + "grad_norm": 0.08420629313557561, + "learning_rate": 8.104711306201305e-06, + "loss": 0.39, + "step": 10799 + }, + { + "epoch": 4.072398190045249, + "grad_norm": 0.08615008562050691, + "learning_rate": 8.098357890275239e-06, + "loss": 0.3798, + "step": 10800 + }, + { + "epoch": 4.072775263951734, + "grad_norm": 0.08501127328119962, + "learning_rate": 8.092006685104672e-06, + "loss": 0.3796, + "step": 10801 + }, + { + "epoch": 4.07315233785822, + "grad_norm": 0.08725560427970395, + "learning_rate": 8.08565769112974e-06, + "loss": 0.3813, + "step": 10802 + }, + { + "epoch": 4.073529411764706, + "grad_norm": 0.0882170542326474, + "learning_rate": 8.07931090879042e-06, + "loss": 0.3828, + "step": 10803 + }, + { + "epoch": 4.073906485671191, + "grad_norm": 0.08204985219800578, + "learning_rate": 8.072966338526544e-06, + "loss": 0.3706, + "step": 10804 + }, + { + "epoch": 4.074283559577677, + "grad_norm": 0.08392281882233595, + "learning_rate": 8.06662398077779e-06, + "loss": 0.3753, + "step": 10805 + }, + { + "epoch": 4.074660633484163, + "grad_norm": 0.09017093686189132, + "learning_rate": 8.06028383598366e-06, + "loss": 0.3854, + "step": 10806 + }, + { + "epoch": 4.075037707390648, + "grad_norm": 0.08216916788704258, + "learning_rate": 8.053945904583526e-06, + "loss": 0.3702, + "step": 10807 + }, + { + "epoch": 4.075414781297134, + "grad_norm": 0.08629312530517537, + "learning_rate": 8.047610187016607e-06, + "loss": 0.3829, + "step": 10808 + }, + { + "epoch": 4.07579185520362, + "grad_norm": 0.08555997454784234, + "learning_rate": 8.04127668372196e-06, + "loss": 0.3741, + "step": 10809 + }, + { + "epoch": 4.076168929110105, + "grad_norm": 0.09354973323961814, + "learning_rate": 8.034945395138495e-06, + "loss": 0.3837, + "step": 10810 + }, + { + "epoch": 4.076546003016591, + "grad_norm": 0.08702407562170128, + "learning_rate": 8.028616321704957e-06, + "loss": 0.3679, + "step": 10811 + }, + { + "epoch": 4.076923076923077, + "grad_norm": 0.0907428450479876, + "learning_rate": 8.022289463859963e-06, + "loss": 0.3807, + "step": 10812 + }, + { + "epoch": 4.077300150829562, + "grad_norm": 0.08804846665206292, + "learning_rate": 8.01596482204193e-06, + "loss": 0.3866, + "step": 10813 + }, + { + "epoch": 4.077677224736048, + "grad_norm": 0.09685520538605569, + "learning_rate": 8.009642396689168e-06, + "loss": 0.3814, + "step": 10814 + }, + { + "epoch": 4.078054298642534, + "grad_norm": 0.09139210853586817, + "learning_rate": 8.003322188239813e-06, + "loss": 0.3754, + "step": 10815 + }, + { + "epoch": 4.078431372549019, + "grad_norm": 0.08787379419883679, + "learning_rate": 7.997004197131847e-06, + "loss": 0.3759, + "step": 10816 + }, + { + "epoch": 4.078808446455505, + "grad_norm": 0.08930085238475412, + "learning_rate": 7.990688423803106e-06, + "loss": 0.3828, + "step": 10817 + }, + { + "epoch": 4.079185520361991, + "grad_norm": 0.09244596812145621, + "learning_rate": 7.984374868691276e-06, + "loss": 0.3787, + "step": 10818 + }, + { + "epoch": 4.079562594268476, + "grad_norm": 0.09195780303412174, + "learning_rate": 7.978063532233857e-06, + "loss": 0.3957, + "step": 10819 + }, + { + "epoch": 4.079939668174962, + "grad_norm": 0.08293756993515218, + "learning_rate": 7.97175441486823e-06, + "loss": 0.3797, + "step": 10820 + }, + { + "epoch": 4.080316742081448, + "grad_norm": 0.08141711602494343, + "learning_rate": 7.965447517031615e-06, + "loss": 0.3708, + "step": 10821 + }, + { + "epoch": 4.080693815987933, + "grad_norm": 0.0897204724361134, + "learning_rate": 7.959142839161073e-06, + "loss": 0.3864, + "step": 10822 + }, + { + "epoch": 4.081070889894419, + "grad_norm": 0.08681710119045441, + "learning_rate": 7.952840381693514e-06, + "loss": 0.3767, + "step": 10823 + }, + { + "epoch": 4.081447963800905, + "grad_norm": 0.08947744303293975, + "learning_rate": 7.946540145065697e-06, + "loss": 0.3917, + "step": 10824 + }, + { + "epoch": 4.08182503770739, + "grad_norm": 0.0793906170636583, + "learning_rate": 7.9402421297142e-06, + "loss": 0.3752, + "step": 10825 + }, + { + "epoch": 4.082202111613876, + "grad_norm": 0.08370009853424698, + "learning_rate": 7.933946336075507e-06, + "loss": 0.3735, + "step": 10826 + }, + { + "epoch": 4.082579185520362, + "grad_norm": 0.09078584580157825, + "learning_rate": 7.927652764585873e-06, + "loss": 0.381, + "step": 10827 + }, + { + "epoch": 4.082956259426847, + "grad_norm": 0.08975583198586955, + "learning_rate": 7.921361415681463e-06, + "loss": 0.3753, + "step": 10828 + }, + { + "epoch": 4.083333333333333, + "grad_norm": 0.08561897436011429, + "learning_rate": 7.915072289798247e-06, + "loss": 0.3622, + "step": 10829 + }, + { + "epoch": 4.083710407239819, + "grad_norm": 0.09044985299516047, + "learning_rate": 7.908785387372075e-06, + "loss": 0.3807, + "step": 10830 + }, + { + "epoch": 4.084087481146304, + "grad_norm": 0.08931197611350858, + "learning_rate": 7.902500708838591e-06, + "loss": 0.3774, + "step": 10831 + }, + { + "epoch": 4.08446455505279, + "grad_norm": 0.09372647724146178, + "learning_rate": 7.896218254633359e-06, + "loss": 0.3824, + "step": 10832 + }, + { + "epoch": 4.084841628959276, + "grad_norm": 0.0909774927451946, + "learning_rate": 7.889938025191713e-06, + "loss": 0.3703, + "step": 10833 + }, + { + "epoch": 4.085218702865761, + "grad_norm": 0.09946859086843635, + "learning_rate": 7.883660020948883e-06, + "loss": 0.3821, + "step": 10834 + }, + { + "epoch": 4.085595776772247, + "grad_norm": 0.08470201257080959, + "learning_rate": 7.87738424233992e-06, + "loss": 0.3835, + "step": 10835 + }, + { + "epoch": 4.085972850678733, + "grad_norm": 0.08259856185289424, + "learning_rate": 7.871110689799741e-06, + "loss": 0.3825, + "step": 10836 + }, + { + "epoch": 4.086349924585218, + "grad_norm": 0.0917291021556982, + "learning_rate": 7.864839363763095e-06, + "loss": 0.3871, + "step": 10837 + }, + { + "epoch": 4.086726998491704, + "grad_norm": 0.08637912415442642, + "learning_rate": 7.858570264664585e-06, + "loss": 0.3836, + "step": 10838 + }, + { + "epoch": 4.08710407239819, + "grad_norm": 0.0960692929156531, + "learning_rate": 7.852303392938636e-06, + "loss": 0.3803, + "step": 10839 + }, + { + "epoch": 4.087481146304675, + "grad_norm": 0.07962657249676021, + "learning_rate": 7.846038749019546e-06, + "loss": 0.3693, + "step": 10840 + }, + { + "epoch": 4.087858220211161, + "grad_norm": 0.09310808201938595, + "learning_rate": 7.83977633334145e-06, + "loss": 0.3856, + "step": 10841 + }, + { + "epoch": 4.088235294117647, + "grad_norm": 0.09321394788279612, + "learning_rate": 7.833516146338329e-06, + "loss": 0.3815, + "step": 10842 + }, + { + "epoch": 4.088612368024132, + "grad_norm": 0.08070012967990169, + "learning_rate": 7.82725818844401e-06, + "loss": 0.3846, + "step": 10843 + }, + { + "epoch": 4.088989441930618, + "grad_norm": 0.08268464405580023, + "learning_rate": 7.821002460092168e-06, + "loss": 0.3785, + "step": 10844 + }, + { + "epoch": 4.089366515837104, + "grad_norm": 0.09106167048285416, + "learning_rate": 7.814748961716292e-06, + "loss": 0.3716, + "step": 10845 + }, + { + "epoch": 4.089743589743589, + "grad_norm": 0.08817161092594604, + "learning_rate": 7.808497693749787e-06, + "loss": 0.3684, + "step": 10846 + }, + { + "epoch": 4.090120663650075, + "grad_norm": 0.0838157843481368, + "learning_rate": 7.802248656625827e-06, + "loss": 0.3786, + "step": 10847 + }, + { + "epoch": 4.090497737556561, + "grad_norm": 0.09295075246814281, + "learning_rate": 7.796001850777477e-06, + "loss": 0.3921, + "step": 10848 + }, + { + "epoch": 4.090874811463046, + "grad_norm": 0.08140389951800325, + "learning_rate": 7.78975727663763e-06, + "loss": 0.3766, + "step": 10849 + }, + { + "epoch": 4.091251885369532, + "grad_norm": 0.08336818488477529, + "learning_rate": 7.783514934639047e-06, + "loss": 0.3786, + "step": 10850 + }, + { + "epoch": 4.091628959276018, + "grad_norm": 0.10159364359263545, + "learning_rate": 7.777274825214283e-06, + "loss": 0.374, + "step": 10851 + }, + { + "epoch": 4.092006033182503, + "grad_norm": 0.08838937903466422, + "learning_rate": 7.771036948795813e-06, + "loss": 0.3814, + "step": 10852 + }, + { + "epoch": 4.092383107088989, + "grad_norm": 0.09146126986715372, + "learning_rate": 7.76480130581588e-06, + "loss": 0.3695, + "step": 10853 + }, + { + "epoch": 4.092760180995475, + "grad_norm": 0.08572855517975254, + "learning_rate": 7.758567896706632e-06, + "loss": 0.3788, + "step": 10854 + }, + { + "epoch": 4.0931372549019605, + "grad_norm": 0.09022614596384124, + "learning_rate": 7.752336721900025e-06, + "loss": 0.3888, + "step": 10855 + }, + { + "epoch": 4.093514328808446, + "grad_norm": 0.08725367334831767, + "learning_rate": 7.74610778182789e-06, + "loss": 0.3858, + "step": 10856 + }, + { + "epoch": 4.093891402714932, + "grad_norm": 0.08675376388584723, + "learning_rate": 7.739881076921855e-06, + "loss": 0.3733, + "step": 10857 + }, + { + "epoch": 4.0942684766214175, + "grad_norm": 0.0835214208002355, + "learning_rate": 7.73365660761347e-06, + "loss": 0.3859, + "step": 10858 + }, + { + "epoch": 4.094645550527903, + "grad_norm": 0.09168024474703329, + "learning_rate": 7.727434374334052e-06, + "loss": 0.3801, + "step": 10859 + }, + { + "epoch": 4.095022624434389, + "grad_norm": 0.09078206355539095, + "learning_rate": 7.721214377514807e-06, + "loss": 0.3757, + "step": 10860 + }, + { + "epoch": 4.0953996983408745, + "grad_norm": 0.08701824921143017, + "learning_rate": 7.714996617586772e-06, + "loss": 0.3826, + "step": 10861 + }, + { + "epoch": 4.09577677224736, + "grad_norm": 0.09052049526484028, + "learning_rate": 7.708781094980846e-06, + "loss": 0.3687, + "step": 10862 + }, + { + "epoch": 4.096153846153846, + "grad_norm": 0.0883997181578327, + "learning_rate": 7.702567810127726e-06, + "loss": 0.3712, + "step": 10863 + }, + { + "epoch": 4.0965309200603315, + "grad_norm": 0.09866136262712827, + "learning_rate": 7.69635676345803e-06, + "loss": 0.3749, + "step": 10864 + }, + { + "epoch": 4.096907993966817, + "grad_norm": 0.07901454230539463, + "learning_rate": 7.690147955402146e-06, + "loss": 0.3691, + "step": 10865 + }, + { + "epoch": 4.097285067873303, + "grad_norm": 0.08890012934413623, + "learning_rate": 7.683941386390352e-06, + "loss": 0.3863, + "step": 10866 + }, + { + "epoch": 4.0976621417797885, + "grad_norm": 0.09308677329805459, + "learning_rate": 7.677737056852752e-06, + "loss": 0.3779, + "step": 10867 + }, + { + "epoch": 4.098039215686274, + "grad_norm": 0.0854111912737278, + "learning_rate": 7.671534967219317e-06, + "loss": 0.383, + "step": 10868 + }, + { + "epoch": 4.09841628959276, + "grad_norm": 0.09160243486184076, + "learning_rate": 7.665335117919812e-06, + "loss": 0.3808, + "step": 10869 + }, + { + "epoch": 4.0987933634992455, + "grad_norm": 0.08969957373169954, + "learning_rate": 7.659137509383928e-06, + "loss": 0.3723, + "step": 10870 + }, + { + "epoch": 4.099170437405731, + "grad_norm": 0.08544347005868469, + "learning_rate": 7.652942142041109e-06, + "loss": 0.3727, + "step": 10871 + }, + { + "epoch": 4.099547511312217, + "grad_norm": 0.08870099554827122, + "learning_rate": 7.646749016320715e-06, + "loss": 0.3807, + "step": 10872 + }, + { + "epoch": 4.0999245852187025, + "grad_norm": 0.0792569343525127, + "learning_rate": 7.640558132651912e-06, + "loss": 0.3691, + "step": 10873 + }, + { + "epoch": 4.100301659125188, + "grad_norm": 0.08226138146023586, + "learning_rate": 7.63436949146374e-06, + "loss": 0.3846, + "step": 10874 + }, + { + "epoch": 4.100678733031674, + "grad_norm": 0.08434773478441704, + "learning_rate": 7.628183093185036e-06, + "loss": 0.3866, + "step": 10875 + }, + { + "epoch": 4.1010558069381595, + "grad_norm": 0.08106571156583774, + "learning_rate": 7.621998938244548e-06, + "loss": 0.3789, + "step": 10876 + }, + { + "epoch": 4.101432880844645, + "grad_norm": 0.07557837894541217, + "learning_rate": 7.615817027070802e-06, + "loss": 0.3764, + "step": 10877 + }, + { + "epoch": 4.101809954751131, + "grad_norm": 0.08388786642700365, + "learning_rate": 7.609637360092215e-06, + "loss": 0.3721, + "step": 10878 + }, + { + "epoch": 4.1021870286576165, + "grad_norm": 0.084688201441769, + "learning_rate": 7.603459937737025e-06, + "loss": 0.3788, + "step": 10879 + }, + { + "epoch": 4.102564102564102, + "grad_norm": 0.08177291480095353, + "learning_rate": 7.597284760433341e-06, + "loss": 0.3733, + "step": 10880 + }, + { + "epoch": 4.102941176470588, + "grad_norm": 0.08158082092749536, + "learning_rate": 7.591111828609059e-06, + "loss": 0.38, + "step": 10881 + }, + { + "epoch": 4.1033182503770735, + "grad_norm": 0.08240004453741627, + "learning_rate": 7.5849411426919975e-06, + "loss": 0.3752, + "step": 10882 + }, + { + "epoch": 4.103695324283559, + "grad_norm": 0.08428080093570828, + "learning_rate": 7.578772703109756e-06, + "loss": 0.3847, + "step": 10883 + }, + { + "epoch": 4.104072398190045, + "grad_norm": 0.08579007827818821, + "learning_rate": 7.572606510289811e-06, + "loss": 0.3787, + "step": 10884 + }, + { + "epoch": 4.1044494720965305, + "grad_norm": 0.08451934883291687, + "learning_rate": 7.5664425646594684e-06, + "loss": 0.3827, + "step": 10885 + }, + { + "epoch": 4.104826546003016, + "grad_norm": 0.08214772346587465, + "learning_rate": 7.560280866645899e-06, + "loss": 0.3737, + "step": 10886 + }, + { + "epoch": 4.105203619909502, + "grad_norm": 0.0877200083883139, + "learning_rate": 7.554121416676068e-06, + "loss": 0.3817, + "step": 10887 + }, + { + "epoch": 4.105580693815988, + "grad_norm": 0.08168802506325609, + "learning_rate": 7.547964215176868e-06, + "loss": 0.3791, + "step": 10888 + }, + { + "epoch": 4.105957767722473, + "grad_norm": 0.09048663851280056, + "learning_rate": 7.541809262574951e-06, + "loss": 0.3739, + "step": 10889 + }, + { + "epoch": 4.106334841628959, + "grad_norm": 0.08581898039178075, + "learning_rate": 7.535656559296858e-06, + "loss": 0.3844, + "step": 10890 + }, + { + "epoch": 4.106711915535445, + "grad_norm": 0.08497098892946242, + "learning_rate": 7.529506105768973e-06, + "loss": 0.3748, + "step": 10891 + }, + { + "epoch": 4.10708898944193, + "grad_norm": 0.08008020809276907, + "learning_rate": 7.523357902417525e-06, + "loss": 0.3757, + "step": 10892 + }, + { + "epoch": 4.107466063348416, + "grad_norm": 0.08214463748927725, + "learning_rate": 7.517211949668546e-06, + "loss": 0.3788, + "step": 10893 + }, + { + "epoch": 4.107843137254902, + "grad_norm": 0.08098042187535762, + "learning_rate": 7.511068247947988e-06, + "loss": 0.3876, + "step": 10894 + }, + { + "epoch": 4.108220211161387, + "grad_norm": 0.07601322414749226, + "learning_rate": 7.504926797681573e-06, + "loss": 0.3763, + "step": 10895 + }, + { + "epoch": 4.108597285067873, + "grad_norm": 0.0783891165947323, + "learning_rate": 7.498787599294908e-06, + "loss": 0.3704, + "step": 10896 + }, + { + "epoch": 4.108974358974359, + "grad_norm": 0.08086555851459752, + "learning_rate": 7.492650653213429e-06, + "loss": 0.3764, + "step": 10897 + }, + { + "epoch": 4.109351432880844, + "grad_norm": 0.08128569209071151, + "learning_rate": 7.4865159598624415e-06, + "loss": 0.3812, + "step": 10898 + }, + { + "epoch": 4.10972850678733, + "grad_norm": 0.07926257009423462, + "learning_rate": 7.480383519667036e-06, + "loss": 0.3755, + "step": 10899 + }, + { + "epoch": 4.110105580693816, + "grad_norm": 0.08001152878901863, + "learning_rate": 7.474253333052228e-06, + "loss": 0.3672, + "step": 10900 + }, + { + "epoch": 4.110482654600301, + "grad_norm": 0.07822243445463357, + "learning_rate": 7.468125400442803e-06, + "loss": 0.3781, + "step": 10901 + }, + { + "epoch": 4.110859728506787, + "grad_norm": 0.08082058515288004, + "learning_rate": 7.4619997222634335e-06, + "loss": 0.3705, + "step": 10902 + }, + { + "epoch": 4.111236802413273, + "grad_norm": 0.08473027476449646, + "learning_rate": 7.45587629893862e-06, + "loss": 0.375, + "step": 10903 + }, + { + "epoch": 4.111613876319758, + "grad_norm": 0.0841488339709486, + "learning_rate": 7.449755130892726e-06, + "loss": 0.383, + "step": 10904 + }, + { + "epoch": 4.111990950226244, + "grad_norm": 0.07928253787244843, + "learning_rate": 7.443636218549906e-06, + "loss": 0.3909, + "step": 10905 + }, + { + "epoch": 4.11236802413273, + "grad_norm": 0.08015357510091979, + "learning_rate": 7.437519562334241e-06, + "loss": 0.3854, + "step": 10906 + }, + { + "epoch": 4.112745098039215, + "grad_norm": 0.08106040760846311, + "learning_rate": 7.431405162669576e-06, + "loss": 0.3881, + "step": 10907 + }, + { + "epoch": 4.113122171945701, + "grad_norm": 0.08050249344782814, + "learning_rate": 7.425293019979643e-06, + "loss": 0.375, + "step": 10908 + }, + { + "epoch": 4.113499245852187, + "grad_norm": 0.08647714239447435, + "learning_rate": 7.419183134688013e-06, + "loss": 0.3794, + "step": 10909 + }, + { + "epoch": 4.113876319758672, + "grad_norm": 0.09044112358665121, + "learning_rate": 7.413075507218104e-06, + "loss": 0.3807, + "step": 10910 + }, + { + "epoch": 4.114253393665158, + "grad_norm": 0.07882123898466692, + "learning_rate": 7.4069701379931325e-06, + "loss": 0.3743, + "step": 10911 + }, + { + "epoch": 4.114630467571644, + "grad_norm": 0.09178590033328668, + "learning_rate": 7.4008670274362445e-06, + "loss": 0.3836, + "step": 10912 + }, + { + "epoch": 4.115007541478129, + "grad_norm": 0.09338222659616137, + "learning_rate": 7.394766175970343e-06, + "loss": 0.3901, + "step": 10913 + }, + { + "epoch": 4.115384615384615, + "grad_norm": 0.08018684608415629, + "learning_rate": 7.38866758401823e-06, + "loss": 0.38, + "step": 10914 + }, + { + "epoch": 4.115761689291101, + "grad_norm": 0.09135639775917669, + "learning_rate": 7.382571252002524e-06, + "loss": 0.385, + "step": 10915 + }, + { + "epoch": 4.116138763197586, + "grad_norm": 0.08443239165088033, + "learning_rate": 7.376477180345709e-06, + "loss": 0.3785, + "step": 10916 + }, + { + "epoch": 4.116515837104072, + "grad_norm": 0.09357734394037806, + "learning_rate": 7.3703853694700655e-06, + "loss": 0.3867, + "step": 10917 + }, + { + "epoch": 4.116892911010558, + "grad_norm": 0.09056355341810146, + "learning_rate": 7.364295819797798e-06, + "loss": 0.3854, + "step": 10918 + }, + { + "epoch": 4.117269984917043, + "grad_norm": 0.08985915290779203, + "learning_rate": 7.35820853175087e-06, + "loss": 0.3768, + "step": 10919 + }, + { + "epoch": 4.117647058823529, + "grad_norm": 0.0930113156287621, + "learning_rate": 7.3521235057511364e-06, + "loss": 0.3737, + "step": 10920 + }, + { + "epoch": 4.118024132730015, + "grad_norm": 0.08643052687640601, + "learning_rate": 7.346040742220286e-06, + "loss": 0.3845, + "step": 10921 + }, + { + "epoch": 4.1184012066365, + "grad_norm": 0.08448723951277026, + "learning_rate": 7.339960241579858e-06, + "loss": 0.3875, + "step": 10922 + }, + { + "epoch": 4.118778280542986, + "grad_norm": 0.0960479801471921, + "learning_rate": 7.333882004251198e-06, + "loss": 0.3756, + "step": 10923 + }, + { + "epoch": 4.119155354449472, + "grad_norm": 0.08861634091370299, + "learning_rate": 7.327806030655557e-06, + "loss": 0.3755, + "step": 10924 + }, + { + "epoch": 4.119532428355957, + "grad_norm": 0.09263481893127475, + "learning_rate": 7.321732321213973e-06, + "loss": 0.3846, + "step": 10925 + }, + { + "epoch": 4.119909502262443, + "grad_norm": 0.09173592776774092, + "learning_rate": 7.3156608763473505e-06, + "loss": 0.3654, + "step": 10926 + }, + { + "epoch": 4.120286576168929, + "grad_norm": 0.08542596215960317, + "learning_rate": 7.309591696476435e-06, + "loss": 0.3773, + "step": 10927 + }, + { + "epoch": 4.120663650075414, + "grad_norm": 0.09378241407096297, + "learning_rate": 7.303524782021831e-06, + "loss": 0.378, + "step": 10928 + }, + { + "epoch": 4.1210407239819, + "grad_norm": 0.08102242306471058, + "learning_rate": 7.297460133403942e-06, + "loss": 0.3749, + "step": 10929 + }, + { + "epoch": 4.121417797888386, + "grad_norm": 0.08499903983841849, + "learning_rate": 7.291397751043074e-06, + "loss": 0.3749, + "step": 10930 + }, + { + "epoch": 4.121794871794871, + "grad_norm": 0.07856266277589129, + "learning_rate": 7.28533763535932e-06, + "loss": 0.368, + "step": 10931 + }, + { + "epoch": 4.122171945701357, + "grad_norm": 0.081003669898577, + "learning_rate": 7.279279786772648e-06, + "loss": 0.372, + "step": 10932 + }, + { + "epoch": 4.122549019607843, + "grad_norm": 0.08663954569938956, + "learning_rate": 7.273224205702867e-06, + "loss": 0.3762, + "step": 10933 + }, + { + "epoch": 4.122926093514328, + "grad_norm": 0.07878691078873122, + "learning_rate": 7.267170892569631e-06, + "loss": 0.373, + "step": 10934 + }, + { + "epoch": 4.123303167420815, + "grad_norm": 0.09155407227565956, + "learning_rate": 7.261119847792395e-06, + "loss": 0.3818, + "step": 10935 + }, + { + "epoch": 4.1236802413273, + "grad_norm": 0.08554299080924792, + "learning_rate": 7.255071071790536e-06, + "loss": 0.3717, + "step": 10936 + }, + { + "epoch": 4.124057315233786, + "grad_norm": 0.08962131738967191, + "learning_rate": 7.249024564983194e-06, + "loss": 0.3782, + "step": 10937 + }, + { + "epoch": 4.124434389140271, + "grad_norm": 0.08769768355090933, + "learning_rate": 7.242980327789402e-06, + "loss": 0.3768, + "step": 10938 + }, + { + "epoch": 4.124811463046758, + "grad_norm": 0.0796281870758157, + "learning_rate": 7.236938360628021e-06, + "loss": 0.3801, + "step": 10939 + }, + { + "epoch": 4.125188536953242, + "grad_norm": 0.09164562857945623, + "learning_rate": 7.230898663917747e-06, + "loss": 0.3868, + "step": 10940 + }, + { + "epoch": 4.125565610859729, + "grad_norm": 0.08267090385476396, + "learning_rate": 7.22486123807713e-06, + "loss": 0.3818, + "step": 10941 + }, + { + "epoch": 4.125942684766214, + "grad_norm": 0.07979962335098582, + "learning_rate": 7.2188260835245684e-06, + "loss": 0.3789, + "step": 10942 + }, + { + "epoch": 4.1263197586727, + "grad_norm": 0.08530208948393753, + "learning_rate": 7.21279320067827e-06, + "loss": 0.3707, + "step": 10943 + }, + { + "epoch": 4.126696832579185, + "grad_norm": 0.07730146317809117, + "learning_rate": 7.206762589956322e-06, + "loss": 0.3946, + "step": 10944 + }, + { + "epoch": 4.127073906485672, + "grad_norm": 0.08825013171781049, + "learning_rate": 7.20073425177664e-06, + "loss": 0.3913, + "step": 10945 + }, + { + "epoch": 4.127450980392156, + "grad_norm": 0.08941014847852516, + "learning_rate": 7.194708186556978e-06, + "loss": 0.3833, + "step": 10946 + }, + { + "epoch": 4.127828054298643, + "grad_norm": 0.08331183888574957, + "learning_rate": 7.188684394714944e-06, + "loss": 0.3532, + "step": 10947 + }, + { + "epoch": 4.128205128205128, + "grad_norm": 0.09902226620276634, + "learning_rate": 7.182662876667987e-06, + "loss": 0.3874, + "step": 10948 + }, + { + "epoch": 4.128582202111614, + "grad_norm": 0.08631637189358649, + "learning_rate": 7.176643632833374e-06, + "loss": 0.3748, + "step": 10949 + }, + { + "epoch": 4.1289592760181, + "grad_norm": 0.08347552143775008, + "learning_rate": 7.170626663628244e-06, + "loss": 0.3899, + "step": 10950 + }, + { + "epoch": 4.129336349924586, + "grad_norm": 0.08464745958237703, + "learning_rate": 7.164611969469563e-06, + "loss": 0.3799, + "step": 10951 + }, + { + "epoch": 4.129713423831071, + "grad_norm": 0.08350250391417202, + "learning_rate": 7.15859955077415e-06, + "loss": 0.3874, + "step": 10952 + }, + { + "epoch": 4.130090497737557, + "grad_norm": 0.08186510802650211, + "learning_rate": 7.152589407958657e-06, + "loss": 0.3823, + "step": 10953 + }, + { + "epoch": 4.130467571644043, + "grad_norm": 0.07800439340408415, + "learning_rate": 7.146581541439594e-06, + "loss": 0.3812, + "step": 10954 + }, + { + "epoch": 4.130844645550528, + "grad_norm": 0.08260832991575792, + "learning_rate": 7.140575951633275e-06, + "loss": 0.3868, + "step": 10955 + }, + { + "epoch": 4.131221719457014, + "grad_norm": 0.0924574253913269, + "learning_rate": 7.134572638955903e-06, + "loss": 0.3669, + "step": 10956 + }, + { + "epoch": 4.1315987933635, + "grad_norm": 0.08509823562567972, + "learning_rate": 7.128571603823489e-06, + "loss": 0.3702, + "step": 10957 + }, + { + "epoch": 4.131975867269985, + "grad_norm": 0.08477782441512523, + "learning_rate": 7.1225728466519115e-06, + "loss": 0.3775, + "step": 10958 + }, + { + "epoch": 4.132352941176471, + "grad_norm": 0.08181214789877614, + "learning_rate": 7.116576367856871e-06, + "loss": 0.3632, + "step": 10959 + }, + { + "epoch": 4.132730015082957, + "grad_norm": 0.08673705520821431, + "learning_rate": 7.110582167853919e-06, + "loss": 0.3921, + "step": 10960 + }, + { + "epoch": 4.133107088989442, + "grad_norm": 0.08961850882632397, + "learning_rate": 7.104590247058452e-06, + "loss": 0.3852, + "step": 10961 + }, + { + "epoch": 4.133484162895928, + "grad_norm": 0.08247188724686856, + "learning_rate": 7.098600605885711e-06, + "loss": 0.3794, + "step": 10962 + }, + { + "epoch": 4.133861236802414, + "grad_norm": 0.085686283458436, + "learning_rate": 7.092613244750755e-06, + "loss": 0.3718, + "step": 10963 + }, + { + "epoch": 4.134238310708899, + "grad_norm": 0.08245794658877736, + "learning_rate": 7.086628164068514e-06, + "loss": 0.3721, + "step": 10964 + }, + { + "epoch": 4.134615384615385, + "grad_norm": 0.08537989856062421, + "learning_rate": 7.080645364253747e-06, + "loss": 0.38, + "step": 10965 + }, + { + "epoch": 4.134992458521871, + "grad_norm": 0.08514913382560066, + "learning_rate": 7.074664845721053e-06, + "loss": 0.3646, + "step": 10966 + }, + { + "epoch": 4.135369532428356, + "grad_norm": 0.08418661848691002, + "learning_rate": 7.068686608884885e-06, + "loss": 0.368, + "step": 10967 + }, + { + "epoch": 4.135746606334842, + "grad_norm": 0.08156565335908278, + "learning_rate": 7.062710654159533e-06, + "loss": 0.3794, + "step": 10968 + }, + { + "epoch": 4.136123680241328, + "grad_norm": 0.07781667403436134, + "learning_rate": 7.056736981959109e-06, + "loss": 0.388, + "step": 10969 + }, + { + "epoch": 4.136500754147813, + "grad_norm": 0.08555081079642525, + "learning_rate": 7.05076559269759e-06, + "loss": 0.3736, + "step": 10970 + }, + { + "epoch": 4.136877828054299, + "grad_norm": 0.09197382410078737, + "learning_rate": 7.044796486788788e-06, + "loss": 0.3748, + "step": 10971 + }, + { + "epoch": 4.137254901960785, + "grad_norm": 0.07986718790977572, + "learning_rate": 7.03882966464636e-06, + "loss": 0.3906, + "step": 10972 + }, + { + "epoch": 4.13763197586727, + "grad_norm": 0.07819485186799, + "learning_rate": 7.032865126683801e-06, + "loss": 0.3777, + "step": 10973 + }, + { + "epoch": 4.138009049773756, + "grad_norm": 0.07606716732936589, + "learning_rate": 7.026902873314445e-06, + "loss": 0.3753, + "step": 10974 + }, + { + "epoch": 4.138386123680242, + "grad_norm": 0.0884895625982429, + "learning_rate": 7.020942904951482e-06, + "loss": 0.3746, + "step": 10975 + }, + { + "epoch": 4.138763197586727, + "grad_norm": 0.0789702427141318, + "learning_rate": 7.014985222007915e-06, + "loss": 0.3749, + "step": 10976 + }, + { + "epoch": 4.139140271493213, + "grad_norm": 0.07900085670467791, + "learning_rate": 7.009029824896614e-06, + "loss": 0.39, + "step": 10977 + }, + { + "epoch": 4.139517345399699, + "grad_norm": 0.08012900776878878, + "learning_rate": 7.003076714030284e-06, + "loss": 0.3758, + "step": 10978 + }, + { + "epoch": 4.139894419306184, + "grad_norm": 0.0859974354017021, + "learning_rate": 6.997125889821466e-06, + "loss": 0.386, + "step": 10979 + }, + { + "epoch": 4.14027149321267, + "grad_norm": 0.0879825541235021, + "learning_rate": 6.9911773526825545e-06, + "loss": 0.3713, + "step": 10980 + }, + { + "epoch": 4.140648567119156, + "grad_norm": 0.07722533263348187, + "learning_rate": 6.985231103025781e-06, + "loss": 0.3758, + "step": 10981 + }, + { + "epoch": 4.141025641025641, + "grad_norm": 0.09420986066556572, + "learning_rate": 6.9792871412632e-06, + "loss": 0.386, + "step": 10982 + }, + { + "epoch": 4.141402714932127, + "grad_norm": 0.08014267288549837, + "learning_rate": 6.973345467806729e-06, + "loss": 0.3696, + "step": 10983 + }, + { + "epoch": 4.141779788838613, + "grad_norm": 0.07562987506132299, + "learning_rate": 6.967406083068122e-06, + "loss": 0.3669, + "step": 10984 + }, + { + "epoch": 4.142156862745098, + "grad_norm": 0.08262492453866348, + "learning_rate": 6.9614689874589746e-06, + "loss": 0.3826, + "step": 10985 + }, + { + "epoch": 4.142533936651584, + "grad_norm": 0.08667409958165846, + "learning_rate": 6.955534181390722e-06, + "loss": 0.3784, + "step": 10986 + }, + { + "epoch": 4.14291101055807, + "grad_norm": 0.07625057592478454, + "learning_rate": 6.949601665274652e-06, + "loss": 0.3635, + "step": 10987 + }, + { + "epoch": 4.143288084464555, + "grad_norm": 0.08608176649070481, + "learning_rate": 6.943671439521855e-06, + "loss": 0.3759, + "step": 10988 + }, + { + "epoch": 4.143665158371041, + "grad_norm": 0.08861622135074428, + "learning_rate": 6.937743504543313e-06, + "loss": 0.3934, + "step": 10989 + }, + { + "epoch": 4.144042232277527, + "grad_norm": 0.08464270681252314, + "learning_rate": 6.931817860749817e-06, + "loss": 0.3656, + "step": 10990 + }, + { + "epoch": 4.144419306184012, + "grad_norm": 0.09328987572344412, + "learning_rate": 6.925894508552011e-06, + "loss": 0.3853, + "step": 10991 + }, + { + "epoch": 4.144796380090498, + "grad_norm": 0.08828836589304184, + "learning_rate": 6.919973448360378e-06, + "loss": 0.3818, + "step": 10992 + }, + { + "epoch": 4.145173453996984, + "grad_norm": 0.087867116222739, + "learning_rate": 6.9140546805852535e-06, + "loss": 0.3784, + "step": 10993 + }, + { + "epoch": 4.1455505279034695, + "grad_norm": 0.07939453642738105, + "learning_rate": 6.908138205636774e-06, + "loss": 0.3806, + "step": 10994 + }, + { + "epoch": 4.145927601809955, + "grad_norm": 0.09274358987567105, + "learning_rate": 6.902224023924984e-06, + "loss": 0.371, + "step": 10995 + }, + { + "epoch": 4.146304675716441, + "grad_norm": 0.09042848285977471, + "learning_rate": 6.896312135859702e-06, + "loss": 0.3707, + "step": 10996 + }, + { + "epoch": 4.1466817496229265, + "grad_norm": 0.08475913257966514, + "learning_rate": 6.890402541850627e-06, + "loss": 0.3838, + "step": 10997 + }, + { + "epoch": 4.147058823529412, + "grad_norm": 0.09571297553067798, + "learning_rate": 6.884495242307285e-06, + "loss": 0.3859, + "step": 10998 + }, + { + "epoch": 4.147435897435898, + "grad_norm": 0.0793453851262338, + "learning_rate": 6.87859023763906e-06, + "loss": 0.3703, + "step": 10999 + }, + { + "epoch": 4.1478129713423835, + "grad_norm": 0.08723427145167717, + "learning_rate": 6.8726875282551356e-06, + "loss": 0.3761, + "step": 11000 + }, + { + "epoch": 4.148190045248869, + "grad_norm": 0.08685204576236093, + "learning_rate": 6.866787114564597e-06, + "loss": 0.3801, + "step": 11001 + }, + { + "epoch": 4.148567119155355, + "grad_norm": 0.08077093044951814, + "learning_rate": 6.860888996976318e-06, + "loss": 0.3788, + "step": 11002 + }, + { + "epoch": 4.1489441930618405, + "grad_norm": 0.07452805017173733, + "learning_rate": 6.85499317589903e-06, + "loss": 0.3688, + "step": 11003 + }, + { + "epoch": 4.149321266968326, + "grad_norm": 0.09008096356182366, + "learning_rate": 6.84909965174132e-06, + "loss": 0.3736, + "step": 11004 + }, + { + "epoch": 4.149698340874812, + "grad_norm": 0.07856214157844241, + "learning_rate": 6.843208424911609e-06, + "loss": 0.3795, + "step": 11005 + }, + { + "epoch": 4.1500754147812975, + "grad_norm": 0.08150496867319655, + "learning_rate": 6.837319495818122e-06, + "loss": 0.3785, + "step": 11006 + }, + { + "epoch": 4.150452488687783, + "grad_norm": 0.07896566727827307, + "learning_rate": 6.8314328648690034e-06, + "loss": 0.3833, + "step": 11007 + }, + { + "epoch": 4.150829562594269, + "grad_norm": 0.07848581403096172, + "learning_rate": 6.825548532472153e-06, + "loss": 0.381, + "step": 11008 + }, + { + "epoch": 4.1512066365007545, + "grad_norm": 0.08246247274464195, + "learning_rate": 6.819666499035364e-06, + "loss": 0.3849, + "step": 11009 + }, + { + "epoch": 4.15158371040724, + "grad_norm": 0.07927497443047395, + "learning_rate": 6.813786764966255e-06, + "loss": 0.3786, + "step": 11010 + }, + { + "epoch": 4.151960784313726, + "grad_norm": 0.08124132312876144, + "learning_rate": 6.807909330672298e-06, + "loss": 0.382, + "step": 11011 + }, + { + "epoch": 4.1523378582202115, + "grad_norm": 0.07754824028235281, + "learning_rate": 6.802034196560763e-06, + "loss": 0.3791, + "step": 11012 + }, + { + "epoch": 4.152714932126697, + "grad_norm": 0.08605926658244337, + "learning_rate": 6.796161363038831e-06, + "loss": 0.3852, + "step": 11013 + }, + { + "epoch": 4.153092006033183, + "grad_norm": 0.08180141100634945, + "learning_rate": 6.790290830513457e-06, + "loss": 0.3717, + "step": 11014 + }, + { + "epoch": 4.1534690799396685, + "grad_norm": 0.08509972213872441, + "learning_rate": 6.784422599391467e-06, + "loss": 0.3786, + "step": 11015 + }, + { + "epoch": 4.153846153846154, + "grad_norm": 0.08737903804090794, + "learning_rate": 6.778556670079535e-06, + "loss": 0.3835, + "step": 11016 + }, + { + "epoch": 4.15422322775264, + "grad_norm": 0.08284561700083183, + "learning_rate": 6.772693042984162e-06, + "loss": 0.3824, + "step": 11017 + }, + { + "epoch": 4.1546003016591255, + "grad_norm": 0.07945734448871712, + "learning_rate": 6.7668317185116725e-06, + "loss": 0.3808, + "step": 11018 + }, + { + "epoch": 4.154977375565611, + "grad_norm": 0.09069052427382714, + "learning_rate": 6.760972697068284e-06, + "loss": 0.3728, + "step": 11019 + }, + { + "epoch": 4.155354449472097, + "grad_norm": 0.07445566191363152, + "learning_rate": 6.755115979059997e-06, + "loss": 0.378, + "step": 11020 + }, + { + "epoch": 4.1557315233785825, + "grad_norm": 0.08452589537006512, + "learning_rate": 6.749261564892684e-06, + "loss": 0.3747, + "step": 11021 + }, + { + "epoch": 4.156108597285068, + "grad_norm": 0.07712495613642477, + "learning_rate": 6.7434094549720545e-06, + "loss": 0.3744, + "step": 11022 + }, + { + "epoch": 4.156485671191554, + "grad_norm": 0.08201452253372839, + "learning_rate": 6.737559649703653e-06, + "loss": 0.3653, + "step": 11023 + }, + { + "epoch": 4.1568627450980395, + "grad_norm": 0.08991413660732948, + "learning_rate": 6.7317121494928506e-06, + "loss": 0.3818, + "step": 11024 + }, + { + "epoch": 4.157239819004525, + "grad_norm": 0.07687782381637644, + "learning_rate": 6.725866954744908e-06, + "loss": 0.3767, + "step": 11025 + }, + { + "epoch": 4.157616892911011, + "grad_norm": 0.07996796293772143, + "learning_rate": 6.720024065864858e-06, + "loss": 0.383, + "step": 11026 + }, + { + "epoch": 4.157993966817497, + "grad_norm": 0.07756341474516863, + "learning_rate": 6.714183483257621e-06, + "loss": 0.3785, + "step": 11027 + }, + { + "epoch": 4.158371040723982, + "grad_norm": 0.07544393763559185, + "learning_rate": 6.708345207327944e-06, + "loss": 0.3813, + "step": 11028 + }, + { + "epoch": 4.158748114630468, + "grad_norm": 0.08172520227396782, + "learning_rate": 6.702509238480428e-06, + "loss": 0.3802, + "step": 11029 + }, + { + "epoch": 4.159125188536954, + "grad_norm": 0.08222601361664354, + "learning_rate": 6.696675577119465e-06, + "loss": 0.3868, + "step": 11030 + }, + { + "epoch": 4.159502262443439, + "grad_norm": 0.07478457462747506, + "learning_rate": 6.690844223649362e-06, + "loss": 0.3745, + "step": 11031 + }, + { + "epoch": 4.159879336349925, + "grad_norm": 0.08043989745369381, + "learning_rate": 6.685015178474201e-06, + "loss": 0.3726, + "step": 11032 + }, + { + "epoch": 4.160256410256411, + "grad_norm": 0.07974282484372358, + "learning_rate": 6.679188441997939e-06, + "loss": 0.3752, + "step": 11033 + }, + { + "epoch": 4.160633484162896, + "grad_norm": 0.07784625177459484, + "learning_rate": 6.6733640146243595e-06, + "loss": 0.3829, + "step": 11034 + }, + { + "epoch": 4.161010558069382, + "grad_norm": 0.08137917313262921, + "learning_rate": 6.667541896757104e-06, + "loss": 0.3846, + "step": 11035 + }, + { + "epoch": 4.161387631975868, + "grad_norm": 0.07788486550836275, + "learning_rate": 6.661722088799614e-06, + "loss": 0.3748, + "step": 11036 + }, + { + "epoch": 4.161764705882353, + "grad_norm": 0.08538184900210093, + "learning_rate": 6.655904591155224e-06, + "loss": 0.3799, + "step": 11037 + }, + { + "epoch": 4.162141779788839, + "grad_norm": 0.08532120859842338, + "learning_rate": 6.650089404227067e-06, + "loss": 0.3767, + "step": 11038 + }, + { + "epoch": 4.162518853695325, + "grad_norm": 0.08565301582275424, + "learning_rate": 6.64427652841813e-06, + "loss": 0.3891, + "step": 11039 + }, + { + "epoch": 4.16289592760181, + "grad_norm": 0.08273462619906181, + "learning_rate": 6.6384659641312425e-06, + "loss": 0.3842, + "step": 11040 + }, + { + "epoch": 4.163273001508296, + "grad_norm": 0.08703964775953758, + "learning_rate": 6.632657711769082e-06, + "loss": 0.3766, + "step": 11041 + }, + { + "epoch": 4.163650075414782, + "grad_norm": 0.07823231949966311, + "learning_rate": 6.6268517717341265e-06, + "loss": 0.384, + "step": 11042 + }, + { + "epoch": 4.164027149321267, + "grad_norm": 0.07974796641120933, + "learning_rate": 6.621048144428761e-06, + "loss": 0.3773, + "step": 11043 + }, + { + "epoch": 4.164404223227753, + "grad_norm": 0.08126356871103475, + "learning_rate": 6.6152468302551395e-06, + "loss": 0.3714, + "step": 11044 + }, + { + "epoch": 4.164781297134239, + "grad_norm": 0.07593826716552823, + "learning_rate": 6.6094478296153054e-06, + "loss": 0.3792, + "step": 11045 + }, + { + "epoch": 4.165158371040724, + "grad_norm": 0.08865696583807954, + "learning_rate": 6.6036511429111135e-06, + "loss": 0.3645, + "step": 11046 + }, + { + "epoch": 4.16553544494721, + "grad_norm": 0.08248243349758351, + "learning_rate": 6.597856770544289e-06, + "loss": 0.3785, + "step": 11047 + }, + { + "epoch": 4.165912518853696, + "grad_norm": 0.08239214345329689, + "learning_rate": 6.5920647129163396e-06, + "loss": 0.3879, + "step": 11048 + }, + { + "epoch": 4.166289592760181, + "grad_norm": 0.08681241764562384, + "learning_rate": 6.5862749704286965e-06, + "loss": 0.3842, + "step": 11049 + }, + { + "epoch": 4.166666666666667, + "grad_norm": 0.08083458847451822, + "learning_rate": 6.58048754348255e-06, + "loss": 0.3699, + "step": 11050 + }, + { + "epoch": 4.167043740573153, + "grad_norm": 0.08665939003921595, + "learning_rate": 6.57470243247897e-06, + "loss": 0.3761, + "step": 11051 + }, + { + "epoch": 4.167420814479638, + "grad_norm": 0.08507131053406759, + "learning_rate": 6.568919637818862e-06, + "loss": 0.3645, + "step": 11052 + }, + { + "epoch": 4.167797888386124, + "grad_norm": 0.09194274187147741, + "learning_rate": 6.563139159902987e-06, + "loss": 0.37, + "step": 11053 + }, + { + "epoch": 4.16817496229261, + "grad_norm": 0.0892540198745446, + "learning_rate": 6.557360999131885e-06, + "loss": 0.3903, + "step": 11054 + }, + { + "epoch": 4.168552036199095, + "grad_norm": 0.09563467035129787, + "learning_rate": 6.551585155906019e-06, + "loss": 0.3705, + "step": 11055 + }, + { + "epoch": 4.168929110105581, + "grad_norm": 0.09318333891872449, + "learning_rate": 6.545811630625629e-06, + "loss": 0.3848, + "step": 11056 + }, + { + "epoch": 4.169306184012067, + "grad_norm": 0.08665364309154716, + "learning_rate": 6.540040423690816e-06, + "loss": 0.392, + "step": 11057 + }, + { + "epoch": 4.169683257918552, + "grad_norm": 0.08478947582899918, + "learning_rate": 6.534271535501524e-06, + "loss": 0.3711, + "step": 11058 + }, + { + "epoch": 4.170060331825038, + "grad_norm": 0.08952792236596724, + "learning_rate": 6.5285049664575425e-06, + "loss": 0.3889, + "step": 11059 + }, + { + "epoch": 4.170437405731524, + "grad_norm": 0.08363685953951747, + "learning_rate": 6.522740716958455e-06, + "loss": 0.3927, + "step": 11060 + }, + { + "epoch": 4.170814479638009, + "grad_norm": 0.08397789367427036, + "learning_rate": 6.51697878740376e-06, + "loss": 0.3673, + "step": 11061 + }, + { + "epoch": 4.171191553544495, + "grad_norm": 0.08088200914132614, + "learning_rate": 6.511219178192729e-06, + "loss": 0.3707, + "step": 11062 + }, + { + "epoch": 4.171568627450981, + "grad_norm": 0.08007344824647186, + "learning_rate": 6.5054618897245e-06, + "loss": 0.3797, + "step": 11063 + }, + { + "epoch": 4.171945701357466, + "grad_norm": 0.08320798609664255, + "learning_rate": 6.4997069223980565e-06, + "loss": 0.3845, + "step": 11064 + }, + { + "epoch": 4.172322775263952, + "grad_norm": 0.0802499263210488, + "learning_rate": 6.493954276612217e-06, + "loss": 0.3727, + "step": 11065 + }, + { + "epoch": 4.172699849170438, + "grad_norm": 0.08242610522132353, + "learning_rate": 6.488203952765606e-06, + "loss": 0.3758, + "step": 11066 + }, + { + "epoch": 4.173076923076923, + "grad_norm": 0.07942052107321707, + "learning_rate": 6.482455951256756e-06, + "loss": 0.3839, + "step": 11067 + }, + { + "epoch": 4.173453996983409, + "grad_norm": 0.08475308328988558, + "learning_rate": 6.476710272483973e-06, + "loss": 0.377, + "step": 11068 + }, + { + "epoch": 4.173831070889895, + "grad_norm": 0.08244098494981006, + "learning_rate": 6.470966916845429e-06, + "loss": 0.3847, + "step": 11069 + }, + { + "epoch": 4.17420814479638, + "grad_norm": 0.0799835405271708, + "learning_rate": 6.465225884739137e-06, + "loss": 0.3865, + "step": 11070 + }, + { + "epoch": 4.174585218702866, + "grad_norm": 0.07445439323947083, + "learning_rate": 6.459487176562955e-06, + "loss": 0.372, + "step": 11071 + }, + { + "epoch": 4.174962292609352, + "grad_norm": 0.0809131256949087, + "learning_rate": 6.453750792714544e-06, + "loss": 0.3769, + "step": 11072 + }, + { + "epoch": 4.175339366515837, + "grad_norm": 0.09363774268500828, + "learning_rate": 6.448016733591465e-06, + "loss": 0.3847, + "step": 11073 + }, + { + "epoch": 4.175716440422323, + "grad_norm": 0.08153278841247716, + "learning_rate": 6.4422849995910575e-06, + "loss": 0.3715, + "step": 11074 + }, + { + "epoch": 4.176093514328809, + "grad_norm": 0.08301578325524255, + "learning_rate": 6.4365555911105334e-06, + "loss": 0.3631, + "step": 11075 + }, + { + "epoch": 4.176470588235294, + "grad_norm": 0.0891745138371161, + "learning_rate": 6.430828508546936e-06, + "loss": 0.3781, + "step": 11076 + }, + { + "epoch": 4.17684766214178, + "grad_norm": 0.08090520723508511, + "learning_rate": 6.425103752297159e-06, + "loss": 0.3807, + "step": 11077 + }, + { + "epoch": 4.177224736048266, + "grad_norm": 0.08541341043305023, + "learning_rate": 6.4193813227578875e-06, + "loss": 0.378, + "step": 11078 + }, + { + "epoch": 4.177601809954751, + "grad_norm": 0.09332708408988352, + "learning_rate": 6.413661220325726e-06, + "loss": 0.3759, + "step": 11079 + }, + { + "epoch": 4.177978883861237, + "grad_norm": 0.08877225060526049, + "learning_rate": 6.407943445397044e-06, + "loss": 0.3849, + "step": 11080 + }, + { + "epoch": 4.178355957767723, + "grad_norm": 0.08448589971199712, + "learning_rate": 6.40222799836808e-06, + "loss": 0.391, + "step": 11081 + }, + { + "epoch": 4.178733031674208, + "grad_norm": 0.09878385336145558, + "learning_rate": 6.396514879634916e-06, + "loss": 0.394, + "step": 11082 + }, + { + "epoch": 4.179110105580694, + "grad_norm": 0.09151299507865468, + "learning_rate": 6.390804089593477e-06, + "loss": 0.3801, + "step": 11083 + }, + { + "epoch": 4.17948717948718, + "grad_norm": 0.08553219390192936, + "learning_rate": 6.3850956286394835e-06, + "loss": 0.3729, + "step": 11084 + }, + { + "epoch": 4.179864253393665, + "grad_norm": 0.09424225907887007, + "learning_rate": 6.379389497168564e-06, + "loss": 0.3864, + "step": 11085 + }, + { + "epoch": 4.180241327300151, + "grad_norm": 0.08810030160771982, + "learning_rate": 6.373685695576113e-06, + "loss": 0.3745, + "step": 11086 + }, + { + "epoch": 4.180618401206637, + "grad_norm": 0.09588951700686303, + "learning_rate": 6.3679842242574356e-06, + "loss": 0.3801, + "step": 11087 + }, + { + "epoch": 4.180995475113122, + "grad_norm": 0.08144686219510754, + "learning_rate": 6.362285083607607e-06, + "loss": 0.3784, + "step": 11088 + }, + { + "epoch": 4.181372549019608, + "grad_norm": 0.09415952186063238, + "learning_rate": 6.356588274021592e-06, + "loss": 0.3771, + "step": 11089 + }, + { + "epoch": 4.181749622926094, + "grad_norm": 0.08708327756691582, + "learning_rate": 6.350893795894162e-06, + "loss": 0.3738, + "step": 11090 + }, + { + "epoch": 4.182126696832579, + "grad_norm": 0.0858778846170663, + "learning_rate": 6.345201649619959e-06, + "loss": 0.3768, + "step": 11091 + }, + { + "epoch": 4.182503770739065, + "grad_norm": 0.09633712426214103, + "learning_rate": 6.339511835593408e-06, + "loss": 0.3747, + "step": 11092 + }, + { + "epoch": 4.182880844645551, + "grad_norm": 0.08265089994211063, + "learning_rate": 6.333824354208848e-06, + "loss": 0.3738, + "step": 11093 + }, + { + "epoch": 4.183257918552036, + "grad_norm": 0.07915182840542949, + "learning_rate": 6.328139205860391e-06, + "loss": 0.3772, + "step": 11094 + }, + { + "epoch": 4.183634992458522, + "grad_norm": 0.08685488886976381, + "learning_rate": 6.322456390942022e-06, + "loss": 0.3717, + "step": 11095 + }, + { + "epoch": 4.184012066365008, + "grad_norm": 0.08191108706628904, + "learning_rate": 6.316775909847548e-06, + "loss": 0.3716, + "step": 11096 + }, + { + "epoch": 4.1843891402714934, + "grad_norm": 0.07659597467228314, + "learning_rate": 6.311097762970639e-06, + "loss": 0.3905, + "step": 11097 + }, + { + "epoch": 4.184766214177979, + "grad_norm": 0.0781371951245906, + "learning_rate": 6.305421950704755e-06, + "loss": 0.3733, + "step": 11098 + }, + { + "epoch": 4.185143288084465, + "grad_norm": 0.07640834725776563, + "learning_rate": 6.299748473443262e-06, + "loss": 0.3742, + "step": 11099 + }, + { + "epoch": 4.1855203619909505, + "grad_norm": 0.08493150677716133, + "learning_rate": 6.294077331579296e-06, + "loss": 0.3757, + "step": 11100 + }, + { + "epoch": 4.185897435897436, + "grad_norm": 0.0761731669012118, + "learning_rate": 6.288408525505874e-06, + "loss": 0.3705, + "step": 11101 + }, + { + "epoch": 4.186274509803922, + "grad_norm": 0.09194169675062464, + "learning_rate": 6.28274205561584e-06, + "loss": 0.3703, + "step": 11102 + }, + { + "epoch": 4.1866515837104075, + "grad_norm": 0.08835777932263114, + "learning_rate": 6.2770779223018865e-06, + "loss": 0.382, + "step": 11103 + }, + { + "epoch": 4.187028657616893, + "grad_norm": 0.09400763804634703, + "learning_rate": 6.271416125956497e-06, + "loss": 0.3718, + "step": 11104 + }, + { + "epoch": 4.187405731523379, + "grad_norm": 0.0893334848099108, + "learning_rate": 6.265756666972076e-06, + "loss": 0.3761, + "step": 11105 + }, + { + "epoch": 4.1877828054298645, + "grad_norm": 0.08413137836834242, + "learning_rate": 6.260099545740783e-06, + "loss": 0.3778, + "step": 11106 + }, + { + "epoch": 4.18815987933635, + "grad_norm": 0.08185164340877549, + "learning_rate": 6.254444762654666e-06, + "loss": 0.3786, + "step": 11107 + }, + { + "epoch": 4.188536953242836, + "grad_norm": 0.07702671646157282, + "learning_rate": 6.2487923181055925e-06, + "loss": 0.3833, + "step": 11108 + }, + { + "epoch": 4.1889140271493215, + "grad_norm": 0.08621134432793866, + "learning_rate": 6.243142212485276e-06, + "loss": 0.3779, + "step": 11109 + }, + { + "epoch": 4.189291101055807, + "grad_norm": 0.09035833243671158, + "learning_rate": 6.237494446185262e-06, + "loss": 0.3753, + "step": 11110 + }, + { + "epoch": 4.189668174962293, + "grad_norm": 0.07902620354273993, + "learning_rate": 6.2318490195969425e-06, + "loss": 0.3711, + "step": 11111 + }, + { + "epoch": 4.1900452488687785, + "grad_norm": 0.07976641274559458, + "learning_rate": 6.226205933111526e-06, + "loss": 0.367, + "step": 11112 + }, + { + "epoch": 4.190422322775264, + "grad_norm": 0.08813053556365022, + "learning_rate": 6.220565187120082e-06, + "loss": 0.3842, + "step": 11113 + }, + { + "epoch": 4.19079939668175, + "grad_norm": 0.08131587545551024, + "learning_rate": 6.214926782013506e-06, + "loss": 0.3762, + "step": 11114 + }, + { + "epoch": 4.1911764705882355, + "grad_norm": 0.07912221078307112, + "learning_rate": 6.209290718182539e-06, + "loss": 0.3721, + "step": 11115 + }, + { + "epoch": 4.191553544494721, + "grad_norm": 0.08793892689234936, + "learning_rate": 6.203656996017753e-06, + "loss": 0.3869, + "step": 11116 + }, + { + "epoch": 4.191930618401207, + "grad_norm": 0.07484820362840686, + "learning_rate": 6.198025615909569e-06, + "loss": 0.3706, + "step": 11117 + }, + { + "epoch": 4.1923076923076925, + "grad_norm": 0.08838618715822866, + "learning_rate": 6.1923965782482165e-06, + "loss": 0.3889, + "step": 11118 + }, + { + "epoch": 4.192684766214178, + "grad_norm": 0.08458517232140175, + "learning_rate": 6.186769883423793e-06, + "loss": 0.3715, + "step": 11119 + }, + { + "epoch": 4.193061840120664, + "grad_norm": 0.09172130029670676, + "learning_rate": 6.181145531826227e-06, + "loss": 0.3857, + "step": 11120 + }, + { + "epoch": 4.1934389140271495, + "grad_norm": 0.08099399501233465, + "learning_rate": 6.175523523845277e-06, + "loss": 0.373, + "step": 11121 + }, + { + "epoch": 4.193815987933635, + "grad_norm": 0.08327028761167718, + "learning_rate": 6.169903859870543e-06, + "loss": 0.3798, + "step": 11122 + }, + { + "epoch": 4.194193061840121, + "grad_norm": 0.08883419208870473, + "learning_rate": 6.1642865402914686e-06, + "loss": 0.3941, + "step": 11123 + }, + { + "epoch": 4.1945701357466065, + "grad_norm": 0.0822190486977687, + "learning_rate": 6.1586715654973296e-06, + "loss": 0.3768, + "step": 11124 + }, + { + "epoch": 4.194947209653092, + "grad_norm": 0.08426081371993535, + "learning_rate": 6.153058935877227e-06, + "loss": 0.3546, + "step": 11125 + }, + { + "epoch": 4.195324283559578, + "grad_norm": 0.09274475384003315, + "learning_rate": 6.147448651820118e-06, + "loss": 0.3811, + "step": 11126 + }, + { + "epoch": 4.1957013574660635, + "grad_norm": 0.07482100808982232, + "learning_rate": 6.141840713714788e-06, + "loss": 0.3703, + "step": 11127 + }, + { + "epoch": 4.196078431372549, + "grad_norm": 0.09250537518256591, + "learning_rate": 6.136235121949865e-06, + "loss": 0.3789, + "step": 11128 + }, + { + "epoch": 4.196455505279035, + "grad_norm": 0.08511616341831917, + "learning_rate": 6.130631876913815e-06, + "loss": 0.3733, + "step": 11129 + }, + { + "epoch": 4.1968325791855206, + "grad_norm": 0.07577666950998378, + "learning_rate": 6.125030978994937e-06, + "loss": 0.3697, + "step": 11130 + }, + { + "epoch": 4.197209653092006, + "grad_norm": 0.08189639761737587, + "learning_rate": 6.11943242858136e-06, + "loss": 0.3777, + "step": 11131 + }, + { + "epoch": 4.197586726998492, + "grad_norm": 0.0885721092282458, + "learning_rate": 6.113836226061063e-06, + "loss": 0.3718, + "step": 11132 + }, + { + "epoch": 4.197963800904978, + "grad_norm": 0.08033299194103205, + "learning_rate": 6.108242371821856e-06, + "loss": 0.3875, + "step": 11133 + }, + { + "epoch": 4.198340874811463, + "grad_norm": 0.09280440086943711, + "learning_rate": 6.102650866251396e-06, + "loss": 0.3838, + "step": 11134 + }, + { + "epoch": 4.198717948717949, + "grad_norm": 0.08316190059141591, + "learning_rate": 6.097061709737158e-06, + "loss": 0.3753, + "step": 11135 + }, + { + "epoch": 4.199095022624435, + "grad_norm": 0.0826878490665616, + "learning_rate": 6.0914749026664835e-06, + "loss": 0.3836, + "step": 11136 + }, + { + "epoch": 4.19947209653092, + "grad_norm": 0.08732831914999956, + "learning_rate": 6.085890445426512e-06, + "loss": 0.3706, + "step": 11137 + }, + { + "epoch": 4.199849170437406, + "grad_norm": 0.08846669784622767, + "learning_rate": 6.08030833840425e-06, + "loss": 0.3734, + "step": 11138 + }, + { + "epoch": 4.200226244343892, + "grad_norm": 0.09078171127474957, + "learning_rate": 6.074728581986536e-06, + "loss": 0.378, + "step": 11139 + }, + { + "epoch": 4.200603318250377, + "grad_norm": 0.08821676261881606, + "learning_rate": 6.06915117656004e-06, + "loss": 0.3865, + "step": 11140 + }, + { + "epoch": 4.200980392156863, + "grad_norm": 0.08355032157992696, + "learning_rate": 6.063576122511268e-06, + "loss": 0.3831, + "step": 11141 + }, + { + "epoch": 4.201357466063349, + "grad_norm": 0.08448502806988188, + "learning_rate": 6.058003420226582e-06, + "loss": 0.3779, + "step": 11142 + }, + { + "epoch": 4.201734539969834, + "grad_norm": 0.07764011728931605, + "learning_rate": 6.05243307009213e-06, + "loss": 0.3838, + "step": 11143 + }, + { + "epoch": 4.20211161387632, + "grad_norm": 0.08132572448176652, + "learning_rate": 6.046865072493973e-06, + "loss": 0.3802, + "step": 11144 + }, + { + "epoch": 4.202488687782806, + "grad_norm": 0.08186858999068533, + "learning_rate": 6.041299427817944e-06, + "loss": 0.3773, + "step": 11145 + }, + { + "epoch": 4.202865761689291, + "grad_norm": 0.07545535504392667, + "learning_rate": 6.035736136449739e-06, + "loss": 0.3797, + "step": 11146 + }, + { + "epoch": 4.203242835595777, + "grad_norm": 0.08079496631336383, + "learning_rate": 6.030175198774899e-06, + "loss": 0.3857, + "step": 11147 + }, + { + "epoch": 4.203619909502263, + "grad_norm": 0.07931491116533626, + "learning_rate": 6.024616615178791e-06, + "loss": 0.3793, + "step": 11148 + }, + { + "epoch": 4.203996983408748, + "grad_norm": 0.07532017019869883, + "learning_rate": 6.0190603860466e-06, + "loss": 0.3815, + "step": 11149 + }, + { + "epoch": 4.204374057315234, + "grad_norm": 0.08159427989673072, + "learning_rate": 6.013506511763401e-06, + "loss": 0.3829, + "step": 11150 + }, + { + "epoch": 4.20475113122172, + "grad_norm": 0.08040703535704229, + "learning_rate": 6.007954992714045e-06, + "loss": 0.3698, + "step": 11151 + }, + { + "epoch": 4.205128205128205, + "grad_norm": 0.07709558616095534, + "learning_rate": 6.002405829283256e-06, + "loss": 0.385, + "step": 11152 + }, + { + "epoch": 4.205505279034691, + "grad_norm": 0.07915314375111619, + "learning_rate": 5.996859021855592e-06, + "loss": 0.3892, + "step": 11153 + }, + { + "epoch": 4.205882352941177, + "grad_norm": 0.08144468510330172, + "learning_rate": 5.991314570815441e-06, + "loss": 0.3804, + "step": 11154 + }, + { + "epoch": 4.206259426847662, + "grad_norm": 0.07338820438388574, + "learning_rate": 5.98577247654701e-06, + "loss": 0.3798, + "step": 11155 + }, + { + "epoch": 4.206636500754148, + "grad_norm": 0.08102728527133846, + "learning_rate": 5.980232739434391e-06, + "loss": 0.3851, + "step": 11156 + }, + { + "epoch": 4.207013574660634, + "grad_norm": 0.08491698978952597, + "learning_rate": 5.97469535986146e-06, + "loss": 0.3794, + "step": 11157 + }, + { + "epoch": 4.207390648567119, + "grad_norm": 0.08174611023536603, + "learning_rate": 5.969160338211959e-06, + "loss": 0.3836, + "step": 11158 + }, + { + "epoch": 4.207767722473605, + "grad_norm": 0.08491131905195058, + "learning_rate": 5.963627674869461e-06, + "loss": 0.3745, + "step": 11159 + }, + { + "epoch": 4.208144796380091, + "grad_norm": 0.07759605410255126, + "learning_rate": 5.9580973702173836e-06, + "loss": 0.3808, + "step": 11160 + }, + { + "epoch": 4.208521870286576, + "grad_norm": 0.07673247033366212, + "learning_rate": 5.95256942463895e-06, + "loss": 0.39, + "step": 11161 + }, + { + "epoch": 4.208898944193062, + "grad_norm": 0.07880282210430431, + "learning_rate": 5.947043838517266e-06, + "loss": 0.3813, + "step": 11162 + }, + { + "epoch": 4.209276018099548, + "grad_norm": 0.07944882654495794, + "learning_rate": 5.941520612235234e-06, + "loss": 0.3738, + "step": 11163 + }, + { + "epoch": 4.209653092006033, + "grad_norm": 0.08399020376532453, + "learning_rate": 5.9359997461756155e-06, + "loss": 0.3898, + "step": 11164 + }, + { + "epoch": 4.210030165912519, + "grad_norm": 0.08340978964467526, + "learning_rate": 5.930481240720998e-06, + "loss": 0.3733, + "step": 11165 + }, + { + "epoch": 4.210407239819005, + "grad_norm": 0.08767142550226559, + "learning_rate": 5.924965096253821e-06, + "loss": 0.3712, + "step": 11166 + }, + { + "epoch": 4.21078431372549, + "grad_norm": 0.09417031066394588, + "learning_rate": 5.919451313156317e-06, + "loss": 0.398, + "step": 11167 + }, + { + "epoch": 4.211161387631976, + "grad_norm": 0.08725685138395493, + "learning_rate": 5.913939891810629e-06, + "loss": 0.3884, + "step": 11168 + }, + { + "epoch": 4.211538461538462, + "grad_norm": 0.0991810061955475, + "learning_rate": 5.908430832598662e-06, + "loss": 0.384, + "step": 11169 + }, + { + "epoch": 4.211915535444947, + "grad_norm": 0.07901528469784104, + "learning_rate": 5.9029241359021975e-06, + "loss": 0.3803, + "step": 11170 + }, + { + "epoch": 4.212292609351433, + "grad_norm": 0.09137739418695524, + "learning_rate": 5.89741980210285e-06, + "loss": 0.3942, + "step": 11171 + }, + { + "epoch": 4.212669683257919, + "grad_norm": 0.08884857242009662, + "learning_rate": 5.891917831582068e-06, + "loss": 0.3778, + "step": 11172 + }, + { + "epoch": 4.213046757164404, + "grad_norm": 0.07613772061998343, + "learning_rate": 5.886418224721109e-06, + "loss": 0.378, + "step": 11173 + }, + { + "epoch": 4.21342383107089, + "grad_norm": 0.08932228534340209, + "learning_rate": 5.880920981901126e-06, + "loss": 0.3828, + "step": 11174 + }, + { + "epoch": 4.213800904977376, + "grad_norm": 0.0896572183109884, + "learning_rate": 5.875426103503046e-06, + "loss": 0.3715, + "step": 11175 + }, + { + "epoch": 4.214177978883861, + "grad_norm": 0.07601558767144886, + "learning_rate": 5.869933589907666e-06, + "loss": 0.3848, + "step": 11176 + }, + { + "epoch": 4.214555052790347, + "grad_norm": 0.0828078009838109, + "learning_rate": 5.86444344149562e-06, + "loss": 0.372, + "step": 11177 + }, + { + "epoch": 4.214932126696833, + "grad_norm": 0.08168136177125651, + "learning_rate": 5.858955658647372e-06, + "loss": 0.3796, + "step": 11178 + }, + { + "epoch": 4.215309200603318, + "grad_norm": 0.0770359984042378, + "learning_rate": 5.853470241743195e-06, + "loss": 0.3606, + "step": 11179 + }, + { + "epoch": 4.215686274509804, + "grad_norm": 0.07545282082952481, + "learning_rate": 5.847987191163263e-06, + "loss": 0.3846, + "step": 11180 + }, + { + "epoch": 4.21606334841629, + "grad_norm": 0.07641928502284533, + "learning_rate": 5.842506507287513e-06, + "loss": 0.3638, + "step": 11181 + }, + { + "epoch": 4.216440422322775, + "grad_norm": 0.07470087628939345, + "learning_rate": 5.8370281904957686e-06, + "loss": 0.3861, + "step": 11182 + }, + { + "epoch": 4.216817496229261, + "grad_norm": 0.08698595255044786, + "learning_rate": 5.831552241167666e-06, + "loss": 0.3732, + "step": 11183 + }, + { + "epoch": 4.217194570135747, + "grad_norm": 0.07198647411149407, + "learning_rate": 5.826078659682695e-06, + "loss": 0.3826, + "step": 11184 + }, + { + "epoch": 4.217571644042232, + "grad_norm": 0.07352387289542732, + "learning_rate": 5.820607446420141e-06, + "loss": 0.38, + "step": 11185 + }, + { + "epoch": 4.217948717948718, + "grad_norm": 0.0825866132326269, + "learning_rate": 5.815138601759192e-06, + "loss": 0.3863, + "step": 11186 + }, + { + "epoch": 4.218325791855204, + "grad_norm": 0.07705427178402957, + "learning_rate": 5.809672126078809e-06, + "loss": 0.3809, + "step": 11187 + }, + { + "epoch": 4.218702865761689, + "grad_norm": 0.08106208905357697, + "learning_rate": 5.80420801975782e-06, + "loss": 0.3768, + "step": 11188 + }, + { + "epoch": 4.219079939668175, + "grad_norm": 0.0778820838861759, + "learning_rate": 5.7987462831748805e-06, + "loss": 0.3777, + "step": 11189 + }, + { + "epoch": 4.219457013574661, + "grad_norm": 0.078182431742442, + "learning_rate": 5.793286916708503e-06, + "loss": 0.3797, + "step": 11190 + }, + { + "epoch": 4.219834087481146, + "grad_norm": 0.07951647483749652, + "learning_rate": 5.787829920736978e-06, + "loss": 0.381, + "step": 11191 + }, + { + "epoch": 4.220211161387632, + "grad_norm": 0.0844731497150725, + "learning_rate": 5.782375295638512e-06, + "loss": 0.3889, + "step": 11192 + }, + { + "epoch": 4.220588235294118, + "grad_norm": 0.07834319900266437, + "learning_rate": 5.776923041791076e-06, + "loss": 0.3787, + "step": 11193 + }, + { + "epoch": 4.220965309200603, + "grad_norm": 0.08610013328756382, + "learning_rate": 5.77147315957252e-06, + "loss": 0.3841, + "step": 11194 + }, + { + "epoch": 4.221342383107089, + "grad_norm": 0.08695140197194029, + "learning_rate": 5.766025649360516e-06, + "loss": 0.3839, + "step": 11195 + }, + { + "epoch": 4.221719457013575, + "grad_norm": 0.07872487256567698, + "learning_rate": 5.760580511532574e-06, + "loss": 0.3794, + "step": 11196 + }, + { + "epoch": 4.22209653092006, + "grad_norm": 0.07876286857644002, + "learning_rate": 5.755137746466015e-06, + "loss": 0.3858, + "step": 11197 + }, + { + "epoch": 4.222473604826546, + "grad_norm": 0.08115539672309982, + "learning_rate": 5.749697354538053e-06, + "loss": 0.3782, + "step": 11198 + }, + { + "epoch": 4.222850678733032, + "grad_norm": 0.07594333369271278, + "learning_rate": 5.744259336125675e-06, + "loss": 0.3785, + "step": 11199 + }, + { + "epoch": 4.223227752639517, + "grad_norm": 0.07722021965847613, + "learning_rate": 5.738823691605743e-06, + "loss": 0.3785, + "step": 11200 + }, + { + "epoch": 4.223604826546003, + "grad_norm": 0.07401398477393627, + "learning_rate": 5.733390421354936e-06, + "loss": 0.3764, + "step": 11201 + }, + { + "epoch": 4.223981900452489, + "grad_norm": 0.07954199023671293, + "learning_rate": 5.727959525749791e-06, + "loss": 0.3786, + "step": 11202 + }, + { + "epoch": 4.2243589743589745, + "grad_norm": 0.07525393980762936, + "learning_rate": 5.7225310051666296e-06, + "loss": 0.3749, + "step": 11203 + }, + { + "epoch": 4.22473604826546, + "grad_norm": 0.0763585836740414, + "learning_rate": 5.717104859981684e-06, + "loss": 0.3795, + "step": 11204 + }, + { + "epoch": 4.225113122171946, + "grad_norm": 0.07968292881335863, + "learning_rate": 5.711681090570955e-06, + "loss": 0.3823, + "step": 11205 + }, + { + "epoch": 4.2254901960784315, + "grad_norm": 0.07864356502316754, + "learning_rate": 5.706259697310313e-06, + "loss": 0.3733, + "step": 11206 + }, + { + "epoch": 4.225867269984917, + "grad_norm": 0.0776864465418244, + "learning_rate": 5.700840680575455e-06, + "loss": 0.3644, + "step": 11207 + }, + { + "epoch": 4.226244343891403, + "grad_norm": 0.07964548808240535, + "learning_rate": 5.695424040741926e-06, + "loss": 0.3766, + "step": 11208 + }, + { + "epoch": 4.2266214177978885, + "grad_norm": 0.07484723156111675, + "learning_rate": 5.690009778185062e-06, + "loss": 0.3684, + "step": 11209 + }, + { + "epoch": 4.226998491704374, + "grad_norm": 0.07262862222918395, + "learning_rate": 5.684597893280104e-06, + "loss": 0.3856, + "step": 11210 + }, + { + "epoch": 4.22737556561086, + "grad_norm": 0.07993953480558277, + "learning_rate": 5.679188386402059e-06, + "loss": 0.3773, + "step": 11211 + }, + { + "epoch": 4.2277526395173455, + "grad_norm": 0.08372093427806365, + "learning_rate": 5.673781257925832e-06, + "loss": 0.3869, + "step": 11212 + }, + { + "epoch": 4.228129713423831, + "grad_norm": 0.07941450880559538, + "learning_rate": 5.66837650822611e-06, + "loss": 0.3807, + "step": 11213 + }, + { + "epoch": 4.228506787330317, + "grad_norm": 0.07782825368556318, + "learning_rate": 5.662974137677442e-06, + "loss": 0.385, + "step": 11214 + }, + { + "epoch": 4.2288838612368025, + "grad_norm": 0.07866540680437634, + "learning_rate": 5.6575741466542125e-06, + "loss": 0.3804, + "step": 11215 + }, + { + "epoch": 4.229260935143288, + "grad_norm": 0.08258977850303256, + "learning_rate": 5.652176535530638e-06, + "loss": 0.3853, + "step": 11216 + }, + { + "epoch": 4.229638009049774, + "grad_norm": 0.07968797669606728, + "learning_rate": 5.646781304680748e-06, + "loss": 0.3731, + "step": 11217 + }, + { + "epoch": 4.2300150829562595, + "grad_norm": 0.07901483748268251, + "learning_rate": 5.641388454478458e-06, + "loss": 0.3723, + "step": 11218 + }, + { + "epoch": 4.230392156862745, + "grad_norm": 0.0735952085969394, + "learning_rate": 5.635997985297463e-06, + "loss": 0.3757, + "step": 11219 + }, + { + "epoch": 4.230769230769231, + "grad_norm": 0.08304671831832787, + "learning_rate": 5.630609897511328e-06, + "loss": 0.3843, + "step": 11220 + }, + { + "epoch": 4.2311463046757165, + "grad_norm": 0.08456203710771613, + "learning_rate": 5.625224191493442e-06, + "loss": 0.3835, + "step": 11221 + }, + { + "epoch": 4.231523378582202, + "grad_norm": 0.07671914518081324, + "learning_rate": 5.619840867617039e-06, + "loss": 0.3899, + "step": 11222 + }, + { + "epoch": 4.231900452488688, + "grad_norm": 0.08030959305313397, + "learning_rate": 5.6144599262551465e-06, + "loss": 0.3702, + "step": 11223 + }, + { + "epoch": 4.2322775263951735, + "grad_norm": 0.08049755558188215, + "learning_rate": 5.609081367780702e-06, + "loss": 0.3797, + "step": 11224 + }, + { + "epoch": 4.232654600301659, + "grad_norm": 0.08235979316105033, + "learning_rate": 5.603705192566406e-06, + "loss": 0.3771, + "step": 11225 + }, + { + "epoch": 4.233031674208145, + "grad_norm": 0.08702499326737699, + "learning_rate": 5.598331400984829e-06, + "loss": 0.386, + "step": 11226 + }, + { + "epoch": 4.2334087481146305, + "grad_norm": 0.07981678889649572, + "learning_rate": 5.592959993408374e-06, + "loss": 0.3768, + "step": 11227 + }, + { + "epoch": 4.233785822021116, + "grad_norm": 0.07820118585103751, + "learning_rate": 5.58759097020928e-06, + "loss": 0.3889, + "step": 11228 + }, + { + "epoch": 4.234162895927602, + "grad_norm": 0.08858365959110646, + "learning_rate": 5.5822243317595895e-06, + "loss": 0.3868, + "step": 11229 + }, + { + "epoch": 4.2345399698340875, + "grad_norm": 0.0770154392624374, + "learning_rate": 5.576860078431243e-06, + "loss": 0.38, + "step": 11230 + }, + { + "epoch": 4.234917043740573, + "grad_norm": 0.08126853463099912, + "learning_rate": 5.5714982105959495e-06, + "loss": 0.375, + "step": 11231 + }, + { + "epoch": 4.235294117647059, + "grad_norm": 0.07521132971244933, + "learning_rate": 5.566138728625294e-06, + "loss": 0.3728, + "step": 11232 + }, + { + "epoch": 4.2356711915535445, + "grad_norm": 0.0786700093662033, + "learning_rate": 5.560781632890679e-06, + "loss": 0.3845, + "step": 11233 + }, + { + "epoch": 4.23604826546003, + "grad_norm": 0.08914344291571462, + "learning_rate": 5.5554269237633585e-06, + "loss": 0.3795, + "step": 11234 + }, + { + "epoch": 4.236425339366516, + "grad_norm": 0.07691421174824846, + "learning_rate": 5.550074601614386e-06, + "loss": 0.3818, + "step": 11235 + }, + { + "epoch": 4.2368024132730016, + "grad_norm": 0.07570473934024734, + "learning_rate": 5.5447246668147e-06, + "loss": 0.3877, + "step": 11236 + }, + { + "epoch": 4.237179487179487, + "grad_norm": 0.08003735523140507, + "learning_rate": 5.539377119735028e-06, + "loss": 0.3777, + "step": 11237 + }, + { + "epoch": 4.237556561085973, + "grad_norm": 0.08087675572350836, + "learning_rate": 5.534031960745951e-06, + "loss": 0.3682, + "step": 11238 + }, + { + "epoch": 4.237933634992459, + "grad_norm": 0.07294751468013168, + "learning_rate": 5.528689190217891e-06, + "loss": 0.3712, + "step": 11239 + }, + { + "epoch": 4.238310708898944, + "grad_norm": 0.08871838872738111, + "learning_rate": 5.523348808521101e-06, + "loss": 0.3804, + "step": 11240 + }, + { + "epoch": 4.23868778280543, + "grad_norm": 0.07415401843597388, + "learning_rate": 5.51801081602564e-06, + "loss": 0.3731, + "step": 11241 + }, + { + "epoch": 4.239064856711916, + "grad_norm": 0.0757270207269462, + "learning_rate": 5.512675213101464e-06, + "loss": 0.3804, + "step": 11242 + }, + { + "epoch": 4.239441930618401, + "grad_norm": 0.07650430394219969, + "learning_rate": 5.507342000118292e-06, + "loss": 0.3797, + "step": 11243 + }, + { + "epoch": 4.239819004524887, + "grad_norm": 0.0749469707325687, + "learning_rate": 5.502011177445728e-06, + "loss": 0.3877, + "step": 11244 + }, + { + "epoch": 4.240196078431373, + "grad_norm": 0.07895252518489784, + "learning_rate": 5.49668274545319e-06, + "loss": 0.3928, + "step": 11245 + }, + { + "epoch": 4.240573152337858, + "grad_norm": 0.07841897147222732, + "learning_rate": 5.49135670450994e-06, + "loss": 0.3744, + "step": 11246 + }, + { + "epoch": 4.240950226244344, + "grad_norm": 0.08596730389423982, + "learning_rate": 5.486033054985043e-06, + "loss": 0.3863, + "step": 11247 + }, + { + "epoch": 4.24132730015083, + "grad_norm": 0.07522727688036208, + "learning_rate": 5.480711797247464e-06, + "loss": 0.3808, + "step": 11248 + }, + { + "epoch": 4.241704374057315, + "grad_norm": 0.08307492081629711, + "learning_rate": 5.475392931665924e-06, + "loss": 0.3749, + "step": 11249 + }, + { + "epoch": 4.242081447963801, + "grad_norm": 0.08046355509259803, + "learning_rate": 5.470076458609037e-06, + "loss": 0.3805, + "step": 11250 + }, + { + "epoch": 4.242458521870287, + "grad_norm": 0.07995438648217318, + "learning_rate": 5.464762378445216e-06, + "loss": 0.3782, + "step": 11251 + }, + { + "epoch": 4.242835595776772, + "grad_norm": 0.0766055867532103, + "learning_rate": 5.459450691542745e-06, + "loss": 0.3666, + "step": 11252 + }, + { + "epoch": 4.243212669683258, + "grad_norm": 0.0796391453487555, + "learning_rate": 5.454141398269683e-06, + "loss": 0.3671, + "step": 11253 + }, + { + "epoch": 4.243589743589744, + "grad_norm": 0.08055998786609295, + "learning_rate": 5.4488344989939955e-06, + "loss": 0.3868, + "step": 11254 + }, + { + "epoch": 4.243966817496229, + "grad_norm": 0.08067416213133621, + "learning_rate": 5.4435299940834275e-06, + "loss": 0.3814, + "step": 11255 + }, + { + "epoch": 4.244343891402715, + "grad_norm": 0.08525524327608985, + "learning_rate": 5.438227883905573e-06, + "loss": 0.3814, + "step": 11256 + }, + { + "epoch": 4.244720965309201, + "grad_norm": 0.080209077088242, + "learning_rate": 5.432928168827874e-06, + "loss": 0.3818, + "step": 11257 + }, + { + "epoch": 4.245098039215686, + "grad_norm": 0.07774436109834965, + "learning_rate": 5.427630849217589e-06, + "loss": 0.3731, + "step": 11258 + }, + { + "epoch": 4.245475113122172, + "grad_norm": 0.07542742478075652, + "learning_rate": 5.422335925441822e-06, + "loss": 0.377, + "step": 11259 + }, + { + "epoch": 4.245852187028658, + "grad_norm": 0.08260650919815465, + "learning_rate": 5.417043397867514e-06, + "loss": 0.3838, + "step": 11260 + }, + { + "epoch": 4.246229260935143, + "grad_norm": 0.08114896307124014, + "learning_rate": 5.411753266861417e-06, + "loss": 0.382, + "step": 11261 + }, + { + "epoch": 4.246606334841629, + "grad_norm": 0.0834729339856979, + "learning_rate": 5.406465532790135e-06, + "loss": 0.3743, + "step": 11262 + }, + { + "epoch": 4.246983408748115, + "grad_norm": 0.08249135995258414, + "learning_rate": 5.401180196020108e-06, + "loss": 0.3745, + "step": 11263 + }, + { + "epoch": 4.2473604826546, + "grad_norm": 0.082731812866161, + "learning_rate": 5.395897256917604e-06, + "loss": 0.3822, + "step": 11264 + }, + { + "epoch": 4.247737556561086, + "grad_norm": 0.07830519072843031, + "learning_rate": 5.390616715848729e-06, + "loss": 0.3795, + "step": 11265 + }, + { + "epoch": 4.248114630467572, + "grad_norm": 0.07667434957207098, + "learning_rate": 5.385338573179422e-06, + "loss": 0.3872, + "step": 11266 + }, + { + "epoch": 4.248491704374057, + "grad_norm": 0.07476697311952248, + "learning_rate": 5.380062829275443e-06, + "loss": 0.3735, + "step": 11267 + }, + { + "epoch": 4.248868778280543, + "grad_norm": 0.08735487269314428, + "learning_rate": 5.3747894845024025e-06, + "loss": 0.3788, + "step": 11268 + }, + { + "epoch": 4.249245852187029, + "grad_norm": 0.07809902415015062, + "learning_rate": 5.369518539225737e-06, + "loss": 0.3758, + "step": 11269 + }, + { + "epoch": 4.249622926093514, + "grad_norm": 0.0800627053796241, + "learning_rate": 5.36424999381072e-06, + "loss": 0.3892, + "step": 11270 + }, + { + "epoch": 4.25, + "grad_norm": 0.08254319191033119, + "learning_rate": 5.358983848622452e-06, + "loss": 0.3786, + "step": 11271 + }, + { + "epoch": 4.250377073906486, + "grad_norm": 0.07329984045402808, + "learning_rate": 5.353720104025884e-06, + "loss": 0.3687, + "step": 11272 + }, + { + "epoch": 4.250754147812971, + "grad_norm": 0.07610045033421363, + "learning_rate": 5.348458760385784e-06, + "loss": 0.3821, + "step": 11273 + }, + { + "epoch": 4.251131221719457, + "grad_norm": 0.07735481866692683, + "learning_rate": 5.343199818066751e-06, + "loss": 0.3741, + "step": 11274 + }, + { + "epoch": 4.251508295625943, + "grad_norm": 0.07731313505411663, + "learning_rate": 5.337943277433231e-06, + "loss": 0.3775, + "step": 11275 + }, + { + "epoch": 4.251885369532428, + "grad_norm": 0.07406718669395923, + "learning_rate": 5.332689138849492e-06, + "loss": 0.3752, + "step": 11276 + }, + { + "epoch": 4.252262443438914, + "grad_norm": 0.07417705861470733, + "learning_rate": 5.3274374026796516e-06, + "loss": 0.3798, + "step": 11277 + }, + { + "epoch": 4.2526395173454, + "grad_norm": 0.0726537723731238, + "learning_rate": 5.322188069287642e-06, + "loss": 0.3745, + "step": 11278 + }, + { + "epoch": 4.253016591251885, + "grad_norm": 0.07885939480875427, + "learning_rate": 5.316941139037246e-06, + "loss": 0.3745, + "step": 11279 + }, + { + "epoch": 4.253393665158371, + "grad_norm": 0.074711877458922, + "learning_rate": 5.311696612292059e-06, + "loss": 0.3782, + "step": 11280 + }, + { + "epoch": 4.253770739064857, + "grad_norm": 0.07334351644904998, + "learning_rate": 5.306454489415531e-06, + "loss": 0.37, + "step": 11281 + }, + { + "epoch": 4.254147812971342, + "grad_norm": 0.0792152304500388, + "learning_rate": 5.301214770770928e-06, + "loss": 0.3779, + "step": 11282 + }, + { + "epoch": 4.254524886877828, + "grad_norm": 0.0752595028877866, + "learning_rate": 5.295977456721368e-06, + "loss": 0.3778, + "step": 11283 + }, + { + "epoch": 4.254901960784314, + "grad_norm": 0.07926122178715117, + "learning_rate": 5.290742547629788e-06, + "loss": 0.3701, + "step": 11284 + }, + { + "epoch": 4.255279034690799, + "grad_norm": 0.08233272932209658, + "learning_rate": 5.285510043858968e-06, + "loss": 0.3692, + "step": 11285 + }, + { + "epoch": 4.255656108597285, + "grad_norm": 0.07206555163007657, + "learning_rate": 5.280279945771503e-06, + "loss": 0.3731, + "step": 11286 + }, + { + "epoch": 4.256033182503771, + "grad_norm": 0.0862018062090147, + "learning_rate": 5.27505225372984e-06, + "loss": 0.3784, + "step": 11287 + }, + { + "epoch": 4.256410256410256, + "grad_norm": 0.0817356041850592, + "learning_rate": 5.269826968096259e-06, + "loss": 0.3843, + "step": 11288 + }, + { + "epoch": 4.256787330316742, + "grad_norm": 0.07790978593365207, + "learning_rate": 5.264604089232861e-06, + "loss": 0.3845, + "step": 11289 + }, + { + "epoch": 4.257164404223228, + "grad_norm": 0.08363132305675884, + "learning_rate": 5.259383617501588e-06, + "loss": 0.3711, + "step": 11290 + }, + { + "epoch": 4.257541478129713, + "grad_norm": 0.08773697661987892, + "learning_rate": 5.2541655532642255e-06, + "loss": 0.3811, + "step": 11291 + }, + { + "epoch": 4.257918552036199, + "grad_norm": 0.08414264387188461, + "learning_rate": 5.248949896882351e-06, + "loss": 0.3806, + "step": 11292 + }, + { + "epoch": 4.258295625942685, + "grad_norm": 0.09292969597658869, + "learning_rate": 5.243736648717446e-06, + "loss": 0.3835, + "step": 11293 + }, + { + "epoch": 4.25867269984917, + "grad_norm": 0.07950219273260202, + "learning_rate": 5.238525809130752e-06, + "loss": 0.3786, + "step": 11294 + }, + { + "epoch": 4.259049773755656, + "grad_norm": 0.0800277024341748, + "learning_rate": 5.233317378483387e-06, + "loss": 0.371, + "step": 11295 + }, + { + "epoch": 4.259426847662142, + "grad_norm": 0.08551165929560416, + "learning_rate": 5.228111357136287e-06, + "loss": 0.3845, + "step": 11296 + }, + { + "epoch": 4.259803921568627, + "grad_norm": 0.07768153977767812, + "learning_rate": 5.222907745450236e-06, + "loss": 0.3674, + "step": 11297 + }, + { + "epoch": 4.260180995475113, + "grad_norm": 0.08646180102256894, + "learning_rate": 5.217706543785817e-06, + "loss": 0.3842, + "step": 11298 + }, + { + "epoch": 4.260558069381599, + "grad_norm": 0.07974751771013217, + "learning_rate": 5.212507752503495e-06, + "loss": 0.3798, + "step": 11299 + }, + { + "epoch": 4.260935143288084, + "grad_norm": 0.08013369062226341, + "learning_rate": 5.207311371963526e-06, + "loss": 0.3831, + "step": 11300 + }, + { + "epoch": 4.26131221719457, + "grad_norm": 0.0851216100434455, + "learning_rate": 5.202117402526017e-06, + "loss": 0.377, + "step": 11301 + }, + { + "epoch": 4.261689291101056, + "grad_norm": 0.07618299746148943, + "learning_rate": 5.196925844550902e-06, + "loss": 0.3824, + "step": 11302 + }, + { + "epoch": 4.262066365007541, + "grad_norm": 0.07813037429294428, + "learning_rate": 5.191736698397969e-06, + "loss": 0.3941, + "step": 11303 + }, + { + "epoch": 4.262443438914027, + "grad_norm": 0.08134716696216737, + "learning_rate": 5.186549964426788e-06, + "loss": 0.3809, + "step": 11304 + }, + { + "epoch": 4.262820512820513, + "grad_norm": 0.07505243592257474, + "learning_rate": 5.181365642996832e-06, + "loss": 0.378, + "step": 11305 + }, + { + "epoch": 4.2631975867269984, + "grad_norm": 0.07914207074329131, + "learning_rate": 5.176183734467346e-06, + "loss": 0.3766, + "step": 11306 + }, + { + "epoch": 4.263574660633484, + "grad_norm": 0.08181601682394032, + "learning_rate": 5.171004239197439e-06, + "loss": 0.3878, + "step": 11307 + }, + { + "epoch": 4.26395173453997, + "grad_norm": 0.07666544518412746, + "learning_rate": 5.165827157546041e-06, + "loss": 0.3673, + "step": 11308 + }, + { + "epoch": 4.2643288084464555, + "grad_norm": 0.07530220439921217, + "learning_rate": 5.160652489871938e-06, + "loss": 0.383, + "step": 11309 + }, + { + "epoch": 4.264705882352941, + "grad_norm": 0.07442174280130419, + "learning_rate": 5.15548023653369e-06, + "loss": 0.3755, + "step": 11310 + }, + { + "epoch": 4.265082956259427, + "grad_norm": 0.07582931065713602, + "learning_rate": 5.150310397889779e-06, + "loss": 0.3687, + "step": 11311 + }, + { + "epoch": 4.2654600301659125, + "grad_norm": 0.07969648037456646, + "learning_rate": 5.145142974298431e-06, + "loss": 0.3807, + "step": 11312 + }, + { + "epoch": 4.265837104072398, + "grad_norm": 0.07686158415792642, + "learning_rate": 5.139977966117764e-06, + "loss": 0.3711, + "step": 11313 + }, + { + "epoch": 4.266214177978884, + "grad_norm": 0.07899642817613797, + "learning_rate": 5.1348153737057e-06, + "loss": 0.3678, + "step": 11314 + }, + { + "epoch": 4.2665912518853695, + "grad_norm": 0.07468358804995452, + "learning_rate": 5.129655197420013e-06, + "loss": 0.3769, + "step": 11315 + }, + { + "epoch": 4.266968325791855, + "grad_norm": 0.08059349505113188, + "learning_rate": 5.124497437618279e-06, + "loss": 0.3835, + "step": 11316 + }, + { + "epoch": 4.267345399698341, + "grad_norm": 0.08066494136253048, + "learning_rate": 5.1193420946579505e-06, + "loss": 0.3776, + "step": 11317 + }, + { + "epoch": 4.2677224736048265, + "grad_norm": 0.07435402595610284, + "learning_rate": 5.114189168896268e-06, + "loss": 0.3793, + "step": 11318 + }, + { + "epoch": 4.268099547511312, + "grad_norm": 0.07882408749600828, + "learning_rate": 5.109038660690333e-06, + "loss": 0.379, + "step": 11319 + }, + { + "epoch": 4.268476621417798, + "grad_norm": 0.0749641081371647, + "learning_rate": 5.103890570397072e-06, + "loss": 0.3695, + "step": 11320 + }, + { + "epoch": 4.2688536953242835, + "grad_norm": 0.07206405934457324, + "learning_rate": 5.098744898373244e-06, + "loss": 0.3798, + "step": 11321 + }, + { + "epoch": 4.269230769230769, + "grad_norm": 0.07435707857501068, + "learning_rate": 5.093601644975428e-06, + "loss": 0.3829, + "step": 11322 + }, + { + "epoch": 4.269607843137255, + "grad_norm": 0.08145419933322506, + "learning_rate": 5.08846081056007e-06, + "loss": 0.3759, + "step": 11323 + }, + { + "epoch": 4.2699849170437405, + "grad_norm": 0.07139159383249091, + "learning_rate": 5.083322395483401e-06, + "loss": 0.3773, + "step": 11324 + }, + { + "epoch": 4.270361990950226, + "grad_norm": 0.07296407768204421, + "learning_rate": 5.078186400101519e-06, + "loss": 0.3778, + "step": 11325 + }, + { + "epoch": 4.270739064856712, + "grad_norm": 0.07371791361454674, + "learning_rate": 5.073052824770348e-06, + "loss": 0.3882, + "step": 11326 + }, + { + "epoch": 4.2711161387631975, + "grad_norm": 0.07486625744715728, + "learning_rate": 5.067921669845639e-06, + "loss": 0.378, + "step": 11327 + }, + { + "epoch": 4.271493212669683, + "grad_norm": 0.07139370497094465, + "learning_rate": 5.0627929356829605e-06, + "loss": 0.3727, + "step": 11328 + }, + { + "epoch": 4.271870286576169, + "grad_norm": 0.07580624406947722, + "learning_rate": 5.057666622637758e-06, + "loss": 0.3829, + "step": 11329 + }, + { + "epoch": 4.2722473604826545, + "grad_norm": 0.07291637216617368, + "learning_rate": 5.0525427310652575e-06, + "loss": 0.3723, + "step": 11330 + }, + { + "epoch": 4.27262443438914, + "grad_norm": 0.07985651573442512, + "learning_rate": 5.0474212613205445e-06, + "loss": 0.3823, + "step": 11331 + }, + { + "epoch": 4.273001508295626, + "grad_norm": 0.07374700451362536, + "learning_rate": 5.042302213758539e-06, + "loss": 0.3794, + "step": 11332 + }, + { + "epoch": 4.2733785822021115, + "grad_norm": 0.07896136543351544, + "learning_rate": 5.037185588733988e-06, + "loss": 0.3844, + "step": 11333 + }, + { + "epoch": 4.273755656108597, + "grad_norm": 0.07907282357932456, + "learning_rate": 5.032071386601449e-06, + "loss": 0.3761, + "step": 11334 + }, + { + "epoch": 4.274132730015083, + "grad_norm": 0.07446758564118235, + "learning_rate": 5.0269596077153625e-06, + "loss": 0.385, + "step": 11335 + }, + { + "epoch": 4.2745098039215685, + "grad_norm": 0.07205120053657375, + "learning_rate": 5.021850252429938e-06, + "loss": 0.3751, + "step": 11336 + }, + { + "epoch": 4.274886877828054, + "grad_norm": 0.07593043233318422, + "learning_rate": 5.0167433210992796e-06, + "loss": 0.3758, + "step": 11337 + }, + { + "epoch": 4.27526395173454, + "grad_norm": 0.07873323069426043, + "learning_rate": 5.011638814077273e-06, + "loss": 0.3859, + "step": 11338 + }, + { + "epoch": 4.2756410256410255, + "grad_norm": 0.07925511349510013, + "learning_rate": 5.006536731717662e-06, + "loss": 0.3972, + "step": 11339 + }, + { + "epoch": 4.276018099547511, + "grad_norm": 0.07332503544221734, + "learning_rate": 5.0014370743740156e-06, + "loss": 0.376, + "step": 11340 + }, + { + "epoch": 4.276395173453997, + "grad_norm": 0.07986897336113674, + "learning_rate": 4.996339842399742e-06, + "loss": 0.3703, + "step": 11341 + }, + { + "epoch": 4.276772247360483, + "grad_norm": 0.07831323908942095, + "learning_rate": 4.9912450361480555e-06, + "loss": 0.3833, + "step": 11342 + }, + { + "epoch": 4.277149321266968, + "grad_norm": 0.08069182144513082, + "learning_rate": 4.986152655972052e-06, + "loss": 0.374, + "step": 11343 + }, + { + "epoch": 4.277526395173454, + "grad_norm": 0.07770018623600188, + "learning_rate": 4.9810627022246036e-06, + "loss": 0.376, + "step": 11344 + }, + { + "epoch": 4.27790346907994, + "grad_norm": 0.08048517751567578, + "learning_rate": 4.975975175258447e-06, + "loss": 0.3816, + "step": 11345 + }, + { + "epoch": 4.278280542986425, + "grad_norm": 0.07593310692376211, + "learning_rate": 4.970890075426145e-06, + "loss": 0.3791, + "step": 11346 + }, + { + "epoch": 4.278657616892911, + "grad_norm": 0.08126317391720526, + "learning_rate": 4.965807403080099e-06, + "loss": 0.3841, + "step": 11347 + }, + { + "epoch": 4.279034690799397, + "grad_norm": 0.08173596762851294, + "learning_rate": 4.960727158572507e-06, + "loss": 0.3821, + "step": 11348 + }, + { + "epoch": 4.279411764705882, + "grad_norm": 0.07435382480502606, + "learning_rate": 4.955649342255462e-06, + "loss": 0.3769, + "step": 11349 + }, + { + "epoch": 4.279788838612368, + "grad_norm": 0.08107260140786365, + "learning_rate": 4.950573954480824e-06, + "loss": 0.3802, + "step": 11350 + }, + { + "epoch": 4.280165912518854, + "grad_norm": 0.08197462532579837, + "learning_rate": 4.945500995600321e-06, + "loss": 0.3789, + "step": 11351 + }, + { + "epoch": 4.280542986425339, + "grad_norm": 0.07335876581358197, + "learning_rate": 4.940430465965507e-06, + "loss": 0.3728, + "step": 11352 + }, + { + "epoch": 4.280920060331825, + "grad_norm": 0.07283272171678841, + "learning_rate": 4.935362365927771e-06, + "loss": 0.3749, + "step": 11353 + }, + { + "epoch": 4.281297134238311, + "grad_norm": 0.08258414688542987, + "learning_rate": 4.930296695838301e-06, + "loss": 0.3752, + "step": 11354 + }, + { + "epoch": 4.281674208144796, + "grad_norm": 0.07476263298864995, + "learning_rate": 4.9252334560481844e-06, + "loss": 0.3869, + "step": 11355 + }, + { + "epoch": 4.282051282051282, + "grad_norm": 0.08116061512201506, + "learning_rate": 4.9201726469082675e-06, + "loss": 0.3706, + "step": 11356 + }, + { + "epoch": 4.282428355957768, + "grad_norm": 0.07666593644921571, + "learning_rate": 4.915114268769272e-06, + "loss": 0.3768, + "step": 11357 + }, + { + "epoch": 4.282805429864253, + "grad_norm": 0.07399154890424549, + "learning_rate": 4.910058321981739e-06, + "loss": 0.3771, + "step": 11358 + }, + { + "epoch": 4.283182503770739, + "grad_norm": 0.07943606388703447, + "learning_rate": 4.905004806896045e-06, + "loss": 0.3831, + "step": 11359 + }, + { + "epoch": 4.283559577677225, + "grad_norm": 0.07802825748294831, + "learning_rate": 4.899953723862374e-06, + "loss": 0.3736, + "step": 11360 + }, + { + "epoch": 4.28393665158371, + "grad_norm": 0.07784124163828884, + "learning_rate": 4.8949050732307915e-06, + "loss": 0.384, + "step": 11361 + }, + { + "epoch": 4.284313725490196, + "grad_norm": 0.0784776623468348, + "learning_rate": 4.889858855351142e-06, + "loss": 0.3721, + "step": 11362 + }, + { + "epoch": 4.284690799396682, + "grad_norm": 0.07388871723465473, + "learning_rate": 4.884815070573128e-06, + "loss": 0.3805, + "step": 11363 + }, + { + "epoch": 4.285067873303167, + "grad_norm": 0.07309765717680439, + "learning_rate": 4.879773719246284e-06, + "loss": 0.3698, + "step": 11364 + }, + { + "epoch": 4.285444947209653, + "grad_norm": 0.07590818608498723, + "learning_rate": 4.874734801719983e-06, + "loss": 0.3811, + "step": 11365 + }, + { + "epoch": 4.285822021116139, + "grad_norm": 0.07886133148288242, + "learning_rate": 4.869698318343381e-06, + "loss": 0.3843, + "step": 11366 + }, + { + "epoch": 4.286199095022624, + "grad_norm": 0.07939491055992695, + "learning_rate": 4.864664269465542e-06, + "loss": 0.3732, + "step": 11367 + }, + { + "epoch": 4.28657616892911, + "grad_norm": 0.08660024991904579, + "learning_rate": 4.859632655435298e-06, + "loss": 0.3787, + "step": 11368 + }, + { + "epoch": 4.286953242835596, + "grad_norm": 0.07285568508581311, + "learning_rate": 4.854603476601343e-06, + "loss": 0.3807, + "step": 11369 + }, + { + "epoch": 4.287330316742081, + "grad_norm": 0.0792722945272364, + "learning_rate": 4.849576733312189e-06, + "loss": 0.3787, + "step": 11370 + }, + { + "epoch": 4.287707390648567, + "grad_norm": 0.07696611264817801, + "learning_rate": 4.844552425916202e-06, + "loss": 0.3831, + "step": 11371 + }, + { + "epoch": 4.288084464555053, + "grad_norm": 0.0751763673355594, + "learning_rate": 4.839530554761531e-06, + "loss": 0.3773, + "step": 11372 + }, + { + "epoch": 4.288461538461538, + "grad_norm": 0.07413392761536963, + "learning_rate": 4.834511120196222e-06, + "loss": 0.3771, + "step": 11373 + }, + { + "epoch": 4.288838612368024, + "grad_norm": 0.07755166049527032, + "learning_rate": 4.829494122568088e-06, + "loss": 0.3889, + "step": 11374 + }, + { + "epoch": 4.28921568627451, + "grad_norm": 0.07468627121682989, + "learning_rate": 4.824479562224817e-06, + "loss": 0.3805, + "step": 11375 + }, + { + "epoch": 4.289592760180995, + "grad_norm": 0.07816604892848306, + "learning_rate": 4.819467439513914e-06, + "loss": 0.3817, + "step": 11376 + }, + { + "epoch": 4.289969834087481, + "grad_norm": 0.07138541487048314, + "learning_rate": 4.814457754782722e-06, + "loss": 0.3822, + "step": 11377 + }, + { + "epoch": 4.290346907993967, + "grad_norm": 0.07504386492102508, + "learning_rate": 4.809450508378377e-06, + "loss": 0.3741, + "step": 11378 + }, + { + "epoch": 4.290723981900452, + "grad_norm": 0.07202824426118584, + "learning_rate": 4.804445700647917e-06, + "loss": 0.3901, + "step": 11379 + }, + { + "epoch": 4.291101055806938, + "grad_norm": 0.073364944768978, + "learning_rate": 4.799443331938145e-06, + "loss": 0.3842, + "step": 11380 + }, + { + "epoch": 4.291478129713424, + "grad_norm": 0.0756784698955519, + "learning_rate": 4.794443402595726e-06, + "loss": 0.3852, + "step": 11381 + }, + { + "epoch": 4.291855203619909, + "grad_norm": 0.07194639280953953, + "learning_rate": 4.78944591296715e-06, + "loss": 0.3721, + "step": 11382 + }, + { + "epoch": 4.292232277526395, + "grad_norm": 0.07731662372200353, + "learning_rate": 4.7844508633987465e-06, + "loss": 0.3706, + "step": 11383 + }, + { + "epoch": 4.292609351432881, + "grad_norm": 0.0829357399807604, + "learning_rate": 4.779458254236651e-06, + "loss": 0.38, + "step": 11384 + }, + { + "epoch": 4.292986425339366, + "grad_norm": 0.07397768047292384, + "learning_rate": 4.774468085826871e-06, + "loss": 0.3673, + "step": 11385 + }, + { + "epoch": 4.293363499245852, + "grad_norm": 0.07561322778342217, + "learning_rate": 4.769480358515202e-06, + "loss": 0.3849, + "step": 11386 + }, + { + "epoch": 4.293740573152338, + "grad_norm": 0.07423159639828525, + "learning_rate": 4.7644950726472905e-06, + "loss": 0.3741, + "step": 11387 + }, + { + "epoch": 4.294117647058823, + "grad_norm": 0.07551918434143014, + "learning_rate": 4.7595122285686215e-06, + "loss": 0.3734, + "step": 11388 + }, + { + "epoch": 4.294494720965309, + "grad_norm": 0.07274612285837201, + "learning_rate": 4.754531826624505e-06, + "loss": 0.3826, + "step": 11389 + }, + { + "epoch": 4.294871794871795, + "grad_norm": 0.07240475337917512, + "learning_rate": 4.74955386716005e-06, + "loss": 0.3666, + "step": 11390 + }, + { + "epoch": 4.29524886877828, + "grad_norm": 0.07727402564055773, + "learning_rate": 4.744578350520264e-06, + "loss": 0.3721, + "step": 11391 + }, + { + "epoch": 4.295625942684766, + "grad_norm": 0.08320896659409191, + "learning_rate": 4.739605277049916e-06, + "loss": 0.378, + "step": 11392 + }, + { + "epoch": 4.296003016591252, + "grad_norm": 0.07636882489374228, + "learning_rate": 4.734634647093646e-06, + "loss": 0.3857, + "step": 11393 + }, + { + "epoch": 4.296380090497737, + "grad_norm": 0.07498211405746723, + "learning_rate": 4.7296664609959165e-06, + "loss": 0.3846, + "step": 11394 + }, + { + "epoch": 4.296757164404223, + "grad_norm": 0.08177751558408919, + "learning_rate": 4.724700719101028e-06, + "loss": 0.3724, + "step": 11395 + }, + { + "epoch": 4.297134238310709, + "grad_norm": 0.07685414642263289, + "learning_rate": 4.719737421753068e-06, + "loss": 0.3673, + "step": 11396 + }, + { + "epoch": 4.297511312217194, + "grad_norm": 0.0765090239703167, + "learning_rate": 4.714776569296029e-06, + "loss": 0.371, + "step": 11397 + }, + { + "epoch": 4.29788838612368, + "grad_norm": 0.08495377157605553, + "learning_rate": 4.709818162073671e-06, + "loss": 0.3753, + "step": 11398 + }, + { + "epoch": 4.298265460030166, + "grad_norm": 0.07893538692626023, + "learning_rate": 4.70486220042961e-06, + "loss": 0.3788, + "step": 11399 + }, + { + "epoch": 4.298642533936651, + "grad_norm": 0.07561159419124051, + "learning_rate": 4.699908684707293e-06, + "loss": 0.3849, + "step": 11400 + }, + { + "epoch": 4.299019607843137, + "grad_norm": 0.08403623434617455, + "learning_rate": 4.6949576152499976e-06, + "loss": 0.3873, + "step": 11401 + }, + { + "epoch": 4.299396681749623, + "grad_norm": 0.07502771476448786, + "learning_rate": 4.69000899240081e-06, + "loss": 0.3774, + "step": 11402 + }, + { + "epoch": 4.299773755656108, + "grad_norm": 0.07075981987934234, + "learning_rate": 4.685062816502695e-06, + "loss": 0.3722, + "step": 11403 + }, + { + "epoch": 4.300150829562594, + "grad_norm": 0.07110231315638862, + "learning_rate": 4.680119087898396e-06, + "loss": 0.3773, + "step": 11404 + }, + { + "epoch": 4.30052790346908, + "grad_norm": 0.0738846765465395, + "learning_rate": 4.6751778069305156e-06, + "loss": 0.3815, + "step": 11405 + }, + { + "epoch": 4.300904977375565, + "grad_norm": 0.06995830145344144, + "learning_rate": 4.670238973941477e-06, + "loss": 0.3792, + "step": 11406 + }, + { + "epoch": 4.301282051282051, + "grad_norm": 0.06893770321486385, + "learning_rate": 4.665302589273543e-06, + "loss": 0.3744, + "step": 11407 + }, + { + "epoch": 4.301659125188537, + "grad_norm": 0.07375358913654946, + "learning_rate": 4.660368653268794e-06, + "loss": 0.3805, + "step": 11408 + }, + { + "epoch": 4.302036199095022, + "grad_norm": 0.07595281307683766, + "learning_rate": 4.655437166269163e-06, + "loss": 0.3832, + "step": 11409 + }, + { + "epoch": 4.302413273001508, + "grad_norm": 0.0690438013580597, + "learning_rate": 4.650508128616378e-06, + "loss": 0.3851, + "step": 11410 + }, + { + "epoch": 4.302790346907994, + "grad_norm": 0.07508149366183282, + "learning_rate": 4.645581540652022e-06, + "loss": 0.3832, + "step": 11411 + }, + { + "epoch": 4.3031674208144794, + "grad_norm": 0.07861130905315794, + "learning_rate": 4.640657402717508e-06, + "loss": 0.3814, + "step": 11412 + }, + { + "epoch": 4.303544494720965, + "grad_norm": 0.07035674540480205, + "learning_rate": 4.635735715154072e-06, + "loss": 0.3812, + "step": 11413 + }, + { + "epoch": 4.303921568627451, + "grad_norm": 0.08585299371285388, + "learning_rate": 4.630816478302782e-06, + "loss": 0.3715, + "step": 11414 + }, + { + "epoch": 4.3042986425339365, + "grad_norm": 0.07477842565805212, + "learning_rate": 4.6258996925045455e-06, + "loss": 0.3773, + "step": 11415 + }, + { + "epoch": 4.304675716440422, + "grad_norm": 0.07793520272265196, + "learning_rate": 4.620985358100076e-06, + "loss": 0.3734, + "step": 11416 + }, + { + "epoch": 4.305052790346908, + "grad_norm": 0.07959428876953716, + "learning_rate": 4.616073475429939e-06, + "loss": 0.3881, + "step": 11417 + }, + { + "epoch": 4.3054298642533935, + "grad_norm": 0.08495175516153158, + "learning_rate": 4.611164044834526e-06, + "loss": 0.3685, + "step": 11418 + }, + { + "epoch": 4.305806938159879, + "grad_norm": 0.08713818472674906, + "learning_rate": 4.606257066654056e-06, + "loss": 0.3708, + "step": 11419 + }, + { + "epoch": 4.306184012066365, + "grad_norm": 0.08487314354675649, + "learning_rate": 4.601352541228576e-06, + "loss": 0.3875, + "step": 11420 + }, + { + "epoch": 4.3065610859728505, + "grad_norm": 0.08209584963169933, + "learning_rate": 4.596450468897966e-06, + "loss": 0.3802, + "step": 11421 + }, + { + "epoch": 4.306938159879336, + "grad_norm": 0.08706360290198434, + "learning_rate": 4.591550850001944e-06, + "loss": 0.3888, + "step": 11422 + }, + { + "epoch": 4.307315233785822, + "grad_norm": 0.07369530526584214, + "learning_rate": 4.586653684880031e-06, + "loss": 0.3793, + "step": 11423 + }, + { + "epoch": 4.3076923076923075, + "grad_norm": 0.07898002384442855, + "learning_rate": 4.581758973871609e-06, + "loss": 0.3797, + "step": 11424 + }, + { + "epoch": 4.308069381598793, + "grad_norm": 0.07517560335151503, + "learning_rate": 4.576866717315871e-06, + "loss": 0.379, + "step": 11425 + }, + { + "epoch": 4.308446455505279, + "grad_norm": 0.07455484619090472, + "learning_rate": 4.57197691555185e-06, + "loss": 0.3804, + "step": 11426 + }, + { + "epoch": 4.3088235294117645, + "grad_norm": 0.07468549967415913, + "learning_rate": 4.567089568918403e-06, + "loss": 0.3707, + "step": 11427 + }, + { + "epoch": 4.30920060331825, + "grad_norm": 0.07493898995070215, + "learning_rate": 4.562204677754234e-06, + "loss": 0.3834, + "step": 11428 + }, + { + "epoch": 4.309577677224736, + "grad_norm": 0.07677485032280686, + "learning_rate": 4.557322242397835e-06, + "loss": 0.3797, + "step": 11429 + }, + { + "epoch": 4.3099547511312215, + "grad_norm": 0.07237851679800646, + "learning_rate": 4.5524422631875665e-06, + "loss": 0.378, + "step": 11430 + }, + { + "epoch": 4.310331825037707, + "grad_norm": 0.07185419214428085, + "learning_rate": 4.547564740461607e-06, + "loss": 0.3867, + "step": 11431 + }, + { + "epoch": 4.310708898944193, + "grad_norm": 0.07315125274632307, + "learning_rate": 4.542689674557967e-06, + "loss": 0.3782, + "step": 11432 + }, + { + "epoch": 4.3110859728506785, + "grad_norm": 0.07236042052225344, + "learning_rate": 4.5378170658144785e-06, + "loss": 0.3824, + "step": 11433 + }, + { + "epoch": 4.311463046757164, + "grad_norm": 0.07390363027448785, + "learning_rate": 4.532946914568826e-06, + "loss": 0.386, + "step": 11434 + }, + { + "epoch": 4.31184012066365, + "grad_norm": 0.06726137972050389, + "learning_rate": 4.528079221158481e-06, + "loss": 0.3819, + "step": 11435 + }, + { + "epoch": 4.3122171945701355, + "grad_norm": 0.07208921493629104, + "learning_rate": 4.52321398592078e-06, + "loss": 0.3733, + "step": 11436 + }, + { + "epoch": 4.312594268476621, + "grad_norm": 0.07786982702891866, + "learning_rate": 4.518351209192888e-06, + "loss": 0.3758, + "step": 11437 + }, + { + "epoch": 4.312971342383107, + "grad_norm": 0.08037593972790857, + "learning_rate": 4.51349089131178e-06, + "loss": 0.3768, + "step": 11438 + }, + { + "epoch": 4.3133484162895925, + "grad_norm": 0.07803861144774477, + "learning_rate": 4.50863303261428e-06, + "loss": 0.3722, + "step": 11439 + }, + { + "epoch": 4.313725490196078, + "grad_norm": 0.07680296825455489, + "learning_rate": 4.503777633437034e-06, + "loss": 0.3811, + "step": 11440 + }, + { + "epoch": 4.314102564102564, + "grad_norm": 0.07938280563083197, + "learning_rate": 4.498924694116502e-06, + "loss": 0.3784, + "step": 11441 + }, + { + "epoch": 4.3144796380090495, + "grad_norm": 0.07923292657285105, + "learning_rate": 4.494074214989015e-06, + "loss": 0.3796, + "step": 11442 + }, + { + "epoch": 4.314856711915535, + "grad_norm": 0.07622102541264558, + "learning_rate": 4.489226196390677e-06, + "loss": 0.3808, + "step": 11443 + }, + { + "epoch": 4.315233785822021, + "grad_norm": 0.08135563106248986, + "learning_rate": 4.484380638657473e-06, + "loss": 0.3774, + "step": 11444 + }, + { + "epoch": 4.3156108597285066, + "grad_norm": 0.07293511487409747, + "learning_rate": 4.479537542125183e-06, + "loss": 0.3834, + "step": 11445 + }, + { + "epoch": 4.315987933634992, + "grad_norm": 0.08137332374778242, + "learning_rate": 4.474696907129446e-06, + "loss": 0.3813, + "step": 11446 + }, + { + "epoch": 4.316365007541478, + "grad_norm": 0.07469204481447027, + "learning_rate": 4.469858734005685e-06, + "loss": 0.3708, + "step": 11447 + }, + { + "epoch": 4.316742081447964, + "grad_norm": 0.0766271344683596, + "learning_rate": 4.465023023089217e-06, + "loss": 0.3778, + "step": 11448 + }, + { + "epoch": 4.317119155354449, + "grad_norm": 0.07965454964894028, + "learning_rate": 4.4601897747151224e-06, + "loss": 0.3796, + "step": 11449 + }, + { + "epoch": 4.317496229260935, + "grad_norm": 0.07498194223214832, + "learning_rate": 4.455358989218352e-06, + "loss": 0.3783, + "step": 11450 + }, + { + "epoch": 4.317873303167421, + "grad_norm": 0.07401541389144324, + "learning_rate": 4.4505306669336786e-06, + "loss": 0.3753, + "step": 11451 + }, + { + "epoch": 4.318250377073906, + "grad_norm": 0.07924921634656207, + "learning_rate": 4.445704808195705e-06, + "loss": 0.3724, + "step": 11452 + }, + { + "epoch": 4.318627450980392, + "grad_norm": 0.07914678723954205, + "learning_rate": 4.4408814133388356e-06, + "loss": 0.383, + "step": 11453 + }, + { + "epoch": 4.319004524886878, + "grad_norm": 0.07428911134204068, + "learning_rate": 4.436060482697357e-06, + "loss": 0.3924, + "step": 11454 + }, + { + "epoch": 4.319381598793363, + "grad_norm": 0.07396962201120644, + "learning_rate": 4.431242016605337e-06, + "loss": 0.3722, + "step": 11455 + }, + { + "epoch": 4.319758672699849, + "grad_norm": 0.07214797547290214, + "learning_rate": 4.426426015396694e-06, + "loss": 0.3843, + "step": 11456 + }, + { + "epoch": 4.320135746606335, + "grad_norm": 0.07884769102017539, + "learning_rate": 4.421612479405175e-06, + "loss": 0.3765, + "step": 11457 + }, + { + "epoch": 4.32051282051282, + "grad_norm": 0.07935469856132998, + "learning_rate": 4.416801408964353e-06, + "loss": 0.373, + "step": 11458 + }, + { + "epoch": 4.320889894419306, + "grad_norm": 0.07357124483456208, + "learning_rate": 4.411992804407632e-06, + "loss": 0.377, + "step": 11459 + }, + { + "epoch": 4.321266968325792, + "grad_norm": 0.07423239506502266, + "learning_rate": 4.407186666068253e-06, + "loss": 0.3862, + "step": 11460 + }, + { + "epoch": 4.321644042232277, + "grad_norm": 0.07339197616635476, + "learning_rate": 4.402382994279246e-06, + "loss": 0.3778, + "step": 11461 + }, + { + "epoch": 4.322021116138763, + "grad_norm": 0.07149785748847726, + "learning_rate": 4.3975817893735415e-06, + "loss": 0.3679, + "step": 11462 + }, + { + "epoch": 4.322398190045249, + "grad_norm": 0.07604909631286685, + "learning_rate": 4.3927830516838336e-06, + "loss": 0.3733, + "step": 11463 + }, + { + "epoch": 4.322775263951734, + "grad_norm": 0.07197309039109506, + "learning_rate": 4.387986781542672e-06, + "loss": 0.3788, + "step": 11464 + }, + { + "epoch": 4.32315233785822, + "grad_norm": 0.07348184083005554, + "learning_rate": 4.383192979282442e-06, + "loss": 0.3824, + "step": 11465 + }, + { + "epoch": 4.323529411764706, + "grad_norm": 0.07730718266298058, + "learning_rate": 4.3784016452353526e-06, + "loss": 0.3823, + "step": 11466 + }, + { + "epoch": 4.323906485671191, + "grad_norm": 0.07733473035589589, + "learning_rate": 4.3736127797334185e-06, + "loss": 0.3818, + "step": 11467 + }, + { + "epoch": 4.324283559577677, + "grad_norm": 0.07024326974883256, + "learning_rate": 4.368826383108533e-06, + "loss": 0.3754, + "step": 11468 + }, + { + "epoch": 4.324660633484163, + "grad_norm": 0.07882161160251057, + "learning_rate": 4.364042455692365e-06, + "loss": 0.387, + "step": 11469 + }, + { + "epoch": 4.325037707390648, + "grad_norm": 0.07725238397654328, + "learning_rate": 4.35926099781645e-06, + "loss": 0.3758, + "step": 11470 + }, + { + "epoch": 4.325414781297134, + "grad_norm": 0.07644126126228766, + "learning_rate": 4.354482009812131e-06, + "loss": 0.3816, + "step": 11471 + }, + { + "epoch": 4.32579185520362, + "grad_norm": 0.07661368520510477, + "learning_rate": 4.349705492010601e-06, + "loss": 0.3697, + "step": 11472 + }, + { + "epoch": 4.326168929110105, + "grad_norm": 0.08114393681029003, + "learning_rate": 4.3449314447428435e-06, + "loss": 0.373, + "step": 11473 + }, + { + "epoch": 4.326546003016591, + "grad_norm": 0.07258809263374111, + "learning_rate": 4.340159868339724e-06, + "loss": 0.3771, + "step": 11474 + }, + { + "epoch": 4.326923076923077, + "grad_norm": 0.07697807768741641, + "learning_rate": 4.335390763131888e-06, + "loss": 0.3754, + "step": 11475 + }, + { + "epoch": 4.327300150829562, + "grad_norm": 0.08434597585403232, + "learning_rate": 4.330624129449837e-06, + "loss": 0.381, + "step": 11476 + }, + { + "epoch": 4.327677224736048, + "grad_norm": 0.07504369947248662, + "learning_rate": 4.3258599676239e-06, + "loss": 0.3768, + "step": 11477 + }, + { + "epoch": 4.328054298642534, + "grad_norm": 0.0788318726106731, + "learning_rate": 4.321098277984228e-06, + "loss": 0.3871, + "step": 11478 + }, + { + "epoch": 4.328431372549019, + "grad_norm": 0.07408576705030154, + "learning_rate": 4.316339060860784e-06, + "loss": 0.3769, + "step": 11479 + }, + { + "epoch": 4.328808446455505, + "grad_norm": 0.07416595564583114, + "learning_rate": 4.311582316583405e-06, + "loss": 0.384, + "step": 11480 + }, + { + "epoch": 4.329185520361991, + "grad_norm": 0.0766472380127915, + "learning_rate": 4.306828045481708e-06, + "loss": 0.3846, + "step": 11481 + }, + { + "epoch": 4.329562594268476, + "grad_norm": 0.07489518787730329, + "learning_rate": 4.302076247885172e-06, + "loss": 0.3751, + "step": 11482 + }, + { + "epoch": 4.329939668174962, + "grad_norm": 0.074587674325697, + "learning_rate": 4.297326924123084e-06, + "loss": 0.3721, + "step": 11483 + }, + { + "epoch": 4.330316742081448, + "grad_norm": 0.07540449786038698, + "learning_rate": 4.292580074524582e-06, + "loss": 0.3866, + "step": 11484 + }, + { + "epoch": 4.330693815987933, + "grad_norm": 0.07630856586892931, + "learning_rate": 4.2878356994185874e-06, + "loss": 0.3722, + "step": 11485 + }, + { + "epoch": 4.331070889894419, + "grad_norm": 0.07960297839152583, + "learning_rate": 4.283093799133924e-06, + "loss": 0.3778, + "step": 11486 + }, + { + "epoch": 4.331447963800905, + "grad_norm": 0.07603165593264986, + "learning_rate": 4.278354373999172e-06, + "loss": 0.3824, + "step": 11487 + }, + { + "epoch": 4.33182503770739, + "grad_norm": 0.0748165897020913, + "learning_rate": 4.273617424342771e-06, + "loss": 0.3821, + "step": 11488 + }, + { + "epoch": 4.332202111613876, + "grad_norm": 0.08383136530487771, + "learning_rate": 4.268882950492996e-06, + "loss": 0.3767, + "step": 11489 + }, + { + "epoch": 4.332579185520362, + "grad_norm": 0.08289362902780209, + "learning_rate": 4.26415095277795e-06, + "loss": 0.3791, + "step": 11490 + }, + { + "epoch": 4.332956259426847, + "grad_norm": 0.0748713101293784, + "learning_rate": 4.259421431525525e-06, + "loss": 0.3676, + "step": 11491 + }, + { + "epoch": 4.333333333333333, + "grad_norm": 0.08715452559185816, + "learning_rate": 4.254694387063514e-06, + "loss": 0.3725, + "step": 11492 + }, + { + "epoch": 4.333710407239819, + "grad_norm": 0.07604253294801788, + "learning_rate": 4.249969819719466e-06, + "loss": 0.3716, + "step": 11493 + }, + { + "epoch": 4.334087481146304, + "grad_norm": 0.07747833738829732, + "learning_rate": 4.245247729820796e-06, + "loss": 0.3785, + "step": 11494 + }, + { + "epoch": 4.33446455505279, + "grad_norm": 0.0873397466085585, + "learning_rate": 4.240528117694749e-06, + "loss": 0.3698, + "step": 11495 + }, + { + "epoch": 4.334841628959276, + "grad_norm": 0.07583599596089727, + "learning_rate": 4.235810983668387e-06, + "loss": 0.3847, + "step": 11496 + }, + { + "epoch": 4.335218702865761, + "grad_norm": 0.07837492702793808, + "learning_rate": 4.231096328068587e-06, + "loss": 0.3828, + "step": 11497 + }, + { + "epoch": 4.335595776772247, + "grad_norm": 0.07956262387694316, + "learning_rate": 4.226384151222105e-06, + "loss": 0.3886, + "step": 11498 + }, + { + "epoch": 4.335972850678733, + "grad_norm": 0.07639263772617595, + "learning_rate": 4.221674453455457e-06, + "loss": 0.3694, + "step": 11499 + }, + { + "epoch": 4.336349924585218, + "grad_norm": 0.0735758367627138, + "learning_rate": 4.216967235095037e-06, + "loss": 0.3651, + "step": 11500 + }, + { + "epoch": 4.336726998491704, + "grad_norm": 0.0750734722834371, + "learning_rate": 4.212262496467045e-06, + "loss": 0.3847, + "step": 11501 + }, + { + "epoch": 4.33710407239819, + "grad_norm": 0.07437200226783858, + "learning_rate": 4.207560237897528e-06, + "loss": 0.3695, + "step": 11502 + }, + { + "epoch": 4.337481146304675, + "grad_norm": 0.08229491956477152, + "learning_rate": 4.2028604597123215e-06, + "loss": 0.3933, + "step": 11503 + }, + { + "epoch": 4.337858220211161, + "grad_norm": 0.07264159066755024, + "learning_rate": 4.198163162237152e-06, + "loss": 0.3801, + "step": 11504 + }, + { + "epoch": 4.338235294117647, + "grad_norm": 0.07574524046250823, + "learning_rate": 4.193468345797511e-06, + "loss": 0.3714, + "step": 11505 + }, + { + "epoch": 4.338612368024132, + "grad_norm": 0.07790433681731915, + "learning_rate": 4.188776010718751e-06, + "loss": 0.3669, + "step": 11506 + }, + { + "epoch": 4.338989441930618, + "grad_norm": 0.07899271965935102, + "learning_rate": 4.184086157326048e-06, + "loss": 0.3819, + "step": 11507 + }, + { + "epoch": 4.339366515837104, + "grad_norm": 0.07406793569976383, + "learning_rate": 4.1793987859444175e-06, + "loss": 0.3701, + "step": 11508 + }, + { + "epoch": 4.339743589743589, + "grad_norm": 0.08284068456670028, + "learning_rate": 4.174713896898661e-06, + "loss": 0.3853, + "step": 11509 + }, + { + "epoch": 4.340120663650075, + "grad_norm": 0.08459181386334451, + "learning_rate": 4.170031490513467e-06, + "loss": 0.3881, + "step": 11510 + }, + { + "epoch": 4.340497737556561, + "grad_norm": 0.07836656867557676, + "learning_rate": 4.165351567113307e-06, + "loss": 0.3775, + "step": 11511 + }, + { + "epoch": 4.340874811463046, + "grad_norm": 0.07947993147876597, + "learning_rate": 4.160674127022493e-06, + "loss": 0.3926, + "step": 11512 + }, + { + "epoch": 4.341251885369532, + "grad_norm": 0.07803056979645154, + "learning_rate": 4.155999170565177e-06, + "loss": 0.3706, + "step": 11513 + }, + { + "epoch": 4.341628959276018, + "grad_norm": 0.07281149137246651, + "learning_rate": 4.151326698065328e-06, + "loss": 0.3715, + "step": 11514 + }, + { + "epoch": 4.342006033182503, + "grad_norm": 0.07758172271799126, + "learning_rate": 4.146656709846731e-06, + "loss": 0.3789, + "step": 11515 + }, + { + "epoch": 4.342383107088989, + "grad_norm": 0.079697205141279, + "learning_rate": 4.141989206233033e-06, + "loss": 0.381, + "step": 11516 + }, + { + "epoch": 4.342760180995475, + "grad_norm": 0.07219035177364563, + "learning_rate": 4.1373241875476686e-06, + "loss": 0.3702, + "step": 11517 + }, + { + "epoch": 4.3431372549019605, + "grad_norm": 0.073859977419051, + "learning_rate": 4.132661654113932e-06, + "loss": 0.3759, + "step": 11518 + }, + { + "epoch": 4.343514328808446, + "grad_norm": 0.07407658011240449, + "learning_rate": 4.1280016062549275e-06, + "loss": 0.371, + "step": 11519 + }, + { + "epoch": 4.343891402714932, + "grad_norm": 0.0729321470745042, + "learning_rate": 4.123344044293598e-06, + "loss": 0.3867, + "step": 11520 + }, + { + "epoch": 4.3442684766214175, + "grad_norm": 0.07151642624505111, + "learning_rate": 4.118688968552693e-06, + "loss": 0.3842, + "step": 11521 + }, + { + "epoch": 4.344645550527903, + "grad_norm": 0.07554759703774733, + "learning_rate": 4.114036379354826e-06, + "loss": 0.3753, + "step": 11522 + }, + { + "epoch": 4.345022624434389, + "grad_norm": 0.07711344509830968, + "learning_rate": 4.109386277022402e-06, + "loss": 0.374, + "step": 11523 + }, + { + "epoch": 4.3453996983408745, + "grad_norm": 0.07132147714946051, + "learning_rate": 4.104738661877674e-06, + "loss": 0.3823, + "step": 11524 + }, + { + "epoch": 4.34577677224736, + "grad_norm": 0.0718081877326546, + "learning_rate": 4.100093534242722e-06, + "loss": 0.3874, + "step": 11525 + }, + { + "epoch": 4.346153846153846, + "grad_norm": 0.07516183484199847, + "learning_rate": 4.0954508944394474e-06, + "loss": 0.3621, + "step": 11526 + }, + { + "epoch": 4.3465309200603315, + "grad_norm": 0.07176197702585006, + "learning_rate": 4.0908107427895635e-06, + "loss": 0.3647, + "step": 11527 + }, + { + "epoch": 4.346907993966817, + "grad_norm": 0.07587889384839275, + "learning_rate": 4.086173079614657e-06, + "loss": 0.3744, + "step": 11528 + }, + { + "epoch": 4.347285067873303, + "grad_norm": 0.07333796844965632, + "learning_rate": 4.081537905236097e-06, + "loss": 0.3639, + "step": 11529 + }, + { + "epoch": 4.3476621417797885, + "grad_norm": 0.07771481521711088, + "learning_rate": 4.076905219975098e-06, + "loss": 0.3825, + "step": 11530 + }, + { + "epoch": 4.348039215686274, + "grad_norm": 0.07276492767356743, + "learning_rate": 4.072275024152706e-06, + "loss": 0.3743, + "step": 11531 + }, + { + "epoch": 4.34841628959276, + "grad_norm": 0.0762004059072537, + "learning_rate": 4.067647318089795e-06, + "loss": 0.3728, + "step": 11532 + }, + { + "epoch": 4.3487933634992455, + "grad_norm": 0.07547511015796735, + "learning_rate": 4.063022102107033e-06, + "loss": 0.377, + "step": 11533 + }, + { + "epoch": 4.349170437405731, + "grad_norm": 0.07714111143892168, + "learning_rate": 4.058399376524982e-06, + "loss": 0.3791, + "step": 11534 + }, + { + "epoch": 4.349547511312217, + "grad_norm": 0.06932090618988089, + "learning_rate": 4.053779141663965e-06, + "loss": 0.3848, + "step": 11535 + }, + { + "epoch": 4.3499245852187025, + "grad_norm": 0.07751019738190139, + "learning_rate": 4.0491613978441706e-06, + "loss": 0.3875, + "step": 11536 + }, + { + "epoch": 4.350301659125188, + "grad_norm": 0.08207646034482462, + "learning_rate": 4.044546145385604e-06, + "loss": 0.3836, + "step": 11537 + }, + { + "epoch": 4.350678733031674, + "grad_norm": 0.07703465875739386, + "learning_rate": 4.039933384608104e-06, + "loss": 0.3793, + "step": 11538 + }, + { + "epoch": 4.3510558069381595, + "grad_norm": 0.07371475350291533, + "learning_rate": 4.035323115831306e-06, + "loss": 0.3764, + "step": 11539 + }, + { + "epoch": 4.351432880844645, + "grad_norm": 0.0749200276959974, + "learning_rate": 4.030715339374731e-06, + "loss": 0.3793, + "step": 11540 + }, + { + "epoch": 4.351809954751131, + "grad_norm": 0.07802447726292698, + "learning_rate": 4.026110055557673e-06, + "loss": 0.387, + "step": 11541 + }, + { + "epoch": 4.3521870286576165, + "grad_norm": 0.07317836285621855, + "learning_rate": 4.02150726469928e-06, + "loss": 0.3795, + "step": 11542 + }, + { + "epoch": 4.352564102564102, + "grad_norm": 0.07800063918952307, + "learning_rate": 4.016906967118517e-06, + "loss": 0.3664, + "step": 11543 + }, + { + "epoch": 4.352941176470588, + "grad_norm": 0.07989322859275784, + "learning_rate": 4.012309163134194e-06, + "loss": 0.3753, + "step": 11544 + }, + { + "epoch": 4.3533182503770735, + "grad_norm": 0.075233112849232, + "learning_rate": 4.00771385306491e-06, + "loss": 0.3794, + "step": 11545 + }, + { + "epoch": 4.353695324283559, + "grad_norm": 0.08817431674620042, + "learning_rate": 4.003121037229148e-06, + "loss": 0.3786, + "step": 11546 + }, + { + "epoch": 4.354072398190045, + "grad_norm": 0.08022844117602417, + "learning_rate": 3.998530715945159e-06, + "loss": 0.3897, + "step": 11547 + }, + { + "epoch": 4.3544494720965305, + "grad_norm": 0.07883397765813517, + "learning_rate": 3.993942889531055e-06, + "loss": 0.3823, + "step": 11548 + }, + { + "epoch": 4.354826546003016, + "grad_norm": 0.07683683062250375, + "learning_rate": 3.989357558304776e-06, + "loss": 0.3748, + "step": 11549 + }, + { + "epoch": 4.355203619909502, + "grad_norm": 0.07919752048435881, + "learning_rate": 3.9847747225840774e-06, + "loss": 0.3768, + "step": 11550 + }, + { + "epoch": 4.355580693815988, + "grad_norm": 0.07785174578625742, + "learning_rate": 3.9801943826865394e-06, + "loss": 0.3712, + "step": 11551 + }, + { + "epoch": 4.355957767722473, + "grad_norm": 0.07305782796871825, + "learning_rate": 3.975616538929594e-06, + "loss": 0.3829, + "step": 11552 + }, + { + "epoch": 4.356334841628959, + "grad_norm": 0.07784199515045154, + "learning_rate": 3.971041191630458e-06, + "loss": 0.3824, + "step": 11553 + }, + { + "epoch": 4.356711915535445, + "grad_norm": 0.07459685311736827, + "learning_rate": 3.966468341106216e-06, + "loss": 0.3831, + "step": 11554 + }, + { + "epoch": 4.35708898944193, + "grad_norm": 0.08606819037129851, + "learning_rate": 3.961897987673751e-06, + "loss": 0.3673, + "step": 11555 + }, + { + "epoch": 4.357466063348416, + "grad_norm": 0.07659455671204268, + "learning_rate": 3.957330131649788e-06, + "loss": 0.3814, + "step": 11556 + }, + { + "epoch": 4.357843137254902, + "grad_norm": 0.0844711674793772, + "learning_rate": 3.952764773350879e-06, + "loss": 0.3877, + "step": 11557 + }, + { + "epoch": 4.358220211161387, + "grad_norm": 0.08404670276855754, + "learning_rate": 3.948201913093406e-06, + "loss": 0.3841, + "step": 11558 + }, + { + "epoch": 4.358597285067873, + "grad_norm": 0.07495100272706806, + "learning_rate": 3.943641551193556e-06, + "loss": 0.3737, + "step": 11559 + }, + { + "epoch": 4.358974358974359, + "grad_norm": 0.0796832485753814, + "learning_rate": 3.93908368796736e-06, + "loss": 0.3782, + "step": 11560 + }, + { + "epoch": 4.359351432880844, + "grad_norm": 0.08358276602632916, + "learning_rate": 3.934528323730682e-06, + "loss": 0.384, + "step": 11561 + }, + { + "epoch": 4.35972850678733, + "grad_norm": 0.07585780261582425, + "learning_rate": 3.929975458799202e-06, + "loss": 0.385, + "step": 11562 + }, + { + "epoch": 4.360105580693816, + "grad_norm": 0.0765081071346615, + "learning_rate": 3.925425093488429e-06, + "loss": 0.3837, + "step": 11563 + }, + { + "epoch": 4.360482654600301, + "grad_norm": 0.08058452427513312, + "learning_rate": 3.920877228113704e-06, + "loss": 0.3793, + "step": 11564 + }, + { + "epoch": 4.360859728506787, + "grad_norm": 0.07372545826357402, + "learning_rate": 3.916331862990177e-06, + "loss": 0.3751, + "step": 11565 + }, + { + "epoch": 4.361236802413273, + "grad_norm": 0.0696132234543786, + "learning_rate": 3.911788998432844e-06, + "loss": 0.3741, + "step": 11566 + }, + { + "epoch": 4.361613876319758, + "grad_norm": 0.08011818550739623, + "learning_rate": 3.907248634756529e-06, + "loss": 0.3784, + "step": 11567 + }, + { + "epoch": 4.361990950226244, + "grad_norm": 0.06941871334467617, + "learning_rate": 3.902710772275864e-06, + "loss": 0.3809, + "step": 11568 + }, + { + "epoch": 4.36236802413273, + "grad_norm": 0.07296809819489049, + "learning_rate": 3.8981754113053275e-06, + "loss": 0.383, + "step": 11569 + }, + { + "epoch": 4.362745098039215, + "grad_norm": 0.07020042531397928, + "learning_rate": 3.893642552159209e-06, + "loss": 0.3803, + "step": 11570 + }, + { + "epoch": 4.363122171945701, + "grad_norm": 0.07523061013031034, + "learning_rate": 3.8891121951516494e-06, + "loss": 0.3753, + "step": 11571 + }, + { + "epoch": 4.363499245852187, + "grad_norm": 0.07897308641393426, + "learning_rate": 3.884584340596571e-06, + "loss": 0.3658, + "step": 11572 + }, + { + "epoch": 4.363876319758672, + "grad_norm": 0.07230997433964807, + "learning_rate": 3.880058988807766e-06, + "loss": 0.3788, + "step": 11573 + }, + { + "epoch": 4.364253393665158, + "grad_norm": 0.07633402665117711, + "learning_rate": 3.875536140098834e-06, + "loss": 0.3779, + "step": 11574 + }, + { + "epoch": 4.364630467571644, + "grad_norm": 0.08135211253615941, + "learning_rate": 3.871015794783204e-06, + "loss": 0.3643, + "step": 11575 + }, + { + "epoch": 4.365007541478129, + "grad_norm": 0.07616950561654502, + "learning_rate": 3.866497953174135e-06, + "loss": 0.3866, + "step": 11576 + }, + { + "epoch": 4.365384615384615, + "grad_norm": 0.07915704234510372, + "learning_rate": 3.861982615584716e-06, + "loss": 0.3864, + "step": 11577 + }, + { + "epoch": 4.365761689291101, + "grad_norm": 0.07490481772523082, + "learning_rate": 3.857469782327839e-06, + "loss": 0.3656, + "step": 11578 + }, + { + "epoch": 4.366138763197586, + "grad_norm": 0.0765140355812303, + "learning_rate": 3.852959453716243e-06, + "loss": 0.3802, + "step": 11579 + }, + { + "epoch": 4.366515837104072, + "grad_norm": 0.06801872513214369, + "learning_rate": 3.848451630062502e-06, + "loss": 0.3735, + "step": 11580 + }, + { + "epoch": 4.366892911010558, + "grad_norm": 0.08116180310801, + "learning_rate": 3.843946311678992e-06, + "loss": 0.3787, + "step": 11581 + }, + { + "epoch": 4.367269984917044, + "grad_norm": 0.0734871321045056, + "learning_rate": 3.83944349887793e-06, + "loss": 0.369, + "step": 11582 + }, + { + "epoch": 4.367647058823529, + "grad_norm": 0.07815463061518431, + "learning_rate": 3.8349431919713655e-06, + "loss": 0.384, + "step": 11583 + }, + { + "epoch": 4.3680241327300156, + "grad_norm": 0.07956356415875013, + "learning_rate": 3.830445391271162e-06, + "loss": 0.3794, + "step": 11584 + }, + { + "epoch": 4.3684012066365, + "grad_norm": 0.07225737953855066, + "learning_rate": 3.825950097089006e-06, + "loss": 0.3835, + "step": 11585 + }, + { + "epoch": 4.368778280542987, + "grad_norm": 0.07438362801032439, + "learning_rate": 3.82145730973642e-06, + "loss": 0.373, + "step": 11586 + }, + { + "epoch": 4.369155354449472, + "grad_norm": 0.07996330503615566, + "learning_rate": 3.816967029524752e-06, + "loss": 0.3809, + "step": 11587 + }, + { + "epoch": 4.369532428355958, + "grad_norm": 0.07446582295404301, + "learning_rate": 3.812479256765178e-06, + "loss": 0.3793, + "step": 11588 + }, + { + "epoch": 4.369909502262443, + "grad_norm": 0.07194408955760587, + "learning_rate": 3.80799399176869e-06, + "loss": 0.3803, + "step": 11589 + }, + { + "epoch": 4.37028657616893, + "grad_norm": 0.07455077592693866, + "learning_rate": 3.8035112348461156e-06, + "loss": 0.3895, + "step": 11590 + }, + { + "epoch": 4.370663650075414, + "grad_norm": 0.07351437752300882, + "learning_rate": 3.7990309863081166e-06, + "loss": 0.3769, + "step": 11591 + }, + { + "epoch": 4.371040723981901, + "grad_norm": 0.07248930823446965, + "learning_rate": 3.794553246465151e-06, + "loss": 0.3766, + "step": 11592 + }, + { + "epoch": 4.371417797888386, + "grad_norm": 0.06920097665002813, + "learning_rate": 3.7900780156275272e-06, + "loss": 0.3697, + "step": 11593 + }, + { + "epoch": 4.371794871794872, + "grad_norm": 0.07291973069032279, + "learning_rate": 3.7856052941053833e-06, + "loss": 0.375, + "step": 11594 + }, + { + "epoch": 4.372171945701357, + "grad_norm": 0.07419547640034335, + "learning_rate": 3.7811350822086713e-06, + "loss": 0.3739, + "step": 11595 + }, + { + "epoch": 4.372549019607844, + "grad_norm": 0.06996362911533557, + "learning_rate": 3.7766673802471697e-06, + "loss": 0.3708, + "step": 11596 + }, + { + "epoch": 4.372926093514328, + "grad_norm": 0.07658897045119523, + "learning_rate": 3.7722021885304984e-06, + "loss": 0.372, + "step": 11597 + }, + { + "epoch": 4.373303167420815, + "grad_norm": 0.07417728268586636, + "learning_rate": 3.767739507368071e-06, + "loss": 0.3789, + "step": 11598 + }, + { + "epoch": 4.3736802413273, + "grad_norm": 0.07599821016101613, + "learning_rate": 3.763279337069161e-06, + "loss": 0.3762, + "step": 11599 + }, + { + "epoch": 4.374057315233786, + "grad_norm": 0.07595900381593762, + "learning_rate": 3.75882167794285e-06, + "loss": 0.387, + "step": 11600 + }, + { + "epoch": 4.374434389140271, + "grad_norm": 0.0770583942125218, + "learning_rate": 3.754366530298046e-06, + "loss": 0.3831, + "step": 11601 + }, + { + "epoch": 4.374811463046758, + "grad_norm": 0.07436767128998502, + "learning_rate": 3.7499138944434976e-06, + "loss": 0.37, + "step": 11602 + }, + { + "epoch": 4.375188536953242, + "grad_norm": 0.07606037288982449, + "learning_rate": 3.745463770687767e-06, + "loss": 0.374, + "step": 11603 + }, + { + "epoch": 4.375565610859729, + "grad_norm": 0.08124484592984911, + "learning_rate": 3.7410161593392347e-06, + "loss": 0.3954, + "step": 11604 + }, + { + "epoch": 4.375942684766214, + "grad_norm": 0.07001596056514423, + "learning_rate": 3.73657106070612e-06, + "loss": 0.3798, + "step": 11605 + }, + { + "epoch": 4.3763197586727, + "grad_norm": 0.07660855472837311, + "learning_rate": 3.7321284750964617e-06, + "loss": 0.3753, + "step": 11606 + }, + { + "epoch": 4.376696832579185, + "grad_norm": 0.07763672344300723, + "learning_rate": 3.7276884028181326e-06, + "loss": 0.3904, + "step": 11607 + }, + { + "epoch": 4.377073906485672, + "grad_norm": 0.071955821106755, + "learning_rate": 3.723250844178825e-06, + "loss": 0.3677, + "step": 11608 + }, + { + "epoch": 4.377450980392156, + "grad_norm": 0.07217645497369155, + "learning_rate": 3.718815799486062e-06, + "loss": 0.3818, + "step": 11609 + }, + { + "epoch": 4.377828054298643, + "grad_norm": 0.07835935021593089, + "learning_rate": 3.7143832690471706e-06, + "loss": 0.3663, + "step": 11610 + }, + { + "epoch": 4.378205128205128, + "grad_norm": 0.08228550775943463, + "learning_rate": 3.709953253169345e-06, + "loss": 0.3775, + "step": 11611 + }, + { + "epoch": 4.378582202111614, + "grad_norm": 0.0738003116345117, + "learning_rate": 3.7055257521595668e-06, + "loss": 0.3815, + "step": 11612 + }, + { + "epoch": 4.378959276018099, + "grad_norm": 0.0813923906358239, + "learning_rate": 3.7011007663246568e-06, + "loss": 0.3798, + "step": 11613 + }, + { + "epoch": 4.379336349924586, + "grad_norm": 0.08026931343197291, + "learning_rate": 3.6966782959712675e-06, + "loss": 0.3811, + "step": 11614 + }, + { + "epoch": 4.37971342383107, + "grad_norm": 0.07977933341327659, + "learning_rate": 3.692258341405879e-06, + "loss": 0.369, + "step": 11615 + }, + { + "epoch": 4.380090497737557, + "grad_norm": 0.07240232597363325, + "learning_rate": 3.6878409029347652e-06, + "loss": 0.3834, + "step": 11616 + }, + { + "epoch": 4.380467571644042, + "grad_norm": 0.07582098167822829, + "learning_rate": 3.683425980864086e-06, + "loss": 0.3738, + "step": 11617 + }, + { + "epoch": 4.380844645550528, + "grad_norm": 0.07639772209676403, + "learning_rate": 3.6790135754997656e-06, + "loss": 0.3787, + "step": 11618 + }, + { + "epoch": 4.381221719457013, + "grad_norm": 0.07738773821414799, + "learning_rate": 3.6746036871475865e-06, + "loss": 0.3755, + "step": 11619 + }, + { + "epoch": 4.3815987933635, + "grad_norm": 0.07783763810628781, + "learning_rate": 3.670196316113153e-06, + "loss": 0.3726, + "step": 11620 + }, + { + "epoch": 4.3819758672699844, + "grad_norm": 0.0700810185700087, + "learning_rate": 3.6657914627018954e-06, + "loss": 0.3782, + "step": 11621 + }, + { + "epoch": 4.382352941176471, + "grad_norm": 0.08204784920307163, + "learning_rate": 3.6613891272190506e-06, + "loss": 0.3707, + "step": 11622 + }, + { + "epoch": 4.382730015082956, + "grad_norm": 0.08079389546266592, + "learning_rate": 3.6569893099697163e-06, + "loss": 0.3753, + "step": 11623 + }, + { + "epoch": 4.383107088989442, + "grad_norm": 0.07091320731242139, + "learning_rate": 3.652592011258782e-06, + "loss": 0.3783, + "step": 11624 + }, + { + "epoch": 4.383484162895928, + "grad_norm": 0.08828627304104933, + "learning_rate": 3.648197231390982e-06, + "loss": 0.3822, + "step": 11625 + }, + { + "epoch": 4.383861236802414, + "grad_norm": 0.07534945342503074, + "learning_rate": 3.6438049706708677e-06, + "loss": 0.3829, + "step": 11626 + }, + { + "epoch": 4.384238310708899, + "grad_norm": 0.07635457302958473, + "learning_rate": 3.639415229402832e-06, + "loss": 0.3793, + "step": 11627 + }, + { + "epoch": 4.384615384615385, + "grad_norm": 0.0747702506618152, + "learning_rate": 3.635028007891048e-06, + "loss": 0.3685, + "step": 11628 + }, + { + "epoch": 4.384992458521871, + "grad_norm": 0.0765653122397985, + "learning_rate": 3.6306433064395895e-06, + "loss": 0.3861, + "step": 11629 + }, + { + "epoch": 4.385369532428356, + "grad_norm": 0.07356220612213421, + "learning_rate": 3.6262611253522795e-06, + "loss": 0.3655, + "step": 11630 + }, + { + "epoch": 4.385746606334842, + "grad_norm": 0.08153880610847944, + "learning_rate": 3.6218814649328083e-06, + "loss": 0.3807, + "step": 11631 + }, + { + "epoch": 4.386123680241328, + "grad_norm": 0.07112737552645298, + "learning_rate": 3.617504325484684e-06, + "loss": 0.3674, + "step": 11632 + }, + { + "epoch": 4.386500754147813, + "grad_norm": 0.07479898746406363, + "learning_rate": 3.6131297073112516e-06, + "loss": 0.3706, + "step": 11633 + }, + { + "epoch": 4.386877828054299, + "grad_norm": 0.08253286668467254, + "learning_rate": 3.6087576107156364e-06, + "loss": 0.3877, + "step": 11634 + }, + { + "epoch": 4.387254901960785, + "grad_norm": 0.0819851322856864, + "learning_rate": 3.604388036000854e-06, + "loss": 0.3727, + "step": 11635 + }, + { + "epoch": 4.38763197586727, + "grad_norm": 0.07112535963923211, + "learning_rate": 3.6000209834696898e-06, + "loss": 0.3791, + "step": 11636 + }, + { + "epoch": 4.388009049773756, + "grad_norm": 0.0780683426978914, + "learning_rate": 3.5956564534247896e-06, + "loss": 0.3646, + "step": 11637 + }, + { + "epoch": 4.388386123680242, + "grad_norm": 0.07379000403986673, + "learning_rate": 3.5912944461686004e-06, + "loss": 0.3754, + "step": 11638 + }, + { + "epoch": 4.388763197586727, + "grad_norm": 0.07130142808817842, + "learning_rate": 3.586934962003423e-06, + "loss": 0.3783, + "step": 11639 + }, + { + "epoch": 4.389140271493213, + "grad_norm": 0.07268445990729602, + "learning_rate": 3.5825780012313403e-06, + "loss": 0.3793, + "step": 11640 + }, + { + "epoch": 4.389517345399699, + "grad_norm": 0.07294802052658025, + "learning_rate": 3.5782235641543106e-06, + "loss": 0.3705, + "step": 11641 + }, + { + "epoch": 4.389894419306184, + "grad_norm": 0.07127059929402145, + "learning_rate": 3.573871651074079e-06, + "loss": 0.3829, + "step": 11642 + }, + { + "epoch": 4.39027149321267, + "grad_norm": 0.07864074766710465, + "learning_rate": 3.569522262292231e-06, + "loss": 0.3797, + "step": 11643 + }, + { + "epoch": 4.390648567119156, + "grad_norm": 0.0758725063773327, + "learning_rate": 3.5651753981101745e-06, + "loss": 0.3846, + "step": 11644 + }, + { + "epoch": 4.391025641025641, + "grad_norm": 0.07607125324633132, + "learning_rate": 3.560831058829153e-06, + "loss": 0.3697, + "step": 11645 + }, + { + "epoch": 4.391402714932127, + "grad_norm": 0.07684083558405236, + "learning_rate": 3.5564892447502053e-06, + "loss": 0.3872, + "step": 11646 + }, + { + "epoch": 4.391779788838613, + "grad_norm": 0.06708827906835671, + "learning_rate": 3.552149956174238e-06, + "loss": 0.3701, + "step": 11647 + }, + { + "epoch": 4.392156862745098, + "grad_norm": 0.07324937285604123, + "learning_rate": 3.5478131934019434e-06, + "loss": 0.3734, + "step": 11648 + }, + { + "epoch": 4.392533936651584, + "grad_norm": 0.07097363746502938, + "learning_rate": 3.5434789567338634e-06, + "loss": 0.3803, + "step": 11649 + }, + { + "epoch": 4.39291101055807, + "grad_norm": 0.07648094478488944, + "learning_rate": 3.5391472464703534e-06, + "loss": 0.3778, + "step": 11650 + }, + { + "epoch": 4.393288084464555, + "grad_norm": 0.07308748284674907, + "learning_rate": 3.534818062911609e-06, + "loss": 0.3782, + "step": 11651 + }, + { + "epoch": 4.393665158371041, + "grad_norm": 0.07118439044264532, + "learning_rate": 3.530491406357608e-06, + "loss": 0.3715, + "step": 11652 + }, + { + "epoch": 4.394042232277527, + "grad_norm": 0.07196010345459163, + "learning_rate": 3.526167277108221e-06, + "loss": 0.3696, + "step": 11653 + }, + { + "epoch": 4.394419306184012, + "grad_norm": 0.0715782160324789, + "learning_rate": 3.521845675463085e-06, + "loss": 0.3732, + "step": 11654 + }, + { + "epoch": 4.394796380090498, + "grad_norm": 0.06883252834022689, + "learning_rate": 3.5175266017216835e-06, + "loss": 0.3757, + "step": 11655 + }, + { + "epoch": 4.395173453996984, + "grad_norm": 0.07765892922202106, + "learning_rate": 3.5132100561833337e-06, + "loss": 0.3866, + "step": 11656 + }, + { + "epoch": 4.3955505279034695, + "grad_norm": 0.07587932146482995, + "learning_rate": 3.508896039147165e-06, + "loss": 0.3867, + "step": 11657 + }, + { + "epoch": 4.395927601809955, + "grad_norm": 0.07255994852491358, + "learning_rate": 3.5045845509121244e-06, + "loss": 0.3826, + "step": 11658 + }, + { + "epoch": 4.396304675716441, + "grad_norm": 0.07559978954342128, + "learning_rate": 3.5002755917770137e-06, + "loss": 0.3743, + "step": 11659 + }, + { + "epoch": 4.3966817496229265, + "grad_norm": 0.08115178295302926, + "learning_rate": 3.4959691620404202e-06, + "loss": 0.3855, + "step": 11660 + }, + { + "epoch": 4.397058823529412, + "grad_norm": 0.07860631843533794, + "learning_rate": 3.49166526200079e-06, + "loss": 0.3802, + "step": 11661 + }, + { + "epoch": 4.397435897435898, + "grad_norm": 0.07317570499188114, + "learning_rate": 3.487363891956368e-06, + "loss": 0.3871, + "step": 11662 + }, + { + "epoch": 4.3978129713423835, + "grad_norm": 0.07813990443543929, + "learning_rate": 3.4830650522052543e-06, + "loss": 0.3856, + "step": 11663 + }, + { + "epoch": 4.398190045248869, + "grad_norm": 0.07688567832920118, + "learning_rate": 3.4787687430453267e-06, + "loss": 0.3926, + "step": 11664 + }, + { + "epoch": 4.398567119155355, + "grad_norm": 0.07294098047500865, + "learning_rate": 3.4744749647743416e-06, + "loss": 0.3944, + "step": 11665 + }, + { + "epoch": 4.3989441930618405, + "grad_norm": 0.07360395993018871, + "learning_rate": 3.470183717689839e-06, + "loss": 0.3784, + "step": 11666 + }, + { + "epoch": 4.399321266968326, + "grad_norm": 0.07571996262452695, + "learning_rate": 3.4658950020892036e-06, + "loss": 0.3719, + "step": 11667 + }, + { + "epoch": 4.399698340874812, + "grad_norm": 0.07949728743934496, + "learning_rate": 3.4616088182696373e-06, + "loss": 0.3771, + "step": 11668 + }, + { + "epoch": 4.4000754147812975, + "grad_norm": 0.07024700415089026, + "learning_rate": 3.4573251665281783e-06, + "loss": 0.3778, + "step": 11669 + }, + { + "epoch": 4.400452488687783, + "grad_norm": 0.07534440129375104, + "learning_rate": 3.4530440471616554e-06, + "loss": 0.3825, + "step": 11670 + }, + { + "epoch": 4.400829562594269, + "grad_norm": 0.07454993217258732, + "learning_rate": 3.4487654604667787e-06, + "loss": 0.3735, + "step": 11671 + }, + { + "epoch": 4.4012066365007545, + "grad_norm": 0.07341772845551403, + "learning_rate": 3.4444894067400304e-06, + "loss": 0.3721, + "step": 11672 + }, + { + "epoch": 4.40158371040724, + "grad_norm": 0.07432815937142086, + "learning_rate": 3.440215886277738e-06, + "loss": 0.3758, + "step": 11673 + }, + { + "epoch": 4.401960784313726, + "grad_norm": 0.07489577050880043, + "learning_rate": 3.435944899376056e-06, + "loss": 0.3854, + "step": 11674 + }, + { + "epoch": 4.4023378582202115, + "grad_norm": 0.06972813292754497, + "learning_rate": 3.4316764463309692e-06, + "loss": 0.382, + "step": 11675 + }, + { + "epoch": 4.402714932126697, + "grad_norm": 0.07250634997172609, + "learning_rate": 3.427410527438255e-06, + "loss": 0.382, + "step": 11676 + }, + { + "epoch": 4.403092006033183, + "grad_norm": 0.07512397511587983, + "learning_rate": 3.423147142993566e-06, + "loss": 0.3834, + "step": 11677 + }, + { + "epoch": 4.4034690799396685, + "grad_norm": 0.06813914673336706, + "learning_rate": 3.4188862932923273e-06, + "loss": 0.3806, + "step": 11678 + }, + { + "epoch": 4.403846153846154, + "grad_norm": 0.07119617033436931, + "learning_rate": 3.4146279786298185e-06, + "loss": 0.3635, + "step": 11679 + }, + { + "epoch": 4.40422322775264, + "grad_norm": 0.07224040194209519, + "learning_rate": 3.410372199301142e-06, + "loss": 0.3794, + "step": 11680 + }, + { + "epoch": 4.4046003016591255, + "grad_norm": 0.07083924434675687, + "learning_rate": 3.4061189556012254e-06, + "loss": 0.3778, + "step": 11681 + }, + { + "epoch": 4.404977375565611, + "grad_norm": 0.07108708484750126, + "learning_rate": 3.4018682478247934e-06, + "loss": 0.3846, + "step": 11682 + }, + { + "epoch": 4.405354449472097, + "grad_norm": 0.06989855012676582, + "learning_rate": 3.3976200762664413e-06, + "loss": 0.3785, + "step": 11683 + }, + { + "epoch": 4.4057315233785825, + "grad_norm": 0.08007226013939059, + "learning_rate": 3.3933744412205427e-06, + "loss": 0.3734, + "step": 11684 + }, + { + "epoch": 4.406108597285068, + "grad_norm": 0.07957695227119317, + "learning_rate": 3.389131342981329e-06, + "loss": 0.3745, + "step": 11685 + }, + { + "epoch": 4.406485671191554, + "grad_norm": 0.07579926689640937, + "learning_rate": 3.38489078184284e-06, + "loss": 0.3831, + "step": 11686 + }, + { + "epoch": 4.4068627450980395, + "grad_norm": 0.06894880442929231, + "learning_rate": 3.380652758098948e-06, + "loss": 0.3692, + "step": 11687 + }, + { + "epoch": 4.407239819004525, + "grad_norm": 0.08727128000221543, + "learning_rate": 3.3764172720433198e-06, + "loss": 0.3815, + "step": 11688 + }, + { + "epoch": 4.407616892911011, + "grad_norm": 0.08361270068886148, + "learning_rate": 3.372184323969507e-06, + "loss": 0.3899, + "step": 11689 + }, + { + "epoch": 4.407993966817497, + "grad_norm": 0.0787352906859804, + "learning_rate": 3.367953914170827e-06, + "loss": 0.3846, + "step": 11690 + }, + { + "epoch": 4.408371040723982, + "grad_norm": 0.06862684515125604, + "learning_rate": 3.363726042940445e-06, + "loss": 0.3735, + "step": 11691 + }, + { + "epoch": 4.408748114630468, + "grad_norm": 0.08258162587953626, + "learning_rate": 3.359500710571357e-06, + "loss": 0.3852, + "step": 11692 + }, + { + "epoch": 4.409125188536954, + "grad_norm": 0.08000764819075543, + "learning_rate": 3.3552779173563698e-06, + "loss": 0.3889, + "step": 11693 + }, + { + "epoch": 4.409502262443439, + "grad_norm": 0.07715747507729143, + "learning_rate": 3.351057663588111e-06, + "loss": 0.3824, + "step": 11694 + }, + { + "epoch": 4.409879336349925, + "grad_norm": 0.0735485846982788, + "learning_rate": 3.3468399495590577e-06, + "loss": 0.3735, + "step": 11695 + }, + { + "epoch": 4.410256410256411, + "grad_norm": 0.08282063381095116, + "learning_rate": 3.3426247755614826e-06, + "loss": 0.3807, + "step": 11696 + }, + { + "epoch": 4.410633484162896, + "grad_norm": 0.07578772127709242, + "learning_rate": 3.3384121418874906e-06, + "loss": 0.3908, + "step": 11697 + }, + { + "epoch": 4.411010558069382, + "grad_norm": 0.07394217679099013, + "learning_rate": 3.3342020488290204e-06, + "loss": 0.3634, + "step": 11698 + }, + { + "epoch": 4.411387631975868, + "grad_norm": 0.07464384909786136, + "learning_rate": 3.3299944966778265e-06, + "loss": 0.3769, + "step": 11699 + }, + { + "epoch": 4.411764705882353, + "grad_norm": 0.0755405221810433, + "learning_rate": 3.325789485725488e-06, + "loss": 0.3706, + "step": 11700 + }, + { + "epoch": 4.412141779788839, + "grad_norm": 0.07192644480936393, + "learning_rate": 3.3215870162634124e-06, + "loss": 0.3773, + "step": 11701 + }, + { + "epoch": 4.412518853695325, + "grad_norm": 0.06761950547935164, + "learning_rate": 3.3173870885828154e-06, + "loss": 0.382, + "step": 11702 + }, + { + "epoch": 4.41289592760181, + "grad_norm": 0.07805164662713109, + "learning_rate": 3.3131897029747617e-06, + "loss": 0.3698, + "step": 11703 + }, + { + "epoch": 4.413273001508296, + "grad_norm": 0.07147680592622584, + "learning_rate": 3.3089948597301124e-06, + "loss": 0.3808, + "step": 11704 + }, + { + "epoch": 4.413650075414782, + "grad_norm": 0.07317409529014324, + "learning_rate": 3.3048025591395815e-06, + "loss": 0.391, + "step": 11705 + }, + { + "epoch": 4.414027149321267, + "grad_norm": 0.07058671353330781, + "learning_rate": 3.3006128014936787e-06, + "loss": 0.3782, + "step": 11706 + }, + { + "epoch": 4.414404223227753, + "grad_norm": 0.07499904432987461, + "learning_rate": 3.296425587082768e-06, + "loss": 0.3714, + "step": 11707 + }, + { + "epoch": 4.414781297134239, + "grad_norm": 0.08075639867672772, + "learning_rate": 3.292240916196994e-06, + "loss": 0.3758, + "step": 11708 + }, + { + "epoch": 4.415158371040724, + "grad_norm": 0.07255640550259511, + "learning_rate": 3.2880587891263784e-06, + "loss": 0.3768, + "step": 11709 + }, + { + "epoch": 4.41553544494721, + "grad_norm": 0.07363347470533156, + "learning_rate": 3.2838792061607162e-06, + "loss": 0.3692, + "step": 11710 + }, + { + "epoch": 4.415912518853696, + "grad_norm": 0.07215346304325371, + "learning_rate": 3.27970216758966e-06, + "loss": 0.3804, + "step": 11711 + }, + { + "epoch": 4.416289592760181, + "grad_norm": 0.06925930641872036, + "learning_rate": 3.2755276737026678e-06, + "loss": 0.3697, + "step": 11712 + }, + { + "epoch": 4.416666666666667, + "grad_norm": 0.07031742468894528, + "learning_rate": 3.2713557247890447e-06, + "loss": 0.3798, + "step": 11713 + }, + { + "epoch": 4.417043740573153, + "grad_norm": 0.07179910493184638, + "learning_rate": 3.2671863211378763e-06, + "loss": 0.3853, + "step": 11714 + }, + { + "epoch": 4.417420814479638, + "grad_norm": 0.07279569858662567, + "learning_rate": 3.263019463038131e-06, + "loss": 0.3776, + "step": 11715 + }, + { + "epoch": 4.417797888386124, + "grad_norm": 0.07145465722539422, + "learning_rate": 3.2588551507785414e-06, + "loss": 0.3769, + "step": 11716 + }, + { + "epoch": 4.41817496229261, + "grad_norm": 0.07229678482592672, + "learning_rate": 3.2546933846477003e-06, + "loss": 0.3914, + "step": 11717 + }, + { + "epoch": 4.418552036199095, + "grad_norm": 0.07311193913932429, + "learning_rate": 3.250534164934016e-06, + "loss": 0.3846, + "step": 11718 + }, + { + "epoch": 4.418929110105581, + "grad_norm": 0.08087100931332905, + "learning_rate": 3.246377491925716e-06, + "loss": 0.3816, + "step": 11719 + }, + { + "epoch": 4.419306184012067, + "grad_norm": 0.07240557739654001, + "learning_rate": 3.2422233659108548e-06, + "loss": 0.3779, + "step": 11720 + }, + { + "epoch": 4.419683257918552, + "grad_norm": 0.0791983861479551, + "learning_rate": 3.2380717871773172e-06, + "loss": 0.3829, + "step": 11721 + }, + { + "epoch": 4.420060331825038, + "grad_norm": 0.07495285091537279, + "learning_rate": 3.2339227560127928e-06, + "loss": 0.3843, + "step": 11722 + }, + { + "epoch": 4.420437405731524, + "grad_norm": 0.07310771677682745, + "learning_rate": 3.2297762727048075e-06, + "loss": 0.3894, + "step": 11723 + }, + { + "epoch": 4.420814479638009, + "grad_norm": 0.07256256325768529, + "learning_rate": 3.225632337540714e-06, + "loss": 0.3694, + "step": 11724 + }, + { + "epoch": 4.421191553544495, + "grad_norm": 0.07946466740232991, + "learning_rate": 3.2214909508076777e-06, + "loss": 0.3895, + "step": 11725 + }, + { + "epoch": 4.421568627450981, + "grad_norm": 0.07397597487625843, + "learning_rate": 3.2173521127926956e-06, + "loss": 0.3942, + "step": 11726 + }, + { + "epoch": 4.421945701357466, + "grad_norm": 0.07529783830205365, + "learning_rate": 3.213215823782596e-06, + "loss": 0.374, + "step": 11727 + }, + { + "epoch": 4.422322775263952, + "grad_norm": 0.0712505912184173, + "learning_rate": 3.209082084063999e-06, + "loss": 0.3833, + "step": 11728 + }, + { + "epoch": 4.422699849170438, + "grad_norm": 0.08171526000829742, + "learning_rate": 3.204950893923382e-06, + "loss": 0.3913, + "step": 11729 + }, + { + "epoch": 4.423076923076923, + "grad_norm": 0.06873673020627953, + "learning_rate": 3.200822253647031e-06, + "loss": 0.3644, + "step": 11730 + }, + { + "epoch": 4.423453996983409, + "grad_norm": 0.06988565547108902, + "learning_rate": 3.1966961635210603e-06, + "loss": 0.3708, + "step": 11731 + }, + { + "epoch": 4.423831070889895, + "grad_norm": 0.0743270724925888, + "learning_rate": 3.192572623831396e-06, + "loss": 0.3816, + "step": 11732 + }, + { + "epoch": 4.42420814479638, + "grad_norm": 0.07347795803139046, + "learning_rate": 3.1884516348638094e-06, + "loss": 0.3761, + "step": 11733 + }, + { + "epoch": 4.424585218702866, + "grad_norm": 0.0723807535490249, + "learning_rate": 3.1843331969038683e-06, + "loss": 0.3688, + "step": 11734 + }, + { + "epoch": 4.424962292609352, + "grad_norm": 0.0751522087820955, + "learning_rate": 3.1802173102369795e-06, + "loss": 0.3798, + "step": 11735 + }, + { + "epoch": 4.425339366515837, + "grad_norm": 0.07419358742398013, + "learning_rate": 3.176103975148368e-06, + "loss": 0.3644, + "step": 11736 + }, + { + "epoch": 4.425716440422323, + "grad_norm": 0.07538889886711304, + "learning_rate": 3.171993191923095e-06, + "loss": 0.3699, + "step": 11737 + }, + { + "epoch": 4.426093514328809, + "grad_norm": 0.07020496374712484, + "learning_rate": 3.1678849608460214e-06, + "loss": 0.3788, + "step": 11738 + }, + { + "epoch": 4.426470588235294, + "grad_norm": 0.06976660985994498, + "learning_rate": 3.163779282201853e-06, + "loss": 0.373, + "step": 11739 + }, + { + "epoch": 4.42684766214178, + "grad_norm": 0.0748992352227549, + "learning_rate": 3.1596761562751132e-06, + "loss": 0.3817, + "step": 11740 + }, + { + "epoch": 4.427224736048266, + "grad_norm": 0.06686774301597098, + "learning_rate": 3.15557558335013e-06, + "loss": 0.3754, + "step": 11741 + }, + { + "epoch": 4.427601809954751, + "grad_norm": 0.07042727185677149, + "learning_rate": 3.1514775637110806e-06, + "loss": 0.3793, + "step": 11742 + }, + { + "epoch": 4.427978883861237, + "grad_norm": 0.0734884774457936, + "learning_rate": 3.1473820976419513e-06, + "loss": 0.3759, + "step": 11743 + }, + { + "epoch": 4.428355957767723, + "grad_norm": 0.07225657227947317, + "learning_rate": 3.1432891854265544e-06, + "loss": 0.3854, + "step": 11744 + }, + { + "epoch": 4.428733031674208, + "grad_norm": 0.07065887602074175, + "learning_rate": 3.139198827348522e-06, + "loss": 0.3829, + "step": 11745 + }, + { + "epoch": 4.429110105580694, + "grad_norm": 0.08412137557029446, + "learning_rate": 3.135111023691324e-06, + "loss": 0.3832, + "step": 11746 + }, + { + "epoch": 4.42948717948718, + "grad_norm": 0.07215197534679811, + "learning_rate": 3.1310257747382276e-06, + "loss": 0.3749, + "step": 11747 + }, + { + "epoch": 4.429864253393665, + "grad_norm": 0.0718857533539663, + "learning_rate": 3.12694308077234e-06, + "loss": 0.3845, + "step": 11748 + }, + { + "epoch": 4.430241327300151, + "grad_norm": 0.07114814932046398, + "learning_rate": 3.12286294207659e-06, + "loss": 0.3771, + "step": 11749 + }, + { + "epoch": 4.430618401206637, + "grad_norm": 0.06875908819955855, + "learning_rate": 3.118785358933729e-06, + "loss": 0.3753, + "step": 11750 + }, + { + "epoch": 4.430995475113122, + "grad_norm": 0.06825162397857276, + "learning_rate": 3.1147103316263316e-06, + "loss": 0.3757, + "step": 11751 + }, + { + "epoch": 4.431372549019608, + "grad_norm": 0.0748313645688335, + "learning_rate": 3.110637860436798e-06, + "loss": 0.3811, + "step": 11752 + }, + { + "epoch": 4.431749622926094, + "grad_norm": 0.06919973667673365, + "learning_rate": 3.10656794564733e-06, + "loss": 0.3669, + "step": 11753 + }, + { + "epoch": 4.432126696832579, + "grad_norm": 0.07070023451045715, + "learning_rate": 3.1025005875399806e-06, + "loss": 0.3863, + "step": 11754 + }, + { + "epoch": 4.432503770739065, + "grad_norm": 0.06596939765083903, + "learning_rate": 3.09843578639661e-06, + "loss": 0.3853, + "step": 11755 + }, + { + "epoch": 4.432880844645551, + "grad_norm": 0.07383700849449108, + "learning_rate": 3.0943735424989074e-06, + "loss": 0.3979, + "step": 11756 + }, + { + "epoch": 4.433257918552036, + "grad_norm": 0.06791485451381349, + "learning_rate": 3.0903138561283862e-06, + "loss": 0.3726, + "step": 11757 + }, + { + "epoch": 4.433634992458522, + "grad_norm": 0.07373363908493322, + "learning_rate": 3.086256727566381e-06, + "loss": 0.3704, + "step": 11758 + }, + { + "epoch": 4.434012066365008, + "grad_norm": 0.07194234169116247, + "learning_rate": 3.082202157094032e-06, + "loss": 0.3904, + "step": 11759 + }, + { + "epoch": 4.4343891402714934, + "grad_norm": 0.07209728832976632, + "learning_rate": 3.0781501449923355e-06, + "loss": 0.3759, + "step": 11760 + }, + { + "epoch": 4.434766214177979, + "grad_norm": 0.06660701205744707, + "learning_rate": 3.074100691542081e-06, + "loss": 0.3839, + "step": 11761 + }, + { + "epoch": 4.435143288084465, + "grad_norm": 0.07023814253842525, + "learning_rate": 3.070053797023893e-06, + "loss": 0.3751, + "step": 11762 + }, + { + "epoch": 4.4355203619909505, + "grad_norm": 0.0732304398919915, + "learning_rate": 3.0660094617182224e-06, + "loss": 0.3777, + "step": 11763 + }, + { + "epoch": 4.435897435897436, + "grad_norm": 0.07458492167489082, + "learning_rate": 3.0619676859053428e-06, + "loss": 0.378, + "step": 11764 + }, + { + "epoch": 4.436274509803922, + "grad_norm": 0.07113641514172374, + "learning_rate": 3.057928469865323e-06, + "loss": 0.38, + "step": 11765 + }, + { + "epoch": 4.4366515837104075, + "grad_norm": 0.07635984997610117, + "learning_rate": 3.0538918138781093e-06, + "loss": 0.3817, + "step": 11766 + }, + { + "epoch": 4.437028657616893, + "grad_norm": 0.07169534399127954, + "learning_rate": 3.04985771822341e-06, + "loss": 0.3796, + "step": 11767 + }, + { + "epoch": 4.437405731523379, + "grad_norm": 0.06762685771353202, + "learning_rate": 3.045826183180802e-06, + "loss": 0.3779, + "step": 11768 + }, + { + "epoch": 4.4377828054298645, + "grad_norm": 0.07169434702115622, + "learning_rate": 3.0417972090296578e-06, + "loss": 0.3791, + "step": 11769 + }, + { + "epoch": 4.43815987933635, + "grad_norm": 0.08144086511602928, + "learning_rate": 3.037770796049189e-06, + "loss": 0.3891, + "step": 11770 + }, + { + "epoch": 4.438536953242836, + "grad_norm": 0.0751562524032492, + "learning_rate": 3.033746944518412e-06, + "loss": 0.3886, + "step": 11771 + }, + { + "epoch": 4.4389140271493215, + "grad_norm": 0.06716621574354043, + "learning_rate": 3.029725654716189e-06, + "loss": 0.3722, + "step": 11772 + }, + { + "epoch": 4.439291101055807, + "grad_norm": 0.06768304124381908, + "learning_rate": 3.025706926921186e-06, + "loss": 0.377, + "step": 11773 + }, + { + "epoch": 4.439668174962293, + "grad_norm": 0.07036064481595235, + "learning_rate": 3.0216907614118906e-06, + "loss": 0.3896, + "step": 11774 + }, + { + "epoch": 4.4400452488687785, + "grad_norm": 0.0684954738596718, + "learning_rate": 3.0176771584666278e-06, + "loss": 0.3887, + "step": 11775 + }, + { + "epoch": 4.440422322775264, + "grad_norm": 0.07221374749607469, + "learning_rate": 3.013666118363543e-06, + "loss": 0.3693, + "step": 11776 + }, + { + "epoch": 4.44079939668175, + "grad_norm": 0.07056707174886342, + "learning_rate": 3.0096576413805746e-06, + "loss": 0.3798, + "step": 11777 + }, + { + "epoch": 4.4411764705882355, + "grad_norm": 0.06490305694516821, + "learning_rate": 3.0056517277955357e-06, + "loss": 0.3699, + "step": 11778 + }, + { + "epoch": 4.441553544494721, + "grad_norm": 0.06716660114996556, + "learning_rate": 3.0016483778860086e-06, + "loss": 0.3916, + "step": 11779 + }, + { + "epoch": 4.441930618401207, + "grad_norm": 0.06936228959840132, + "learning_rate": 2.9976475919294334e-06, + "loss": 0.3788, + "step": 11780 + }, + { + "epoch": 4.4423076923076925, + "grad_norm": 0.07277348821124141, + "learning_rate": 2.99364937020306e-06, + "loss": 0.3819, + "step": 11781 + }, + { + "epoch": 4.442684766214178, + "grad_norm": 0.07460286782141448, + "learning_rate": 2.989653712983964e-06, + "loss": 0.3804, + "step": 11782 + }, + { + "epoch": 4.443061840120664, + "grad_norm": 0.06715137646523445, + "learning_rate": 2.9856606205490268e-06, + "loss": 0.3743, + "step": 11783 + }, + { + "epoch": 4.4434389140271495, + "grad_norm": 0.06999952833251791, + "learning_rate": 2.981670093174991e-06, + "loss": 0.3713, + "step": 11784 + }, + { + "epoch": 4.443815987933635, + "grad_norm": 0.07109882986497873, + "learning_rate": 2.977682131138373e-06, + "loss": 0.3757, + "step": 11785 + }, + { + "epoch": 4.444193061840121, + "grad_norm": 0.0756646672492643, + "learning_rate": 2.973696734715548e-06, + "loss": 0.3778, + "step": 11786 + }, + { + "epoch": 4.4445701357466065, + "grad_norm": 0.07118124296815971, + "learning_rate": 2.9697139041826937e-06, + "loss": 0.3722, + "step": 11787 + }, + { + "epoch": 4.444947209653092, + "grad_norm": 0.0746684409662003, + "learning_rate": 2.9657336398158308e-06, + "loss": 0.3739, + "step": 11788 + }, + { + "epoch": 4.445324283559578, + "grad_norm": 0.0737313100537276, + "learning_rate": 2.9617559418907605e-06, + "loss": 0.3709, + "step": 11789 + }, + { + "epoch": 4.4457013574660635, + "grad_norm": 0.0747436821504828, + "learning_rate": 2.957780810683164e-06, + "loss": 0.3889, + "step": 11790 + }, + { + "epoch": 4.446078431372549, + "grad_norm": 0.06950277938819692, + "learning_rate": 2.9538082464684927e-06, + "loss": 0.3743, + "step": 11791 + }, + { + "epoch": 4.446455505279035, + "grad_norm": 0.07425385061266955, + "learning_rate": 2.9498382495220546e-06, + "loss": 0.3879, + "step": 11792 + }, + { + "epoch": 4.4468325791855206, + "grad_norm": 0.07068981354029581, + "learning_rate": 2.945870820118959e-06, + "loss": 0.3676, + "step": 11793 + }, + { + "epoch": 4.447209653092006, + "grad_norm": 0.07117445629826663, + "learning_rate": 2.9419059585341593e-06, + "loss": 0.3675, + "step": 11794 + }, + { + "epoch": 4.447586726998492, + "grad_norm": 0.07413982058339086, + "learning_rate": 2.9379436650423866e-06, + "loss": 0.3837, + "step": 11795 + }, + { + "epoch": 4.447963800904978, + "grad_norm": 0.06818194905465999, + "learning_rate": 2.9339839399182613e-06, + "loss": 0.3669, + "step": 11796 + }, + { + "epoch": 4.448340874811463, + "grad_norm": 0.07135923465277277, + "learning_rate": 2.930026783436164e-06, + "loss": 0.3776, + "step": 11797 + }, + { + "epoch": 4.448717948717949, + "grad_norm": 0.07176653186056714, + "learning_rate": 2.9260721958703288e-06, + "loss": 0.3827, + "step": 11798 + }, + { + "epoch": 4.449095022624435, + "grad_norm": 0.07146468068412096, + "learning_rate": 2.922120177494807e-06, + "loss": 0.3867, + "step": 11799 + }, + { + "epoch": 4.44947209653092, + "grad_norm": 0.06988112342696386, + "learning_rate": 2.9181707285834735e-06, + "loss": 0.3792, + "step": 11800 + }, + { + "epoch": 4.449849170437406, + "grad_norm": 0.07094234225877115, + "learning_rate": 2.914223849410007e-06, + "loss": 0.3748, + "step": 11801 + }, + { + "epoch": 4.450226244343892, + "grad_norm": 0.07427745749851454, + "learning_rate": 2.9102795402479446e-06, + "loss": 0.3697, + "step": 11802 + }, + { + "epoch": 4.450603318250377, + "grad_norm": 0.07073631435250019, + "learning_rate": 2.9063378013706047e-06, + "loss": 0.3907, + "step": 11803 + }, + { + "epoch": 4.450980392156863, + "grad_norm": 0.06683846464433234, + "learning_rate": 2.902398633051151e-06, + "loss": 0.3777, + "step": 11804 + }, + { + "epoch": 4.451357466063349, + "grad_norm": 0.06771491588475455, + "learning_rate": 2.8984620355625704e-06, + "loss": 0.37, + "step": 11805 + }, + { + "epoch": 4.451734539969834, + "grad_norm": 0.07457895814112364, + "learning_rate": 2.8945280091776663e-06, + "loss": 0.3846, + "step": 11806 + }, + { + "epoch": 4.45211161387632, + "grad_norm": 0.07799110581172286, + "learning_rate": 2.8905965541690473e-06, + "loss": 0.3743, + "step": 11807 + }, + { + "epoch": 4.452488687782806, + "grad_norm": 0.07470979631466658, + "learning_rate": 2.88666767080918e-06, + "loss": 0.3799, + "step": 11808 + }, + { + "epoch": 4.452865761689291, + "grad_norm": 0.07555192859633061, + "learning_rate": 2.8827413593703225e-06, + "loss": 0.3812, + "step": 11809 + }, + { + "epoch": 4.453242835595777, + "grad_norm": 0.07055030890923744, + "learning_rate": 2.8788176201245632e-06, + "loss": 0.3706, + "step": 11810 + }, + { + "epoch": 4.453619909502263, + "grad_norm": 0.07963038394525278, + "learning_rate": 2.874896453343814e-06, + "loss": 0.376, + "step": 11811 + }, + { + "epoch": 4.453996983408748, + "grad_norm": 0.0767058756161269, + "learning_rate": 2.8709778592998217e-06, + "loss": 0.3808, + "step": 11812 + }, + { + "epoch": 4.454374057315234, + "grad_norm": 0.06779101336794248, + "learning_rate": 2.8670618382641115e-06, + "loss": 0.3817, + "step": 11813 + }, + { + "epoch": 4.45475113122172, + "grad_norm": 0.07232608646113964, + "learning_rate": 2.863148390508097e-06, + "loss": 0.3846, + "step": 11814 + }, + { + "epoch": 4.455128205128205, + "grad_norm": 0.07429536936688916, + "learning_rate": 2.8592375163029483e-06, + "loss": 0.3786, + "step": 11815 + }, + { + "epoch": 4.455505279034691, + "grad_norm": 0.07379199194536824, + "learning_rate": 2.8553292159197022e-06, + "loss": 0.377, + "step": 11816 + }, + { + "epoch": 4.455882352941177, + "grad_norm": 0.07113103980546481, + "learning_rate": 2.8514234896291904e-06, + "loss": 0.3786, + "step": 11817 + }, + { + "epoch": 4.456259426847662, + "grad_norm": 0.06880428938714918, + "learning_rate": 2.8475203377020855e-06, + "loss": 0.3856, + "step": 11818 + }, + { + "epoch": 4.456636500754148, + "grad_norm": 0.07344583645559638, + "learning_rate": 2.8436197604088555e-06, + "loss": 0.3729, + "step": 11819 + }, + { + "epoch": 4.457013574660634, + "grad_norm": 0.0728595344281853, + "learning_rate": 2.8397217580198265e-06, + "loss": 0.3792, + "step": 11820 + }, + { + "epoch": 4.457390648567119, + "grad_norm": 0.07483468707475976, + "learning_rate": 2.8358263308051205e-06, + "loss": 0.3849, + "step": 11821 + }, + { + "epoch": 4.457767722473605, + "grad_norm": 0.06688700330841174, + "learning_rate": 2.831933479034681e-06, + "loss": 0.3789, + "step": 11822 + }, + { + "epoch": 4.458144796380091, + "grad_norm": 0.07208717842127293, + "learning_rate": 2.8280432029782833e-06, + "loss": 0.3606, + "step": 11823 + }, + { + "epoch": 4.458521870286576, + "grad_norm": 0.07830062393126157, + "learning_rate": 2.824155502905526e-06, + "loss": 0.3891, + "step": 11824 + }, + { + "epoch": 4.458898944193062, + "grad_norm": 0.07465010619585992, + "learning_rate": 2.8202703790858054e-06, + "loss": 0.3766, + "step": 11825 + }, + { + "epoch": 4.459276018099548, + "grad_norm": 0.07005293009155002, + "learning_rate": 2.8163878317883787e-06, + "loss": 0.3823, + "step": 11826 + }, + { + "epoch": 4.459653092006033, + "grad_norm": 0.073344036474334, + "learning_rate": 2.812507861282292e-06, + "loss": 0.3763, + "step": 11827 + }, + { + "epoch": 4.460030165912519, + "grad_norm": 0.07523210322044259, + "learning_rate": 2.808630467836424e-06, + "loss": 0.3938, + "step": 11828 + }, + { + "epoch": 4.460407239819005, + "grad_norm": 0.07033585786581481, + "learning_rate": 2.8047556517194753e-06, + "loss": 0.383, + "step": 11829 + }, + { + "epoch": 4.46078431372549, + "grad_norm": 0.06802034191060752, + "learning_rate": 2.800883413199973e-06, + "loss": 0.3756, + "step": 11830 + }, + { + "epoch": 4.461161387631976, + "grad_norm": 0.07129811755650373, + "learning_rate": 2.7970137525462448e-06, + "loss": 0.374, + "step": 11831 + }, + { + "epoch": 4.461538461538462, + "grad_norm": 0.06887201395974515, + "learning_rate": 2.793146670026472e-06, + "loss": 0.3708, + "step": 11832 + }, + { + "epoch": 4.461915535444947, + "grad_norm": 0.07128834342992159, + "learning_rate": 2.7892821659086268e-06, + "loss": 0.3831, + "step": 11833 + }, + { + "epoch": 4.462292609351433, + "grad_norm": 0.07063101132452433, + "learning_rate": 2.7854202404605303e-06, + "loss": 0.3778, + "step": 11834 + }, + { + "epoch": 4.462669683257919, + "grad_norm": 0.06905251111249486, + "learning_rate": 2.781560893949795e-06, + "loss": 0.3694, + "step": 11835 + }, + { + "epoch": 4.463046757164404, + "grad_norm": 0.07297667793169946, + "learning_rate": 2.7777041266438785e-06, + "loss": 0.3866, + "step": 11836 + }, + { + "epoch": 4.46342383107089, + "grad_norm": 0.06885948850650221, + "learning_rate": 2.773849938810047e-06, + "loss": 0.3819, + "step": 11837 + }, + { + "epoch": 4.463800904977376, + "grad_norm": 0.08093611318083217, + "learning_rate": 2.7699983307154064e-06, + "loss": 0.3756, + "step": 11838 + }, + { + "epoch": 4.464177978883861, + "grad_norm": 0.07100584248529551, + "learning_rate": 2.7661493026268416e-06, + "loss": 0.375, + "step": 11839 + }, + { + "epoch": 4.464555052790347, + "grad_norm": 0.0695049455552249, + "learning_rate": 2.7623028548111164e-06, + "loss": 0.3926, + "step": 11840 + }, + { + "epoch": 4.464932126696833, + "grad_norm": 0.07587472205282894, + "learning_rate": 2.758458987534769e-06, + "loss": 0.3819, + "step": 11841 + }, + { + "epoch": 4.465309200603318, + "grad_norm": 0.0703571381852136, + "learning_rate": 2.754617701064182e-06, + "loss": 0.3729, + "step": 11842 + }, + { + "epoch": 4.465686274509804, + "grad_norm": 0.06577459674918643, + "learning_rate": 2.7507789956655508e-06, + "loss": 0.3725, + "step": 11843 + }, + { + "epoch": 4.46606334841629, + "grad_norm": 0.06719197592275306, + "learning_rate": 2.7469428716049027e-06, + "loss": 0.3795, + "step": 11844 + }, + { + "epoch": 4.466440422322775, + "grad_norm": 0.06948556591322691, + "learning_rate": 2.7431093291480528e-06, + "loss": 0.3745, + "step": 11845 + }, + { + "epoch": 4.466817496229261, + "grad_norm": 0.06923485837693295, + "learning_rate": 2.739278368560698e-06, + "loss": 0.3803, + "step": 11846 + }, + { + "epoch": 4.467194570135747, + "grad_norm": 0.0763101029778858, + "learning_rate": 2.7354499901082897e-06, + "loss": 0.3877, + "step": 11847 + }, + { + "epoch": 4.467571644042232, + "grad_norm": 0.06932232433439796, + "learning_rate": 2.7316241940561482e-06, + "loss": 0.3626, + "step": 11848 + }, + { + "epoch": 4.467948717948718, + "grad_norm": 0.06865064626108983, + "learning_rate": 2.7278009806693908e-06, + "loss": 0.3834, + "step": 11849 + }, + { + "epoch": 4.468325791855204, + "grad_norm": 0.07403295648496215, + "learning_rate": 2.7239803502129735e-06, + "loss": 0.3739, + "step": 11850 + }, + { + "epoch": 4.468702865761689, + "grad_norm": 0.07377052308487288, + "learning_rate": 2.720162302951641e-06, + "loss": 0.3687, + "step": 11851 + }, + { + "epoch": 4.469079939668175, + "grad_norm": 0.07004749074007699, + "learning_rate": 2.7163468391500035e-06, + "loss": 0.3792, + "step": 11852 + }, + { + "epoch": 4.469457013574661, + "grad_norm": 0.06913157358949404, + "learning_rate": 2.7125339590724543e-06, + "loss": 0.3769, + "step": 11853 + }, + { + "epoch": 4.469834087481146, + "grad_norm": 0.07178977043021215, + "learning_rate": 2.7087236629832304e-06, + "loss": 0.3764, + "step": 11854 + }, + { + "epoch": 4.470211161387632, + "grad_norm": 0.07057569527213556, + "learning_rate": 2.7049159511463784e-06, + "loss": 0.3766, + "step": 11855 + }, + { + "epoch": 4.470588235294118, + "grad_norm": 0.06852709652024556, + "learning_rate": 2.7011108238257723e-06, + "loss": 0.3846, + "step": 11856 + }, + { + "epoch": 4.470965309200603, + "grad_norm": 0.069072856561324, + "learning_rate": 2.6973082812850938e-06, + "loss": 0.378, + "step": 11857 + }, + { + "epoch": 4.471342383107089, + "grad_norm": 0.06943912518638265, + "learning_rate": 2.6935083237878744e-06, + "loss": 0.3687, + "step": 11858 + }, + { + "epoch": 4.471719457013575, + "grad_norm": 0.07759690421260612, + "learning_rate": 2.689710951597433e-06, + "loss": 0.3737, + "step": 11859 + }, + { + "epoch": 4.47209653092006, + "grad_norm": 0.06917275492336282, + "learning_rate": 2.6859161649769273e-06, + "loss": 0.367, + "step": 11860 + }, + { + "epoch": 4.472473604826546, + "grad_norm": 0.07109480405907544, + "learning_rate": 2.6821239641893337e-06, + "loss": 0.37, + "step": 11861 + }, + { + "epoch": 4.472850678733032, + "grad_norm": 0.07681865066486165, + "learning_rate": 2.6783343494974555e-06, + "loss": 0.3672, + "step": 11862 + }, + { + "epoch": 4.473227752639517, + "grad_norm": 0.07212994063489639, + "learning_rate": 2.6745473211638916e-06, + "loss": 0.3788, + "step": 11863 + }, + { + "epoch": 4.473604826546003, + "grad_norm": 0.07127718548181954, + "learning_rate": 2.670762879451103e-06, + "loss": 0.3811, + "step": 11864 + }, + { + "epoch": 4.473981900452489, + "grad_norm": 0.07643700572789608, + "learning_rate": 2.6669810246213247e-06, + "loss": 0.3809, + "step": 11865 + }, + { + "epoch": 4.4743589743589745, + "grad_norm": 0.07397108287895747, + "learning_rate": 2.663201756936653e-06, + "loss": 0.3765, + "step": 11866 + }, + { + "epoch": 4.47473604826546, + "grad_norm": 0.0726577770276491, + "learning_rate": 2.659425076658977e-06, + "loss": 0.3793, + "step": 11867 + }, + { + "epoch": 4.475113122171946, + "grad_norm": 0.06855843271931275, + "learning_rate": 2.6556509840500244e-06, + "loss": 0.3689, + "step": 11868 + }, + { + "epoch": 4.4754901960784315, + "grad_norm": 0.07414381272193367, + "learning_rate": 2.651879479371333e-06, + "loss": 0.3781, + "step": 11869 + }, + { + "epoch": 4.475867269984917, + "grad_norm": 0.06818911968231407, + "learning_rate": 2.6481105628842762e-06, + "loss": 0.3856, + "step": 11870 + }, + { + "epoch": 4.476244343891403, + "grad_norm": 0.06868188722189665, + "learning_rate": 2.6443442348500183e-06, + "loss": 0.379, + "step": 11871 + }, + { + "epoch": 4.4766214177978885, + "grad_norm": 0.06615372190864503, + "learning_rate": 2.640580495529568e-06, + "loss": 0.3816, + "step": 11872 + }, + { + "epoch": 4.476998491704374, + "grad_norm": 0.0652357894108916, + "learning_rate": 2.636819345183752e-06, + "loss": 0.3776, + "step": 11873 + }, + { + "epoch": 4.47737556561086, + "grad_norm": 0.0673113080394965, + "learning_rate": 2.6330607840732157e-06, + "loss": 0.3768, + "step": 11874 + }, + { + "epoch": 4.4777526395173455, + "grad_norm": 0.07241331160806051, + "learning_rate": 2.629304812458422e-06, + "loss": 0.3727, + "step": 11875 + }, + { + "epoch": 4.478129713423831, + "grad_norm": 0.07212903793921671, + "learning_rate": 2.6255514305996643e-06, + "loss": 0.3851, + "step": 11876 + }, + { + "epoch": 4.478506787330317, + "grad_norm": 0.06686135354503636, + "learning_rate": 2.6218006387570328e-06, + "loss": 0.374, + "step": 11877 + }, + { + "epoch": 4.4788838612368025, + "grad_norm": 0.06935026126674897, + "learning_rate": 2.618052437190466e-06, + "loss": 0.3773, + "step": 11878 + }, + { + "epoch": 4.479260935143288, + "grad_norm": 0.06780301954226366, + "learning_rate": 2.6143068261597073e-06, + "loss": 0.3716, + "step": 11879 + }, + { + "epoch": 4.479638009049774, + "grad_norm": 0.07153154214308935, + "learning_rate": 2.610563805924322e-06, + "loss": 0.3738, + "step": 11880 + }, + { + "epoch": 4.4800150829562595, + "grad_norm": 0.06953166717657919, + "learning_rate": 2.606823376743699e-06, + "loss": 0.376, + "step": 11881 + }, + { + "epoch": 4.480392156862745, + "grad_norm": 0.06570305206361712, + "learning_rate": 2.6030855388770573e-06, + "loss": 0.3823, + "step": 11882 + }, + { + "epoch": 4.480769230769231, + "grad_norm": 0.06760180378094942, + "learning_rate": 2.5993502925834115e-06, + "loss": 0.3693, + "step": 11883 + }, + { + "epoch": 4.4811463046757165, + "grad_norm": 0.06721274276812536, + "learning_rate": 2.5956176381216123e-06, + "loss": 0.3746, + "step": 11884 + }, + { + "epoch": 4.481523378582202, + "grad_norm": 0.06978648900310146, + "learning_rate": 2.5918875757503383e-06, + "loss": 0.3821, + "step": 11885 + }, + { + "epoch": 4.481900452488688, + "grad_norm": 0.07016572090636183, + "learning_rate": 2.58816010572807e-06, + "loss": 0.3823, + "step": 11886 + }, + { + "epoch": 4.4822775263951735, + "grad_norm": 0.07037222651109212, + "learning_rate": 2.5844352283131223e-06, + "loss": 0.3852, + "step": 11887 + }, + { + "epoch": 4.482654600301659, + "grad_norm": 0.0671773969281017, + "learning_rate": 2.5807129437636304e-06, + "loss": 0.3853, + "step": 11888 + }, + { + "epoch": 4.483031674208145, + "grad_norm": 0.0696130101320053, + "learning_rate": 2.5769932523375428e-06, + "loss": 0.3756, + "step": 11889 + }, + { + "epoch": 4.4834087481146305, + "grad_norm": 0.07295450363252147, + "learning_rate": 2.5732761542926233e-06, + "loss": 0.3801, + "step": 11890 + }, + { + "epoch": 4.483785822021116, + "grad_norm": 0.0732361492028192, + "learning_rate": 2.5695616498864696e-06, + "loss": 0.3707, + "step": 11891 + }, + { + "epoch": 4.484162895927602, + "grad_norm": 0.06812508959321602, + "learning_rate": 2.5658497393764894e-06, + "loss": 0.3719, + "step": 11892 + }, + { + "epoch": 4.4845399698340875, + "grad_norm": 0.07104342675704124, + "learning_rate": 2.562140423019921e-06, + "loss": 0.3761, + "step": 11893 + }, + { + "epoch": 4.484917043740573, + "grad_norm": 0.07085502169098153, + "learning_rate": 2.5584337010738127e-06, + "loss": 0.378, + "step": 11894 + }, + { + "epoch": 4.485294117647059, + "grad_norm": 0.06883120225121431, + "learning_rate": 2.5547295737950475e-06, + "loss": 0.381, + "step": 11895 + }, + { + "epoch": 4.4856711915535445, + "grad_norm": 0.07240307513361967, + "learning_rate": 2.5510280414402998e-06, + "loss": 0.3871, + "step": 11896 + }, + { + "epoch": 4.48604826546003, + "grad_norm": 0.06869992147330192, + "learning_rate": 2.547329104266094e-06, + "loss": 0.3688, + "step": 11897 + }, + { + "epoch": 4.486425339366516, + "grad_norm": 0.0703649500155186, + "learning_rate": 2.5436327625287583e-06, + "loss": 0.3831, + "step": 11898 + }, + { + "epoch": 4.4868024132730016, + "grad_norm": 0.06966743312664435, + "learning_rate": 2.5399390164844475e-06, + "loss": 0.3717, + "step": 11899 + }, + { + "epoch": 4.487179487179487, + "grad_norm": 0.07441776847450046, + "learning_rate": 2.536247866389139e-06, + "loss": 0.3754, + "step": 11900 + }, + { + "epoch": 4.487556561085973, + "grad_norm": 0.06888545526518121, + "learning_rate": 2.532559312498628e-06, + "loss": 0.3837, + "step": 11901 + }, + { + "epoch": 4.487933634992459, + "grad_norm": 0.06665703438619647, + "learning_rate": 2.5288733550685195e-06, + "loss": 0.3768, + "step": 11902 + }, + { + "epoch": 4.488310708898944, + "grad_norm": 0.06663080902185442, + "learning_rate": 2.5251899943542447e-06, + "loss": 0.3734, + "step": 11903 + }, + { + "epoch": 4.48868778280543, + "grad_norm": 0.0736593616234989, + "learning_rate": 2.5215092306110654e-06, + "loss": 0.3788, + "step": 11904 + }, + { + "epoch": 4.489064856711916, + "grad_norm": 0.07343168359561669, + "learning_rate": 2.5178310640940582e-06, + "loss": 0.375, + "step": 11905 + }, + { + "epoch": 4.489441930618401, + "grad_norm": 0.06647807543450918, + "learning_rate": 2.514155495058104e-06, + "loss": 0.3824, + "step": 11906 + }, + { + "epoch": 4.489819004524887, + "grad_norm": 0.0690717216659908, + "learning_rate": 2.5104825237579354e-06, + "loss": 0.3825, + "step": 11907 + }, + { + "epoch": 4.490196078431373, + "grad_norm": 0.06838650728196739, + "learning_rate": 2.5068121504480615e-06, + "loss": 0.3845, + "step": 11908 + }, + { + "epoch": 4.490573152337858, + "grad_norm": 0.08051339767683162, + "learning_rate": 2.50314437538286e-06, + "loss": 0.3816, + "step": 11909 + }, + { + "epoch": 4.490950226244344, + "grad_norm": 0.0705315680115325, + "learning_rate": 2.4994791988164878e-06, + "loss": 0.3802, + "step": 11910 + }, + { + "epoch": 4.49132730015083, + "grad_norm": 0.06899194970890384, + "learning_rate": 2.4958166210029468e-06, + "loss": 0.3703, + "step": 11911 + }, + { + "epoch": 4.491704374057315, + "grad_norm": 0.06780973109026656, + "learning_rate": 2.492156642196042e-06, + "loss": 0.3747, + "step": 11912 + }, + { + "epoch": 4.492081447963801, + "grad_norm": 0.06962887395883711, + "learning_rate": 2.488499262649424e-06, + "loss": 0.3642, + "step": 11913 + }, + { + "epoch": 4.492458521870287, + "grad_norm": 0.06515678107531998, + "learning_rate": 2.484844482616522e-06, + "loss": 0.3716, + "step": 11914 + }, + { + "epoch": 4.492835595776772, + "grad_norm": 0.06786057713216236, + "learning_rate": 2.48119230235063e-06, + "loss": 0.3701, + "step": 11915 + }, + { + "epoch": 4.493212669683258, + "grad_norm": 0.06809811748284617, + "learning_rate": 2.477542722104831e-06, + "loss": 0.3761, + "step": 11916 + }, + { + "epoch": 4.493589743589744, + "grad_norm": 0.06968194225916668, + "learning_rate": 2.473895742132033e-06, + "loss": 0.3894, + "step": 11917 + }, + { + "epoch": 4.493966817496229, + "grad_norm": 0.06673558113393148, + "learning_rate": 2.470251362684981e-06, + "loss": 0.3733, + "step": 11918 + }, + { + "epoch": 4.494343891402715, + "grad_norm": 0.06797492300611954, + "learning_rate": 2.4666095840162197e-06, + "loss": 0.3885, + "step": 11919 + }, + { + "epoch": 4.494720965309201, + "grad_norm": 0.07533392093792643, + "learning_rate": 2.4629704063781155e-06, + "loss": 0.3773, + "step": 11920 + }, + { + "epoch": 4.495098039215686, + "grad_norm": 0.07087449398520089, + "learning_rate": 2.4593338300228763e-06, + "loss": 0.3862, + "step": 11921 + }, + { + "epoch": 4.495475113122172, + "grad_norm": 0.06835051732692812, + "learning_rate": 2.4556998552025002e-06, + "loss": 0.3741, + "step": 11922 + }, + { + "epoch": 4.495852187028658, + "grad_norm": 0.06760436490193986, + "learning_rate": 2.4520684821688213e-06, + "loss": 0.3789, + "step": 11923 + }, + { + "epoch": 4.496229260935143, + "grad_norm": 0.06911329598655688, + "learning_rate": 2.4484397111734913e-06, + "loss": 0.377, + "step": 11924 + }, + { + "epoch": 4.496606334841629, + "grad_norm": 0.06925150414940841, + "learning_rate": 2.4448135424679896e-06, + "loss": 0.3693, + "step": 11925 + }, + { + "epoch": 4.496983408748115, + "grad_norm": 0.0710050026700081, + "learning_rate": 2.4411899763035818e-06, + "loss": 0.3913, + "step": 11926 + }, + { + "epoch": 4.4973604826546, + "grad_norm": 0.06659810286534108, + "learning_rate": 2.437569012931409e-06, + "loss": 0.3694, + "step": 11927 + }, + { + "epoch": 4.497737556561086, + "grad_norm": 0.06886986309991995, + "learning_rate": 2.4339506526023773e-06, + "loss": 0.376, + "step": 11928 + }, + { + "epoch": 4.498114630467572, + "grad_norm": 0.0738411065268422, + "learning_rate": 2.4303348955672456e-06, + "loss": 0.3768, + "step": 11929 + }, + { + "epoch": 4.498491704374057, + "grad_norm": 0.06914005787208533, + "learning_rate": 2.426721742076583e-06, + "loss": 0.3874, + "step": 11930 + }, + { + "epoch": 4.498868778280543, + "grad_norm": 0.0703156249111487, + "learning_rate": 2.4231111923807804e-06, + "loss": 0.3802, + "step": 11931 + }, + { + "epoch": 4.499245852187029, + "grad_norm": 0.07228635465908372, + "learning_rate": 2.4195032467300285e-06, + "loss": 0.3722, + "step": 11932 + }, + { + "epoch": 4.499622926093514, + "grad_norm": 0.07210293189049236, + "learning_rate": 2.4158979053743805e-06, + "loss": 0.3758, + "step": 11933 + }, + { + "epoch": 4.5, + "grad_norm": 0.07119717474281162, + "learning_rate": 2.4122951685636674e-06, + "loss": 0.3754, + "step": 11934 + }, + { + "epoch": 4.500377073906486, + "grad_norm": 0.07113591201341883, + "learning_rate": 2.408695036547557e-06, + "loss": 0.3747, + "step": 11935 + }, + { + "epoch": 4.500754147812971, + "grad_norm": 0.07549328016717143, + "learning_rate": 2.405097509575538e-06, + "loss": 0.3839, + "step": 11936 + }, + { + "epoch": 4.501131221719457, + "grad_norm": 0.07288736460688297, + "learning_rate": 2.4015025878969176e-06, + "loss": 0.3638, + "step": 11937 + }, + { + "epoch": 4.501508295625943, + "grad_norm": 0.06800481241865522, + "learning_rate": 2.3979102717608084e-06, + "loss": 0.3782, + "step": 11938 + }, + { + "epoch": 4.501885369532428, + "grad_norm": 0.06862644612364183, + "learning_rate": 2.3943205614161744e-06, + "loss": 0.3999, + "step": 11939 + }, + { + "epoch": 4.502262443438914, + "grad_norm": 0.07245175890968224, + "learning_rate": 2.390733457111769e-06, + "loss": 0.3727, + "step": 11940 + }, + { + "epoch": 4.5026395173454, + "grad_norm": 0.06991742925497875, + "learning_rate": 2.3871489590961707e-06, + "loss": 0.3817, + "step": 11941 + }, + { + "epoch": 4.503016591251885, + "grad_norm": 0.07064905400748059, + "learning_rate": 2.383567067617789e-06, + "loss": 0.3824, + "step": 11942 + }, + { + "epoch": 4.503393665158371, + "grad_norm": 0.0738987181671365, + "learning_rate": 2.3799877829248486e-06, + "loss": 0.382, + "step": 11943 + }, + { + "epoch": 4.503770739064857, + "grad_norm": 0.07175516821344208, + "learning_rate": 2.3764111052653726e-06, + "loss": 0.3796, + "step": 11944 + }, + { + "epoch": 4.504147812971342, + "grad_norm": 0.0704053734655661, + "learning_rate": 2.372837034887248e-06, + "loss": 0.3821, + "step": 11945 + }, + { + "epoch": 4.504524886877828, + "grad_norm": 0.0721335480114925, + "learning_rate": 2.3692655720381376e-06, + "loss": 0.3741, + "step": 11946 + }, + { + "epoch": 4.504901960784314, + "grad_norm": 0.06922786058360968, + "learning_rate": 2.365696716965542e-06, + "loss": 0.384, + "step": 11947 + }, + { + "epoch": 4.505279034690799, + "grad_norm": 0.07300407789627876, + "learning_rate": 2.3621304699167837e-06, + "loss": 0.3733, + "step": 11948 + }, + { + "epoch": 4.505656108597285, + "grad_norm": 0.07597410356355172, + "learning_rate": 2.358566831139002e-06, + "loss": 0.3779, + "step": 11949 + }, + { + "epoch": 4.506033182503771, + "grad_norm": 0.0722639590895696, + "learning_rate": 2.355005800879142e-06, + "loss": 0.3796, + "step": 11950 + }, + { + "epoch": 4.506410256410256, + "grad_norm": 0.0694895294291155, + "learning_rate": 2.351447379384002e-06, + "loss": 0.3796, + "step": 11951 + }, + { + "epoch": 4.506787330316742, + "grad_norm": 0.06975396300678809, + "learning_rate": 2.3478915669001577e-06, + "loss": 0.3861, + "step": 11952 + }, + { + "epoch": 4.507164404223228, + "grad_norm": 0.06753965593494027, + "learning_rate": 2.3443383636740304e-06, + "loss": 0.3798, + "step": 11953 + }, + { + "epoch": 4.507541478129713, + "grad_norm": 0.0725125332889866, + "learning_rate": 2.3407877699518534e-06, + "loss": 0.3829, + "step": 11954 + }, + { + "epoch": 4.507918552036199, + "grad_norm": 0.0727441732736675, + "learning_rate": 2.337239785979679e-06, + "loss": 0.3821, + "step": 11955 + }, + { + "epoch": 4.508295625942685, + "grad_norm": 0.06872794320509906, + "learning_rate": 2.333694412003382e-06, + "loss": 0.3712, + "step": 11956 + }, + { + "epoch": 4.50867269984917, + "grad_norm": 0.06794606153122712, + "learning_rate": 2.3301516482686595e-06, + "loss": 0.3724, + "step": 11957 + }, + { + "epoch": 4.509049773755656, + "grad_norm": 0.07037015239456508, + "learning_rate": 2.3266114950209986e-06, + "loss": 0.3834, + "step": 11958 + }, + { + "epoch": 4.509426847662142, + "grad_norm": 0.07298837514597421, + "learning_rate": 2.323073952505759e-06, + "loss": 0.3861, + "step": 11959 + }, + { + "epoch": 4.509803921568627, + "grad_norm": 0.07870767813261985, + "learning_rate": 2.3195390209680736e-06, + "loss": 0.3841, + "step": 11960 + }, + { + "epoch": 4.510180995475113, + "grad_norm": 0.06694256032893209, + "learning_rate": 2.3160067006529062e-06, + "loss": 0.3861, + "step": 11961 + }, + { + "epoch": 4.510558069381599, + "grad_norm": 0.06815118214660126, + "learning_rate": 2.3124769918050484e-06, + "loss": 0.3718, + "step": 11962 + }, + { + "epoch": 4.510935143288084, + "grad_norm": 0.0763609106795003, + "learning_rate": 2.308949894669117e-06, + "loss": 0.3801, + "step": 11963 + }, + { + "epoch": 4.51131221719457, + "grad_norm": 0.06799828376846358, + "learning_rate": 2.3054254094895077e-06, + "loss": 0.3936, + "step": 11964 + }, + { + "epoch": 4.511689291101056, + "grad_norm": 0.06896815688771091, + "learning_rate": 2.301903536510497e-06, + "loss": 0.3805, + "step": 11965 + }, + { + "epoch": 4.512066365007541, + "grad_norm": 0.06479102336700558, + "learning_rate": 2.298384275976129e-06, + "loss": 0.3688, + "step": 11966 + }, + { + "epoch": 4.512443438914027, + "grad_norm": 0.0739875748761395, + "learning_rate": 2.294867628130288e-06, + "loss": 0.3824, + "step": 11967 + }, + { + "epoch": 4.512820512820513, + "grad_norm": 0.07895564339542177, + "learning_rate": 2.2913535932166785e-06, + "loss": 0.3846, + "step": 11968 + }, + { + "epoch": 4.5131975867269984, + "grad_norm": 0.07266756806943848, + "learning_rate": 2.2878421714788202e-06, + "loss": 0.3846, + "step": 11969 + }, + { + "epoch": 4.513574660633484, + "grad_norm": 0.07104970988346475, + "learning_rate": 2.2843333631600383e-06, + "loss": 0.3838, + "step": 11970 + }, + { + "epoch": 4.51395173453997, + "grad_norm": 0.0697809239929829, + "learning_rate": 2.280827168503512e-06, + "loss": 0.3789, + "step": 11971 + }, + { + "epoch": 4.5143288084464555, + "grad_norm": 0.07270075304307246, + "learning_rate": 2.277323587752198e-06, + "loss": 0.3854, + "step": 11972 + }, + { + "epoch": 4.514705882352941, + "grad_norm": 0.07240306039938262, + "learning_rate": 2.2738226211489024e-06, + "loss": 0.3804, + "step": 11973 + }, + { + "epoch": 4.515082956259427, + "grad_norm": 0.07248221904692051, + "learning_rate": 2.270324268936235e-06, + "loss": 0.3761, + "step": 11974 + }, + { + "epoch": 4.5154600301659125, + "grad_norm": 0.06735485032737597, + "learning_rate": 2.266828531356637e-06, + "loss": 0.388, + "step": 11975 + }, + { + "epoch": 4.515837104072398, + "grad_norm": 0.07072772466389263, + "learning_rate": 2.2633354086523383e-06, + "loss": 0.3868, + "step": 11976 + }, + { + "epoch": 4.516214177978884, + "grad_norm": 0.0664079391841141, + "learning_rate": 2.2598449010654334e-06, + "loss": 0.367, + "step": 11977 + }, + { + "epoch": 4.5165912518853695, + "grad_norm": 0.0770771378361771, + "learning_rate": 2.2563570088377995e-06, + "loss": 0.3854, + "step": 11978 + }, + { + "epoch": 4.516968325791855, + "grad_norm": 0.0680524923760498, + "learning_rate": 2.252871732211146e-06, + "loss": 0.3831, + "step": 11979 + }, + { + "epoch": 4.517345399698341, + "grad_norm": 0.06913947003251321, + "learning_rate": 2.2493890714269993e-06, + "loss": 0.3807, + "step": 11980 + }, + { + "epoch": 4.5177224736048265, + "grad_norm": 0.07384418317275551, + "learning_rate": 2.2459090267267093e-06, + "loss": 0.3675, + "step": 11981 + }, + { + "epoch": 4.518099547511312, + "grad_norm": 0.07068743470639592, + "learning_rate": 2.2424315983514246e-06, + "loss": 0.3722, + "step": 11982 + }, + { + "epoch": 4.518476621417798, + "grad_norm": 0.06599523454275096, + "learning_rate": 2.2389567865421526e-06, + "loss": 0.3702, + "step": 11983 + }, + { + "epoch": 4.5188536953242835, + "grad_norm": 0.06848039898082628, + "learning_rate": 2.235484591539674e-06, + "loss": 0.3755, + "step": 11984 + }, + { + "epoch": 4.519230769230769, + "grad_norm": 0.07992392876131346, + "learning_rate": 2.2320150135846184e-06, + "loss": 0.3735, + "step": 11985 + }, + { + "epoch": 4.519607843137255, + "grad_norm": 0.07530370892462429, + "learning_rate": 2.2285480529174207e-06, + "loss": 0.381, + "step": 11986 + }, + { + "epoch": 4.5199849170437405, + "grad_norm": 0.0657490901396663, + "learning_rate": 2.2250837097783464e-06, + "loss": 0.3776, + "step": 11987 + }, + { + "epoch": 4.520361990950226, + "grad_norm": 0.07394042687892481, + "learning_rate": 2.2216219844074514e-06, + "loss": 0.3693, + "step": 11988 + }, + { + "epoch": 4.520739064856712, + "grad_norm": 0.07880706900376626, + "learning_rate": 2.218162877044656e-06, + "loss": 0.383, + "step": 11989 + }, + { + "epoch": 4.5211161387631975, + "grad_norm": 0.07242251037656747, + "learning_rate": 2.2147063879296527e-06, + "loss": 0.3809, + "step": 11990 + }, + { + "epoch": 4.521493212669683, + "grad_norm": 0.068390790417748, + "learning_rate": 2.211252517301983e-06, + "loss": 0.3672, + "step": 11991 + }, + { + "epoch": 4.521870286576169, + "grad_norm": 0.07134908719219052, + "learning_rate": 2.2078012654009974e-06, + "loss": 0.3811, + "step": 11992 + }, + { + "epoch": 4.5222473604826545, + "grad_norm": 0.07671069793174783, + "learning_rate": 2.204352632465865e-06, + "loss": 0.3793, + "step": 11993 + }, + { + "epoch": 4.52262443438914, + "grad_norm": 0.06635356331891586, + "learning_rate": 2.200906618735559e-06, + "loss": 0.3709, + "step": 11994 + }, + { + "epoch": 4.523001508295626, + "grad_norm": 0.06800233906327889, + "learning_rate": 2.1974632244489102e-06, + "loss": 0.374, + "step": 11995 + }, + { + "epoch": 4.5233785822021115, + "grad_norm": 0.07145219709265904, + "learning_rate": 2.194022449844524e-06, + "loss": 0.3737, + "step": 11996 + }, + { + "epoch": 4.523755656108597, + "grad_norm": 0.07044418783006769, + "learning_rate": 2.1905842951608445e-06, + "loss": 0.3786, + "step": 11997 + }, + { + "epoch": 4.524132730015083, + "grad_norm": 0.06897316982518252, + "learning_rate": 2.1871487606361396e-06, + "loss": 0.3775, + "step": 11998 + }, + { + "epoch": 4.5245098039215685, + "grad_norm": 0.06932722528757755, + "learning_rate": 2.1837158465084896e-06, + "loss": 0.3861, + "step": 11999 + }, + { + "epoch": 4.524886877828054, + "grad_norm": 0.06479079608592851, + "learning_rate": 2.1802855530157795e-06, + "loss": 0.366, + "step": 12000 + }, + { + "epoch": 4.52526395173454, + "grad_norm": 0.06975898996167552, + "learning_rate": 2.176857880395744e-06, + "loss": 0.3739, + "step": 12001 + }, + { + "epoch": 4.5256410256410255, + "grad_norm": 0.06783905601251641, + "learning_rate": 2.173432828885904e-06, + "loss": 0.3845, + "step": 12002 + }, + { + "epoch": 4.526018099547511, + "grad_norm": 0.0700355036085928, + "learning_rate": 2.1700103987236166e-06, + "loss": 0.378, + "step": 12003 + }, + { + "epoch": 4.526395173453997, + "grad_norm": 0.06730011367121232, + "learning_rate": 2.1665905901460515e-06, + "loss": 0.3743, + "step": 12004 + }, + { + "epoch": 4.526772247360483, + "grad_norm": 0.06739476017484342, + "learning_rate": 2.163173403390206e-06, + "loss": 0.3985, + "step": 12005 + }, + { + "epoch": 4.527149321266968, + "grad_norm": 0.07425928875064709, + "learning_rate": 2.1597588386928734e-06, + "loss": 0.3871, + "step": 12006 + }, + { + "epoch": 4.527526395173454, + "grad_norm": 0.06769350001831458, + "learning_rate": 2.1563468962907e-06, + "loss": 0.3777, + "step": 12007 + }, + { + "epoch": 4.52790346907994, + "grad_norm": 0.06890934579332166, + "learning_rate": 2.152937576420109e-06, + "loss": 0.3862, + "step": 12008 + }, + { + "epoch": 4.528280542986425, + "grad_norm": 0.06889844941951513, + "learning_rate": 2.1495308793173784e-06, + "loss": 0.3774, + "step": 12009 + }, + { + "epoch": 4.528657616892911, + "grad_norm": 0.06897151551420326, + "learning_rate": 2.1461268052185826e-06, + "loss": 0.3738, + "step": 12010 + }, + { + "epoch": 4.529034690799397, + "grad_norm": 0.07039115209063132, + "learning_rate": 2.1427253543596295e-06, + "loss": 0.391, + "step": 12011 + }, + { + "epoch": 4.529411764705882, + "grad_norm": 0.06422316068390191, + "learning_rate": 2.1393265269762197e-06, + "loss": 0.3805, + "step": 12012 + }, + { + "epoch": 4.529788838612368, + "grad_norm": 0.06672297092401283, + "learning_rate": 2.1359303233039118e-06, + "loss": 0.3798, + "step": 12013 + }, + { + "epoch": 4.530165912518854, + "grad_norm": 0.06713044397361945, + "learning_rate": 2.1325367435780376e-06, + "loss": 0.3752, + "step": 12014 + }, + { + "epoch": 4.530542986425339, + "grad_norm": 0.07129529406459749, + "learning_rate": 2.129145788033782e-06, + "loss": 0.3745, + "step": 12015 + }, + { + "epoch": 4.530920060331825, + "grad_norm": 0.07068966093218285, + "learning_rate": 2.1257574569061257e-06, + "loss": 0.3818, + "step": 12016 + }, + { + "epoch": 4.531297134238311, + "grad_norm": 0.07072477378564618, + "learning_rate": 2.12237175042989e-06, + "loss": 0.3807, + "step": 12017 + }, + { + "epoch": 4.531674208144796, + "grad_norm": 0.07500332659705539, + "learning_rate": 2.118988668839692e-06, + "loss": 0.3847, + "step": 12018 + }, + { + "epoch": 4.532051282051282, + "grad_norm": 0.0738149004870201, + "learning_rate": 2.115608212369984e-06, + "loss": 0.3705, + "step": 12019 + }, + { + "epoch": 4.532428355957768, + "grad_norm": 0.06570355152572747, + "learning_rate": 2.112230381255018e-06, + "loss": 0.3771, + "step": 12020 + }, + { + "epoch": 4.532805429864253, + "grad_norm": 0.06785669773269357, + "learning_rate": 2.1088551757288833e-06, + "loss": 0.3799, + "step": 12021 + }, + { + "epoch": 4.533182503770739, + "grad_norm": 0.07560812943664964, + "learning_rate": 2.105482596025472e-06, + "loss": 0.39, + "step": 12022 + }, + { + "epoch": 4.533559577677225, + "grad_norm": 0.06712630279131387, + "learning_rate": 2.1021126423785044e-06, + "loss": 0.3682, + "step": 12023 + }, + { + "epoch": 4.53393665158371, + "grad_norm": 0.06909444682655323, + "learning_rate": 2.0987453150215177e-06, + "loss": 0.3851, + "step": 12024 + }, + { + "epoch": 4.534313725490196, + "grad_norm": 0.07445738708242272, + "learning_rate": 2.095380614187863e-06, + "loss": 0.3921, + "step": 12025 + }, + { + "epoch": 4.534690799396682, + "grad_norm": 0.06806104496114626, + "learning_rate": 2.0920185401107097e-06, + "loss": 0.3784, + "step": 12026 + }, + { + "epoch": 4.535067873303167, + "grad_norm": 0.06681186397298303, + "learning_rate": 2.0886590930230445e-06, + "loss": 0.38, + "step": 12027 + }, + { + "epoch": 4.535444947209653, + "grad_norm": 0.07090544967755898, + "learning_rate": 2.0853022731576767e-06, + "loss": 0.3826, + "step": 12028 + }, + { + "epoch": 4.535822021116139, + "grad_norm": 0.07018356561256671, + "learning_rate": 2.0819480807472337e-06, + "loss": 0.3794, + "step": 12029 + }, + { + "epoch": 4.536199095022624, + "grad_norm": 0.0692521408284158, + "learning_rate": 2.078596516024156e-06, + "loss": 0.3853, + "step": 12030 + }, + { + "epoch": 4.53657616892911, + "grad_norm": 0.07065646654615622, + "learning_rate": 2.075247579220707e-06, + "loss": 0.3734, + "step": 12031 + }, + { + "epoch": 4.536953242835596, + "grad_norm": 0.07204181745075486, + "learning_rate": 2.0719012705689543e-06, + "loss": 0.3965, + "step": 12032 + }, + { + "epoch": 4.537330316742081, + "grad_norm": 0.06285431294954767, + "learning_rate": 2.068557590300806e-06, + "loss": 0.373, + "step": 12033 + }, + { + "epoch": 4.537707390648567, + "grad_norm": 0.06927140869117826, + "learning_rate": 2.065216538647965e-06, + "loss": 0.3792, + "step": 12034 + }, + { + "epoch": 4.538084464555053, + "grad_norm": 0.06816551043902333, + "learning_rate": 2.0618781158419753e-06, + "loss": 0.3745, + "step": 12035 + }, + { + "epoch": 4.538461538461538, + "grad_norm": 0.06722052052879117, + "learning_rate": 2.0585423221141807e-06, + "loss": 0.3767, + "step": 12036 + }, + { + "epoch": 4.538838612368024, + "grad_norm": 0.06656565704258673, + "learning_rate": 2.0552091576957477e-06, + "loss": 0.3782, + "step": 12037 + }, + { + "epoch": 4.53921568627451, + "grad_norm": 0.06792892678194726, + "learning_rate": 2.0518786228176692e-06, + "loss": 0.3777, + "step": 12038 + }, + { + "epoch": 4.539592760180995, + "grad_norm": 0.07228344229351223, + "learning_rate": 2.0485507177107335e-06, + "loss": 0.3814, + "step": 12039 + }, + { + "epoch": 4.539969834087481, + "grad_norm": 0.06951374913873103, + "learning_rate": 2.045225442605574e-06, + "loss": 0.3847, + "step": 12040 + }, + { + "epoch": 4.540346907993967, + "grad_norm": 0.06882832307352091, + "learning_rate": 2.0419027977326244e-06, + "loss": 0.3804, + "step": 12041 + }, + { + "epoch": 4.540723981900452, + "grad_norm": 0.06777475365352907, + "learning_rate": 2.0385827833221404e-06, + "loss": 0.3753, + "step": 12042 + }, + { + "epoch": 4.541101055806938, + "grad_norm": 0.06870972797085957, + "learning_rate": 2.0352653996042006e-06, + "loss": 0.3846, + "step": 12043 + }, + { + "epoch": 4.541478129713424, + "grad_norm": 0.07212216211935824, + "learning_rate": 2.031950646808696e-06, + "loss": 0.3741, + "step": 12044 + }, + { + "epoch": 4.541855203619909, + "grad_norm": 0.06890703548987798, + "learning_rate": 2.0286385251653318e-06, + "loss": 0.3707, + "step": 12045 + }, + { + "epoch": 4.542232277526395, + "grad_norm": 0.0683415282565879, + "learning_rate": 2.025329034903636e-06, + "loss": 0.3802, + "step": 12046 + }, + { + "epoch": 4.542609351432881, + "grad_norm": 0.06746521621522263, + "learning_rate": 2.0220221762529535e-06, + "loss": 0.3787, + "step": 12047 + }, + { + "epoch": 4.542986425339366, + "grad_norm": 0.06695937678392568, + "learning_rate": 2.0187179494424524e-06, + "loss": 0.3757, + "step": 12048 + }, + { + "epoch": 4.543363499245852, + "grad_norm": 0.07005346766861599, + "learning_rate": 2.0154163547011006e-06, + "loss": 0.3715, + "step": 12049 + }, + { + "epoch": 4.543740573152338, + "grad_norm": 0.07354791864003396, + "learning_rate": 2.0121173922577154e-06, + "loss": 0.3769, + "step": 12050 + }, + { + "epoch": 4.544117647058823, + "grad_norm": 0.06886969231674525, + "learning_rate": 2.008821062340891e-06, + "loss": 0.3683, + "step": 12051 + }, + { + "epoch": 4.544494720965309, + "grad_norm": 0.06660978318422991, + "learning_rate": 2.0055273651790676e-06, + "loss": 0.3795, + "step": 12052 + }, + { + "epoch": 4.544871794871795, + "grad_norm": 0.06608840117475229, + "learning_rate": 2.0022363010004976e-06, + "loss": 0.3791, + "step": 12053 + }, + { + "epoch": 4.54524886877828, + "grad_norm": 0.07245297145587169, + "learning_rate": 1.9989478700332522e-06, + "loss": 0.3901, + "step": 12054 + }, + { + "epoch": 4.545625942684766, + "grad_norm": 0.06695603753219116, + "learning_rate": 1.9956620725052067e-06, + "loss": 0.3662, + "step": 12055 + }, + { + "epoch": 4.546003016591252, + "grad_norm": 0.06681889801147047, + "learning_rate": 1.992378908644077e-06, + "loss": 0.3809, + "step": 12056 + }, + { + "epoch": 4.546380090497737, + "grad_norm": 0.06566005676551821, + "learning_rate": 1.9890983786773655e-06, + "loss": 0.3827, + "step": 12057 + }, + { + "epoch": 4.546757164404223, + "grad_norm": 0.0657984737849434, + "learning_rate": 1.9858204828324323e-06, + "loss": 0.3731, + "step": 12058 + }, + { + "epoch": 4.547134238310709, + "grad_norm": 0.07134424902405205, + "learning_rate": 1.982545221336416e-06, + "loss": 0.3843, + "step": 12059 + }, + { + "epoch": 4.547511312217194, + "grad_norm": 0.0660987411752322, + "learning_rate": 1.979272594416295e-06, + "loss": 0.3774, + "step": 12060 + }, + { + "epoch": 4.54788838612368, + "grad_norm": 0.06850968402249023, + "learning_rate": 1.9760026022988567e-06, + "loss": 0.3721, + "step": 12061 + }, + { + "epoch": 4.548265460030166, + "grad_norm": 0.06756960699152993, + "learning_rate": 1.9727352452107194e-06, + "loss": 0.3833, + "step": 12062 + }, + { + "epoch": 4.548642533936651, + "grad_norm": 0.06675238918595001, + "learning_rate": 1.969470523378285e-06, + "loss": 0.3805, + "step": 12063 + }, + { + "epoch": 4.549019607843137, + "grad_norm": 0.06836646516664077, + "learning_rate": 1.966208437027821e-06, + "loss": 0.3734, + "step": 12064 + }, + { + "epoch": 4.549396681749623, + "grad_norm": 0.0676981459227693, + "learning_rate": 1.9629489863853736e-06, + "loss": 0.3706, + "step": 12065 + }, + { + "epoch": 4.549773755656108, + "grad_norm": 0.06709158377786006, + "learning_rate": 1.9596921716768235e-06, + "loss": 0.3863, + "step": 12066 + }, + { + "epoch": 4.550150829562594, + "grad_norm": 0.07148720498436612, + "learning_rate": 1.9564379931278623e-06, + "loss": 0.3777, + "step": 12067 + }, + { + "epoch": 4.55052790346908, + "grad_norm": 0.07588665717734672, + "learning_rate": 1.9531864509640065e-06, + "loss": 0.3847, + "step": 12068 + }, + { + "epoch": 4.550904977375565, + "grad_norm": 0.07247941869452189, + "learning_rate": 1.949937545410574e-06, + "loss": 0.387, + "step": 12069 + }, + { + "epoch": 4.551282051282051, + "grad_norm": 0.0634081141624829, + "learning_rate": 1.9466912766927315e-06, + "loss": 0.3691, + "step": 12070 + }, + { + "epoch": 4.551659125188537, + "grad_norm": 0.06694443686666297, + "learning_rate": 1.9434476450354236e-06, + "loss": 0.366, + "step": 12071 + }, + { + "epoch": 4.552036199095022, + "grad_norm": 0.06633396965991427, + "learning_rate": 1.940206650663434e-06, + "loss": 0.3697, + "step": 12072 + }, + { + "epoch": 4.552413273001508, + "grad_norm": 0.0669135349383411, + "learning_rate": 1.936968293801371e-06, + "loss": 0.3693, + "step": 12073 + }, + { + "epoch": 4.552790346907994, + "grad_norm": 0.06676465389866988, + "learning_rate": 1.9337325746736457e-06, + "loss": 0.3799, + "step": 12074 + }, + { + "epoch": 4.5531674208144794, + "grad_norm": 0.06778813876373231, + "learning_rate": 1.930499493504474e-06, + "loss": 0.3768, + "step": 12075 + }, + { + "epoch": 4.553544494720965, + "grad_norm": 0.06762822530699186, + "learning_rate": 1.9272690505179347e-06, + "loss": 0.3774, + "step": 12076 + }, + { + "epoch": 4.553921568627451, + "grad_norm": 0.06774227386185533, + "learning_rate": 1.9240412459378753e-06, + "loss": 0.379, + "step": 12077 + }, + { + "epoch": 4.5542986425339365, + "grad_norm": 0.06770528086603515, + "learning_rate": 1.9208160799879838e-06, + "loss": 0.3723, + "step": 12078 + }, + { + "epoch": 4.554675716440422, + "grad_norm": 0.06442598766316136, + "learning_rate": 1.917593552891761e-06, + "loss": 0.3839, + "step": 12079 + }, + { + "epoch": 4.555052790346908, + "grad_norm": 0.06915951855833141, + "learning_rate": 1.914373664872522e-06, + "loss": 0.3781, + "step": 12080 + }, + { + "epoch": 4.5554298642533935, + "grad_norm": 0.06742115838500985, + "learning_rate": 1.9111564161534124e-06, + "loss": 0.3721, + "step": 12081 + }, + { + "epoch": 4.555806938159879, + "grad_norm": 0.0705459612641424, + "learning_rate": 1.9079418069573787e-06, + "loss": 0.3774, + "step": 12082 + }, + { + "epoch": 4.556184012066365, + "grad_norm": 0.0655665841250233, + "learning_rate": 1.9047298375071844e-06, + "loss": 0.3871, + "step": 12083 + }, + { + "epoch": 4.5565610859728505, + "grad_norm": 0.07026856487190776, + "learning_rate": 1.9015205080254296e-06, + "loss": 0.3799, + "step": 12084 + }, + { + "epoch": 4.556938159879336, + "grad_norm": 0.07021488227246922, + "learning_rate": 1.898313818734505e-06, + "loss": 0.3879, + "step": 12085 + }, + { + "epoch": 4.557315233785822, + "grad_norm": 0.06451828951632642, + "learning_rate": 1.8951097698566335e-06, + "loss": 0.3813, + "step": 12086 + }, + { + "epoch": 4.5576923076923075, + "grad_norm": 0.06440367954267401, + "learning_rate": 1.891908361613859e-06, + "loss": 0.3742, + "step": 12087 + }, + { + "epoch": 4.558069381598793, + "grad_norm": 0.06787713009994308, + "learning_rate": 1.8887095942280398e-06, + "loss": 0.3835, + "step": 12088 + }, + { + "epoch": 4.558446455505279, + "grad_norm": 0.06795189968942635, + "learning_rate": 1.8855134679208299e-06, + "loss": 0.3693, + "step": 12089 + }, + { + "epoch": 4.5588235294117645, + "grad_norm": 0.0687030934786879, + "learning_rate": 1.8823199829137406e-06, + "loss": 0.3827, + "step": 12090 + }, + { + "epoch": 4.55920060331825, + "grad_norm": 0.06685242011647569, + "learning_rate": 1.8791291394280575e-06, + "loss": 0.3782, + "step": 12091 + }, + { + "epoch": 4.559577677224736, + "grad_norm": 0.06661282319864839, + "learning_rate": 1.8759409376849147e-06, + "loss": 0.3846, + "step": 12092 + }, + { + "epoch": 4.5599547511312215, + "grad_norm": 0.06982614121674756, + "learning_rate": 1.8727553779052465e-06, + "loss": 0.383, + "step": 12093 + }, + { + "epoch": 4.560331825037707, + "grad_norm": 0.06396673924890017, + "learning_rate": 1.8695724603098187e-06, + "loss": 0.3704, + "step": 12094 + }, + { + "epoch": 4.560708898944193, + "grad_norm": 0.0657466681705329, + "learning_rate": 1.8663921851191879e-06, + "loss": 0.3632, + "step": 12095 + }, + { + "epoch": 4.5610859728506785, + "grad_norm": 0.06787967624458428, + "learning_rate": 1.8632145525537648e-06, + "loss": 0.3818, + "step": 12096 + }, + { + "epoch": 4.561463046757164, + "grad_norm": 0.06790608327795647, + "learning_rate": 1.860039562833742e-06, + "loss": 0.3719, + "step": 12097 + }, + { + "epoch": 4.56184012066365, + "grad_norm": 0.06601019813303953, + "learning_rate": 1.8568672161791435e-06, + "loss": 0.3844, + "step": 12098 + }, + { + "epoch": 4.5622171945701355, + "grad_norm": 0.06798261474252308, + "learning_rate": 1.8536975128098156e-06, + "loss": 0.3806, + "step": 12099 + }, + { + "epoch": 4.562594268476621, + "grad_norm": 0.0665357732175128, + "learning_rate": 1.8505304529454182e-06, + "loss": 0.387, + "step": 12100 + }, + { + "epoch": 4.562971342383107, + "grad_norm": 0.07380801165437972, + "learning_rate": 1.8473660368054158e-06, + "loss": 0.3867, + "step": 12101 + }, + { + "epoch": 4.5633484162895925, + "grad_norm": 0.0670307240843095, + "learning_rate": 1.844204264609113e-06, + "loss": 0.385, + "step": 12102 + }, + { + "epoch": 4.563725490196078, + "grad_norm": 0.07045035421362673, + "learning_rate": 1.8410451365756056e-06, + "loss": 0.3833, + "step": 12103 + }, + { + "epoch": 4.564102564102564, + "grad_norm": 0.06650578367597698, + "learning_rate": 1.8378886529238204e-06, + "loss": 0.3816, + "step": 12104 + }, + { + "epoch": 4.5644796380090495, + "grad_norm": 0.06856032279298892, + "learning_rate": 1.8347348138725074e-06, + "loss": 0.3864, + "step": 12105 + }, + { + "epoch": 4.564856711915535, + "grad_norm": 0.06547279424884783, + "learning_rate": 1.8315836196402203e-06, + "loss": 0.3793, + "step": 12106 + }, + { + "epoch": 4.565233785822021, + "grad_norm": 0.06904919302852469, + "learning_rate": 1.8284350704453224e-06, + "loss": 0.3805, + "step": 12107 + }, + { + "epoch": 4.5656108597285066, + "grad_norm": 0.07037838987344003, + "learning_rate": 1.8252891665060302e-06, + "loss": 0.3739, + "step": 12108 + }, + { + "epoch": 4.565987933634992, + "grad_norm": 0.06637386274771227, + "learning_rate": 1.8221459080403291e-06, + "loss": 0.3697, + "step": 12109 + }, + { + "epoch": 4.566365007541478, + "grad_norm": 0.06659600267137915, + "learning_rate": 1.8190052952660542e-06, + "loss": 0.38, + "step": 12110 + }, + { + "epoch": 4.566742081447964, + "grad_norm": 0.06706381426590916, + "learning_rate": 1.8158673284008443e-06, + "loss": 0.3716, + "step": 12111 + }, + { + "epoch": 4.567119155354449, + "grad_norm": 0.06630890040411203, + "learning_rate": 1.81273200766217e-06, + "loss": 0.3707, + "step": 12112 + }, + { + "epoch": 4.567496229260935, + "grad_norm": 0.06522152811076722, + "learning_rate": 1.80959933326728e-06, + "loss": 0.3699, + "step": 12113 + }, + { + "epoch": 4.567873303167421, + "grad_norm": 0.0663077277336769, + "learning_rate": 1.8064693054332937e-06, + "loss": 0.371, + "step": 12114 + }, + { + "epoch": 4.568250377073906, + "grad_norm": 0.07010527571993944, + "learning_rate": 1.8033419243771e-06, + "loss": 0.3755, + "step": 12115 + }, + { + "epoch": 4.568627450980392, + "grad_norm": 0.06526150513415642, + "learning_rate": 1.800217190315432e-06, + "loss": 0.3732, + "step": 12116 + }, + { + "epoch": 4.569004524886878, + "grad_norm": 0.06470388556365263, + "learning_rate": 1.7970951034648276e-06, + "loss": 0.3782, + "step": 12117 + }, + { + "epoch": 4.569381598793363, + "grad_norm": 0.06688836700168006, + "learning_rate": 1.7939756640416517e-06, + "loss": 0.3712, + "step": 12118 + }, + { + "epoch": 4.569758672699849, + "grad_norm": 0.06984127439187895, + "learning_rate": 1.7908588722620646e-06, + "loss": 0.3854, + "step": 12119 + }, + { + "epoch": 4.570135746606335, + "grad_norm": 0.06966722777439215, + "learning_rate": 1.7877447283420756e-06, + "loss": 0.3851, + "step": 12120 + }, + { + "epoch": 4.57051282051282, + "grad_norm": 0.06503903729888938, + "learning_rate": 1.7846332324974814e-06, + "loss": 0.3775, + "step": 12121 + }, + { + "epoch": 4.570889894419306, + "grad_norm": 0.06736514840900355, + "learning_rate": 1.7815243849439047e-06, + "loss": 0.3805, + "step": 12122 + }, + { + "epoch": 4.571266968325792, + "grad_norm": 0.06713802901069037, + "learning_rate": 1.7784181858967864e-06, + "loss": 0.386, + "step": 12123 + }, + { + "epoch": 4.571644042232277, + "grad_norm": 0.0691438687381477, + "learning_rate": 1.7753146355713902e-06, + "loss": 0.3702, + "step": 12124 + }, + { + "epoch": 4.572021116138763, + "grad_norm": 0.0653119210796453, + "learning_rate": 1.7722137341827793e-06, + "loss": 0.3812, + "step": 12125 + }, + { + "epoch": 4.572398190045249, + "grad_norm": 0.07050131489578237, + "learning_rate": 1.7691154819458534e-06, + "loss": 0.382, + "step": 12126 + }, + { + "epoch": 4.572775263951734, + "grad_norm": 0.06619261976566646, + "learning_rate": 1.7660198790753113e-06, + "loss": 0.3721, + "step": 12127 + }, + { + "epoch": 4.57315233785822, + "grad_norm": 0.06643203908570428, + "learning_rate": 1.7629269257856796e-06, + "loss": 0.3604, + "step": 12128 + }, + { + "epoch": 4.573529411764706, + "grad_norm": 0.06467598551519925, + "learning_rate": 1.7598366222912933e-06, + "loss": 0.3785, + "step": 12129 + }, + { + "epoch": 4.573906485671191, + "grad_norm": 0.06748718048749702, + "learning_rate": 1.7567489688063144e-06, + "loss": 0.3734, + "step": 12130 + }, + { + "epoch": 4.574283559577677, + "grad_norm": 0.06525656178318241, + "learning_rate": 1.7536639655447007e-06, + "loss": 0.3731, + "step": 12131 + }, + { + "epoch": 4.574660633484163, + "grad_norm": 0.0698920447224868, + "learning_rate": 1.7505816127202635e-06, + "loss": 0.3835, + "step": 12132 + }, + { + "epoch": 4.575037707390648, + "grad_norm": 0.06803669627295417, + "learning_rate": 1.7475019105465829e-06, + "loss": 0.3761, + "step": 12133 + }, + { + "epoch": 4.575414781297134, + "grad_norm": 0.06552206334926744, + "learning_rate": 1.7444248592370926e-06, + "loss": 0.3654, + "step": 12134 + }, + { + "epoch": 4.57579185520362, + "grad_norm": 0.06545895075171287, + "learning_rate": 1.7413504590050266e-06, + "loss": 0.3771, + "step": 12135 + }, + { + "epoch": 4.576168929110105, + "grad_norm": 0.07116460622274169, + "learning_rate": 1.7382787100634457e-06, + "loss": 0.3754, + "step": 12136 + }, + { + "epoch": 4.576546003016591, + "grad_norm": 0.06653365073070991, + "learning_rate": 1.7352096126251972e-06, + "loss": 0.3796, + "step": 12137 + }, + { + "epoch": 4.576923076923077, + "grad_norm": 0.06661465363264149, + "learning_rate": 1.7321431669029953e-06, + "loss": 0.3728, + "step": 12138 + }, + { + "epoch": 4.577300150829562, + "grad_norm": 0.07311419305066146, + "learning_rate": 1.7290793731093237e-06, + "loss": 0.3944, + "step": 12139 + }, + { + "epoch": 4.577677224736048, + "grad_norm": 0.06570986608763783, + "learning_rate": 1.726018231456501e-06, + "loss": 0.3874, + "step": 12140 + }, + { + "epoch": 4.578054298642534, + "grad_norm": 0.06757792207348245, + "learning_rate": 1.722959742156669e-06, + "loss": 0.3667, + "step": 12141 + }, + { + "epoch": 4.578431372549019, + "grad_norm": 0.0736708021309376, + "learning_rate": 1.7199039054217825e-06, + "loss": 0.373, + "step": 12142 + }, + { + "epoch": 4.578808446455505, + "grad_norm": 0.07135384150941716, + "learning_rate": 1.7168507214635876e-06, + "loss": 0.3729, + "step": 12143 + }, + { + "epoch": 4.579185520361991, + "grad_norm": 0.06934709819987389, + "learning_rate": 1.713800190493693e-06, + "loss": 0.3785, + "step": 12144 + }, + { + "epoch": 4.579562594268476, + "grad_norm": 0.0695921781774161, + "learning_rate": 1.710752312723476e-06, + "loss": 0.3731, + "step": 12145 + }, + { + "epoch": 4.579939668174962, + "grad_norm": 0.06654285129271834, + "learning_rate": 1.7077070883641677e-06, + "loss": 0.3847, + "step": 12146 + }, + { + "epoch": 4.580316742081448, + "grad_norm": 0.07810553398201496, + "learning_rate": 1.7046645176267863e-06, + "loss": 0.3765, + "step": 12147 + }, + { + "epoch": 4.580693815987933, + "grad_norm": 0.06847017753282766, + "learning_rate": 1.7016246007221937e-06, + "loss": 0.3806, + "step": 12148 + }, + { + "epoch": 4.581070889894419, + "grad_norm": 0.06642820140280861, + "learning_rate": 1.6985873378610395e-06, + "loss": 0.3774, + "step": 12149 + }, + { + "epoch": 4.581447963800905, + "grad_norm": 0.06491352904565144, + "learning_rate": 1.6955527292538133e-06, + "loss": 0.3778, + "step": 12150 + }, + { + "epoch": 4.58182503770739, + "grad_norm": 0.06907041866005909, + "learning_rate": 1.6925207751108042e-06, + "loss": 0.3785, + "step": 12151 + }, + { + "epoch": 4.582202111613876, + "grad_norm": 0.06975081310859947, + "learning_rate": 1.6894914756421289e-06, + "loss": 0.3767, + "step": 12152 + }, + { + "epoch": 4.582579185520362, + "grad_norm": 0.06714997972217039, + "learning_rate": 1.6864648310577125e-06, + "loss": 0.3826, + "step": 12153 + }, + { + "epoch": 4.582956259426847, + "grad_norm": 0.06347067341721133, + "learning_rate": 1.6834408415673031e-06, + "loss": 0.3689, + "step": 12154 + }, + { + "epoch": 4.583333333333333, + "grad_norm": 0.06983279655083746, + "learning_rate": 1.6804195073804442e-06, + "loss": 0.3774, + "step": 12155 + }, + { + "epoch": 4.583710407239819, + "grad_norm": 0.06864293847111066, + "learning_rate": 1.677400828706537e-06, + "loss": 0.3775, + "step": 12156 + }, + { + "epoch": 4.584087481146304, + "grad_norm": 0.0701716251424793, + "learning_rate": 1.6743848057547517e-06, + "loss": 0.383, + "step": 12157 + }, + { + "epoch": 4.58446455505279, + "grad_norm": 0.06371971817778031, + "learning_rate": 1.6713714387341084e-06, + "loss": 0.3667, + "step": 12158 + }, + { + "epoch": 4.584841628959276, + "grad_norm": 0.06963821970693643, + "learning_rate": 1.6683607278534265e-06, + "loss": 0.3829, + "step": 12159 + }, + { + "epoch": 4.585218702865761, + "grad_norm": 0.06802362250922121, + "learning_rate": 1.6653526733213522e-06, + "loss": 0.384, + "step": 12160 + }, + { + "epoch": 4.585595776772247, + "grad_norm": 0.06756533396406322, + "learning_rate": 1.6623472753463232e-06, + "loss": 0.3737, + "step": 12161 + }, + { + "epoch": 4.585972850678733, + "grad_norm": 0.06706376234810492, + "learning_rate": 1.659344534136631e-06, + "loss": 0.3745, + "step": 12162 + }, + { + "epoch": 4.586349924585218, + "grad_norm": 0.06532865236389086, + "learning_rate": 1.656344449900349e-06, + "loss": 0.3811, + "step": 12163 + }, + { + "epoch": 4.586726998491704, + "grad_norm": 0.06674584951480018, + "learning_rate": 1.6533470228453908e-06, + "loss": 0.3798, + "step": 12164 + }, + { + "epoch": 4.58710407239819, + "grad_norm": 0.07039592096738477, + "learning_rate": 1.6503522531794658e-06, + "loss": 0.3774, + "step": 12165 + }, + { + "epoch": 4.587481146304675, + "grad_norm": 0.06651489136018099, + "learning_rate": 1.6473601411101147e-06, + "loss": 0.3862, + "step": 12166 + }, + { + "epoch": 4.587858220211161, + "grad_norm": 0.06464360391170848, + "learning_rate": 1.6443706868446874e-06, + "loss": 0.3808, + "step": 12167 + }, + { + "epoch": 4.588235294117647, + "grad_norm": 0.06680935369018595, + "learning_rate": 1.6413838905903556e-06, + "loss": 0.3741, + "step": 12168 + }, + { + "epoch": 4.588612368024132, + "grad_norm": 0.06478785002766706, + "learning_rate": 1.6383997525540917e-06, + "loss": 0.371, + "step": 12169 + }, + { + "epoch": 4.588989441930618, + "grad_norm": 0.06210815651206789, + "learning_rate": 1.6354182729426992e-06, + "loss": 0.3835, + "step": 12170 + }, + { + "epoch": 4.589366515837104, + "grad_norm": 0.06895727680912506, + "learning_rate": 1.6324394519627862e-06, + "loss": 0.3703, + "step": 12171 + }, + { + "epoch": 4.589743589743589, + "grad_norm": 0.0630391755784093, + "learning_rate": 1.6294632898207918e-06, + "loss": 0.3778, + "step": 12172 + }, + { + "epoch": 4.590120663650075, + "grad_norm": 0.06892517708014716, + "learning_rate": 1.6264897867229556e-06, + "loss": 0.3807, + "step": 12173 + }, + { + "epoch": 4.590497737556561, + "grad_norm": 0.06634841398273003, + "learning_rate": 1.6235189428753439e-06, + "loss": 0.3775, + "step": 12174 + }, + { + "epoch": 4.590874811463046, + "grad_norm": 0.07656527373325506, + "learning_rate": 1.6205507584838232e-06, + "loss": 0.3863, + "step": 12175 + }, + { + "epoch": 4.591251885369532, + "grad_norm": 0.06527846639163727, + "learning_rate": 1.6175852337540953e-06, + "loss": 0.3762, + "step": 12176 + }, + { + "epoch": 4.591628959276018, + "grad_norm": 0.0653105512644942, + "learning_rate": 1.6146223688916674e-06, + "loss": 0.3737, + "step": 12177 + }, + { + "epoch": 4.592006033182503, + "grad_norm": 0.0647264772306168, + "learning_rate": 1.6116621641018548e-06, + "loss": 0.3783, + "step": 12178 + }, + { + "epoch": 4.592383107088989, + "grad_norm": 0.06782443597839465, + "learning_rate": 1.6087046195898092e-06, + "loss": 0.3767, + "step": 12179 + }, + { + "epoch": 4.592760180995475, + "grad_norm": 0.06660592730205846, + "learning_rate": 1.6057497355604824e-06, + "loss": 0.3704, + "step": 12180 + }, + { + "epoch": 4.5931372549019605, + "grad_norm": 0.06503319683722747, + "learning_rate": 1.6027975122186434e-06, + "loss": 0.3866, + "step": 12181 + }, + { + "epoch": 4.593514328808446, + "grad_norm": 0.0637898255083224, + "learning_rate": 1.5998479497688712e-06, + "loss": 0.3694, + "step": 12182 + }, + { + "epoch": 4.593891402714932, + "grad_norm": 0.06563154686036748, + "learning_rate": 1.59690104841558e-06, + "loss": 0.3815, + "step": 12183 + }, + { + "epoch": 4.5942684766214175, + "grad_norm": 0.0678651908326703, + "learning_rate": 1.5939568083629842e-06, + "loss": 0.3741, + "step": 12184 + }, + { + "epoch": 4.594645550527903, + "grad_norm": 0.0618854349189333, + "learning_rate": 1.5910152298151116e-06, + "loss": 0.3691, + "step": 12185 + }, + { + "epoch": 4.595022624434389, + "grad_norm": 0.06523872804231798, + "learning_rate": 1.588076312975817e-06, + "loss": 0.3819, + "step": 12186 + }, + { + "epoch": 4.5953996983408745, + "grad_norm": 0.06691133597846284, + "learning_rate": 1.5851400580487642e-06, + "loss": 0.3763, + "step": 12187 + }, + { + "epoch": 4.59577677224736, + "grad_norm": 0.06398620402423327, + "learning_rate": 1.5822064652374302e-06, + "loss": 0.3677, + "step": 12188 + }, + { + "epoch": 4.596153846153846, + "grad_norm": 0.06890768445308347, + "learning_rate": 1.5792755347451106e-06, + "loss": 0.3843, + "step": 12189 + }, + { + "epoch": 4.5965309200603315, + "grad_norm": 0.06824488567490991, + "learning_rate": 1.576347266774918e-06, + "loss": 0.3633, + "step": 12190 + }, + { + "epoch": 4.596907993966817, + "grad_norm": 0.07079658704905682, + "learning_rate": 1.5734216615297748e-06, + "loss": 0.3846, + "step": 12191 + }, + { + "epoch": 4.597285067873303, + "grad_norm": 0.0664914827973549, + "learning_rate": 1.5704987192124299e-06, + "loss": 0.3802, + "step": 12192 + }, + { + "epoch": 4.5976621417797885, + "grad_norm": 0.06750186249689295, + "learning_rate": 1.5675784400254412e-06, + "loss": 0.3798, + "step": 12193 + }, + { + "epoch": 4.598039215686274, + "grad_norm": 0.06535985241873742, + "learning_rate": 1.5646608241711714e-06, + "loss": 0.3706, + "step": 12194 + }, + { + "epoch": 4.59841628959276, + "grad_norm": 0.06794416049847926, + "learning_rate": 1.561745871851814e-06, + "loss": 0.3798, + "step": 12195 + }, + { + "epoch": 4.5987933634992455, + "grad_norm": 0.06602728461025108, + "learning_rate": 1.5588335832693723e-06, + "loss": 0.3858, + "step": 12196 + }, + { + "epoch": 4.599170437405731, + "grad_norm": 0.06699520901347808, + "learning_rate": 1.555923958625667e-06, + "loss": 0.3823, + "step": 12197 + }, + { + "epoch": 4.599547511312217, + "grad_norm": 0.06595294873434106, + "learning_rate": 1.553016998122332e-06, + "loss": 0.3735, + "step": 12198 + }, + { + "epoch": 4.5999245852187025, + "grad_norm": 0.06560466540724573, + "learning_rate": 1.5501127019608198e-06, + "loss": 0.373, + "step": 12199 + }, + { + "epoch": 4.600301659125188, + "grad_norm": 0.06427930985229115, + "learning_rate": 1.5472110703423871e-06, + "loss": 0.3795, + "step": 12200 + }, + { + "epoch": 4.600678733031674, + "grad_norm": 0.06852277686276739, + "learning_rate": 1.5443121034681218e-06, + "loss": 0.3771, + "step": 12201 + }, + { + "epoch": 4.6010558069381595, + "grad_norm": 0.06946545485516654, + "learning_rate": 1.5414158015389169e-06, + "loss": 0.3822, + "step": 12202 + }, + { + "epoch": 4.601432880844645, + "grad_norm": 0.06873590917410026, + "learning_rate": 1.5385221647554826e-06, + "loss": 0.3732, + "step": 12203 + }, + { + "epoch": 4.601809954751131, + "grad_norm": 0.06363460928619248, + "learning_rate": 1.5356311933183522e-06, + "loss": 0.3814, + "step": 12204 + }, + { + "epoch": 4.6021870286576165, + "grad_norm": 0.0650070867300524, + "learning_rate": 1.5327428874278584e-06, + "loss": 0.3594, + "step": 12205 + }, + { + "epoch": 4.602564102564102, + "grad_norm": 0.06854756986170507, + "learning_rate": 1.5298572472841611e-06, + "loss": 0.3867, + "step": 12206 + }, + { + "epoch": 4.602941176470588, + "grad_norm": 0.06665103987137419, + "learning_rate": 1.5269742730872384e-06, + "loss": 0.3816, + "step": 12207 + }, + { + "epoch": 4.6033182503770735, + "grad_norm": 0.06728601583843495, + "learning_rate": 1.5240939650368724e-06, + "loss": 0.3798, + "step": 12208 + }, + { + "epoch": 4.603695324283559, + "grad_norm": 0.06991467512327366, + "learning_rate": 1.5212163233326637e-06, + "loss": 0.3777, + "step": 12209 + }, + { + "epoch": 4.604072398190045, + "grad_norm": 0.06463657020992591, + "learning_rate": 1.5183413481740305e-06, + "loss": 0.3806, + "step": 12210 + }, + { + "epoch": 4.6044494720965305, + "grad_norm": 0.06522947880562528, + "learning_rate": 1.5154690397602135e-06, + "loss": 0.3746, + "step": 12211 + }, + { + "epoch": 4.604826546003016, + "grad_norm": 0.06394318843619631, + "learning_rate": 1.512599398290253e-06, + "loss": 0.3722, + "step": 12212 + }, + { + "epoch": 4.605203619909502, + "grad_norm": 0.06885112696979394, + "learning_rate": 1.5097324239630217e-06, + "loss": 0.3821, + "step": 12213 + }, + { + "epoch": 4.605580693815988, + "grad_norm": 0.06721752043858095, + "learning_rate": 1.506868116977187e-06, + "loss": 0.3921, + "step": 12214 + }, + { + "epoch": 4.605957767722473, + "grad_norm": 0.06559513490668854, + "learning_rate": 1.5040064775312479e-06, + "loss": 0.3726, + "step": 12215 + }, + { + "epoch": 4.606334841628959, + "grad_norm": 0.06727494904347305, + "learning_rate": 1.501147505823517e-06, + "loss": 0.3783, + "step": 12216 + }, + { + "epoch": 4.606711915535445, + "grad_norm": 0.06881825514843617, + "learning_rate": 1.4982912020521156e-06, + "loss": 0.3897, + "step": 12217 + }, + { + "epoch": 4.60708898944193, + "grad_norm": 0.06646115460263059, + "learning_rate": 1.495437566414979e-06, + "loss": 0.3695, + "step": 12218 + }, + { + "epoch": 4.607466063348416, + "grad_norm": 0.06607683466635476, + "learning_rate": 1.4925865991098732e-06, + "loss": 0.3715, + "step": 12219 + }, + { + "epoch": 4.607843137254902, + "grad_norm": 0.06406670206113868, + "learning_rate": 1.4897383003343514e-06, + "loss": 0.381, + "step": 12220 + }, + { + "epoch": 4.608220211161387, + "grad_norm": 0.0647185597663571, + "learning_rate": 1.4868926702858156e-06, + "loss": 0.37, + "step": 12221 + }, + { + "epoch": 4.608597285067873, + "grad_norm": 0.06426745449487432, + "learning_rate": 1.4840497091614503e-06, + "loss": 0.3735, + "step": 12222 + }, + { + "epoch": 4.608974358974359, + "grad_norm": 0.06801197701289684, + "learning_rate": 1.4812094171582758e-06, + "loss": 0.3738, + "step": 12223 + }, + { + "epoch": 4.609351432880844, + "grad_norm": 0.06872435407437558, + "learning_rate": 1.4783717944731258e-06, + "loss": 0.3766, + "step": 12224 + }, + { + "epoch": 4.609728506787331, + "grad_norm": 0.0665558087595542, + "learning_rate": 1.4755368413026428e-06, + "loss": 0.3779, + "step": 12225 + }, + { + "epoch": 4.610105580693816, + "grad_norm": 0.06673390527005503, + "learning_rate": 1.4727045578432787e-06, + "loss": 0.3776, + "step": 12226 + }, + { + "epoch": 4.610482654600302, + "grad_norm": 0.06641645355730158, + "learning_rate": 1.4698749442913252e-06, + "loss": 0.3912, + "step": 12227 + }, + { + "epoch": 4.610859728506787, + "grad_norm": 0.0711520202977343, + "learning_rate": 1.4670480008428568e-06, + "loss": 0.3775, + "step": 12228 + }, + { + "epoch": 4.6112368024132735, + "grad_norm": 0.06914185370848333, + "learning_rate": 1.4642237276937787e-06, + "loss": 0.3764, + "step": 12229 + }, + { + "epoch": 4.611613876319758, + "grad_norm": 0.06407858273856987, + "learning_rate": 1.461402125039819e-06, + "loss": 0.3745, + "step": 12230 + }, + { + "epoch": 4.611990950226245, + "grad_norm": 0.06904698181402935, + "learning_rate": 1.4585831930765148e-06, + "loss": 0.3767, + "step": 12231 + }, + { + "epoch": 4.61236802413273, + "grad_norm": 0.06862663120159268, + "learning_rate": 1.4557669319991985e-06, + "loss": 0.3777, + "step": 12232 + }, + { + "epoch": 4.612745098039216, + "grad_norm": 0.06966215513372305, + "learning_rate": 1.4529533420030517e-06, + "loss": 0.3661, + "step": 12233 + }, + { + "epoch": 4.613122171945701, + "grad_norm": 0.06758896252480152, + "learning_rate": 1.450142423283043e-06, + "loss": 0.371, + "step": 12234 + }, + { + "epoch": 4.6134992458521875, + "grad_norm": 0.06793802318453107, + "learning_rate": 1.4473341760339721e-06, + "loss": 0.3724, + "step": 12235 + }, + { + "epoch": 4.613876319758672, + "grad_norm": 0.0654077653011697, + "learning_rate": 1.4445286004504478e-06, + "loss": 0.382, + "step": 12236 + }, + { + "epoch": 4.614253393665159, + "grad_norm": 0.06626782281502797, + "learning_rate": 1.4417256967268967e-06, + "loss": 0.3852, + "step": 12237 + }, + { + "epoch": 4.614630467571644, + "grad_norm": 0.06669147219332773, + "learning_rate": 1.4389254650575458e-06, + "loss": 0.373, + "step": 12238 + }, + { + "epoch": 4.61500754147813, + "grad_norm": 0.06615007172881064, + "learning_rate": 1.4361279056364618e-06, + "loss": 0.3854, + "step": 12239 + }, + { + "epoch": 4.615384615384615, + "grad_norm": 0.06749568333850889, + "learning_rate": 1.4333330186575079e-06, + "loss": 0.3919, + "step": 12240 + }, + { + "epoch": 4.6157616892911015, + "grad_norm": 0.06536391029681125, + "learning_rate": 1.4305408043143643e-06, + "loss": 0.3787, + "step": 12241 + }, + { + "epoch": 4.616138763197586, + "grad_norm": 0.0647086372875106, + "learning_rate": 1.4277512628005297e-06, + "loss": 0.3769, + "step": 12242 + }, + { + "epoch": 4.616515837104073, + "grad_norm": 0.06913223203871302, + "learning_rate": 1.4249643943093294e-06, + "loss": 0.3738, + "step": 12243 + }, + { + "epoch": 4.616892911010558, + "grad_norm": 0.06448025101164255, + "learning_rate": 1.4221801990338669e-06, + "loss": 0.3848, + "step": 12244 + }, + { + "epoch": 4.617269984917044, + "grad_norm": 0.06306563177715047, + "learning_rate": 1.4193986771671076e-06, + "loss": 0.3662, + "step": 12245 + }, + { + "epoch": 4.617647058823529, + "grad_norm": 0.062229172758125895, + "learning_rate": 1.4166198289017952e-06, + "loss": 0.3822, + "step": 12246 + }, + { + "epoch": 4.6180241327300156, + "grad_norm": 0.06602280871724488, + "learning_rate": 1.4138436544305046e-06, + "loss": 0.3739, + "step": 12247 + }, + { + "epoch": 4.6184012066365, + "grad_norm": 0.06530194785010147, + "learning_rate": 1.411070153945624e-06, + "loss": 0.3815, + "step": 12248 + }, + { + "epoch": 4.618778280542987, + "grad_norm": 0.06663613689957434, + "learning_rate": 1.4082993276393597e-06, + "loss": 0.3789, + "step": 12249 + }, + { + "epoch": 4.619155354449472, + "grad_norm": 0.06867828031291574, + "learning_rate": 1.4055311757037093e-06, + "loss": 0.3798, + "step": 12250 + }, + { + "epoch": 4.619532428355958, + "grad_norm": 0.06558446084443258, + "learning_rate": 1.4027656983305237e-06, + "loss": 0.3781, + "step": 12251 + }, + { + "epoch": 4.619909502262443, + "grad_norm": 0.06908128027476054, + "learning_rate": 1.4000028957114365e-06, + "loss": 0.3748, + "step": 12252 + }, + { + "epoch": 4.62028657616893, + "grad_norm": 0.06306611520880692, + "learning_rate": 1.397242768037912e-06, + "loss": 0.3754, + "step": 12253 + }, + { + "epoch": 4.620663650075414, + "grad_norm": 0.06743534846577579, + "learning_rate": 1.3944853155012194e-06, + "loss": 0.367, + "step": 12254 + }, + { + "epoch": 4.621040723981901, + "grad_norm": 0.06947598064503764, + "learning_rate": 1.3917305382924595e-06, + "loss": 0.3768, + "step": 12255 + }, + { + "epoch": 4.621417797888386, + "grad_norm": 0.06617390879747902, + "learning_rate": 1.388978436602515e-06, + "loss": 0.3726, + "step": 12256 + }, + { + "epoch": 4.621794871794872, + "grad_norm": 0.06494193542772751, + "learning_rate": 1.386229010622131e-06, + "loss": 0.3765, + "step": 12257 + }, + { + "epoch": 4.622171945701357, + "grad_norm": 0.06792097593920163, + "learning_rate": 1.3834822605418175e-06, + "loss": 0.3801, + "step": 12258 + }, + { + "epoch": 4.622549019607844, + "grad_norm": 0.0671570146223989, + "learning_rate": 1.3807381865519332e-06, + "loss": 0.382, + "step": 12259 + }, + { + "epoch": 4.622926093514328, + "grad_norm": 0.06578848784977828, + "learning_rate": 1.3779967888426326e-06, + "loss": 0.3736, + "step": 12260 + }, + { + "epoch": 4.623303167420815, + "grad_norm": 0.06661661650422666, + "learning_rate": 1.3752580676039061e-06, + "loss": 0.389, + "step": 12261 + }, + { + "epoch": 4.6236802413273, + "grad_norm": 0.07073056512750797, + "learning_rate": 1.3725220230255265e-06, + "loss": 0.3729, + "step": 12262 + }, + { + "epoch": 4.624057315233786, + "grad_norm": 0.06523800311633704, + "learning_rate": 1.369788655297115e-06, + "loss": 0.3815, + "step": 12263 + }, + { + "epoch": 4.624434389140271, + "grad_norm": 0.06637851316062818, + "learning_rate": 1.3670579646080807e-06, + "loss": 0.3742, + "step": 12264 + }, + { + "epoch": 4.624811463046758, + "grad_norm": 0.06499773775778547, + "learning_rate": 1.3643299511476627e-06, + "loss": 0.3698, + "step": 12265 + }, + { + "epoch": 4.625188536953242, + "grad_norm": 0.06554086047704688, + "learning_rate": 1.36160461510491e-06, + "loss": 0.3892, + "step": 12266 + }, + { + "epoch": 4.625565610859729, + "grad_norm": 0.06461995205655734, + "learning_rate": 1.358881956668685e-06, + "loss": 0.3889, + "step": 12267 + }, + { + "epoch": 4.625942684766214, + "grad_norm": 0.06432022954826877, + "learning_rate": 1.3561619760276634e-06, + "loss": 0.3769, + "step": 12268 + }, + { + "epoch": 4.6263197586727, + "grad_norm": 0.06523123613434424, + "learning_rate": 1.3534446733703433e-06, + "loss": 0.3741, + "step": 12269 + }, + { + "epoch": 4.626696832579185, + "grad_norm": 0.06487715383253655, + "learning_rate": 1.3507300488850272e-06, + "loss": 0.3781, + "step": 12270 + }, + { + "epoch": 4.627073906485672, + "grad_norm": 0.06788842130880202, + "learning_rate": 1.348018102759836e-06, + "loss": 0.375, + "step": 12271 + }, + { + "epoch": 4.627450980392156, + "grad_norm": 0.06680106981453054, + "learning_rate": 1.3453088351827038e-06, + "loss": 0.3699, + "step": 12272 + }, + { + "epoch": 4.627828054298643, + "grad_norm": 0.0694515232542582, + "learning_rate": 1.3426022463413913e-06, + "loss": 0.3766, + "step": 12273 + }, + { + "epoch": 4.628205128205128, + "grad_norm": 0.06604089329925683, + "learning_rate": 1.339898336423442e-06, + "loss": 0.3674, + "step": 12274 + }, + { + "epoch": 4.628582202111614, + "grad_norm": 0.06738253358112391, + "learning_rate": 1.337197105616257e-06, + "loss": 0.3862, + "step": 12275 + }, + { + "epoch": 4.628959276018099, + "grad_norm": 0.0723587822676707, + "learning_rate": 1.3344985541070111e-06, + "loss": 0.3724, + "step": 12276 + }, + { + "epoch": 4.629336349924586, + "grad_norm": 0.06972714669676923, + "learning_rate": 1.3318026820827235e-06, + "loss": 0.3807, + "step": 12277 + }, + { + "epoch": 4.62971342383107, + "grad_norm": 0.06503895676895646, + "learning_rate": 1.3291094897302092e-06, + "loss": 0.3759, + "step": 12278 + }, + { + "epoch": 4.630090497737557, + "grad_norm": 0.06789617908043744, + "learning_rate": 1.3264189772361103e-06, + "loss": 0.3693, + "step": 12279 + }, + { + "epoch": 4.630467571644042, + "grad_norm": 0.06610381185213553, + "learning_rate": 1.323731144786864e-06, + "loss": 0.3692, + "step": 12280 + }, + { + "epoch": 4.630844645550528, + "grad_norm": 0.07121849162218712, + "learning_rate": 1.3210459925687525e-06, + "loss": 0.3822, + "step": 12281 + }, + { + "epoch": 4.631221719457013, + "grad_norm": 0.06536227950009682, + "learning_rate": 1.3183635207678402e-06, + "loss": 0.3794, + "step": 12282 + }, + { + "epoch": 4.6315987933635, + "grad_norm": 0.06832527297814772, + "learning_rate": 1.3156837295700275e-06, + "loss": 0.3732, + "step": 12283 + }, + { + "epoch": 4.6319758672699844, + "grad_norm": 0.06724617541858945, + "learning_rate": 1.3130066191610192e-06, + "loss": 0.3874, + "step": 12284 + }, + { + "epoch": 4.632352941176471, + "grad_norm": 0.06453307522313138, + "learning_rate": 1.3103321897263421e-06, + "loss": 0.3781, + "step": 12285 + }, + { + "epoch": 4.632730015082956, + "grad_norm": 0.06665847513402098, + "learning_rate": 1.307660441451315e-06, + "loss": 0.3856, + "step": 12286 + }, + { + "epoch": 4.633107088989442, + "grad_norm": 0.06429196932237699, + "learning_rate": 1.3049913745211095e-06, + "loss": 0.3702, + "step": 12287 + }, + { + "epoch": 4.633484162895927, + "grad_norm": 0.0662307156445099, + "learning_rate": 1.3023249891206756e-06, + "loss": 0.3701, + "step": 12288 + }, + { + "epoch": 4.633861236802414, + "grad_norm": 0.06444985907162792, + "learning_rate": 1.2996612854347946e-06, + "loss": 0.3753, + "step": 12289 + }, + { + "epoch": 4.6342383107088985, + "grad_norm": 0.06666444965045952, + "learning_rate": 1.297000263648056e-06, + "loss": 0.3796, + "step": 12290 + }, + { + "epoch": 4.634615384615385, + "grad_norm": 0.06420784477103336, + "learning_rate": 1.2943419239448773e-06, + "loss": 0.3817, + "step": 12291 + }, + { + "epoch": 4.63499245852187, + "grad_norm": 0.06704424100646081, + "learning_rate": 1.2916862665094622e-06, + "loss": 0.3675, + "step": 12292 + }, + { + "epoch": 4.635369532428356, + "grad_norm": 0.06721828256301661, + "learning_rate": 1.289033291525863e-06, + "loss": 0.3739, + "step": 12293 + }, + { + "epoch": 4.635746606334841, + "grad_norm": 0.0668523812576002, + "learning_rate": 1.2863829991779198e-06, + "loss": 0.3675, + "step": 12294 + }, + { + "epoch": 4.636123680241328, + "grad_norm": 0.0636891362442142, + "learning_rate": 1.28373538964929e-06, + "loss": 0.3766, + "step": 12295 + }, + { + "epoch": 4.6365007541478125, + "grad_norm": 0.07159865392242576, + "learning_rate": 1.2810904631234577e-06, + "loss": 0.3818, + "step": 12296 + }, + { + "epoch": 4.636877828054299, + "grad_norm": 0.0680100143165257, + "learning_rate": 1.278448219783721e-06, + "loss": 0.3776, + "step": 12297 + }, + { + "epoch": 4.637254901960784, + "grad_norm": 0.0694911501983506, + "learning_rate": 1.2758086598131646e-06, + "loss": 0.3732, + "step": 12298 + }, + { + "epoch": 4.63763197586727, + "grad_norm": 0.06360603147922243, + "learning_rate": 1.2731717833947266e-06, + "loss": 0.3726, + "step": 12299 + }, + { + "epoch": 4.638009049773755, + "grad_norm": 0.06531405932424993, + "learning_rate": 1.2705375907111317e-06, + "loss": 0.3785, + "step": 12300 + }, + { + "epoch": 4.638386123680242, + "grad_norm": 0.06445070011386912, + "learning_rate": 1.2679060819449274e-06, + "loss": 0.3864, + "step": 12301 + }, + { + "epoch": 4.6387631975867265, + "grad_norm": 0.06847066715323775, + "learning_rate": 1.2652772572784788e-06, + "loss": 0.3925, + "step": 12302 + }, + { + "epoch": 4.639140271493213, + "grad_norm": 0.0677964998150309, + "learning_rate": 1.2626511168939604e-06, + "loss": 0.3846, + "step": 12303 + }, + { + "epoch": 4.639517345399698, + "grad_norm": 0.0642921502844503, + "learning_rate": 1.2600276609733507e-06, + "loss": 0.37, + "step": 12304 + }, + { + "epoch": 4.639894419306184, + "grad_norm": 0.06671494174072932, + "learning_rate": 1.2574068896984737e-06, + "loss": 0.3816, + "step": 12305 + }, + { + "epoch": 4.640271493212669, + "grad_norm": 0.07261403116163576, + "learning_rate": 1.2547888032509259e-06, + "loss": 0.3824, + "step": 12306 + }, + { + "epoch": 4.640648567119156, + "grad_norm": 0.06841737948088443, + "learning_rate": 1.252173401812149e-06, + "loss": 0.3796, + "step": 12307 + }, + { + "epoch": 4.641025641025641, + "grad_norm": 0.06381380296763764, + "learning_rate": 1.249560685563389e-06, + "loss": 0.377, + "step": 12308 + }, + { + "epoch": 4.641402714932127, + "grad_norm": 0.06602167350630965, + "learning_rate": 1.2469506546857014e-06, + "loss": 0.3825, + "step": 12309 + }, + { + "epoch": 4.641779788838613, + "grad_norm": 0.06623559721349342, + "learning_rate": 1.2443433093599545e-06, + "loss": 0.3863, + "step": 12310 + }, + { + "epoch": 4.642156862745098, + "grad_norm": 0.06458571836064529, + "learning_rate": 1.2417386497668482e-06, + "loss": 0.366, + "step": 12311 + }, + { + "epoch": 4.642533936651584, + "grad_norm": 0.06538488110136641, + "learning_rate": 1.2391366760868696e-06, + "loss": 0.3775, + "step": 12312 + }, + { + "epoch": 4.64291101055807, + "grad_norm": 0.06697279737710468, + "learning_rate": 1.2365373885003363e-06, + "loss": 0.3764, + "step": 12313 + }, + { + "epoch": 4.643288084464555, + "grad_norm": 0.06610495753916566, + "learning_rate": 1.2339407871873842e-06, + "loss": 0.3714, + "step": 12314 + }, + { + "epoch": 4.643665158371041, + "grad_norm": 0.06785503891713303, + "learning_rate": 1.2313468723279453e-06, + "loss": 0.3727, + "step": 12315 + }, + { + "epoch": 4.644042232277527, + "grad_norm": 0.06548616112008931, + "learning_rate": 1.2287556441017823e-06, + "loss": 0.3756, + "step": 12316 + }, + { + "epoch": 4.644419306184012, + "grad_norm": 0.06857347253853309, + "learning_rate": 1.226167102688467e-06, + "loss": 0.3803, + "step": 12317 + }, + { + "epoch": 4.644796380090498, + "grad_norm": 0.06355948303747377, + "learning_rate": 1.2235812482673759e-06, + "loss": 0.3716, + "step": 12318 + }, + { + "epoch": 4.645173453996984, + "grad_norm": 0.06307730778795541, + "learning_rate": 1.220998081017708e-06, + "loss": 0.3646, + "step": 12319 + }, + { + "epoch": 4.6455505279034695, + "grad_norm": 0.06500608061195826, + "learning_rate": 1.2184176011184756e-06, + "loss": 0.3697, + "step": 12320 + }, + { + "epoch": 4.645927601809955, + "grad_norm": 0.06947986568728255, + "learning_rate": 1.215839808748509e-06, + "loss": 0.3703, + "step": 12321 + }, + { + "epoch": 4.646304675716441, + "grad_norm": 0.06438730601516693, + "learning_rate": 1.2132647040864386e-06, + "loss": 0.3768, + "step": 12322 + }, + { + "epoch": 4.6466817496229265, + "grad_norm": 0.06371058220675228, + "learning_rate": 1.210692287310722e-06, + "loss": 0.3865, + "step": 12323 + }, + { + "epoch": 4.647058823529412, + "grad_norm": 0.06454587659847091, + "learning_rate": 1.2081225585996248e-06, + "loss": 0.3783, + "step": 12324 + }, + { + "epoch": 4.647435897435898, + "grad_norm": 0.06621446182580971, + "learning_rate": 1.205555518131223e-06, + "loss": 0.3783, + "step": 12325 + }, + { + "epoch": 4.6478129713423835, + "grad_norm": 0.06383864554565608, + "learning_rate": 1.2029911660834138e-06, + "loss": 0.3743, + "step": 12326 + }, + { + "epoch": 4.648190045248869, + "grad_norm": 0.06968487130145797, + "learning_rate": 1.200429502633904e-06, + "loss": 0.3751, + "step": 12327 + }, + { + "epoch": 4.648567119155355, + "grad_norm": 0.06495791573769902, + "learning_rate": 1.1978705279602143e-06, + "loss": 0.3681, + "step": 12328 + }, + { + "epoch": 4.6489441930618405, + "grad_norm": 0.06571506597956144, + "learning_rate": 1.1953142422396824e-06, + "loss": 0.3743, + "step": 12329 + }, + { + "epoch": 4.649321266968326, + "grad_norm": 0.06305167123023092, + "learning_rate": 1.1927606456494467e-06, + "loss": 0.3833, + "step": 12330 + }, + { + "epoch": 4.649698340874812, + "grad_norm": 0.06414986511175197, + "learning_rate": 1.1902097383664858e-06, + "loss": 0.3741, + "step": 12331 + }, + { + "epoch": 4.6500754147812975, + "grad_norm": 0.06610967246166782, + "learning_rate": 1.1876615205675602e-06, + "loss": 0.3899, + "step": 12332 + }, + { + "epoch": 4.650452488687783, + "grad_norm": 0.06308729668812625, + "learning_rate": 1.1851159924292666e-06, + "loss": 0.3642, + "step": 12333 + }, + { + "epoch": 4.650829562594269, + "grad_norm": 0.06696774749373373, + "learning_rate": 1.1825731541280016e-06, + "loss": 0.3746, + "step": 12334 + }, + { + "epoch": 4.6512066365007545, + "grad_norm": 0.07168023246128956, + "learning_rate": 1.180033005839989e-06, + "loss": 0.3874, + "step": 12335 + }, + { + "epoch": 4.65158371040724, + "grad_norm": 0.06563684945489923, + "learning_rate": 1.1774955477412563e-06, + "loss": 0.3667, + "step": 12336 + }, + { + "epoch": 4.651960784313726, + "grad_norm": 0.06786118761736606, + "learning_rate": 1.1749607800076502e-06, + "loss": 0.3733, + "step": 12337 + }, + { + "epoch": 4.6523378582202115, + "grad_norm": 0.06481810140775765, + "learning_rate": 1.1724287028148206e-06, + "loss": 0.3824, + "step": 12338 + }, + { + "epoch": 4.652714932126697, + "grad_norm": 0.06864552592443805, + "learning_rate": 1.169899316338241e-06, + "loss": 0.3724, + "step": 12339 + }, + { + "epoch": 4.653092006033183, + "grad_norm": 0.06328338303128002, + "learning_rate": 1.1673726207531976e-06, + "loss": 0.3781, + "step": 12340 + }, + { + "epoch": 4.6534690799396685, + "grad_norm": 0.0650729546411624, + "learning_rate": 1.1648486162347861e-06, + "loss": 0.3838, + "step": 12341 + }, + { + "epoch": 4.653846153846154, + "grad_norm": 0.06663768375613191, + "learning_rate": 1.1623273029579195e-06, + "loss": 0.3752, + "step": 12342 + }, + { + "epoch": 4.65422322775264, + "grad_norm": 0.06444679950368999, + "learning_rate": 1.1598086810973253e-06, + "loss": 0.382, + "step": 12343 + }, + { + "epoch": 4.6546003016591255, + "grad_norm": 0.06592393403199522, + "learning_rate": 1.1572927508275345e-06, + "loss": 0.3714, + "step": 12344 + }, + { + "epoch": 4.654977375565611, + "grad_norm": 0.06568068287601256, + "learning_rate": 1.1547795123229012e-06, + "loss": 0.3746, + "step": 12345 + }, + { + "epoch": 4.655354449472097, + "grad_norm": 0.06630737877289976, + "learning_rate": 1.1522689657575926e-06, + "loss": 0.3744, + "step": 12346 + }, + { + "epoch": 4.6557315233785825, + "grad_norm": 0.06518141211510786, + "learning_rate": 1.1497611113055896e-06, + "loss": 0.3817, + "step": 12347 + }, + { + "epoch": 4.656108597285068, + "grad_norm": 0.06262122922153755, + "learning_rate": 1.1472559491406775e-06, + "loss": 0.3756, + "step": 12348 + }, + { + "epoch": 4.656485671191554, + "grad_norm": 0.06944246422578208, + "learning_rate": 1.1447534794364735e-06, + "loss": 0.383, + "step": 12349 + }, + { + "epoch": 4.6568627450980395, + "grad_norm": 0.06736142998036848, + "learning_rate": 1.1422537023663805e-06, + "loss": 0.3655, + "step": 12350 + }, + { + "epoch": 4.657239819004525, + "grad_norm": 0.06457187432493396, + "learning_rate": 1.1397566181036424e-06, + "loss": 0.3701, + "step": 12351 + }, + { + "epoch": 4.657616892911011, + "grad_norm": 0.06189840168493678, + "learning_rate": 1.1372622268212985e-06, + "loss": 0.3712, + "step": 12352 + }, + { + "epoch": 4.657993966817497, + "grad_norm": 0.06419500429082357, + "learning_rate": 1.134770528692215e-06, + "loss": 0.3821, + "step": 12353 + }, + { + "epoch": 4.658371040723982, + "grad_norm": 0.06429267609842729, + "learning_rate": 1.1322815238890584e-06, + "loss": 0.3715, + "step": 12354 + }, + { + "epoch": 4.658748114630468, + "grad_norm": 0.06403397810733369, + "learning_rate": 1.129795212584317e-06, + "loss": 0.377, + "step": 12355 + }, + { + "epoch": 4.659125188536954, + "grad_norm": 0.06322518383823353, + "learning_rate": 1.1273115949502934e-06, + "loss": 0.3764, + "step": 12356 + }, + { + "epoch": 4.659502262443439, + "grad_norm": 0.06361444962794566, + "learning_rate": 1.124830671159094e-06, + "loss": 0.3742, + "step": 12357 + }, + { + "epoch": 4.659879336349925, + "grad_norm": 0.0679104114306562, + "learning_rate": 1.1223524413826437e-06, + "loss": 0.3775, + "step": 12358 + }, + { + "epoch": 4.660256410256411, + "grad_norm": 0.06841426599723534, + "learning_rate": 1.119876905792685e-06, + "loss": 0.3709, + "step": 12359 + }, + { + "epoch": 4.660633484162896, + "grad_norm": 0.06207649795439087, + "learning_rate": 1.1174040645607742e-06, + "loss": 0.3721, + "step": 12360 + }, + { + "epoch": 4.661010558069382, + "grad_norm": 0.06291435350455388, + "learning_rate": 1.1149339178582674e-06, + "loss": 0.3795, + "step": 12361 + }, + { + "epoch": 4.661387631975868, + "grad_norm": 0.06603125773424096, + "learning_rate": 1.1124664658563566e-06, + "loss": 0.3907, + "step": 12362 + }, + { + "epoch": 4.661764705882353, + "grad_norm": 0.06802470876995911, + "learning_rate": 1.1100017087260205e-06, + "loss": 0.3715, + "step": 12363 + }, + { + "epoch": 4.662141779788839, + "grad_norm": 0.06875852574472063, + "learning_rate": 1.107539646638074e-06, + "loss": 0.3901, + "step": 12364 + }, + { + "epoch": 4.662518853695325, + "grad_norm": 0.06728557397243293, + "learning_rate": 1.105080279763131e-06, + "loss": 0.3779, + "step": 12365 + }, + { + "epoch": 4.66289592760181, + "grad_norm": 0.06462354585955075, + "learning_rate": 1.1026236082716245e-06, + "loss": 0.382, + "step": 12366 + }, + { + "epoch": 4.663273001508296, + "grad_norm": 0.06224112131548713, + "learning_rate": 1.1001696323337962e-06, + "loss": 0.393, + "step": 12367 + }, + { + "epoch": 4.663650075414782, + "grad_norm": 0.0628921185649842, + "learning_rate": 1.0977183521197188e-06, + "loss": 0.3813, + "step": 12368 + }, + { + "epoch": 4.664027149321267, + "grad_norm": 0.06662710761735813, + "learning_rate": 1.0952697677992385e-06, + "loss": 0.3754, + "step": 12369 + }, + { + "epoch": 4.664404223227753, + "grad_norm": 0.0701759329922936, + "learning_rate": 1.0928238795420687e-06, + "loss": 0.4005, + "step": 12370 + }, + { + "epoch": 4.664781297134239, + "grad_norm": 0.0650826811617318, + "learning_rate": 1.090380687517687e-06, + "loss": 0.3788, + "step": 12371 + }, + { + "epoch": 4.665158371040724, + "grad_norm": 0.06358925927494428, + "learning_rate": 1.0879401918954114e-06, + "loss": 0.3767, + "step": 12372 + }, + { + "epoch": 4.66553544494721, + "grad_norm": 0.06252929110708538, + "learning_rate": 1.0855023928443642e-06, + "loss": 0.3835, + "step": 12373 + }, + { + "epoch": 4.665912518853696, + "grad_norm": 0.06403678289759099, + "learning_rate": 1.0830672905334906e-06, + "loss": 0.3797, + "step": 12374 + }, + { + "epoch": 4.666289592760181, + "grad_norm": 0.06727116088769582, + "learning_rate": 1.0806348851315263e-06, + "loss": 0.3789, + "step": 12375 + }, + { + "epoch": 4.666666666666667, + "grad_norm": 0.06270487157281883, + "learning_rate": 1.0782051768070477e-06, + "loss": 0.3742, + "step": 12376 + }, + { + "epoch": 4.667043740573153, + "grad_norm": 0.06434150981888127, + "learning_rate": 1.0757781657284227e-06, + "loss": 0.3719, + "step": 12377 + }, + { + "epoch": 4.667420814479638, + "grad_norm": 0.06639867488724771, + "learning_rate": 1.0733538520638498e-06, + "loss": 0.381, + "step": 12378 + }, + { + "epoch": 4.667797888386124, + "grad_norm": 0.06472185997535791, + "learning_rate": 1.0709322359813234e-06, + "loss": 0.377, + "step": 12379 + }, + { + "epoch": 4.66817496229261, + "grad_norm": 0.06287712549761261, + "learning_rate": 1.0685133176486694e-06, + "loss": 0.3724, + "step": 12380 + }, + { + "epoch": 4.668552036199095, + "grad_norm": 0.0639419583497015, + "learning_rate": 1.0660970972335005e-06, + "loss": 0.3804, + "step": 12381 + }, + { + "epoch": 4.668929110105581, + "grad_norm": 0.06965945725492473, + "learning_rate": 1.063683574903278e-06, + "loss": 0.3726, + "step": 12382 + }, + { + "epoch": 4.669306184012067, + "grad_norm": 0.06383385947840538, + "learning_rate": 1.061272750825242e-06, + "loss": 0.3748, + "step": 12383 + }, + { + "epoch": 4.669683257918552, + "grad_norm": 0.0662916899237351, + "learning_rate": 1.0588646251664626e-06, + "loss": 0.3844, + "step": 12384 + }, + { + "epoch": 4.670060331825038, + "grad_norm": 0.0681060012251757, + "learning_rate": 1.0564591980938288e-06, + "loss": 0.3847, + "step": 12385 + }, + { + "epoch": 4.670437405731524, + "grad_norm": 0.06355286438731497, + "learning_rate": 1.0540564697740341e-06, + "loss": 0.3682, + "step": 12386 + }, + { + "epoch": 4.670814479638009, + "grad_norm": 0.06641533551765215, + "learning_rate": 1.0516564403735673e-06, + "loss": 0.383, + "step": 12387 + }, + { + "epoch": 4.671191553544495, + "grad_norm": 0.06645331144215184, + "learning_rate": 1.0492591100587756e-06, + "loss": 0.381, + "step": 12388 + }, + { + "epoch": 4.671568627450981, + "grad_norm": 0.0674952744458814, + "learning_rate": 1.0468644789957704e-06, + "loss": 0.3783, + "step": 12389 + }, + { + "epoch": 4.671945701357466, + "grad_norm": 0.06572154239043539, + "learning_rate": 1.044472547350508e-06, + "loss": 0.3869, + "step": 12390 + }, + { + "epoch": 4.672322775263952, + "grad_norm": 0.06327947598111065, + "learning_rate": 1.0420833152887443e-06, + "loss": 0.3772, + "step": 12391 + }, + { + "epoch": 4.672699849170438, + "grad_norm": 0.06953648519824192, + "learning_rate": 1.039696782976054e-06, + "loss": 0.3793, + "step": 12392 + }, + { + "epoch": 4.673076923076923, + "grad_norm": 0.06337616201871421, + "learning_rate": 1.0373129505778156e-06, + "loss": 0.3706, + "step": 12393 + }, + { + "epoch": 4.673453996983409, + "grad_norm": 0.06251175426185562, + "learning_rate": 1.0349318182592305e-06, + "loss": 0.3867, + "step": 12394 + }, + { + "epoch": 4.673831070889895, + "grad_norm": 0.07255638432810894, + "learning_rate": 1.0325533861853088e-06, + "loss": 0.3759, + "step": 12395 + }, + { + "epoch": 4.67420814479638, + "grad_norm": 0.06787183025184707, + "learning_rate": 1.0301776545208742e-06, + "loss": 0.3721, + "step": 12396 + }, + { + "epoch": 4.674585218702866, + "grad_norm": 0.06328985658901293, + "learning_rate": 1.0278046234305594e-06, + "loss": 0.3778, + "step": 12397 + }, + { + "epoch": 4.674962292609352, + "grad_norm": 0.06789305803024656, + "learning_rate": 1.0254342930788198e-06, + "loss": 0.3904, + "step": 12398 + }, + { + "epoch": 4.675339366515837, + "grad_norm": 0.07088316382944852, + "learning_rate": 1.0230666636299059e-06, + "loss": 0.3832, + "step": 12399 + }, + { + "epoch": 4.675716440422323, + "grad_norm": 0.06664426854450539, + "learning_rate": 1.0207017352479086e-06, + "loss": 0.3764, + "step": 12400 + }, + { + "epoch": 4.676093514328809, + "grad_norm": 0.06104838006124205, + "learning_rate": 1.0183395080967017e-06, + "loss": 0.369, + "step": 12401 + }, + { + "epoch": 4.676470588235294, + "grad_norm": 0.06895018095202979, + "learning_rate": 1.015979982339994e-06, + "loss": 0.374, + "step": 12402 + }, + { + "epoch": 4.67684766214178, + "grad_norm": 0.0669361089365177, + "learning_rate": 1.0136231581412902e-06, + "loss": 0.3672, + "step": 12403 + }, + { + "epoch": 4.677224736048266, + "grad_norm": 0.06839196986361323, + "learning_rate": 1.0112690356639266e-06, + "loss": 0.3823, + "step": 12404 + }, + { + "epoch": 4.677601809954751, + "grad_norm": 0.06434049166334084, + "learning_rate": 1.0089176150710257e-06, + "loss": 0.381, + "step": 12405 + }, + { + "epoch": 4.677978883861237, + "grad_norm": 0.06409510941186407, + "learning_rate": 1.0065688965255592e-06, + "loss": 0.359, + "step": 12406 + }, + { + "epoch": 4.678355957767723, + "grad_norm": 0.0688334780467525, + "learning_rate": 1.004222880190273e-06, + "loss": 0.3811, + "step": 12407 + }, + { + "epoch": 4.678733031674208, + "grad_norm": 0.06668638642990052, + "learning_rate": 1.0018795662277569e-06, + "loss": 0.3907, + "step": 12408 + }, + { + "epoch": 4.679110105580694, + "grad_norm": 0.06475282056027909, + "learning_rate": 9.99538954800392e-07, + "loss": 0.368, + "step": 12409 + }, + { + "epoch": 4.67948717948718, + "grad_norm": 0.06472603960609893, + "learning_rate": 9.972010460703907e-07, + "loss": 0.3701, + "step": 12410 + }, + { + "epoch": 4.679864253393665, + "grad_norm": 0.06336895192959571, + "learning_rate": 9.948658401997524e-07, + "loss": 0.3776, + "step": 12411 + }, + { + "epoch": 4.680241327300151, + "grad_norm": 0.06510710187104456, + "learning_rate": 9.92533337350321e-07, + "loss": 0.376, + "step": 12412 + }, + { + "epoch": 4.680618401206637, + "grad_norm": 0.06223873478678553, + "learning_rate": 9.902035376837272e-07, + "loss": 0.3769, + "step": 12413 + }, + { + "epoch": 4.680995475113122, + "grad_norm": 0.07091632534942344, + "learning_rate": 9.878764413614284e-07, + "loss": 0.386, + "step": 12414 + }, + { + "epoch": 4.681372549019608, + "grad_norm": 0.06473537347088082, + "learning_rate": 9.855520485446867e-07, + "loss": 0.3782, + "step": 12415 + }, + { + "epoch": 4.681749622926094, + "grad_norm": 0.0662089519633685, + "learning_rate": 9.83230359394587e-07, + "loss": 0.3823, + "step": 12416 + }, + { + "epoch": 4.682126696832579, + "grad_norm": 0.06438810346535742, + "learning_rate": 9.809113740720043e-07, + "loss": 0.3826, + "step": 12417 + }, + { + "epoch": 4.682503770739065, + "grad_norm": 0.06293131472806826, + "learning_rate": 9.78595092737664e-07, + "loss": 0.3858, + "step": 12418 + }, + { + "epoch": 4.682880844645551, + "grad_norm": 0.06597420271856942, + "learning_rate": 9.762815155520688e-07, + "loss": 0.3718, + "step": 12419 + }, + { + "epoch": 4.683257918552036, + "grad_norm": 0.06503293245179301, + "learning_rate": 9.739706426755479e-07, + "loss": 0.3837, + "step": 12420 + }, + { + "epoch": 4.683634992458522, + "grad_norm": 0.0647336088538657, + "learning_rate": 9.71662474268249e-07, + "loss": 0.3664, + "step": 12421 + }, + { + "epoch": 4.684012066365008, + "grad_norm": 0.06527086669576586, + "learning_rate": 9.69357010490124e-07, + "loss": 0.3667, + "step": 12422 + }, + { + "epoch": 4.6843891402714934, + "grad_norm": 0.06333268378286813, + "learning_rate": 9.670542515009296e-07, + "loss": 0.3791, + "step": 12423 + }, + { + "epoch": 4.684766214177979, + "grad_norm": 0.06751979732444893, + "learning_rate": 9.647541974602625e-07, + "loss": 0.3846, + "step": 12424 + }, + { + "epoch": 4.685143288084465, + "grad_norm": 0.06327462690792793, + "learning_rate": 9.624568485275021e-07, + "loss": 0.3753, + "step": 12425 + }, + { + "epoch": 4.6855203619909505, + "grad_norm": 0.06470877230268787, + "learning_rate": 9.601622048618542e-07, + "loss": 0.3677, + "step": 12426 + }, + { + "epoch": 4.685897435897436, + "grad_norm": 0.062386956204359995, + "learning_rate": 9.578702666223383e-07, + "loss": 0.3789, + "step": 12427 + }, + { + "epoch": 4.686274509803922, + "grad_norm": 0.0621545748365082, + "learning_rate": 9.55581033967783e-07, + "loss": 0.3852, + "step": 12428 + }, + { + "epoch": 4.6866515837104075, + "grad_norm": 0.06532645997754873, + "learning_rate": 9.532945070568256e-07, + "loss": 0.396, + "step": 12429 + }, + { + "epoch": 4.687028657616893, + "grad_norm": 0.06447697895675504, + "learning_rate": 9.510106860479307e-07, + "loss": 0.3785, + "step": 12430 + }, + { + "epoch": 4.687405731523379, + "grad_norm": 0.0672321791639533, + "learning_rate": 9.487295710993583e-07, + "loss": 0.3817, + "step": 12431 + }, + { + "epoch": 4.6877828054298645, + "grad_norm": 0.06440044885759197, + "learning_rate": 9.464511623691819e-07, + "loss": 0.3758, + "step": 12432 + }, + { + "epoch": 4.68815987933635, + "grad_norm": 0.06490062021316144, + "learning_rate": 9.441754600153063e-07, + "loss": 0.3771, + "step": 12433 + }, + { + "epoch": 4.688536953242836, + "grad_norm": 0.06304219798813748, + "learning_rate": 9.419024641954277e-07, + "loss": 0.3768, + "step": 12434 + }, + { + "epoch": 4.6889140271493215, + "grad_norm": 0.06473249883893459, + "learning_rate": 9.3963217506706e-07, + "loss": 0.3775, + "step": 12435 + }, + { + "epoch": 4.689291101055807, + "grad_norm": 0.0646582909496829, + "learning_rate": 9.373645927875397e-07, + "loss": 0.3696, + "step": 12436 + }, + { + "epoch": 4.689668174962293, + "grad_norm": 0.06302818247321668, + "learning_rate": 9.350997175140031e-07, + "loss": 0.375, + "step": 12437 + }, + { + "epoch": 4.6900452488687785, + "grad_norm": 0.06463326804787609, + "learning_rate": 9.328375494034048e-07, + "loss": 0.3754, + "step": 12438 + }, + { + "epoch": 4.690422322775264, + "grad_norm": 0.06365265267045275, + "learning_rate": 9.305780886125127e-07, + "loss": 0.378, + "step": 12439 + }, + { + "epoch": 4.69079939668175, + "grad_norm": 0.06761782690935082, + "learning_rate": 9.283213352979081e-07, + "loss": 0.3805, + "step": 12440 + }, + { + "epoch": 4.6911764705882355, + "grad_norm": 0.06821170151523634, + "learning_rate": 9.260672896159728e-07, + "loss": 0.3845, + "step": 12441 + }, + { + "epoch": 4.691553544494721, + "grad_norm": 0.06492588841869028, + "learning_rate": 9.238159517229195e-07, + "loss": 0.39, + "step": 12442 + }, + { + "epoch": 4.691930618401207, + "grad_norm": 0.0632815097775746, + "learning_rate": 9.215673217747567e-07, + "loss": 0.3701, + "step": 12443 + }, + { + "epoch": 4.6923076923076925, + "grad_norm": 0.06494645494309145, + "learning_rate": 9.193213999273199e-07, + "loss": 0.3984, + "step": 12444 + }, + { + "epoch": 4.692684766214178, + "grad_norm": 0.06302275195506704, + "learning_rate": 9.170781863362444e-07, + "loss": 0.3732, + "step": 12445 + }, + { + "epoch": 4.693061840120664, + "grad_norm": 0.061198988691661356, + "learning_rate": 9.148376811569837e-07, + "loss": 0.3768, + "step": 12446 + }, + { + "epoch": 4.6934389140271495, + "grad_norm": 0.06432392917602148, + "learning_rate": 9.125998845448003e-07, + "loss": 0.3708, + "step": 12447 + }, + { + "epoch": 4.693815987933635, + "grad_norm": 0.06656911012065865, + "learning_rate": 9.10364796654779e-07, + "loss": 0.3821, + "step": 12448 + }, + { + "epoch": 4.694193061840121, + "grad_norm": 0.06344327709985388, + "learning_rate": 9.081324176418005e-07, + "loss": 0.3829, + "step": 12449 + }, + { + "epoch": 4.6945701357466065, + "grad_norm": 0.06070576309995074, + "learning_rate": 9.059027476605764e-07, + "loss": 0.3776, + "step": 12450 + }, + { + "epoch": 4.694947209653092, + "grad_norm": 0.06389740575600122, + "learning_rate": 9.036757868656098e-07, + "loss": 0.3743, + "step": 12451 + }, + { + "epoch": 4.695324283559578, + "grad_norm": 0.06344251592574499, + "learning_rate": 9.014515354112352e-07, + "loss": 0.3809, + "step": 12452 + }, + { + "epoch": 4.6957013574660635, + "grad_norm": 0.06299641140741043, + "learning_rate": 8.992299934515869e-07, + "loss": 0.3872, + "step": 12453 + }, + { + "epoch": 4.696078431372549, + "grad_norm": 0.06466889533534415, + "learning_rate": 8.970111611406263e-07, + "loss": 0.3857, + "step": 12454 + }, + { + "epoch": 4.696455505279035, + "grad_norm": 0.06332189239922666, + "learning_rate": 8.94795038632097e-07, + "loss": 0.3685, + "step": 12455 + }, + { + "epoch": 4.6968325791855206, + "grad_norm": 0.06378347532019557, + "learning_rate": 8.925816260795916e-07, + "loss": 0.3667, + "step": 12456 + }, + { + "epoch": 4.697209653092006, + "grad_norm": 0.06538460865094035, + "learning_rate": 8.9037092363649e-07, + "loss": 0.3921, + "step": 12457 + }, + { + "epoch": 4.697586726998492, + "grad_norm": 0.06397881485682813, + "learning_rate": 8.881629314559892e-07, + "loss": 0.3732, + "step": 12458 + }, + { + "epoch": 4.697963800904978, + "grad_norm": 0.06420807750343341, + "learning_rate": 8.859576496911093e-07, + "loss": 0.3789, + "step": 12459 + }, + { + "epoch": 4.698340874811463, + "grad_norm": 0.0671337099886924, + "learning_rate": 8.837550784946702e-07, + "loss": 0.3768, + "step": 12460 + }, + { + "epoch": 4.698717948717949, + "grad_norm": 0.06822206331718245, + "learning_rate": 8.815552180193054e-07, + "loss": 0.373, + "step": 12461 + }, + { + "epoch": 4.699095022624435, + "grad_norm": 0.06604868523199671, + "learning_rate": 8.793580684174707e-07, + "loss": 0.3706, + "step": 12462 + }, + { + "epoch": 4.69947209653092, + "grad_norm": 0.06272773050969654, + "learning_rate": 8.771636298414177e-07, + "loss": 0.3808, + "step": 12463 + }, + { + "epoch": 4.699849170437406, + "grad_norm": 0.06194251058515201, + "learning_rate": 8.749719024432246e-07, + "loss": 0.383, + "step": 12464 + }, + { + "epoch": 4.700226244343892, + "grad_norm": 0.06250824389719602, + "learning_rate": 8.727828863747745e-07, + "loss": 0.3817, + "step": 12465 + }, + { + "epoch": 4.700603318250377, + "grad_norm": 0.0692107732457983, + "learning_rate": 8.705965817877681e-07, + "loss": 0.3728, + "step": 12466 + }, + { + "epoch": 4.700980392156863, + "grad_norm": 0.07008717122112058, + "learning_rate": 8.684129888337023e-07, + "loss": 0.3852, + "step": 12467 + }, + { + "epoch": 4.701357466063349, + "grad_norm": 0.06415417791987924, + "learning_rate": 8.662321076639135e-07, + "loss": 0.3679, + "step": 12468 + }, + { + "epoch": 4.701734539969834, + "grad_norm": 0.06643934631135705, + "learning_rate": 8.640539384295254e-07, + "loss": 0.3729, + "step": 12469 + }, + { + "epoch": 4.70211161387632, + "grad_norm": 0.06415933864082951, + "learning_rate": 8.618784812814885e-07, + "loss": 0.3727, + "step": 12470 + }, + { + "epoch": 4.702488687782806, + "grad_norm": 0.06533611228526355, + "learning_rate": 8.59705736370553e-07, + "loss": 0.3712, + "step": 12471 + }, + { + "epoch": 4.702865761689291, + "grad_norm": 0.06706990880504309, + "learning_rate": 8.575357038473009e-07, + "loss": 0.3894, + "step": 12472 + }, + { + "epoch": 4.703242835595777, + "grad_norm": 0.06666209107373008, + "learning_rate": 8.553683838620919e-07, + "loss": 0.3774, + "step": 12473 + }, + { + "epoch": 4.703619909502263, + "grad_norm": 0.06776444584138357, + "learning_rate": 8.532037765651435e-07, + "loss": 0.3835, + "step": 12474 + }, + { + "epoch": 4.703996983408748, + "grad_norm": 0.06485775923021068, + "learning_rate": 8.51041882106447e-07, + "loss": 0.3825, + "step": 12475 + }, + { + "epoch": 4.704374057315234, + "grad_norm": 0.06608677783067013, + "learning_rate": 8.488827006358247e-07, + "loss": 0.3853, + "step": 12476 + }, + { + "epoch": 4.70475113122172, + "grad_norm": 0.06553691104081164, + "learning_rate": 8.46726232302899e-07, + "loss": 0.3654, + "step": 12477 + }, + { + "epoch": 4.705128205128205, + "grad_norm": 0.06840970090257477, + "learning_rate": 8.44572477257124e-07, + "loss": 0.3811, + "step": 12478 + }, + { + "epoch": 4.705505279034691, + "grad_norm": 0.07024306662198317, + "learning_rate": 8.424214356477401e-07, + "loss": 0.381, + "step": 12479 + }, + { + "epoch": 4.705882352941177, + "grad_norm": 0.06732696886833292, + "learning_rate": 8.402731076238191e-07, + "loss": 0.3834, + "step": 12480 + }, + { + "epoch": 4.706259426847662, + "grad_norm": 0.06660446619331387, + "learning_rate": 8.381274933342376e-07, + "loss": 0.3751, + "step": 12481 + }, + { + "epoch": 4.706636500754148, + "grad_norm": 0.0651736374022283, + "learning_rate": 8.359845929276856e-07, + "loss": 0.3858, + "step": 12482 + }, + { + "epoch": 4.707013574660634, + "grad_norm": 0.06339750788045923, + "learning_rate": 8.338444065526619e-07, + "loss": 0.39, + "step": 12483 + }, + { + "epoch": 4.707390648567119, + "grad_norm": 0.0691531044030618, + "learning_rate": 8.317069343574791e-07, + "loss": 0.3834, + "step": 12484 + }, + { + "epoch": 4.707767722473605, + "grad_norm": 0.06537117740141067, + "learning_rate": 8.295721764902631e-07, + "loss": 0.3831, + "step": 12485 + }, + { + "epoch": 4.708144796380091, + "grad_norm": 0.06661070922143975, + "learning_rate": 8.274401330989535e-07, + "loss": 0.3825, + "step": 12486 + }, + { + "epoch": 4.708521870286576, + "grad_norm": 0.06497265684542862, + "learning_rate": 8.253108043312985e-07, + "loss": 0.3765, + "step": 12487 + }, + { + "epoch": 4.708898944193062, + "grad_norm": 0.06617293330499892, + "learning_rate": 8.231841903348514e-07, + "loss": 0.3775, + "step": 12488 + }, + { + "epoch": 4.709276018099548, + "grad_norm": 0.06775398465183588, + "learning_rate": 8.21060291256992e-07, + "loss": 0.3779, + "step": 12489 + }, + { + "epoch": 4.709653092006033, + "grad_norm": 0.06708090293715249, + "learning_rate": 8.189391072449048e-07, + "loss": 0.3844, + "step": 12490 + }, + { + "epoch": 4.710030165912519, + "grad_norm": 0.06517117551171826, + "learning_rate": 8.16820638445579e-07, + "loss": 0.3828, + "step": 12491 + }, + { + "epoch": 4.710407239819005, + "grad_norm": 0.06440255985541952, + "learning_rate": 8.147048850058347e-07, + "loss": 0.3733, + "step": 12492 + }, + { + "epoch": 4.71078431372549, + "grad_norm": 0.0698340408186121, + "learning_rate": 8.125918470722794e-07, + "loss": 0.3794, + "step": 12493 + }, + { + "epoch": 4.711161387631976, + "grad_norm": 0.062153049502654725, + "learning_rate": 8.104815247913511e-07, + "loss": 0.3765, + "step": 12494 + }, + { + "epoch": 4.711538461538462, + "grad_norm": 0.0651147247164011, + "learning_rate": 8.083739183092931e-07, + "loss": 0.379, + "step": 12495 + }, + { + "epoch": 4.711915535444947, + "grad_norm": 0.06422836572045824, + "learning_rate": 8.062690277721575e-07, + "loss": 0.3753, + "step": 12496 + }, + { + "epoch": 4.712292609351433, + "grad_norm": 0.06364874223413458, + "learning_rate": 8.04166853325814e-07, + "loss": 0.3846, + "step": 12497 + }, + { + "epoch": 4.712669683257919, + "grad_norm": 0.06342964745275119, + "learning_rate": 8.020673951159419e-07, + "loss": 0.3741, + "step": 12498 + }, + { + "epoch": 4.713046757164404, + "grad_norm": 0.06299054364992439, + "learning_rate": 7.999706532880336e-07, + "loss": 0.3853, + "step": 12499 + }, + { + "epoch": 4.71342383107089, + "grad_norm": 0.06482258822964199, + "learning_rate": 7.978766279873861e-07, + "loss": 0.3856, + "step": 12500 + }, + { + "epoch": 4.713800904977376, + "grad_norm": 0.06710002685572403, + "learning_rate": 7.957853193591147e-07, + "loss": 0.3752, + "step": 12501 + }, + { + "epoch": 4.714177978883861, + "grad_norm": 0.06378332628727673, + "learning_rate": 7.936967275481477e-07, + "loss": 0.3899, + "step": 12502 + }, + { + "epoch": 4.714555052790347, + "grad_norm": 0.06136861640179308, + "learning_rate": 7.916108526992183e-07, + "loss": 0.3802, + "step": 12503 + }, + { + "epoch": 4.714932126696833, + "grad_norm": 0.06593496343985668, + "learning_rate": 7.89527694956882e-07, + "loss": 0.3923, + "step": 12504 + }, + { + "epoch": 4.715309200603318, + "grad_norm": 0.06351233040196413, + "learning_rate": 7.874472544654988e-07, + "loss": 0.3794, + "step": 12505 + }, + { + "epoch": 4.715686274509804, + "grad_norm": 0.06696557282196346, + "learning_rate": 7.853695313692378e-07, + "loss": 0.3746, + "step": 12506 + }, + { + "epoch": 4.71606334841629, + "grad_norm": 0.07032226740763331, + "learning_rate": 7.832945258120861e-07, + "loss": 0.3803, + "step": 12507 + }, + { + "epoch": 4.716440422322775, + "grad_norm": 0.06683197777437706, + "learning_rate": 7.812222379378353e-07, + "loss": 0.3751, + "step": 12508 + }, + { + "epoch": 4.716817496229261, + "grad_norm": 0.06533410263844185, + "learning_rate": 7.791526678900996e-07, + "loss": 0.365, + "step": 12509 + }, + { + "epoch": 4.717194570135747, + "grad_norm": 0.06876426779831665, + "learning_rate": 7.770858158122929e-07, + "loss": 0.3724, + "step": 12510 + }, + { + "epoch": 4.717571644042232, + "grad_norm": 0.06450502395789694, + "learning_rate": 7.750216818476519e-07, + "loss": 0.3732, + "step": 12511 + }, + { + "epoch": 4.717948717948718, + "grad_norm": 0.06890670259832404, + "learning_rate": 7.729602661392133e-07, + "loss": 0.3776, + "step": 12512 + }, + { + "epoch": 4.718325791855204, + "grad_norm": 0.06852358353206756, + "learning_rate": 7.709015688298316e-07, + "loss": 0.3823, + "step": 12513 + }, + { + "epoch": 4.718702865761689, + "grad_norm": 0.06757450301015237, + "learning_rate": 7.688455900621749e-07, + "loss": 0.3796, + "step": 12514 + }, + { + "epoch": 4.719079939668175, + "grad_norm": 0.06354057414139237, + "learning_rate": 7.66792329978725e-07, + "loss": 0.3834, + "step": 12515 + }, + { + "epoch": 4.719457013574661, + "grad_norm": 0.0660689984208763, + "learning_rate": 7.647417887217634e-07, + "loss": 0.3745, + "step": 12516 + }, + { + "epoch": 4.719834087481146, + "grad_norm": 0.0696639308729101, + "learning_rate": 7.626939664333988e-07, + "loss": 0.3826, + "step": 12517 + }, + { + "epoch": 4.720211161387632, + "grad_norm": 0.06542819317028871, + "learning_rate": 7.606488632555309e-07, + "loss": 0.3674, + "step": 12518 + }, + { + "epoch": 4.720588235294118, + "grad_norm": 0.06353959800928512, + "learning_rate": 7.586064793298998e-07, + "loss": 0.3722, + "step": 12519 + }, + { + "epoch": 4.720965309200603, + "grad_norm": 0.06514711700370268, + "learning_rate": 7.565668147980276e-07, + "loss": 0.3807, + "step": 12520 + }, + { + "epoch": 4.721342383107089, + "grad_norm": 0.06634759562990215, + "learning_rate": 7.545298698012681e-07, + "loss": 0.3833, + "step": 12521 + }, + { + "epoch": 4.721719457013575, + "grad_norm": 0.06710436254047839, + "learning_rate": 7.524956444807796e-07, + "loss": 0.3831, + "step": 12522 + }, + { + "epoch": 4.72209653092006, + "grad_norm": 0.06576515743094043, + "learning_rate": 7.504641389775292e-07, + "loss": 0.3797, + "step": 12523 + }, + { + "epoch": 4.722473604826546, + "grad_norm": 0.06533138283401867, + "learning_rate": 7.484353534322975e-07, + "loss": 0.3837, + "step": 12524 + }, + { + "epoch": 4.722850678733032, + "grad_norm": 0.060825593271919345, + "learning_rate": 7.464092879856832e-07, + "loss": 0.3738, + "step": 12525 + }, + { + "epoch": 4.723227752639517, + "grad_norm": 0.06212019714435804, + "learning_rate": 7.443859427780853e-07, + "loss": 0.369, + "step": 12526 + }, + { + "epoch": 4.723604826546003, + "grad_norm": 0.06468627377445377, + "learning_rate": 7.423653179497203e-07, + "loss": 0.3858, + "step": 12527 + }, + { + "epoch": 4.723981900452489, + "grad_norm": 0.0640125587799117, + "learning_rate": 7.403474136406186e-07, + "loss": 0.3794, + "step": 12528 + }, + { + "epoch": 4.7243589743589745, + "grad_norm": 0.06671416255539457, + "learning_rate": 7.383322299906192e-07, + "loss": 0.3739, + "step": 12529 + }, + { + "epoch": 4.72473604826546, + "grad_norm": 0.06443838777041966, + "learning_rate": 7.363197671393662e-07, + "loss": 0.3671, + "step": 12530 + }, + { + "epoch": 4.725113122171946, + "grad_norm": 0.06470129366822251, + "learning_rate": 7.343100252263347e-07, + "loss": 0.3753, + "step": 12531 + }, + { + "epoch": 4.7254901960784315, + "grad_norm": 0.06471929417328173, + "learning_rate": 7.323030043907864e-07, + "loss": 0.3783, + "step": 12532 + }, + { + "epoch": 4.725867269984917, + "grad_norm": 0.061953285376805195, + "learning_rate": 7.302987047718057e-07, + "loss": 0.3688, + "step": 12533 + }, + { + "epoch": 4.726244343891403, + "grad_norm": 0.06455313615303589, + "learning_rate": 7.28297126508295e-07, + "loss": 0.3729, + "step": 12534 + }, + { + "epoch": 4.7266214177978885, + "grad_norm": 0.07315111990729496, + "learning_rate": 7.262982697389653e-07, + "loss": 0.3762, + "step": 12535 + }, + { + "epoch": 4.726998491704374, + "grad_norm": 0.061005769771494134, + "learning_rate": 7.243021346023193e-07, + "loss": 0.3779, + "step": 12536 + }, + { + "epoch": 4.72737556561086, + "grad_norm": 0.0643678574122543, + "learning_rate": 7.223087212367086e-07, + "loss": 0.3766, + "step": 12537 + }, + { + "epoch": 4.7277526395173455, + "grad_norm": 0.061790963572880775, + "learning_rate": 7.203180297802581e-07, + "loss": 0.3753, + "step": 12538 + }, + { + "epoch": 4.728129713423831, + "grad_norm": 0.06322366716707498, + "learning_rate": 7.183300603709243e-07, + "loss": 0.3744, + "step": 12539 + }, + { + "epoch": 4.728506787330317, + "grad_norm": 0.06668810159537084, + "learning_rate": 7.163448131464812e-07, + "loss": 0.3875, + "step": 12540 + }, + { + "epoch": 4.7288838612368025, + "grad_norm": 0.06351806707120579, + "learning_rate": 7.143622882444945e-07, + "loss": 0.3839, + "step": 12541 + }, + { + "epoch": 4.729260935143288, + "grad_norm": 0.06333907685375527, + "learning_rate": 7.12382485802352e-07, + "loss": 0.3821, + "step": 12542 + }, + { + "epoch": 4.729638009049774, + "grad_norm": 0.0623741989033195, + "learning_rate": 7.104054059572596e-07, + "loss": 0.3836, + "step": 12543 + }, + { + "epoch": 4.7300150829562595, + "grad_norm": 0.06115765775480125, + "learning_rate": 7.084310488462231e-07, + "loss": 0.3768, + "step": 12544 + }, + { + "epoch": 4.730392156862745, + "grad_norm": 0.062354265292741594, + "learning_rate": 7.064594146060621e-07, + "loss": 0.366, + "step": 12545 + }, + { + "epoch": 4.730769230769231, + "grad_norm": 0.06511720440169858, + "learning_rate": 7.044905033734096e-07, + "loss": 0.3992, + "step": 12546 + }, + { + "epoch": 4.7311463046757165, + "grad_norm": 0.06388545808611854, + "learning_rate": 7.025243152847117e-07, + "loss": 0.3843, + "step": 12547 + }, + { + "epoch": 4.731523378582202, + "grad_norm": 0.06337388321177834, + "learning_rate": 7.005608504762196e-07, + "loss": 0.3809, + "step": 12548 + }, + { + "epoch": 4.731900452488688, + "grad_norm": 0.06389027000581687, + "learning_rate": 6.986001090840022e-07, + "loss": 0.3753, + "step": 12549 + }, + { + "epoch": 4.7322775263951735, + "grad_norm": 0.0623939922915225, + "learning_rate": 6.966420912439376e-07, + "loss": 0.3762, + "step": 12550 + }, + { + "epoch": 4.732654600301659, + "grad_norm": 0.06318813087005516, + "learning_rate": 6.94686797091717e-07, + "loss": 0.3752, + "step": 12551 + }, + { + "epoch": 4.733031674208145, + "grad_norm": 0.06790693980239063, + "learning_rate": 6.927342267628323e-07, + "loss": 0.3742, + "step": 12552 + }, + { + "epoch": 4.7334087481146305, + "grad_norm": 0.06577041501261885, + "learning_rate": 6.907843803926062e-07, + "loss": 0.3794, + "step": 12553 + }, + { + "epoch": 4.733785822021116, + "grad_norm": 0.06217315364615101, + "learning_rate": 6.888372581161485e-07, + "loss": 0.3789, + "step": 12554 + }, + { + "epoch": 4.734162895927602, + "grad_norm": 0.06893425248630457, + "learning_rate": 6.868928600684043e-07, + "loss": 0.3756, + "step": 12555 + }, + { + "epoch": 4.7345399698340875, + "grad_norm": 0.06226173283318329, + "learning_rate": 6.849511863841152e-07, + "loss": 0.393, + "step": 12556 + }, + { + "epoch": 4.734917043740573, + "grad_norm": 0.06796292066534236, + "learning_rate": 6.83012237197831e-07, + "loss": 0.3591, + "step": 12557 + }, + { + "epoch": 4.735294117647059, + "grad_norm": 0.0626302214217893, + "learning_rate": 6.810760126439287e-07, + "loss": 0.3734, + "step": 12558 + }, + { + "epoch": 4.7356711915535445, + "grad_norm": 0.06452876544922377, + "learning_rate": 6.791425128565854e-07, + "loss": 0.3745, + "step": 12559 + }, + { + "epoch": 4.73604826546003, + "grad_norm": 0.06275986975652718, + "learning_rate": 6.772117379697829e-07, + "loss": 0.3799, + "step": 12560 + }, + { + "epoch": 4.736425339366516, + "grad_norm": 0.06329588867952816, + "learning_rate": 6.752836881173297e-07, + "loss": 0.3752, + "step": 12561 + }, + { + "epoch": 4.7368024132730016, + "grad_norm": 0.06113038280687033, + "learning_rate": 6.733583634328388e-07, + "loss": 0.3757, + "step": 12562 + }, + { + "epoch": 4.737179487179487, + "grad_norm": 0.06526152209068589, + "learning_rate": 6.714357640497282e-07, + "loss": 0.3894, + "step": 12563 + }, + { + "epoch": 4.737556561085973, + "grad_norm": 0.06369144118605043, + "learning_rate": 6.695158901012333e-07, + "loss": 0.3732, + "step": 12564 + }, + { + "epoch": 4.737933634992459, + "grad_norm": 0.06246748707871949, + "learning_rate": 6.675987417204033e-07, + "loss": 0.3846, + "step": 12565 + }, + { + "epoch": 4.738310708898944, + "grad_norm": 0.06153290424085981, + "learning_rate": 6.656843190400874e-07, + "loss": 0.3702, + "step": 12566 + }, + { + "epoch": 4.73868778280543, + "grad_norm": 0.06637754129885569, + "learning_rate": 6.637726221929664e-07, + "loss": 0.3811, + "step": 12567 + }, + { + "epoch": 4.739064856711916, + "grad_norm": 0.06202291902891447, + "learning_rate": 6.618636513115073e-07, + "loss": 0.3665, + "step": 12568 + }, + { + "epoch": 4.739441930618401, + "grad_norm": 0.06355848084192317, + "learning_rate": 6.599574065280046e-07, + "loss": 0.3742, + "step": 12569 + }, + { + "epoch": 4.739819004524887, + "grad_norm": 0.06493107230386097, + "learning_rate": 6.580538879745613e-07, + "loss": 0.3937, + "step": 12570 + }, + { + "epoch": 4.740196078431373, + "grad_norm": 0.0637497965615915, + "learning_rate": 6.561530957830898e-07, + "loss": 0.373, + "step": 12571 + }, + { + "epoch": 4.740573152337858, + "grad_norm": 0.06262093744501013, + "learning_rate": 6.542550300853023e-07, + "loss": 0.3699, + "step": 12572 + }, + { + "epoch": 4.740950226244344, + "grad_norm": 0.06420608406346227, + "learning_rate": 6.523596910127472e-07, + "loss": 0.3796, + "step": 12573 + }, + { + "epoch": 4.74132730015083, + "grad_norm": 0.06206584470638393, + "learning_rate": 6.504670786967638e-07, + "loss": 0.3793, + "step": 12574 + }, + { + "epoch": 4.741704374057315, + "grad_norm": 0.06171857154174153, + "learning_rate": 6.485771932685048e-07, + "loss": 0.364, + "step": 12575 + }, + { + "epoch": 4.742081447963801, + "grad_norm": 0.06585509684808362, + "learning_rate": 6.466900348589456e-07, + "loss": 0.3832, + "step": 12576 + }, + { + "epoch": 4.742458521870287, + "grad_norm": 0.06275837904216454, + "learning_rate": 6.448056035988615e-07, + "loss": 0.3777, + "step": 12577 + }, + { + "epoch": 4.742835595776772, + "grad_norm": 0.06314212695451563, + "learning_rate": 6.42923899618837e-07, + "loss": 0.3752, + "step": 12578 + }, + { + "epoch": 4.743212669683258, + "grad_norm": 0.06442337278318243, + "learning_rate": 6.410449230492832e-07, + "loss": 0.3789, + "step": 12579 + }, + { + "epoch": 4.743589743589744, + "grad_norm": 0.0631511369774543, + "learning_rate": 6.391686740203939e-07, + "loss": 0.3665, + "step": 12580 + }, + { + "epoch": 4.743966817496229, + "grad_norm": 0.06428407592701373, + "learning_rate": 6.372951526622162e-07, + "loss": 0.3704, + "step": 12581 + }, + { + "epoch": 4.744343891402715, + "grad_norm": 0.06472496168977858, + "learning_rate": 6.354243591045617e-07, + "loss": 0.3841, + "step": 12582 + }, + { + "epoch": 4.744720965309201, + "grad_norm": 0.06410676466988811, + "learning_rate": 6.335562934770822e-07, + "loss": 0.3856, + "step": 12583 + }, + { + "epoch": 4.745098039215686, + "grad_norm": 0.061638302340295686, + "learning_rate": 6.316909559092343e-07, + "loss": 0.3684, + "step": 12584 + }, + { + "epoch": 4.745475113122172, + "grad_norm": 0.06310386869552718, + "learning_rate": 6.298283465302879e-07, + "loss": 0.3858, + "step": 12585 + }, + { + "epoch": 4.745852187028658, + "grad_norm": 0.06550593020094375, + "learning_rate": 6.279684654693086e-07, + "loss": 0.3653, + "step": 12586 + }, + { + "epoch": 4.746229260935143, + "grad_norm": 0.06402632333087621, + "learning_rate": 6.261113128551977e-07, + "loss": 0.3773, + "step": 12587 + }, + { + "epoch": 4.746606334841629, + "grad_norm": 0.06657757377762241, + "learning_rate": 6.242568888166479e-07, + "loss": 0.3809, + "step": 12588 + }, + { + "epoch": 4.746983408748115, + "grad_norm": 0.060645805962333856, + "learning_rate": 6.224051934821651e-07, + "loss": 0.3732, + "step": 12589 + }, + { + "epoch": 4.7473604826546, + "grad_norm": 0.06307220524293139, + "learning_rate": 6.20556226980078e-07, + "loss": 0.3632, + "step": 12590 + }, + { + "epoch": 4.747737556561086, + "grad_norm": 0.061840660868304254, + "learning_rate": 6.187099894385151e-07, + "loss": 0.3689, + "step": 12591 + }, + { + "epoch": 4.748114630467572, + "grad_norm": 0.06089780320002786, + "learning_rate": 6.168664809854141e-07, + "loss": 0.3877, + "step": 12592 + }, + { + "epoch": 4.748491704374057, + "grad_norm": 0.06288459729082777, + "learning_rate": 6.150257017485395e-07, + "loss": 0.3678, + "step": 12593 + }, + { + "epoch": 4.748868778280543, + "grad_norm": 0.06375009769067794, + "learning_rate": 6.13187651855447e-07, + "loss": 0.3751, + "step": 12594 + }, + { + "epoch": 4.749245852187029, + "grad_norm": 0.06349151599813681, + "learning_rate": 6.113523314335102e-07, + "loss": 0.3763, + "step": 12595 + }, + { + "epoch": 4.749622926093514, + "grad_norm": 0.06394368739536584, + "learning_rate": 6.095197406099207e-07, + "loss": 0.3818, + "step": 12596 + }, + { + "epoch": 4.75, + "grad_norm": 0.060413551002760475, + "learning_rate": 6.076898795116792e-07, + "loss": 0.3723, + "step": 12597 + }, + { + "epoch": 4.750377073906486, + "grad_norm": 0.06224766095233858, + "learning_rate": 6.058627482655821e-07, + "loss": 0.3797, + "step": 12598 + }, + { + "epoch": 4.750754147812971, + "grad_norm": 0.061696819789446015, + "learning_rate": 6.04038346998257e-07, + "loss": 0.3645, + "step": 12599 + }, + { + "epoch": 4.751131221719457, + "grad_norm": 0.06229415406904467, + "learning_rate": 6.022166758361314e-07, + "loss": 0.3674, + "step": 12600 + }, + { + "epoch": 4.751508295625943, + "grad_norm": 0.0649796951810784, + "learning_rate": 6.003977349054424e-07, + "loss": 0.3855, + "step": 12601 + }, + { + "epoch": 4.751885369532428, + "grad_norm": 0.06370645197004476, + "learning_rate": 5.985815243322402e-07, + "loss": 0.3764, + "step": 12602 + }, + { + "epoch": 4.752262443438914, + "grad_norm": 0.0640479886747522, + "learning_rate": 5.967680442423973e-07, + "loss": 0.3801, + "step": 12603 + }, + { + "epoch": 4.7526395173454, + "grad_norm": 0.06004920696987049, + "learning_rate": 5.949572947615689e-07, + "loss": 0.3794, + "step": 12604 + }, + { + "epoch": 4.753016591251885, + "grad_norm": 0.0658042629127195, + "learning_rate": 5.931492760152546e-07, + "loss": 0.3828, + "step": 12605 + }, + { + "epoch": 4.753393665158371, + "grad_norm": 0.06423888458371724, + "learning_rate": 5.913439881287408e-07, + "loss": 0.3787, + "step": 12606 + }, + { + "epoch": 4.753770739064857, + "grad_norm": 0.06474207051879241, + "learning_rate": 5.89541431227132e-07, + "loss": 0.3761, + "step": 12607 + }, + { + "epoch": 4.754147812971342, + "grad_norm": 0.0631832033253422, + "learning_rate": 5.877416054353413e-07, + "loss": 0.3734, + "step": 12608 + }, + { + "epoch": 4.754524886877828, + "grad_norm": 0.061063098578425894, + "learning_rate": 5.859445108781047e-07, + "loss": 0.3728, + "step": 12609 + }, + { + "epoch": 4.754901960784314, + "grad_norm": 0.06247816036253549, + "learning_rate": 5.841501476799494e-07, + "loss": 0.3723, + "step": 12610 + }, + { + "epoch": 4.755279034690799, + "grad_norm": 0.06301419773750185, + "learning_rate": 5.82358515965229e-07, + "loss": 0.3816, + "step": 12611 + }, + { + "epoch": 4.755656108597285, + "grad_norm": 0.06318299271690206, + "learning_rate": 5.805696158580975e-07, + "loss": 0.384, + "step": 12612 + }, + { + "epoch": 4.756033182503771, + "grad_norm": 0.06310794986453108, + "learning_rate": 5.787834474825315e-07, + "loss": 0.3857, + "step": 12613 + }, + { + "epoch": 4.756410256410256, + "grad_norm": 0.06230066894225239, + "learning_rate": 5.770000109623031e-07, + "loss": 0.3683, + "step": 12614 + }, + { + "epoch": 4.756787330316742, + "grad_norm": 0.062791454760351, + "learning_rate": 5.752193064210066e-07, + "loss": 0.3723, + "step": 12615 + }, + { + "epoch": 4.757164404223228, + "grad_norm": 0.060008963293648916, + "learning_rate": 5.734413339820367e-07, + "loss": 0.373, + "step": 12616 + }, + { + "epoch": 4.757541478129713, + "grad_norm": 0.06093982454796239, + "learning_rate": 5.716660937686192e-07, + "loss": 0.3681, + "step": 12617 + }, + { + "epoch": 4.757918552036199, + "grad_norm": 0.06325985880884348, + "learning_rate": 5.698935859037624e-07, + "loss": 0.3826, + "step": 12618 + }, + { + "epoch": 4.758295625942685, + "grad_norm": 0.06392551545058178, + "learning_rate": 5.681238105103104e-07, + "loss": 0.3821, + "step": 12619 + }, + { + "epoch": 4.75867269984917, + "grad_norm": 0.06308267807515862, + "learning_rate": 5.663567677108983e-07, + "loss": 0.3679, + "step": 12620 + }, + { + "epoch": 4.759049773755656, + "grad_norm": 0.06553220849568177, + "learning_rate": 5.645924576279882e-07, + "loss": 0.3717, + "step": 12621 + }, + { + "epoch": 4.759426847662142, + "grad_norm": 0.059737782645953814, + "learning_rate": 5.628308803838378e-07, + "loss": 0.386, + "step": 12622 + }, + { + "epoch": 4.759803921568627, + "grad_norm": 0.06322893477563396, + "learning_rate": 5.610720361005317e-07, + "loss": 0.3836, + "step": 12623 + }, + { + "epoch": 4.760180995475113, + "grad_norm": 0.0625788933653023, + "learning_rate": 5.593159248999502e-07, + "loss": 0.37, + "step": 12624 + }, + { + "epoch": 4.760558069381599, + "grad_norm": 0.06439246703162699, + "learning_rate": 5.575625469037871e-07, + "loss": 0.3722, + "step": 12625 + }, + { + "epoch": 4.760935143288084, + "grad_norm": 0.0619960428766917, + "learning_rate": 5.55811902233554e-07, + "loss": 0.377, + "step": 12626 + }, + { + "epoch": 4.76131221719457, + "grad_norm": 0.06464044608751718, + "learning_rate": 5.540639910105761e-07, + "loss": 0.3758, + "step": 12627 + }, + { + "epoch": 4.761689291101056, + "grad_norm": 0.06225263336669254, + "learning_rate": 5.523188133559653e-07, + "loss": 0.3733, + "step": 12628 + }, + { + "epoch": 4.762066365007541, + "grad_norm": 0.06465653515942377, + "learning_rate": 5.505763693906785e-07, + "loss": 0.3758, + "step": 12629 + }, + { + "epoch": 4.762443438914027, + "grad_norm": 0.06014598056187544, + "learning_rate": 5.488366592354543e-07, + "loss": 0.3686, + "step": 12630 + }, + { + "epoch": 4.762820512820513, + "grad_norm": 0.06441054625232959, + "learning_rate": 5.470996830108588e-07, + "loss": 0.379, + "step": 12631 + }, + { + "epoch": 4.7631975867269984, + "grad_norm": 0.06341554415811106, + "learning_rate": 5.453654408372578e-07, + "loss": 0.391, + "step": 12632 + }, + { + "epoch": 4.763574660633484, + "grad_norm": 0.06332001364184911, + "learning_rate": 5.43633932834835e-07, + "loss": 0.3756, + "step": 12633 + }, + { + "epoch": 4.76395173453997, + "grad_norm": 0.06006909726290168, + "learning_rate": 5.41905159123588e-07, + "loss": 0.39, + "step": 12634 + }, + { + "epoch": 4.7643288084464555, + "grad_norm": 0.06239787215284799, + "learning_rate": 5.401791198233098e-07, + "loss": 0.3742, + "step": 12635 + }, + { + "epoch": 4.764705882352941, + "grad_norm": 0.0628159771483597, + "learning_rate": 5.384558150536201e-07, + "loss": 0.3758, + "step": 12636 + }, + { + "epoch": 4.765082956259427, + "grad_norm": 0.06151562508191504, + "learning_rate": 5.367352449339391e-07, + "loss": 0.3775, + "step": 12637 + }, + { + "epoch": 4.7654600301659125, + "grad_norm": 0.060618641182205855, + "learning_rate": 5.350174095835004e-07, + "loss": 0.3782, + "step": 12638 + }, + { + "epoch": 4.765837104072398, + "grad_norm": 0.06259393138405271, + "learning_rate": 5.333023091213508e-07, + "loss": 0.368, + "step": 12639 + }, + { + "epoch": 4.766214177978884, + "grad_norm": 0.06547717932251053, + "learning_rate": 5.315899436663418e-07, + "loss": 0.389, + "step": 12640 + }, + { + "epoch": 4.7665912518853695, + "grad_norm": 0.06202977702916007, + "learning_rate": 5.298803133371478e-07, + "loss": 0.374, + "step": 12641 + }, + { + "epoch": 4.766968325791855, + "grad_norm": 0.0638042244783596, + "learning_rate": 5.281734182522336e-07, + "loss": 0.3716, + "step": 12642 + }, + { + "epoch": 4.767345399698341, + "grad_norm": 0.06302220745306386, + "learning_rate": 5.264692585298914e-07, + "loss": 0.383, + "step": 12643 + }, + { + "epoch": 4.7677224736048265, + "grad_norm": 0.06412489635135696, + "learning_rate": 5.247678342882135e-07, + "loss": 0.3799, + "step": 12644 + }, + { + "epoch": 4.768099547511312, + "grad_norm": 0.06493169882577429, + "learning_rate": 5.230691456451143e-07, + "loss": 0.3844, + "step": 12645 + }, + { + "epoch": 4.768476621417798, + "grad_norm": 0.06632591394025694, + "learning_rate": 5.213731927183041e-07, + "loss": 0.3776, + "step": 12646 + }, + { + "epoch": 4.7688536953242835, + "grad_norm": 0.0632076185136448, + "learning_rate": 5.196799756253157e-07, + "loss": 0.3725, + "step": 12647 + }, + { + "epoch": 4.769230769230769, + "grad_norm": 0.06692851614451979, + "learning_rate": 5.179894944834863e-07, + "loss": 0.3811, + "step": 12648 + }, + { + "epoch": 4.769607843137255, + "grad_norm": 0.06399610975825208, + "learning_rate": 5.163017494099665e-07, + "loss": 0.3801, + "step": 12649 + }, + { + "epoch": 4.7699849170437405, + "grad_norm": 0.06632860591665136, + "learning_rate": 5.146167405217073e-07, + "loss": 0.3845, + "step": 12650 + }, + { + "epoch": 4.770361990950226, + "grad_norm": 0.06200433142878391, + "learning_rate": 5.129344679354908e-07, + "loss": 0.3761, + "step": 12651 + }, + { + "epoch": 4.770739064856712, + "grad_norm": 0.06224337983583363, + "learning_rate": 5.11254931767886e-07, + "loss": 0.3828, + "step": 12652 + }, + { + "epoch": 4.7711161387631975, + "grad_norm": 0.06394314147812415, + "learning_rate": 5.095781321352889e-07, + "loss": 0.3745, + "step": 12653 + }, + { + "epoch": 4.771493212669683, + "grad_norm": 0.0619848976801205, + "learning_rate": 5.079040691539039e-07, + "loss": 0.377, + "step": 12654 + }, + { + "epoch": 4.771870286576169, + "grad_norm": 0.06216012959207825, + "learning_rate": 5.062327429397318e-07, + "loss": 0.382, + "step": 12655 + }, + { + "epoch": 4.7722473604826545, + "grad_norm": 0.0602277174578973, + "learning_rate": 5.045641536086044e-07, + "loss": 0.3888, + "step": 12656 + }, + { + "epoch": 4.77262443438914, + "grad_norm": 0.06138483156023221, + "learning_rate": 5.028983012761446e-07, + "loss": 0.3779, + "step": 12657 + }, + { + "epoch": 4.773001508295626, + "grad_norm": 0.06643001411470337, + "learning_rate": 5.012351860577979e-07, + "loss": 0.3756, + "step": 12658 + }, + { + "epoch": 4.7733785822021115, + "grad_norm": 0.06088726659131683, + "learning_rate": 4.995748080688189e-07, + "loss": 0.3687, + "step": 12659 + }, + { + "epoch": 4.773755656108597, + "grad_norm": 0.06496895618699122, + "learning_rate": 4.979171674242711e-07, + "loss": 0.3771, + "step": 12660 + }, + { + "epoch": 4.774132730015083, + "grad_norm": 0.06275560950351362, + "learning_rate": 4.962622642390269e-07, + "loss": 0.377, + "step": 12661 + }, + { + "epoch": 4.7745098039215685, + "grad_norm": 0.06198097131276206, + "learning_rate": 4.946100986277635e-07, + "loss": 0.3908, + "step": 12662 + }, + { + "epoch": 4.774886877828054, + "grad_norm": 0.06375960673100682, + "learning_rate": 4.929606707049806e-07, + "loss": 0.363, + "step": 12663 + }, + { + "epoch": 4.77526395173454, + "grad_norm": 0.060463389786173896, + "learning_rate": 4.913139805849776e-07, + "loss": 0.3853, + "step": 12664 + }, + { + "epoch": 4.7756410256410255, + "grad_norm": 0.06423804002094968, + "learning_rate": 4.896700283818767e-07, + "loss": 0.377, + "step": 12665 + }, + { + "epoch": 4.776018099547511, + "grad_norm": 0.06136539702134079, + "learning_rate": 4.880288142095957e-07, + "loss": 0.382, + "step": 12666 + }, + { + "epoch": 4.776395173453997, + "grad_norm": 0.06201575023698914, + "learning_rate": 4.863903381818657e-07, + "loss": 0.3806, + "step": 12667 + }, + { + "epoch": 4.776772247360483, + "grad_norm": 0.06250668167906595, + "learning_rate": 4.84754600412245e-07, + "loss": 0.3791, + "step": 12668 + }, + { + "epoch": 4.777149321266968, + "grad_norm": 0.06076373081092714, + "learning_rate": 4.831216010140738e-07, + "loss": 0.3768, + "step": 12669 + }, + { + "epoch": 4.777526395173454, + "grad_norm": 0.06043044166276422, + "learning_rate": 4.814913401005283e-07, + "loss": 0.3817, + "step": 12670 + }, + { + "epoch": 4.77790346907994, + "grad_norm": 0.06614009147772815, + "learning_rate": 4.798638177845805e-07, + "loss": 0.3653, + "step": 12671 + }, + { + "epoch": 4.778280542986425, + "grad_norm": 0.06256644185939718, + "learning_rate": 4.782390341790155e-07, + "loss": 0.365, + "step": 12672 + }, + { + "epoch": 4.778657616892911, + "grad_norm": 0.06428414941773036, + "learning_rate": 4.766169893964234e-07, + "loss": 0.3773, + "step": 12673 + }, + { + "epoch": 4.779034690799397, + "grad_norm": 0.0614272863284949, + "learning_rate": 4.749976835492254e-07, + "loss": 0.3913, + "step": 12674 + }, + { + "epoch": 4.779411764705882, + "grad_norm": 0.06146505961540112, + "learning_rate": 4.7338111674962495e-07, + "loss": 0.3771, + "step": 12675 + }, + { + "epoch": 4.779788838612368, + "grad_norm": 0.06223434413367209, + "learning_rate": 4.7176728910965255e-07, + "loss": 0.3812, + "step": 12676 + }, + { + "epoch": 4.780165912518854, + "grad_norm": 0.06248990992967046, + "learning_rate": 4.7015620074114753e-07, + "loss": 0.3707, + "step": 12677 + }, + { + "epoch": 4.780542986425339, + "grad_norm": 0.06377897530460055, + "learning_rate": 4.68547851755754e-07, + "loss": 0.381, + "step": 12678 + }, + { + "epoch": 4.780920060331825, + "grad_norm": 0.06062099943162645, + "learning_rate": 4.6694224226492946e-07, + "loss": 0.3672, + "step": 12679 + }, + { + "epoch": 4.781297134238311, + "grad_norm": 0.06466084999885077, + "learning_rate": 4.6533937237994045e-07, + "loss": 0.3745, + "step": 12680 + }, + { + "epoch": 4.781674208144796, + "grad_norm": 0.06460166089630438, + "learning_rate": 4.637392422118669e-07, + "loss": 0.3762, + "step": 12681 + }, + { + "epoch": 4.782051282051282, + "grad_norm": 0.06433744227895383, + "learning_rate": 4.621418518715981e-07, + "loss": 0.3791, + "step": 12682 + }, + { + "epoch": 4.782428355957768, + "grad_norm": 0.06255350052281604, + "learning_rate": 4.605472014698231e-07, + "loss": 0.3722, + "step": 12683 + }, + { + "epoch": 4.782805429864253, + "grad_norm": 0.06479898240224523, + "learning_rate": 4.589552911170625e-07, + "loss": 0.3827, + "step": 12684 + }, + { + "epoch": 4.783182503770739, + "grad_norm": 0.064488611856928, + "learning_rate": 4.5736612092362355e-07, + "loss": 0.3828, + "step": 12685 + }, + { + "epoch": 4.783559577677225, + "grad_norm": 0.06626833662629186, + "learning_rate": 4.5577969099964034e-07, + "loss": 0.3776, + "step": 12686 + }, + { + "epoch": 4.78393665158371, + "grad_norm": 0.06258016863255358, + "learning_rate": 4.5419600145504725e-07, + "loss": 0.3794, + "step": 12687 + }, + { + "epoch": 4.784313725490196, + "grad_norm": 0.062080545580646414, + "learning_rate": 4.52615052399592e-07, + "loss": 0.3774, + "step": 12688 + }, + { + "epoch": 4.784690799396682, + "grad_norm": 0.06239095415512569, + "learning_rate": 4.5103684394283584e-07, + "loss": 0.381, + "step": 12689 + }, + { + "epoch": 4.785067873303167, + "grad_norm": 0.0631300288509875, + "learning_rate": 4.494613761941491e-07, + "loss": 0.3795, + "step": 12690 + }, + { + "epoch": 4.785444947209653, + "grad_norm": 0.06111088528298053, + "learning_rate": 4.4788864926270657e-07, + "loss": 0.3727, + "step": 12691 + }, + { + "epoch": 4.785822021116139, + "grad_norm": 0.06330361959487145, + "learning_rate": 4.4631866325749675e-07, + "loss": 0.3823, + "step": 12692 + }, + { + "epoch": 4.786199095022624, + "grad_norm": 0.06700775674643412, + "learning_rate": 4.447514182873214e-07, + "loss": 0.3803, + "step": 12693 + }, + { + "epoch": 4.78657616892911, + "grad_norm": 0.06844305764160229, + "learning_rate": 4.43186914460787e-07, + "loss": 0.3883, + "step": 12694 + }, + { + "epoch": 4.786953242835596, + "grad_norm": 0.06440735655350889, + "learning_rate": 4.416251518863091e-07, + "loss": 0.3939, + "step": 12695 + }, + { + "epoch": 4.787330316742081, + "grad_norm": 0.0654686424060069, + "learning_rate": 4.4006613067212545e-07, + "loss": 0.3775, + "step": 12696 + }, + { + "epoch": 4.787707390648567, + "grad_norm": 0.0627300648414614, + "learning_rate": 4.385098509262653e-07, + "loss": 0.3817, + "step": 12697 + }, + { + "epoch": 4.788084464555053, + "grad_norm": 0.06233471878409344, + "learning_rate": 4.369563127565846e-07, + "loss": 0.3801, + "step": 12698 + }, + { + "epoch": 4.788461538461538, + "grad_norm": 0.06593738238177593, + "learning_rate": 4.354055162707349e-07, + "loss": 0.3798, + "step": 12699 + }, + { + "epoch": 4.788838612368024, + "grad_norm": 0.060923333424515565, + "learning_rate": 4.338574615761948e-07, + "loss": 0.3787, + "step": 12700 + }, + { + "epoch": 4.78921568627451, + "grad_norm": 0.06765445101317333, + "learning_rate": 4.3231214878023396e-07, + "loss": 0.3885, + "step": 12701 + }, + { + "epoch": 4.789592760180995, + "grad_norm": 0.06472423774101672, + "learning_rate": 4.3076957798994454e-07, + "loss": 0.3673, + "step": 12702 + }, + { + "epoch": 4.789969834087481, + "grad_norm": 0.062147486574276545, + "learning_rate": 4.2922974931222326e-07, + "loss": 0.3828, + "step": 12703 + }, + { + "epoch": 4.790346907993967, + "grad_norm": 0.06247335124208972, + "learning_rate": 4.2769266285378476e-07, + "loss": 0.3811, + "step": 12704 + }, + { + "epoch": 4.790723981900452, + "grad_norm": 0.0646344685783474, + "learning_rate": 4.2615831872114377e-07, + "loss": 0.3697, + "step": 12705 + }, + { + "epoch": 4.791101055806938, + "grad_norm": 0.06526965149756231, + "learning_rate": 4.2462671702062865e-07, + "loss": 0.3745, + "step": 12706 + }, + { + "epoch": 4.791478129713424, + "grad_norm": 0.06267657658645708, + "learning_rate": 4.2309785785837666e-07, + "loss": 0.3826, + "step": 12707 + }, + { + "epoch": 4.791855203619909, + "grad_norm": 0.06301496448169072, + "learning_rate": 4.215717413403431e-07, + "loss": 0.3859, + "step": 12708 + }, + { + "epoch": 4.792232277526395, + "grad_norm": 0.060334737935814946, + "learning_rate": 4.2004836757227887e-07, + "loss": 0.3805, + "step": 12709 + }, + { + "epoch": 4.792609351432881, + "grad_norm": 0.06297345948085051, + "learning_rate": 4.185277366597573e-07, + "loss": 0.3779, + "step": 12710 + }, + { + "epoch": 4.792986425339366, + "grad_norm": 0.06074714330303157, + "learning_rate": 4.17009848708152e-07, + "loss": 0.3859, + "step": 12711 + }, + { + "epoch": 4.793363499245852, + "grad_norm": 0.065460407612675, + "learning_rate": 4.154947038226542e-07, + "loss": 0.3789, + "step": 12712 + }, + { + "epoch": 4.793740573152338, + "grad_norm": 0.06284708016648312, + "learning_rate": 4.139823021082645e-07, + "loss": 0.3791, + "step": 12713 + }, + { + "epoch": 4.794117647058823, + "grad_norm": 0.06320178116635426, + "learning_rate": 4.124726436697879e-07, + "loss": 0.3821, + "step": 12714 + }, + { + "epoch": 4.794494720965309, + "grad_norm": 0.06236550263709039, + "learning_rate": 4.109657286118429e-07, + "loss": 0.3725, + "step": 12715 + }, + { + "epoch": 4.794871794871795, + "grad_norm": 0.06102416101802158, + "learning_rate": 4.0946155703885716e-07, + "loss": 0.368, + "step": 12716 + }, + { + "epoch": 4.79524886877828, + "grad_norm": 0.06158577071572598, + "learning_rate": 4.079601290550672e-07, + "loss": 0.3913, + "step": 12717 + }, + { + "epoch": 4.795625942684766, + "grad_norm": 0.06394228319860311, + "learning_rate": 4.064614447645232e-07, + "loss": 0.3891, + "step": 12718 + }, + { + "epoch": 4.796003016591252, + "grad_norm": 0.0635157293212624, + "learning_rate": 4.0496550427108427e-07, + "loss": 0.3731, + "step": 12719 + }, + { + "epoch": 4.796380090497737, + "grad_norm": 0.06688981107779794, + "learning_rate": 4.034723076784097e-07, + "loss": 0.3845, + "step": 12720 + }, + { + "epoch": 4.796757164404223, + "grad_norm": 0.06212562903236907, + "learning_rate": 4.0198185508998567e-07, + "loss": 0.3852, + "step": 12721 + }, + { + "epoch": 4.797134238310709, + "grad_norm": 0.06127991298183857, + "learning_rate": 4.0049414660909833e-07, + "loss": 0.3728, + "step": 12722 + }, + { + "epoch": 4.797511312217194, + "grad_norm": 0.06356742163533984, + "learning_rate": 3.9900918233883866e-07, + "loss": 0.3672, + "step": 12723 + }, + { + "epoch": 4.79788838612368, + "grad_norm": 0.06238911197467474, + "learning_rate": 3.975269623821154e-07, + "loss": 0.3904, + "step": 12724 + }, + { + "epoch": 4.798265460030166, + "grad_norm": 0.06562421749660229, + "learning_rate": 3.9604748684165084e-07, + "loss": 0.3919, + "step": 12725 + }, + { + "epoch": 4.798642533936651, + "grad_norm": 0.06509544121446872, + "learning_rate": 3.9457075581996294e-07, + "loss": 0.3708, + "step": 12726 + }, + { + "epoch": 4.799019607843137, + "grad_norm": 0.06417121742011411, + "learning_rate": 3.9309676941939214e-07, + "loss": 0.3744, + "step": 12727 + }, + { + "epoch": 4.799396681749623, + "grad_norm": 0.06115456551682931, + "learning_rate": 3.916255277420833e-07, + "loss": 0.3905, + "step": 12728 + }, + { + "epoch": 4.799773755656108, + "grad_norm": 0.06283047534718197, + "learning_rate": 3.901570308899949e-07, + "loss": 0.3851, + "step": 12729 + }, + { + "epoch": 4.800150829562594, + "grad_norm": 0.06053971583891752, + "learning_rate": 3.8869127896489e-07, + "loss": 0.3667, + "step": 12730 + }, + { + "epoch": 4.80052790346908, + "grad_norm": 0.060636509441784905, + "learning_rate": 3.8722827206834067e-07, + "loss": 0.3757, + "step": 12731 + }, + { + "epoch": 4.800904977375565, + "grad_norm": 0.06086960915532099, + "learning_rate": 3.857680103017414e-07, + "loss": 0.3706, + "step": 12732 + }, + { + "epoch": 4.801282051282051, + "grad_norm": 0.059746018378561964, + "learning_rate": 3.8431049376627785e-07, + "loss": 0.372, + "step": 12733 + }, + { + "epoch": 4.801659125188537, + "grad_norm": 0.06096627780981307, + "learning_rate": 3.8285572256295813e-07, + "loss": 0.3695, + "step": 12734 + }, + { + "epoch": 4.802036199095022, + "grad_norm": 0.06255595257503782, + "learning_rate": 3.814036967925949e-07, + "loss": 0.3772, + "step": 12735 + }, + { + "epoch": 4.802413273001508, + "grad_norm": 0.0662188634738061, + "learning_rate": 3.799544165558189e-07, + "loss": 0.3767, + "step": 12736 + }, + { + "epoch": 4.802790346907994, + "grad_norm": 0.06374333402758527, + "learning_rate": 3.7850788195305633e-07, + "loss": 0.3727, + "step": 12737 + }, + { + "epoch": 4.8031674208144794, + "grad_norm": 0.06352463683945057, + "learning_rate": 3.770640930845515e-07, + "loss": 0.3697, + "step": 12738 + }, + { + "epoch": 4.803544494720965, + "grad_norm": 0.06316105387764247, + "learning_rate": 3.756230500503577e-07, + "loss": 0.3769, + "step": 12739 + }, + { + "epoch": 4.803921568627451, + "grad_norm": 0.06224800578414204, + "learning_rate": 3.7418475295034615e-07, + "loss": 0.3749, + "step": 12740 + }, + { + "epoch": 4.8042986425339365, + "grad_norm": 0.06249416840510016, + "learning_rate": 3.727492018841794e-07, + "loss": 0.3907, + "step": 12741 + }, + { + "epoch": 4.804675716440422, + "grad_norm": 0.06076456668982696, + "learning_rate": 3.7131639695134227e-07, + "loss": 0.3897, + "step": 12742 + }, + { + "epoch": 4.805052790346908, + "grad_norm": 0.061447471594407006, + "learning_rate": 3.698863382511331e-07, + "loss": 0.3813, + "step": 12743 + }, + { + "epoch": 4.8054298642533935, + "grad_norm": 0.06143758900642915, + "learning_rate": 3.6845902588264593e-07, + "loss": 0.3721, + "step": 12744 + }, + { + "epoch": 4.805806938159879, + "grad_norm": 0.06525386061222929, + "learning_rate": 3.6703445994479283e-07, + "loss": 0.3901, + "step": 12745 + }, + { + "epoch": 4.806184012066365, + "grad_norm": 0.06358260390571549, + "learning_rate": 3.6561264053629917e-07, + "loss": 0.3728, + "step": 12746 + }, + { + "epoch": 4.8065610859728505, + "grad_norm": 0.06383915526186977, + "learning_rate": 3.6419356775569516e-07, + "loss": 0.3867, + "step": 12747 + }, + { + "epoch": 4.806938159879336, + "grad_norm": 0.06276437899880011, + "learning_rate": 3.627772417013198e-07, + "loss": 0.3819, + "step": 12748 + }, + { + "epoch": 4.807315233785822, + "grad_norm": 0.06209281976150338, + "learning_rate": 3.6136366247132126e-07, + "loss": 0.3714, + "step": 12749 + }, + { + "epoch": 4.8076923076923075, + "grad_norm": 0.06201504341258286, + "learning_rate": 3.599528301636612e-07, + "loss": 0.3788, + "step": 12750 + }, + { + "epoch": 4.808069381598793, + "grad_norm": 0.06258994729326393, + "learning_rate": 3.5854474487611035e-07, + "loss": 0.3714, + "step": 12751 + }, + { + "epoch": 4.808446455505279, + "grad_norm": 0.0620518832341447, + "learning_rate": 3.5713940670624834e-07, + "loss": 0.3775, + "step": 12752 + }, + { + "epoch": 4.8088235294117645, + "grad_norm": 0.0607271254113276, + "learning_rate": 3.557368157514596e-07, + "loss": 0.3767, + "step": 12753 + }, + { + "epoch": 4.80920060331825, + "grad_norm": 0.06393902540471495, + "learning_rate": 3.543369721089507e-07, + "loss": 0.3787, + "step": 12754 + }, + { + "epoch": 4.809577677224736, + "grad_norm": 0.06188386415955179, + "learning_rate": 3.5293987587571963e-07, + "loss": 0.3685, + "step": 12755 + }, + { + "epoch": 4.8099547511312215, + "grad_norm": 0.061518428988936544, + "learning_rate": 3.515455271485868e-07, + "loss": 0.3771, + "step": 12756 + }, + { + "epoch": 4.810331825037707, + "grad_norm": 0.0628813527188942, + "learning_rate": 3.501539260241815e-07, + "loss": 0.3731, + "step": 12757 + }, + { + "epoch": 4.810708898944193, + "grad_norm": 0.06189826463727026, + "learning_rate": 3.487650725989422e-07, + "loss": 0.3947, + "step": 12758 + }, + { + "epoch": 4.8110859728506785, + "grad_norm": 0.0627068563256359, + "learning_rate": 3.473789669691119e-07, + "loss": 0.3681, + "step": 12759 + }, + { + "epoch": 4.811463046757164, + "grad_norm": 0.06591656802796302, + "learning_rate": 3.4599560923074703e-07, + "loss": 0.3787, + "step": 12760 + }, + { + "epoch": 4.81184012066365, + "grad_norm": 0.05909600939839513, + "learning_rate": 3.4461499947971323e-07, + "loss": 0.3688, + "step": 12761 + }, + { + "epoch": 4.8122171945701355, + "grad_norm": 0.061724123774033385, + "learning_rate": 3.43237137811685e-07, + "loss": 0.3762, + "step": 12762 + }, + { + "epoch": 4.812594268476621, + "grad_norm": 0.06309455381311131, + "learning_rate": 3.41862024322146e-07, + "loss": 0.3773, + "step": 12763 + }, + { + "epoch": 4.812971342383107, + "grad_norm": 0.0642946346754901, + "learning_rate": 3.4048965910639775e-07, + "loss": 0.3828, + "step": 12764 + }, + { + "epoch": 4.8133484162895925, + "grad_norm": 0.0636772518409318, + "learning_rate": 3.391200422595331e-07, + "loss": 0.3792, + "step": 12765 + }, + { + "epoch": 4.813725490196078, + "grad_norm": 0.06215249274139015, + "learning_rate": 3.3775317387647167e-07, + "loss": 0.3733, + "step": 12766 + }, + { + "epoch": 4.814102564102564, + "grad_norm": 0.05987532456438769, + "learning_rate": 3.3638905405193764e-07, + "loss": 0.378, + "step": 12767 + }, + { + "epoch": 4.8144796380090495, + "grad_norm": 0.06266840589259845, + "learning_rate": 3.350276828804555e-07, + "loss": 0.379, + "step": 12768 + }, + { + "epoch": 4.814856711915535, + "grad_norm": 0.06260898028742058, + "learning_rate": 3.3366906045637635e-07, + "loss": 0.382, + "step": 12769 + }, + { + "epoch": 4.815233785822021, + "grad_norm": 0.062525593125908, + "learning_rate": 3.3231318687384716e-07, + "loss": 0.3872, + "step": 12770 + }, + { + "epoch": 4.8156108597285066, + "grad_norm": 0.0638046904581455, + "learning_rate": 3.309600622268239e-07, + "loss": 0.3771, + "step": 12771 + }, + { + "epoch": 4.815987933634992, + "grad_norm": 0.06312659414316493, + "learning_rate": 3.296096866090892e-07, + "loss": 0.3709, + "step": 12772 + }, + { + "epoch": 4.816365007541478, + "grad_norm": 0.0627523814404354, + "learning_rate": 3.282620601142084e-07, + "loss": 0.3812, + "step": 12773 + }, + { + "epoch": 4.816742081447964, + "grad_norm": 0.05989341803456775, + "learning_rate": 3.2691718283558213e-07, + "loss": 0.3808, + "step": 12774 + }, + { + "epoch": 4.817119155354449, + "grad_norm": 0.059744616719829974, + "learning_rate": 3.2557505486640717e-07, + "loss": 0.3706, + "step": 12775 + }, + { + "epoch": 4.817496229260935, + "grad_norm": 0.06239593953393785, + "learning_rate": 3.24235676299689e-07, + "loss": 0.3832, + "step": 12776 + }, + { + "epoch": 4.817873303167421, + "grad_norm": 0.06019058698954475, + "learning_rate": 3.2289904722823784e-07, + "loss": 0.3756, + "step": 12777 + }, + { + "epoch": 4.818250377073906, + "grad_norm": 0.06489339767342675, + "learning_rate": 3.215651677446996e-07, + "loss": 0.387, + "step": 12778 + }, + { + "epoch": 4.818627450980392, + "grad_norm": 0.0592755486315811, + "learning_rate": 3.202340379414981e-07, + "loss": 0.3693, + "step": 12779 + }, + { + "epoch": 4.819004524886878, + "grad_norm": 0.06198112038477701, + "learning_rate": 3.1890565791087955e-07, + "loss": 0.3744, + "step": 12780 + }, + { + "epoch": 4.819381598793363, + "grad_norm": 0.06474296965521152, + "learning_rate": 3.1758002774489926e-07, + "loss": 0.3733, + "step": 12781 + }, + { + "epoch": 4.819758672699849, + "grad_norm": 0.06010453185218893, + "learning_rate": 3.1625714753543036e-07, + "loss": 0.376, + "step": 12782 + }, + { + "epoch": 4.820135746606335, + "grad_norm": 0.06377113859954982, + "learning_rate": 3.149370173741373e-07, + "loss": 0.3726, + "step": 12783 + }, + { + "epoch": 4.82051282051282, + "grad_norm": 0.06525593252019381, + "learning_rate": 3.136196373525113e-07, + "loss": 0.3815, + "step": 12784 + }, + { + "epoch": 4.820889894419306, + "grad_norm": 0.06472162451417983, + "learning_rate": 3.123050075618439e-07, + "loss": 0.3798, + "step": 12785 + }, + { + "epoch": 4.821266968325792, + "grad_norm": 0.06081384759200091, + "learning_rate": 3.1099312809323104e-07, + "loss": 0.3745, + "step": 12786 + }, + { + "epoch": 4.821644042232277, + "grad_norm": 0.059936995123208374, + "learning_rate": 3.096839990375955e-07, + "loss": 0.3727, + "step": 12787 + }, + { + "epoch": 4.822021116138763, + "grad_norm": 0.059508065521857265, + "learning_rate": 3.0837762048565143e-07, + "loss": 0.3773, + "step": 12788 + }, + { + "epoch": 4.822398190045249, + "grad_norm": 0.06163913544028371, + "learning_rate": 3.070739925279309e-07, + "loss": 0.3876, + "step": 12789 + }, + { + "epoch": 4.822775263951734, + "grad_norm": 0.061669120191050524, + "learning_rate": 3.057731152547794e-07, + "loss": 0.3709, + "step": 12790 + }, + { + "epoch": 4.82315233785822, + "grad_norm": 0.06257503783649908, + "learning_rate": 3.0447498875633807e-07, + "loss": 0.3685, + "step": 12791 + }, + { + "epoch": 4.823529411764706, + "grad_norm": 0.06091929232948833, + "learning_rate": 3.031796131225706e-07, + "loss": 0.3711, + "step": 12792 + }, + { + "epoch": 4.823906485671191, + "grad_norm": 0.06108767267821017, + "learning_rate": 3.0188698844324513e-07, + "loss": 0.3886, + "step": 12793 + }, + { + "epoch": 4.824283559577677, + "grad_norm": 0.06009311344680644, + "learning_rate": 3.00597114807939e-07, + "loss": 0.3799, + "step": 12794 + }, + { + "epoch": 4.824660633484163, + "grad_norm": 0.06225391486431192, + "learning_rate": 2.993099923060383e-07, + "loss": 0.3756, + "step": 12795 + }, + { + "epoch": 4.825037707390648, + "grad_norm": 0.06201651715754873, + "learning_rate": 2.980256210267429e-07, + "loss": 0.3745, + "step": 12796 + }, + { + "epoch": 4.825414781297134, + "grad_norm": 0.060888525237286516, + "learning_rate": 2.967440010590572e-07, + "loss": 0.3742, + "step": 12797 + }, + { + "epoch": 4.82579185520362, + "grad_norm": 0.06038056446404498, + "learning_rate": 2.9546513249179454e-07, + "loss": 0.3808, + "step": 12798 + }, + { + "epoch": 4.826168929110105, + "grad_norm": 0.062310757021960816, + "learning_rate": 2.941890154135818e-07, + "loss": 0.3804, + "step": 12799 + }, + { + "epoch": 4.826546003016591, + "grad_norm": 0.06096570756647357, + "learning_rate": 2.9291564991284603e-07, + "loss": 0.3757, + "step": 12800 + }, + { + "epoch": 4.826923076923077, + "grad_norm": 0.06229707859868202, + "learning_rate": 2.916450360778411e-07, + "loss": 0.3752, + "step": 12801 + }, + { + "epoch": 4.827300150829562, + "grad_norm": 0.0605868263902541, + "learning_rate": 2.9037717399661214e-07, + "loss": 0.3687, + "step": 12802 + }, + { + "epoch": 4.827677224736048, + "grad_norm": 0.06270612591480128, + "learning_rate": 2.891120637570222e-07, + "loss": 0.3827, + "step": 12803 + }, + { + "epoch": 4.828054298642534, + "grad_norm": 0.06268260179158446, + "learning_rate": 2.878497054467433e-07, + "loss": 0.3816, + "step": 12804 + }, + { + "epoch": 4.828431372549019, + "grad_norm": 0.06235214267520107, + "learning_rate": 2.865900991532522e-07, + "loss": 0.3747, + "step": 12805 + }, + { + "epoch": 4.828808446455505, + "grad_norm": 0.06438552120381516, + "learning_rate": 2.853332449638435e-07, + "loss": 0.3807, + "step": 12806 + }, + { + "epoch": 4.829185520361991, + "grad_norm": 0.062088238671238985, + "learning_rate": 2.8407914296561645e-07, + "loss": 0.3695, + "step": 12807 + }, + { + "epoch": 4.829562594268476, + "grad_norm": 0.06392929692118504, + "learning_rate": 2.8282779324547485e-07, + "loss": 0.376, + "step": 12808 + }, + { + "epoch": 4.829939668174962, + "grad_norm": 0.06448282782271056, + "learning_rate": 2.8157919589014036e-07, + "loss": 0.3705, + "step": 12809 + }, + { + "epoch": 4.830316742081448, + "grad_norm": 0.061519263446178744, + "learning_rate": 2.80333350986135e-07, + "loss": 0.373, + "step": 12810 + }, + { + "epoch": 4.830693815987933, + "grad_norm": 0.06051494313426353, + "learning_rate": 2.790902586197941e-07, + "loss": 0.3748, + "step": 12811 + }, + { + "epoch": 4.831070889894419, + "grad_norm": 0.06284505790323407, + "learning_rate": 2.7784991887727095e-07, + "loss": 0.3651, + "step": 12812 + }, + { + "epoch": 4.831447963800905, + "grad_norm": 0.06481677349782754, + "learning_rate": 2.766123318445102e-07, + "loss": 0.3892, + "step": 12813 + }, + { + "epoch": 4.83182503770739, + "grad_norm": 0.06326993621330283, + "learning_rate": 2.753774976072832e-07, + "loss": 0.3749, + "step": 12814 + }, + { + "epoch": 4.832202111613876, + "grad_norm": 0.06185974243845965, + "learning_rate": 2.741454162511614e-07, + "loss": 0.388, + "step": 12815 + }, + { + "epoch": 4.832579185520362, + "grad_norm": 0.06297733180645226, + "learning_rate": 2.729160878615211e-07, + "loss": 0.3824, + "step": 12816 + }, + { + "epoch": 4.832956259426847, + "grad_norm": 0.06105097798149068, + "learning_rate": 2.7168951252356525e-07, + "loss": 0.3761, + "step": 12817 + }, + { + "epoch": 4.833333333333333, + "grad_norm": 0.061291374715391075, + "learning_rate": 2.704656903222791e-07, + "loss": 0.382, + "step": 12818 + }, + { + "epoch": 4.833710407239819, + "grad_norm": 0.06056528127735537, + "learning_rate": 2.6924462134248377e-07, + "loss": 0.3759, + "step": 12819 + }, + { + "epoch": 4.834087481146304, + "grad_norm": 0.0603703252480822, + "learning_rate": 2.68026305668796e-07, + "loss": 0.3817, + "step": 12820 + }, + { + "epoch": 4.83446455505279, + "grad_norm": 0.06028004332236738, + "learning_rate": 2.6681074338564594e-07, + "loss": 0.3882, + "step": 12821 + }, + { + "epoch": 4.834841628959276, + "grad_norm": 0.06033185923185457, + "learning_rate": 2.655979345772597e-07, + "loss": 0.3699, + "step": 12822 + }, + { + "epoch": 4.835218702865761, + "grad_norm": 0.06101766069238032, + "learning_rate": 2.643878793276988e-07, + "loss": 0.3796, + "step": 12823 + }, + { + "epoch": 4.835595776772247, + "grad_norm": 0.06164462327361216, + "learning_rate": 2.6318057772081184e-07, + "loss": 0.3664, + "step": 12824 + }, + { + "epoch": 4.835972850678733, + "grad_norm": 0.06587417150061527, + "learning_rate": 2.6197602984026517e-07, + "loss": 0.3823, + "step": 12825 + }, + { + "epoch": 4.836349924585218, + "grad_norm": 0.06155278537121624, + "learning_rate": 2.607742357695298e-07, + "loss": 0.3761, + "step": 12826 + }, + { + "epoch": 4.836726998491704, + "grad_norm": 0.06264610074431878, + "learning_rate": 2.5957519559189015e-07, + "loss": 0.3881, + "step": 12827 + }, + { + "epoch": 4.83710407239819, + "grad_norm": 0.06172100055593348, + "learning_rate": 2.5837890939043986e-07, + "loss": 0.3711, + "step": 12828 + }, + { + "epoch": 4.837481146304675, + "grad_norm": 0.061395804436930435, + "learning_rate": 2.5718537724808145e-07, + "loss": 0.383, + "step": 12829 + }, + { + "epoch": 4.837858220211161, + "grad_norm": 0.06291025280428651, + "learning_rate": 2.5599459924752654e-07, + "loss": 0.3817, + "step": 12830 + }, + { + "epoch": 4.838235294117647, + "grad_norm": 0.06178703913545708, + "learning_rate": 2.548065754712914e-07, + "loss": 0.3826, + "step": 12831 + }, + { + "epoch": 4.838612368024132, + "grad_norm": 0.060979504225146024, + "learning_rate": 2.5362130600170565e-07, + "loss": 0.379, + "step": 12832 + }, + { + "epoch": 4.838989441930618, + "grad_norm": 0.06221157740271734, + "learning_rate": 2.5243879092090804e-07, + "loss": 0.3712, + "step": 12833 + }, + { + "epoch": 4.839366515837104, + "grad_norm": 0.06184150131865484, + "learning_rate": 2.5125903031084644e-07, + "loss": 0.366, + "step": 12834 + }, + { + "epoch": 4.839743589743589, + "grad_norm": 0.06033355496693628, + "learning_rate": 2.5008202425328197e-07, + "loss": 0.3757, + "step": 12835 + }, + { + "epoch": 4.840120663650075, + "grad_norm": 0.06193438005653686, + "learning_rate": 2.489077728297673e-07, + "loss": 0.3816, + "step": 12836 + }, + { + "epoch": 4.840497737556561, + "grad_norm": 0.061998148886102125, + "learning_rate": 2.477362761216906e-07, + "loss": 0.3696, + "step": 12837 + }, + { + "epoch": 4.840874811463046, + "grad_norm": 0.063246367337493, + "learning_rate": 2.465675342102269e-07, + "loss": 0.3712, + "step": 12838 + }, + { + "epoch": 4.841251885369532, + "grad_norm": 0.06079399799293539, + "learning_rate": 2.4540154717637377e-07, + "loss": 0.3745, + "step": 12839 + }, + { + "epoch": 4.841628959276018, + "grad_norm": 0.06409883753452619, + "learning_rate": 2.4423831510092863e-07, + "loss": 0.3861, + "step": 12840 + }, + { + "epoch": 4.842006033182503, + "grad_norm": 0.06180800195900741, + "learning_rate": 2.4307783806450714e-07, + "loss": 0.376, + "step": 12841 + }, + { + "epoch": 4.842383107088989, + "grad_norm": 0.06540903487163668, + "learning_rate": 2.4192011614752485e-07, + "loss": 0.3727, + "step": 12842 + }, + { + "epoch": 4.842760180995475, + "grad_norm": 0.06204735615965021, + "learning_rate": 2.4076514943021543e-07, + "loss": 0.3845, + "step": 12843 + }, + { + "epoch": 4.8431372549019605, + "grad_norm": 0.0631949113701398, + "learning_rate": 2.3961293799261263e-07, + "loss": 0.3867, + "step": 12844 + }, + { + "epoch": 4.843514328808446, + "grad_norm": 0.06201619516519761, + "learning_rate": 2.3846348191456816e-07, + "loss": 0.3869, + "step": 12845 + }, + { + "epoch": 4.843891402714932, + "grad_norm": 0.06001765807003923, + "learning_rate": 2.3731678127573377e-07, + "loss": 0.3675, + "step": 12846 + }, + { + "epoch": 4.8442684766214175, + "grad_norm": 0.06424693617037512, + "learning_rate": 2.361728361555793e-07, + "loss": 0.3869, + "step": 12847 + }, + { + "epoch": 4.844645550527903, + "grad_norm": 0.06277649976593612, + "learning_rate": 2.3503164663337463e-07, + "loss": 0.3761, + "step": 12848 + }, + { + "epoch": 4.845022624434389, + "grad_norm": 0.061388350389341755, + "learning_rate": 2.3389321278820764e-07, + "loss": 0.3866, + "step": 12849 + }, + { + "epoch": 4.8453996983408745, + "grad_norm": 0.06143420408197139, + "learning_rate": 2.3275753469896635e-07, + "loss": 0.3758, + "step": 12850 + }, + { + "epoch": 4.84577677224736, + "grad_norm": 0.061026994720169155, + "learning_rate": 2.316246124443522e-07, + "loss": 0.3787, + "step": 12851 + }, + { + "epoch": 4.846153846153846, + "grad_norm": 0.06586665502998172, + "learning_rate": 2.3049444610288462e-07, + "loss": 0.3828, + "step": 12852 + }, + { + "epoch": 4.8465309200603315, + "grad_norm": 0.059732849728992246, + "learning_rate": 2.293670357528699e-07, + "loss": 0.3669, + "step": 12853 + }, + { + "epoch": 4.846907993966817, + "grad_norm": 0.06093221207578884, + "learning_rate": 2.2824238147244548e-07, + "loss": 0.3762, + "step": 12854 + }, + { + "epoch": 4.847285067873303, + "grad_norm": 0.06285105848734181, + "learning_rate": 2.2712048333954906e-07, + "loss": 0.3859, + "step": 12855 + }, + { + "epoch": 4.8476621417797885, + "grad_norm": 0.06200962841749516, + "learning_rate": 2.2600134143192288e-07, + "loss": 0.3761, + "step": 12856 + }, + { + "epoch": 4.848039215686274, + "grad_norm": 0.06049172926889099, + "learning_rate": 2.2488495582712266e-07, + "loss": 0.379, + "step": 12857 + }, + { + "epoch": 4.84841628959276, + "grad_norm": 0.06065867380814893, + "learning_rate": 2.2377132660251321e-07, + "loss": 0.3774, + "step": 12858 + }, + { + "epoch": 4.8487933634992455, + "grad_norm": 0.05948063293289098, + "learning_rate": 2.2266045383527723e-07, + "loss": 0.3806, + "step": 12859 + }, + { + "epoch": 4.849170437405731, + "grad_norm": 0.06525656198189331, + "learning_rate": 2.2155233760237982e-07, + "loss": 0.3838, + "step": 12860 + }, + { + "epoch": 4.849547511312217, + "grad_norm": 0.060619578496868234, + "learning_rate": 2.204469779806262e-07, + "loss": 0.381, + "step": 12861 + }, + { + "epoch": 4.8499245852187025, + "grad_norm": 0.06221868160023281, + "learning_rate": 2.1934437504661288e-07, + "loss": 0.3834, + "step": 12862 + }, + { + "epoch": 4.850301659125188, + "grad_norm": 0.06400959191128995, + "learning_rate": 2.1824452887674984e-07, + "loss": 0.3774, + "step": 12863 + }, + { + "epoch": 4.850678733031674, + "grad_norm": 0.06468540339979115, + "learning_rate": 2.171474395472517e-07, + "loss": 0.3801, + "step": 12864 + }, + { + "epoch": 4.8510558069381595, + "grad_norm": 0.06509679141412532, + "learning_rate": 2.1605310713415096e-07, + "loss": 0.3807, + "step": 12865 + }, + { + "epoch": 4.851432880844645, + "grad_norm": 0.0603206054268597, + "learning_rate": 2.1496153171328026e-07, + "loss": 0.382, + "step": 12866 + }, + { + "epoch": 4.851809954751131, + "grad_norm": 0.061953348221681935, + "learning_rate": 2.138727133602858e-07, + "loss": 0.3847, + "step": 12867 + }, + { + "epoch": 4.8521870286576165, + "grad_norm": 0.06124772927985402, + "learning_rate": 2.1278665215062278e-07, + "loss": 0.3771, + "step": 12868 + }, + { + "epoch": 4.852564102564102, + "grad_norm": 0.059447456680828174, + "learning_rate": 2.117033481595554e-07, + "loss": 0.3904, + "step": 12869 + }, + { + "epoch": 4.852941176470588, + "grad_norm": 0.06145656291474265, + "learning_rate": 2.1062280146215252e-07, + "loss": 0.3721, + "step": 12870 + }, + { + "epoch": 4.8533182503770735, + "grad_norm": 0.06151339465567302, + "learning_rate": 2.0954501213329647e-07, + "loss": 0.3858, + "step": 12871 + }, + { + "epoch": 4.853695324283559, + "grad_norm": 0.06048827122910177, + "learning_rate": 2.0846998024767417e-07, + "loss": 0.3627, + "step": 12872 + }, + { + "epoch": 4.854072398190045, + "grad_norm": 0.06172639894827678, + "learning_rate": 2.0739770587979047e-07, + "loss": 0.3773, + "step": 12873 + }, + { + "epoch": 4.8544494720965305, + "grad_norm": 0.06274908178119007, + "learning_rate": 2.0632818910395036e-07, + "loss": 0.375, + "step": 12874 + }, + { + "epoch": 4.854826546003016, + "grad_norm": 0.06187327967419252, + "learning_rate": 2.0526142999426792e-07, + "loss": 0.3687, + "step": 12875 + }, + { + "epoch": 4.855203619909502, + "grad_norm": 0.06214453765534232, + "learning_rate": 2.0419742862467063e-07, + "loss": 0.3767, + "step": 12876 + }, + { + "epoch": 4.855580693815988, + "grad_norm": 0.06216885384262871, + "learning_rate": 2.031361850688951e-07, + "loss": 0.3729, + "step": 12877 + }, + { + "epoch": 4.855957767722473, + "grad_norm": 0.062238296986942866, + "learning_rate": 2.0207769940047806e-07, + "loss": 0.3857, + "step": 12878 + }, + { + "epoch": 4.856334841628959, + "grad_norm": 0.06284346971430016, + "learning_rate": 2.010219716927786e-07, + "loss": 0.3881, + "step": 12879 + }, + { + "epoch": 4.856711915535445, + "grad_norm": 0.0647885708563927, + "learning_rate": 1.9996900201895153e-07, + "loss": 0.3733, + "step": 12880 + }, + { + "epoch": 4.85708898944193, + "grad_norm": 0.06078617375175199, + "learning_rate": 1.9891879045197404e-07, + "loss": 0.3822, + "step": 12881 + }, + { + "epoch": 4.857466063348416, + "grad_norm": 0.05981431361859295, + "learning_rate": 1.9787133706461902e-07, + "loss": 0.3786, + "step": 12882 + }, + { + "epoch": 4.857843137254902, + "grad_norm": 0.0590476215707285, + "learning_rate": 1.9682664192947733e-07, + "loss": 0.3801, + "step": 12883 + }, + { + "epoch": 4.858220211161387, + "grad_norm": 0.06077746323653223, + "learning_rate": 1.9578470511893988e-07, + "loss": 0.383, + "step": 12884 + }, + { + "epoch": 4.858597285067873, + "grad_norm": 0.060352183094547165, + "learning_rate": 1.9474552670522006e-07, + "loss": 0.3769, + "step": 12885 + }, + { + "epoch": 4.858974358974359, + "grad_norm": 0.06131120936633522, + "learning_rate": 1.9370910676032694e-07, + "loss": 0.3779, + "step": 12886 + }, + { + "epoch": 4.859351432880844, + "grad_norm": 0.05918789128440163, + "learning_rate": 1.9267544535608307e-07, + "loss": 0.3668, + "step": 12887 + }, + { + "epoch": 4.859728506787331, + "grad_norm": 0.06145511594940317, + "learning_rate": 1.9164454256412445e-07, + "loss": 0.3873, + "step": 12888 + }, + { + "epoch": 4.860105580693816, + "grad_norm": 0.06239546927782679, + "learning_rate": 1.906163984558873e-07, + "loss": 0.3874, + "step": 12889 + }, + { + "epoch": 4.860482654600302, + "grad_norm": 0.05901530677953186, + "learning_rate": 1.8959101310262129e-07, + "loss": 0.3751, + "step": 12890 + }, + { + "epoch": 4.860859728506787, + "grad_norm": 0.06041106965204149, + "learning_rate": 1.8856838657538956e-07, + "loss": 0.3778, + "step": 12891 + }, + { + "epoch": 4.8612368024132735, + "grad_norm": 0.061336030120452266, + "learning_rate": 1.8754851894505544e-07, + "loss": 0.3817, + "step": 12892 + }, + { + "epoch": 4.861613876319758, + "grad_norm": 0.06293767833753108, + "learning_rate": 1.8653141028229572e-07, + "loss": 0.3782, + "step": 12893 + }, + { + "epoch": 4.861990950226245, + "grad_norm": 0.060949165011086284, + "learning_rate": 1.855170606575918e-07, + "loss": 0.3699, + "step": 12894 + }, + { + "epoch": 4.86236802413273, + "grad_norm": 0.06304400646729004, + "learning_rate": 1.84505470141243e-07, + "loss": 0.3886, + "step": 12895 + }, + { + "epoch": 4.862745098039216, + "grad_norm": 0.06508552742012401, + "learning_rate": 1.834966388033488e-07, + "loss": 0.373, + "step": 12896 + }, + { + "epoch": 4.863122171945701, + "grad_norm": 0.06082283073126495, + "learning_rate": 1.824905667138177e-07, + "loss": 0.3876, + "step": 12897 + }, + { + "epoch": 4.8634992458521875, + "grad_norm": 0.06255991585063846, + "learning_rate": 1.814872539423762e-07, + "loss": 0.3874, + "step": 12898 + }, + { + "epoch": 4.863876319758672, + "grad_norm": 0.06112806879669722, + "learning_rate": 1.8048670055854643e-07, + "loss": 0.3875, + "step": 12899 + }, + { + "epoch": 4.864253393665159, + "grad_norm": 0.060681412400210265, + "learning_rate": 1.7948890663166852e-07, + "loss": 0.3702, + "step": 12900 + }, + { + "epoch": 4.864630467571644, + "grad_norm": 0.06410187993295688, + "learning_rate": 1.7849387223089154e-07, + "loss": 0.3819, + "step": 12901 + }, + { + "epoch": 4.86500754147813, + "grad_norm": 0.06296060936595017, + "learning_rate": 1.7750159742516037e-07, + "loss": 0.3789, + "step": 12902 + }, + { + "epoch": 4.865384615384615, + "grad_norm": 0.06113156270689443, + "learning_rate": 1.7651208228325112e-07, + "loss": 0.3724, + "step": 12903 + }, + { + "epoch": 4.8657616892911015, + "grad_norm": 0.05930969493193463, + "learning_rate": 1.7552532687373113e-07, + "loss": 0.3675, + "step": 12904 + }, + { + "epoch": 4.866138763197586, + "grad_norm": 0.05903293668128273, + "learning_rate": 1.745413312649813e-07, + "loss": 0.3726, + "step": 12905 + }, + { + "epoch": 4.866515837104073, + "grad_norm": 0.06110817720030767, + "learning_rate": 1.7356009552519148e-07, + "loss": 0.3797, + "step": 12906 + }, + { + "epoch": 4.866892911010558, + "grad_norm": 0.06002413708988983, + "learning_rate": 1.725816197223651e-07, + "loss": 0.3669, + "step": 12907 + }, + { + "epoch": 4.867269984917044, + "grad_norm": 0.06154138855447487, + "learning_rate": 1.7160590392429678e-07, + "loss": 0.3776, + "step": 12908 + }, + { + "epoch": 4.867647058823529, + "grad_norm": 0.05992182452993451, + "learning_rate": 1.706329481986213e-07, + "loss": 0.374, + "step": 12909 + }, + { + "epoch": 4.8680241327300156, + "grad_norm": 0.06356313339241468, + "learning_rate": 1.69662752612747e-07, + "loss": 0.3865, + "step": 12910 + }, + { + "epoch": 4.8684012066365, + "grad_norm": 0.06120072840129025, + "learning_rate": 1.6869531723391786e-07, + "loss": 0.3673, + "step": 12911 + }, + { + "epoch": 4.868778280542987, + "grad_norm": 0.058746179874874155, + "learning_rate": 1.6773064212917356e-07, + "loss": 0.3785, + "step": 12912 + }, + { + "epoch": 4.869155354449472, + "grad_norm": 0.06126772922722759, + "learning_rate": 1.6676872736536286e-07, + "loss": 0.3684, + "step": 12913 + }, + { + "epoch": 4.869532428355958, + "grad_norm": 0.05981000535805328, + "learning_rate": 1.6580957300914358e-07, + "loss": 0.3745, + "step": 12914 + }, + { + "epoch": 4.869909502262443, + "grad_norm": 0.05881542715414412, + "learning_rate": 1.6485317912699582e-07, + "loss": 0.3687, + "step": 12915 + }, + { + "epoch": 4.87028657616893, + "grad_norm": 0.06151224727454762, + "learning_rate": 1.638995457851822e-07, + "loss": 0.3726, + "step": 12916 + }, + { + "epoch": 4.870663650075414, + "grad_norm": 0.06232271201048319, + "learning_rate": 1.629486730498009e-07, + "loss": 0.3739, + "step": 12917 + }, + { + "epoch": 4.871040723981901, + "grad_norm": 0.06092581929776029, + "learning_rate": 1.62000560986737e-07, + "loss": 0.388, + "step": 12918 + }, + { + "epoch": 4.871417797888386, + "grad_norm": 0.06244409887164382, + "learning_rate": 1.610552096617024e-07, + "loss": 0.365, + "step": 12919 + }, + { + "epoch": 4.871794871794872, + "grad_norm": 0.06314489885238463, + "learning_rate": 1.6011261914020028e-07, + "loss": 0.3939, + "step": 12920 + }, + { + "epoch": 4.872171945701357, + "grad_norm": 0.060850524326217285, + "learning_rate": 1.591727894875561e-07, + "loss": 0.3837, + "step": 12921 + }, + { + "epoch": 4.872549019607844, + "grad_norm": 0.06377850248225918, + "learning_rate": 1.5823572076890004e-07, + "loss": 0.3736, + "step": 12922 + }, + { + "epoch": 4.872926093514328, + "grad_norm": 0.06105570689841416, + "learning_rate": 1.573014130491668e-07, + "loss": 0.3771, + "step": 12923 + }, + { + "epoch": 4.873303167420815, + "grad_norm": 0.06473794989226293, + "learning_rate": 1.5636986639310903e-07, + "loss": 0.389, + "step": 12924 + }, + { + "epoch": 4.8736802413273, + "grad_norm": 0.061060974553717846, + "learning_rate": 1.5544108086527954e-07, + "loss": 0.3811, + "step": 12925 + }, + { + "epoch": 4.874057315233786, + "grad_norm": 0.0626749080872161, + "learning_rate": 1.5451505653003573e-07, + "loss": 0.3754, + "step": 12926 + }, + { + "epoch": 4.874434389140271, + "grad_norm": 0.0627535139250332, + "learning_rate": 1.535917934515574e-07, + "loss": 0.3709, + "step": 12927 + }, + { + "epoch": 4.874811463046758, + "grad_norm": 0.0618678927700881, + "learning_rate": 1.5267129169382443e-07, + "loss": 0.3817, + "step": 12928 + }, + { + "epoch": 4.875188536953242, + "grad_norm": 0.059720712398320365, + "learning_rate": 1.5175355132062586e-07, + "loss": 0.3642, + "step": 12929 + }, + { + "epoch": 4.875565610859729, + "grad_norm": 0.059264077028543215, + "learning_rate": 1.508385723955641e-07, + "loss": 0.3783, + "step": 12930 + }, + { + "epoch": 4.875942684766214, + "grad_norm": 0.06182689445199336, + "learning_rate": 1.4992635498203735e-07, + "loss": 0.3751, + "step": 12931 + }, + { + "epoch": 4.8763197586727, + "grad_norm": 0.06260652754900027, + "learning_rate": 1.4901689914327056e-07, + "loss": 0.3775, + "step": 12932 + }, + { + "epoch": 4.876696832579185, + "grad_norm": 0.06289293974652746, + "learning_rate": 1.4811020494228444e-07, + "loss": 0.3656, + "step": 12933 + }, + { + "epoch": 4.877073906485672, + "grad_norm": 0.061648720457793516, + "learning_rate": 1.4720627244191322e-07, + "loss": 0.3798, + "step": 12934 + }, + { + "epoch": 4.877450980392156, + "grad_norm": 0.061072243825700076, + "learning_rate": 1.4630510170480006e-07, + "loss": 0.3773, + "step": 12935 + }, + { + "epoch": 4.877828054298643, + "grad_norm": 0.06240697837076471, + "learning_rate": 1.4540669279338836e-07, + "loss": 0.3778, + "step": 12936 + }, + { + "epoch": 4.878205128205128, + "grad_norm": 0.05915475376123434, + "learning_rate": 1.445110457699439e-07, + "loss": 0.3655, + "step": 12937 + }, + { + "epoch": 4.878582202111614, + "grad_norm": 0.06115309829380249, + "learning_rate": 1.4361816069653256e-07, + "loss": 0.3805, + "step": 12938 + }, + { + "epoch": 4.878959276018099, + "grad_norm": 0.05895594187074362, + "learning_rate": 1.427280376350293e-07, + "loss": 0.3778, + "step": 12939 + }, + { + "epoch": 4.879336349924586, + "grad_norm": 0.0602695618138386, + "learning_rate": 1.4184067664712253e-07, + "loss": 0.3717, + "step": 12940 + }, + { + "epoch": 4.87971342383107, + "grad_norm": 0.06073741552951578, + "learning_rate": 1.4095607779430088e-07, + "loss": 0.3714, + "step": 12941 + }, + { + "epoch": 4.880090497737557, + "grad_norm": 0.06098516673239015, + "learning_rate": 1.4007424113786638e-07, + "loss": 0.394, + "step": 12942 + }, + { + "epoch": 4.880467571644042, + "grad_norm": 0.060644181354461986, + "learning_rate": 1.3919516673893462e-07, + "loss": 0.3733, + "step": 12943 + }, + { + "epoch": 4.880844645550528, + "grad_norm": 0.05935646992547634, + "learning_rate": 1.3831885465841688e-07, + "loss": 0.3694, + "step": 12944 + }, + { + "epoch": 4.881221719457013, + "grad_norm": 0.061436624893114376, + "learning_rate": 1.3744530495704677e-07, + "loss": 0.3798, + "step": 12945 + }, + { + "epoch": 4.8815987933635, + "grad_norm": 0.06275205085199727, + "learning_rate": 1.3657451769536255e-07, + "loss": 0.3736, + "step": 12946 + }, + { + "epoch": 4.8819758672699844, + "grad_norm": 0.062398229436634076, + "learning_rate": 1.3570649293370265e-07, + "loss": 0.3831, + "step": 12947 + }, + { + "epoch": 4.882352941176471, + "grad_norm": 0.06163681042079879, + "learning_rate": 1.3484123073222332e-07, + "loss": 0.3759, + "step": 12948 + }, + { + "epoch": 4.882730015082956, + "grad_norm": 0.06034323241027124, + "learning_rate": 1.3397873115088557e-07, + "loss": 0.3791, + "step": 12949 + }, + { + "epoch": 4.883107088989442, + "grad_norm": 0.062033278389691957, + "learning_rate": 1.331189942494593e-07, + "loss": 0.3753, + "step": 12950 + }, + { + "epoch": 4.883484162895927, + "grad_norm": 0.06653913629183895, + "learning_rate": 1.3226202008752798e-07, + "loss": 0.3798, + "step": 12951 + }, + { + "epoch": 4.883861236802414, + "grad_norm": 0.060304109901625914, + "learning_rate": 1.3140780872447522e-07, + "loss": 0.3743, + "step": 12952 + }, + { + "epoch": 4.8842383107088985, + "grad_norm": 0.06180060284429538, + "learning_rate": 1.3055636021949814e-07, + "loss": 0.3727, + "step": 12953 + }, + { + "epoch": 4.884615384615385, + "grad_norm": 0.05958858042599691, + "learning_rate": 1.2970767463160284e-07, + "loss": 0.3751, + "step": 12954 + }, + { + "epoch": 4.88499245852187, + "grad_norm": 0.061082884198996816, + "learning_rate": 1.288617520195956e-07, + "loss": 0.3661, + "step": 12955 + }, + { + "epoch": 4.885369532428356, + "grad_norm": 0.059957749169272594, + "learning_rate": 1.2801859244210512e-07, + "loss": 0.3829, + "step": 12956 + }, + { + "epoch": 4.885746606334841, + "grad_norm": 0.0610796344686669, + "learning_rate": 1.2717819595756465e-07, + "loss": 0.3819, + "step": 12957 + }, + { + "epoch": 4.886123680241328, + "grad_norm": 0.06365498248912521, + "learning_rate": 1.2634056262420312e-07, + "loss": 0.3907, + "step": 12958 + }, + { + "epoch": 4.8865007541478125, + "grad_norm": 0.05970254071821968, + "learning_rate": 1.2550569250007637e-07, + "loss": 0.3763, + "step": 12959 + }, + { + "epoch": 4.886877828054299, + "grad_norm": 0.05959818419867367, + "learning_rate": 1.2467358564303144e-07, + "loss": 0.3768, + "step": 12960 + }, + { + "epoch": 4.887254901960784, + "grad_norm": 0.06267880319595488, + "learning_rate": 1.2384424211074218e-07, + "loss": 0.3694, + "step": 12961 + }, + { + "epoch": 4.88763197586727, + "grad_norm": 0.060173484096215, + "learning_rate": 1.2301766196067378e-07, + "loss": 0.3856, + "step": 12962 + }, + { + "epoch": 4.888009049773755, + "grad_norm": 0.06266760483348943, + "learning_rate": 1.2219384525010925e-07, + "loss": 0.3875, + "step": 12963 + }, + { + "epoch": 4.888386123680242, + "grad_norm": 0.060758485191281485, + "learning_rate": 1.2137279203614517e-07, + "loss": 0.3774, + "step": 12964 + }, + { + "epoch": 4.8887631975867265, + "grad_norm": 0.06085390424739995, + "learning_rate": 1.2055450237566936e-07, + "loss": 0.3878, + "step": 12965 + }, + { + "epoch": 4.889140271493213, + "grad_norm": 0.0599274247184929, + "learning_rate": 1.1973897632539645e-07, + "loss": 0.3782, + "step": 12966 + }, + { + "epoch": 4.889517345399698, + "grad_norm": 0.06283113381339164, + "learning_rate": 1.1892621394183679e-07, + "loss": 0.3828, + "step": 12967 + }, + { + "epoch": 4.889894419306184, + "grad_norm": 0.06389162845132078, + "learning_rate": 1.1811621528131867e-07, + "loss": 0.3754, + "step": 12968 + }, + { + "epoch": 4.890271493212669, + "grad_norm": 0.05960265112866246, + "learning_rate": 1.1730898039997052e-07, + "loss": 0.3712, + "step": 12969 + }, + { + "epoch": 4.890648567119156, + "grad_norm": 0.060635264067026164, + "learning_rate": 1.1650450935372982e-07, + "loss": 0.379, + "step": 12970 + }, + { + "epoch": 4.891025641025641, + "grad_norm": 0.06097658220520353, + "learning_rate": 1.1570280219835639e-07, + "loss": 0.3757, + "step": 12971 + }, + { + "epoch": 4.891402714932127, + "grad_norm": 0.062109299788908945, + "learning_rate": 1.1490385898940137e-07, + "loss": 0.3717, + "step": 12972 + }, + { + "epoch": 4.891779788838613, + "grad_norm": 0.06239030854890467, + "learning_rate": 1.1410767978222493e-07, + "loss": 0.3896, + "step": 12973 + }, + { + "epoch": 4.892156862745098, + "grad_norm": 0.06048136893941982, + "learning_rate": 1.1331426463200957e-07, + "loss": 0.365, + "step": 12974 + }, + { + "epoch": 4.892533936651584, + "grad_norm": 0.061222521826145834, + "learning_rate": 1.1252361359373798e-07, + "loss": 0.3777, + "step": 12975 + }, + { + "epoch": 4.89291101055807, + "grad_norm": 0.060759569335900175, + "learning_rate": 1.117357267222019e-07, + "loss": 0.379, + "step": 12976 + }, + { + "epoch": 4.893288084464555, + "grad_norm": 0.059602062722183755, + "learning_rate": 1.1095060407199321e-07, + "loss": 0.3762, + "step": 12977 + }, + { + "epoch": 4.893665158371041, + "grad_norm": 0.06668626099185394, + "learning_rate": 1.101682456975306e-07, + "loss": 0.3791, + "step": 12978 + }, + { + "epoch": 4.894042232277527, + "grad_norm": 0.06294020544688413, + "learning_rate": 1.0938865165302847e-07, + "loss": 0.3728, + "step": 12979 + }, + { + "epoch": 4.894419306184012, + "grad_norm": 0.06169374981506676, + "learning_rate": 1.0861182199250586e-07, + "loss": 0.383, + "step": 12980 + }, + { + "epoch": 4.894796380090498, + "grad_norm": 0.06278716440633643, + "learning_rate": 1.0783775676979968e-07, + "loss": 0.381, + "step": 12981 + }, + { + "epoch": 4.895173453996984, + "grad_norm": 0.06138869828567679, + "learning_rate": 1.0706645603855592e-07, + "loss": 0.3798, + "step": 12982 + }, + { + "epoch": 4.8955505279034695, + "grad_norm": 0.06088215063941225, + "learning_rate": 1.0629791985221627e-07, + "loss": 0.3742, + "step": 12983 + }, + { + "epoch": 4.895927601809955, + "grad_norm": 0.06195324160684908, + "learning_rate": 1.0553214826405367e-07, + "loss": 0.3653, + "step": 12984 + }, + { + "epoch": 4.896304675716441, + "grad_norm": 0.06223947202055671, + "learning_rate": 1.0476914132711901e-07, + "loss": 0.3682, + "step": 12985 + }, + { + "epoch": 4.8966817496229265, + "grad_norm": 0.06498912255953392, + "learning_rate": 1.040088990942989e-07, + "loss": 0.3926, + "step": 12986 + }, + { + "epoch": 4.897058823529412, + "grad_norm": 0.06134437626598712, + "learning_rate": 1.0325142161827561e-07, + "loss": 0.3761, + "step": 12987 + }, + { + "epoch": 4.897435897435898, + "grad_norm": 0.059565337228864944, + "learning_rate": 1.0249670895153607e-07, + "loss": 0.3739, + "step": 12988 + }, + { + "epoch": 4.8978129713423835, + "grad_norm": 0.05991292604887733, + "learning_rate": 1.0174476114638954e-07, + "loss": 0.3871, + "step": 12989 + }, + { + "epoch": 4.898190045248869, + "grad_norm": 0.060647392532477355, + "learning_rate": 1.00995578254941e-07, + "loss": 0.3722, + "step": 12990 + }, + { + "epoch": 4.898567119155355, + "grad_norm": 0.059760084380022886, + "learning_rate": 1.0024916032910892e-07, + "loss": 0.375, + "step": 12991 + }, + { + "epoch": 4.8989441930618405, + "grad_norm": 0.060929581862040415, + "learning_rate": 9.950550742061638e-08, + "loss": 0.3769, + "step": 12992 + }, + { + "epoch": 4.899321266968326, + "grad_norm": 0.06290155083044682, + "learning_rate": 9.876461958100436e-08, + "loss": 0.3807, + "step": 12993 + }, + { + "epoch": 4.899698340874812, + "grad_norm": 0.06058559958330754, + "learning_rate": 9.802649686160958e-08, + "loss": 0.3697, + "step": 12994 + }, + { + "epoch": 4.9000754147812975, + "grad_norm": 0.061566891742378324, + "learning_rate": 9.729113931358225e-08, + "loss": 0.3786, + "step": 12995 + }, + { + "epoch": 4.900452488687783, + "grad_norm": 0.06108249750543188, + "learning_rate": 9.655854698789047e-08, + "loss": 0.3721, + "step": 12996 + }, + { + "epoch": 4.900829562594269, + "grad_norm": 0.061997760663403165, + "learning_rate": 9.582871993529364e-08, + "loss": 0.3882, + "step": 12997 + }, + { + "epoch": 4.9012066365007545, + "grad_norm": 0.0645296795501361, + "learning_rate": 9.510165820637352e-08, + "loss": 0.375, + "step": 12998 + }, + { + "epoch": 4.90158371040724, + "grad_norm": 0.06357742302931982, + "learning_rate": 9.437736185151203e-08, + "loss": 0.3862, + "step": 12999 + }, + { + "epoch": 4.901960784313726, + "grad_norm": 0.05924217150274281, + "learning_rate": 9.365583092090458e-08, + "loss": 0.3763, + "step": 13000 + }, + { + "epoch": 4.9023378582202115, + "grad_norm": 0.05861313824810529, + "learning_rate": 9.293706546455117e-08, + "loss": 0.3784, + "step": 13001 + }, + { + "epoch": 4.902714932126697, + "grad_norm": 0.06132510528521585, + "learning_rate": 9.222106553226084e-08, + "loss": 0.3865, + "step": 13002 + }, + { + "epoch": 4.903092006033183, + "grad_norm": 0.05963805582138981, + "learning_rate": 9.150783117365169e-08, + "loss": 0.3697, + "step": 13003 + }, + { + "epoch": 4.9034690799396685, + "grad_norm": 0.058593215221798146, + "learning_rate": 9.079736243815085e-08, + "loss": 0.3829, + "step": 13004 + }, + { + "epoch": 4.903846153846154, + "grad_norm": 0.0642032530978967, + "learning_rate": 9.008965937499447e-08, + "loss": 0.3855, + "step": 13005 + }, + { + "epoch": 4.90422322775264, + "grad_norm": 0.06254208311226767, + "learning_rate": 8.938472203322335e-08, + "loss": 0.3859, + "step": 13006 + }, + { + "epoch": 4.9046003016591255, + "grad_norm": 0.06201490291973551, + "learning_rate": 8.868255046169172e-08, + "loss": 0.3805, + "step": 13007 + }, + { + "epoch": 4.904977375565611, + "grad_norm": 0.06287555180864478, + "learning_rate": 8.798314470905845e-08, + "loss": 0.3802, + "step": 13008 + }, + { + "epoch": 4.905354449472097, + "grad_norm": 0.06129303954214162, + "learning_rate": 8.728650482379142e-08, + "loss": 0.382, + "step": 13009 + }, + { + "epoch": 4.9057315233785825, + "grad_norm": 0.06143651646432337, + "learning_rate": 8.659263085416315e-08, + "loss": 0.3774, + "step": 13010 + }, + { + "epoch": 4.906108597285068, + "grad_norm": 0.060642032331985854, + "learning_rate": 8.590152284826847e-08, + "loss": 0.3806, + "step": 13011 + }, + { + "epoch": 4.906485671191554, + "grad_norm": 0.06174383732657246, + "learning_rate": 8.521318085398911e-08, + "loss": 0.3777, + "step": 13012 + }, + { + "epoch": 4.9068627450980395, + "grad_norm": 0.060969325822300476, + "learning_rate": 8.452760491903356e-08, + "loss": 0.3849, + "step": 13013 + }, + { + "epoch": 4.907239819004525, + "grad_norm": 0.060876243580369443, + "learning_rate": 8.384479509091492e-08, + "loss": 0.3871, + "step": 13014 + }, + { + "epoch": 4.907616892911011, + "grad_norm": 0.061103498051807884, + "learning_rate": 8.316475141693758e-08, + "loss": 0.3697, + "step": 13015 + }, + { + "epoch": 4.907993966817497, + "grad_norm": 0.05932796223811527, + "learning_rate": 8.248747394424161e-08, + "loss": 0.3833, + "step": 13016 + }, + { + "epoch": 4.908371040723982, + "grad_norm": 0.06106724203768966, + "learning_rate": 8.181296271975836e-08, + "loss": 0.384, + "step": 13017 + }, + { + "epoch": 4.908748114630468, + "grad_norm": 0.06159953742157264, + "learning_rate": 8.114121779022821e-08, + "loss": 0.3823, + "step": 13018 + }, + { + "epoch": 4.909125188536954, + "grad_norm": 0.060097855694842064, + "learning_rate": 8.04722392022006e-08, + "loss": 0.3852, + "step": 13019 + }, + { + "epoch": 4.909502262443439, + "grad_norm": 0.06056920357380981, + "learning_rate": 7.980602700204287e-08, + "loss": 0.3681, + "step": 13020 + }, + { + "epoch": 4.909879336349925, + "grad_norm": 0.060440306079438774, + "learning_rate": 7.914258123591367e-08, + "loss": 0.3745, + "step": 13021 + }, + { + "epoch": 4.910256410256411, + "grad_norm": 0.05989051272836247, + "learning_rate": 7.848190194979844e-08, + "loss": 0.372, + "step": 13022 + }, + { + "epoch": 4.910633484162896, + "grad_norm": 0.06077614379991945, + "learning_rate": 7.782398918947387e-08, + "loss": 0.3882, + "step": 13023 + }, + { + "epoch": 4.911010558069382, + "grad_norm": 0.06012251848324608, + "learning_rate": 7.716884300053906e-08, + "loss": 0.3951, + "step": 13024 + }, + { + "epoch": 4.911387631975868, + "grad_norm": 0.05955999138856424, + "learning_rate": 7.651646342838881e-08, + "loss": 0.3826, + "step": 13025 + }, + { + "epoch": 4.911764705882353, + "grad_norm": 0.0612651546922771, + "learning_rate": 7.586685051823584e-08, + "loss": 0.393, + "step": 13026 + }, + { + "epoch": 4.912141779788839, + "grad_norm": 0.0625395378703693, + "learning_rate": 7.522000431509746e-08, + "loss": 0.3825, + "step": 13027 + }, + { + "epoch": 4.912518853695325, + "grad_norm": 0.06268609556831164, + "learning_rate": 7.45759248638045e-08, + "loss": 0.3904, + "step": 13028 + }, + { + "epoch": 4.91289592760181, + "grad_norm": 0.06186885036791157, + "learning_rate": 7.393461220897902e-08, + "loss": 0.3782, + "step": 13029 + }, + { + "epoch": 4.913273001508296, + "grad_norm": 0.0615750467848537, + "learning_rate": 7.329606639507436e-08, + "loss": 0.3763, + "step": 13030 + }, + { + "epoch": 4.913650075414782, + "grad_norm": 0.05966377439599183, + "learning_rate": 7.266028746633514e-08, + "loss": 0.3793, + "step": 13031 + }, + { + "epoch": 4.914027149321267, + "grad_norm": 0.06182776784777951, + "learning_rate": 7.202727546682387e-08, + "loss": 0.3794, + "step": 13032 + }, + { + "epoch": 4.914404223227753, + "grad_norm": 0.06060963095249824, + "learning_rate": 7.139703044040769e-08, + "loss": 0.38, + "step": 13033 + }, + { + "epoch": 4.914781297134239, + "grad_norm": 0.06031088545286043, + "learning_rate": 7.076955243075833e-08, + "loss": 0.3788, + "step": 13034 + }, + { + "epoch": 4.915158371040724, + "grad_norm": 0.05926376314071955, + "learning_rate": 7.014484148136546e-08, + "loss": 0.3696, + "step": 13035 + }, + { + "epoch": 4.91553544494721, + "grad_norm": 0.06170087126281927, + "learning_rate": 6.952289763551445e-08, + "loss": 0.3804, + "step": 13036 + }, + { + "epoch": 4.915912518853696, + "grad_norm": 0.06229613675015601, + "learning_rate": 6.890372093630859e-08, + "loss": 0.3811, + "step": 13037 + }, + { + "epoch": 4.916289592760181, + "grad_norm": 0.06133768249482097, + "learning_rate": 6.828731142665579e-08, + "loss": 0.3811, + "step": 13038 + }, + { + "epoch": 4.916666666666667, + "grad_norm": 0.06034660318541198, + "learning_rate": 6.767366914927298e-08, + "loss": 0.3895, + "step": 13039 + }, + { + "epoch": 4.917043740573153, + "grad_norm": 0.059592675870952204, + "learning_rate": 6.706279414668615e-08, + "loss": 0.3779, + "step": 13040 + }, + { + "epoch": 4.917420814479638, + "grad_norm": 0.06015918836500664, + "learning_rate": 6.645468646122589e-08, + "loss": 0.3745, + "step": 13041 + }, + { + "epoch": 4.917797888386124, + "grad_norm": 0.06218148462544663, + "learning_rate": 6.58493461350318e-08, + "loss": 0.3862, + "step": 13042 + }, + { + "epoch": 4.91817496229261, + "grad_norm": 0.06084368958537087, + "learning_rate": 6.524677321006146e-08, + "loss": 0.3811, + "step": 13043 + }, + { + "epoch": 4.918552036199095, + "grad_norm": 0.06183055675490761, + "learning_rate": 6.46469677280681e-08, + "loss": 0.3797, + "step": 13044 + }, + { + "epoch": 4.918929110105581, + "grad_norm": 0.06332528631938793, + "learning_rate": 6.404992973061408e-08, + "loss": 0.3707, + "step": 13045 + }, + { + "epoch": 4.919306184012067, + "grad_norm": 0.0603459940237691, + "learning_rate": 6.345565925907515e-08, + "loss": 0.3726, + "step": 13046 + }, + { + "epoch": 4.919683257918552, + "grad_norm": 0.06001777684241117, + "learning_rate": 6.286415635464061e-08, + "loss": 0.3832, + "step": 13047 + }, + { + "epoch": 4.920060331825038, + "grad_norm": 0.060561690470929276, + "learning_rate": 6.227542105829098e-08, + "loss": 0.3896, + "step": 13048 + }, + { + "epoch": 4.920437405731524, + "grad_norm": 0.0611712535597029, + "learning_rate": 6.168945341083366e-08, + "loss": 0.3828, + "step": 13049 + }, + { + "epoch": 4.920814479638009, + "grad_norm": 0.06053868073364607, + "learning_rate": 6.11062534528717e-08, + "loss": 0.3797, + "step": 13050 + }, + { + "epoch": 4.921191553544495, + "grad_norm": 0.06053269365333679, + "learning_rate": 6.05258212248172e-08, + "loss": 0.3693, + "step": 13051 + }, + { + "epoch": 4.921568627450981, + "grad_norm": 0.05923921503081157, + "learning_rate": 5.994815676690024e-08, + "loss": 0.3853, + "step": 13052 + }, + { + "epoch": 4.921945701357466, + "grad_norm": 0.062480123086690015, + "learning_rate": 5.937326011914657e-08, + "loss": 0.3917, + "step": 13053 + }, + { + "epoch": 4.922322775263952, + "grad_norm": 0.06099730318834268, + "learning_rate": 5.8801131321404306e-08, + "loss": 0.3747, + "step": 13054 + }, + { + "epoch": 4.922699849170438, + "grad_norm": 0.059971059773129755, + "learning_rate": 5.823177041330841e-08, + "loss": 0.3722, + "step": 13055 + }, + { + "epoch": 4.923076923076923, + "grad_norm": 0.06088224376288011, + "learning_rate": 5.766517743432953e-08, + "loss": 0.3752, + "step": 13056 + }, + { + "epoch": 4.923453996983409, + "grad_norm": 0.06087955938177899, + "learning_rate": 5.7101352423720724e-08, + "loss": 0.3808, + "step": 13057 + }, + { + "epoch": 4.923831070889895, + "grad_norm": 0.06190632529852801, + "learning_rate": 5.654029542055739e-08, + "loss": 0.3826, + "step": 13058 + }, + { + "epoch": 4.92420814479638, + "grad_norm": 0.06196807694191833, + "learning_rate": 5.598200646372398e-08, + "loss": 0.39, + "step": 13059 + }, + { + "epoch": 4.924585218702866, + "grad_norm": 0.06117514598972959, + "learning_rate": 5.542648559190955e-08, + "loss": 0.3822, + "step": 13060 + }, + { + "epoch": 4.924962292609352, + "grad_norm": 0.06041508579823845, + "learning_rate": 5.487373284360775e-08, + "loss": 0.3796, + "step": 13061 + }, + { + "epoch": 4.925339366515837, + "grad_norm": 0.061902699566673075, + "learning_rate": 5.432374825712128e-08, + "loss": 0.3888, + "step": 13062 + }, + { + "epoch": 4.925716440422323, + "grad_norm": 0.060423559950118534, + "learning_rate": 5.377653187057075e-08, + "loss": 0.3833, + "step": 13063 + }, + { + "epoch": 4.926093514328809, + "grad_norm": 0.05884634742383607, + "learning_rate": 5.3232083721872493e-08, + "loss": 0.3758, + "step": 13064 + }, + { + "epoch": 4.926470588235294, + "grad_norm": 0.061074403817104626, + "learning_rate": 5.2690403848760785e-08, + "loss": 0.374, + "step": 13065 + }, + { + "epoch": 4.92684766214178, + "grad_norm": 0.06039504704576671, + "learning_rate": 5.2151492288770035e-08, + "loss": 0.3832, + "step": 13066 + }, + { + "epoch": 4.927224736048266, + "grad_norm": 0.060872904467041765, + "learning_rate": 5.16153490792437e-08, + "loss": 0.3774, + "step": 13067 + }, + { + "epoch": 4.927601809954751, + "grad_norm": 0.06191712329412054, + "learning_rate": 5.108197425734318e-08, + "loss": 0.3908, + "step": 13068 + }, + { + "epoch": 4.927978883861237, + "grad_norm": 0.06242813794802749, + "learning_rate": 5.055136786002557e-08, + "loss": 0.3799, + "step": 13069 + }, + { + "epoch": 4.928355957767723, + "grad_norm": 0.06209610899656442, + "learning_rate": 5.002352992406589e-08, + "loss": 0.3928, + "step": 13070 + }, + { + "epoch": 4.928733031674208, + "grad_norm": 0.062092776913603036, + "learning_rate": 4.94984604860349e-08, + "loss": 0.378, + "step": 13071 + }, + { + "epoch": 4.929110105580694, + "grad_norm": 0.06257508704724785, + "learning_rate": 4.897615958233015e-08, + "loss": 0.3851, + "step": 13072 + }, + { + "epoch": 4.92948717948718, + "grad_norm": 0.060480967233579146, + "learning_rate": 4.8456627249136026e-08, + "loss": 0.3852, + "step": 13073 + }, + { + "epoch": 4.929864253393665, + "grad_norm": 0.061847442901884375, + "learning_rate": 4.7939863522463716e-08, + "loss": 0.3891, + "step": 13074 + }, + { + "epoch": 4.930241327300151, + "grad_norm": 0.060235365393053145, + "learning_rate": 4.742586843812014e-08, + "loss": 0.3781, + "step": 13075 + }, + { + "epoch": 4.930618401206637, + "grad_norm": 0.05876589601431315, + "learning_rate": 4.691464203172569e-08, + "loss": 0.3677, + "step": 13076 + }, + { + "epoch": 4.930995475113122, + "grad_norm": 0.061757103339261224, + "learning_rate": 4.6406184338709806e-08, + "loss": 0.383, + "step": 13077 + }, + { + "epoch": 4.931372549019608, + "grad_norm": 0.06162294820036236, + "learning_rate": 4.590049539430652e-08, + "loss": 0.3772, + "step": 13078 + }, + { + "epoch": 4.931749622926094, + "grad_norm": 0.06123507379262996, + "learning_rate": 4.539757523355892e-08, + "loss": 0.3624, + "step": 13079 + }, + { + "epoch": 4.932126696832579, + "grad_norm": 0.062390721758297435, + "learning_rate": 4.4897423891319126e-08, + "loss": 0.3741, + "step": 13080 + }, + { + "epoch": 4.932503770739065, + "grad_norm": 0.06340651668104454, + "learning_rate": 4.4400041402248294e-08, + "loss": 0.386, + "step": 13081 + }, + { + "epoch": 4.932880844645551, + "grad_norm": 0.061938283167602366, + "learning_rate": 4.3905427800812195e-08, + "loss": 0.371, + "step": 13082 + }, + { + "epoch": 4.933257918552036, + "grad_norm": 0.060163205604770176, + "learning_rate": 4.341358312129007e-08, + "loss": 0.3733, + "step": 13083 + }, + { + "epoch": 4.933634992458522, + "grad_norm": 0.06237049540334635, + "learning_rate": 4.2924507397765766e-08, + "loss": 0.3891, + "step": 13084 + }, + { + "epoch": 4.934012066365008, + "grad_norm": 0.05924910197473463, + "learning_rate": 4.243820066413218e-08, + "loss": 0.3717, + "step": 13085 + }, + { + "epoch": 4.9343891402714934, + "grad_norm": 0.06021164122008437, + "learning_rate": 4.195466295408679e-08, + "loss": 0.3785, + "step": 13086 + }, + { + "epoch": 4.934766214177979, + "grad_norm": 0.06210638283338799, + "learning_rate": 4.147389430114057e-08, + "loss": 0.3827, + "step": 13087 + }, + { + "epoch": 4.935143288084465, + "grad_norm": 0.060960220659600856, + "learning_rate": 4.099589473861354e-08, + "loss": 0.3753, + "step": 13088 + }, + { + "epoch": 4.9355203619909505, + "grad_norm": 0.06147151270787934, + "learning_rate": 4.052066429962587e-08, + "loss": 0.3761, + "step": 13089 + }, + { + "epoch": 4.935897435897436, + "grad_norm": 0.06175837371211539, + "learning_rate": 4.004820301711121e-08, + "loss": 0.3756, + "step": 13090 + }, + { + "epoch": 4.936274509803922, + "grad_norm": 0.059910162264159406, + "learning_rate": 3.9578510923812265e-08, + "loss": 0.3777, + "step": 13091 + }, + { + "epoch": 4.9366515837104075, + "grad_norm": 0.05868910992885308, + "learning_rate": 3.9111588052276327e-08, + "loss": 0.3833, + "step": 13092 + }, + { + "epoch": 4.937028657616893, + "grad_norm": 0.060443840553590554, + "learning_rate": 3.864743443486418e-08, + "loss": 0.3703, + "step": 13093 + }, + { + "epoch": 4.937405731523379, + "grad_norm": 0.0599343643598136, + "learning_rate": 3.818605010373677e-08, + "loss": 0.3761, + "step": 13094 + }, + { + "epoch": 4.9377828054298645, + "grad_norm": 0.05913057693376895, + "learning_rate": 3.772743509087296e-08, + "loss": 0.3776, + "step": 13095 + }, + { + "epoch": 4.93815987933635, + "grad_norm": 0.06081573342377783, + "learning_rate": 3.7271589428047315e-08, + "loss": 0.3767, + "step": 13096 + }, + { + "epoch": 4.938536953242836, + "grad_norm": 0.0601563906569475, + "learning_rate": 3.6818513146856806e-08, + "loss": 0.3895, + "step": 13097 + }, + { + "epoch": 4.9389140271493215, + "grad_norm": 0.06137508319246621, + "learning_rate": 3.636820627869853e-08, + "loss": 0.3796, + "step": 13098 + }, + { + "epoch": 4.939291101055807, + "grad_norm": 0.058998885259656866, + "learning_rate": 3.592066885476975e-08, + "loss": 0.3726, + "step": 13099 + }, + { + "epoch": 4.939668174962293, + "grad_norm": 0.06240138378054263, + "learning_rate": 3.547590090609454e-08, + "loss": 0.3799, + "step": 13100 + }, + { + "epoch": 4.9400452488687785, + "grad_norm": 0.060451733912247535, + "learning_rate": 3.503390246348826e-08, + "loss": 0.3731, + "step": 13101 + }, + { + "epoch": 4.940422322775264, + "grad_norm": 0.0605940827485194, + "learning_rate": 3.459467355758861e-08, + "loss": 0.3791, + "step": 13102 + }, + { + "epoch": 4.94079939668175, + "grad_norm": 0.058951070877394665, + "learning_rate": 3.415821421882459e-08, + "loss": 0.3673, + "step": 13103 + }, + { + "epoch": 4.9411764705882355, + "grad_norm": 0.058428403887209844, + "learning_rate": 3.3724524477447564e-08, + "loss": 0.3852, + "step": 13104 + }, + { + "epoch": 4.941553544494721, + "grad_norm": 0.060717152699247254, + "learning_rate": 3.329360436350903e-08, + "loss": 0.3846, + "step": 13105 + }, + { + "epoch": 4.941930618401207, + "grad_norm": 0.05998864852332163, + "learning_rate": 3.286545390687401e-08, + "loss": 0.3776, + "step": 13106 + }, + { + "epoch": 4.9423076923076925, + "grad_norm": 0.05983208280110172, + "learning_rate": 3.244007313721653e-08, + "loss": 0.388, + "step": 13107 + }, + { + "epoch": 4.942684766214178, + "grad_norm": 0.058715931637570294, + "learning_rate": 3.2017462084006355e-08, + "loss": 0.3833, + "step": 13108 + }, + { + "epoch": 4.943061840120664, + "grad_norm": 0.062126820216479625, + "learning_rate": 3.159762077653561e-08, + "loss": 0.3817, + "step": 13109 + }, + { + "epoch": 4.9434389140271495, + "grad_norm": 0.06079992622401111, + "learning_rate": 3.118054924389657e-08, + "loss": 0.3799, + "step": 13110 + }, + { + "epoch": 4.943815987933635, + "grad_norm": 0.061037914326394896, + "learning_rate": 3.076624751499502e-08, + "loss": 0.3828, + "step": 13111 + }, + { + "epoch": 4.944193061840121, + "grad_norm": 0.060019872646715926, + "learning_rate": 3.035471561854131e-08, + "loss": 0.3694, + "step": 13112 + }, + { + "epoch": 4.9445701357466065, + "grad_norm": 0.06014757483372469, + "learning_rate": 2.9945953583050415e-08, + "loss": 0.3869, + "step": 13113 + }, + { + "epoch": 4.944947209653092, + "grad_norm": 0.06289813932310154, + "learning_rate": 2.9539961436850783e-08, + "loss": 0.3784, + "step": 13114 + }, + { + "epoch": 4.945324283559578, + "grad_norm": 0.060662569233608896, + "learning_rate": 2.913673920807991e-08, + "loss": 0.3768, + "step": 13115 + }, + { + "epoch": 4.9457013574660635, + "grad_norm": 0.060354231910726294, + "learning_rate": 2.873628692467545e-08, + "loss": 0.3755, + "step": 13116 + }, + { + "epoch": 4.946078431372549, + "grad_norm": 0.060649062000315106, + "learning_rate": 2.8338604614397412e-08, + "loss": 0.3738, + "step": 13117 + }, + { + "epoch": 4.946455505279035, + "grad_norm": 0.06112244501121036, + "learning_rate": 2.79436923047971e-08, + "loss": 0.3681, + "step": 13118 + }, + { + "epoch": 4.9468325791855206, + "grad_norm": 0.061247263142236524, + "learning_rate": 2.755155002323928e-08, + "loss": 0.3801, + "step": 13119 + }, + { + "epoch": 4.947209653092006, + "grad_norm": 0.05904507176886716, + "learning_rate": 2.7162177796911105e-08, + "loss": 0.3784, + "step": 13120 + }, + { + "epoch": 4.947586726998492, + "grad_norm": 0.060325883850536395, + "learning_rate": 2.6775575652782105e-08, + "loss": 0.3783, + "step": 13121 + }, + { + "epoch": 4.947963800904978, + "grad_norm": 0.059914749469769595, + "learning_rate": 2.6391743617653067e-08, + "loss": 0.3805, + "step": 13122 + }, + { + "epoch": 4.948340874811463, + "grad_norm": 0.06063248132044652, + "learning_rate": 2.601068171811605e-08, + "loss": 0.3854, + "step": 13123 + }, + { + "epoch": 4.948717948717949, + "grad_norm": 0.0586370204128017, + "learning_rate": 2.563238998058548e-08, + "loss": 0.3716, + "step": 13124 + }, + { + "epoch": 4.949095022624435, + "grad_norm": 0.06023070592061471, + "learning_rate": 2.5256868431271507e-08, + "loss": 0.3771, + "step": 13125 + }, + { + "epoch": 4.94947209653092, + "grad_norm": 0.05961722529854077, + "learning_rate": 2.4884117096197757e-08, + "loss": 0.3733, + "step": 13126 + }, + { + "epoch": 4.949849170437406, + "grad_norm": 0.059569184221623044, + "learning_rate": 2.4514136001201338e-08, + "loss": 0.382, + "step": 13127 + }, + { + "epoch": 4.950226244343892, + "grad_norm": 0.061145794533663145, + "learning_rate": 2.414692517191508e-08, + "loss": 0.3766, + "step": 13128 + }, + { + "epoch": 4.950603318250377, + "grad_norm": 0.05888434520571922, + "learning_rate": 2.3782484633785295e-08, + "loss": 0.3745, + "step": 13129 + }, + { + "epoch": 4.950980392156863, + "grad_norm": 0.06151454684183804, + "learning_rate": 2.342081441207178e-08, + "loss": 0.3776, + "step": 13130 + }, + { + "epoch": 4.951357466063349, + "grad_norm": 0.06022089986750238, + "learning_rate": 2.3061914531838925e-08, + "loss": 0.3683, + "step": 13131 + }, + { + "epoch": 4.951734539969834, + "grad_norm": 0.05963480366994346, + "learning_rate": 2.2705785017955728e-08, + "loss": 0.3724, + "step": 13132 + }, + { + "epoch": 4.95211161387632, + "grad_norm": 0.06262831341719824, + "learning_rate": 2.2352425895100228e-08, + "loss": 0.3666, + "step": 13133 + }, + { + "epoch": 4.952488687782806, + "grad_norm": 0.05872264137102727, + "learning_rate": 2.20018371877595e-08, + "loss": 0.3826, + "step": 13134 + }, + { + "epoch": 4.952865761689291, + "grad_norm": 0.06064862207159234, + "learning_rate": 2.165401892023411e-08, + "loss": 0.3828, + "step": 13135 + }, + { + "epoch": 4.953242835595777, + "grad_norm": 0.05942819404697303, + "learning_rate": 2.1308971116624777e-08, + "loss": 0.3786, + "step": 13136 + }, + { + "epoch": 4.953619909502263, + "grad_norm": 0.06542873265824271, + "learning_rate": 2.0966693800841266e-08, + "loss": 0.375, + "step": 13137 + }, + { + "epoch": 4.953996983408748, + "grad_norm": 0.061027288123210253, + "learning_rate": 2.062718699660238e-08, + "loss": 0.3885, + "step": 13138 + }, + { + "epoch": 4.954374057315234, + "grad_norm": 0.06005818724809713, + "learning_rate": 2.02904507274404e-08, + "loss": 0.3733, + "step": 13139 + }, + { + "epoch": 4.95475113122172, + "grad_norm": 0.06211053742781201, + "learning_rate": 1.995648501668779e-08, + "loss": 0.3871, + "step": 13140 + }, + { + "epoch": 4.955128205128205, + "grad_norm": 0.06013805545287572, + "learning_rate": 1.9625289887490462e-08, + "loss": 0.3609, + "step": 13141 + }, + { + "epoch": 4.955505279034691, + "grad_norm": 0.059246796357484484, + "learning_rate": 1.9296865362794515e-08, + "loss": 0.3753, + "step": 13142 + }, + { + "epoch": 4.955882352941177, + "grad_norm": 0.06240513454590453, + "learning_rate": 1.897121146536396e-08, + "loss": 0.3719, + "step": 13143 + }, + { + "epoch": 4.956259426847662, + "grad_norm": 0.0603978560994562, + "learning_rate": 1.8648328217767407e-08, + "loss": 0.3844, + "step": 13144 + }, + { + "epoch": 4.956636500754148, + "grad_norm": 0.05857133833905266, + "learning_rate": 1.8328215642378077e-08, + "loss": 0.3829, + "step": 13145 + }, + { + "epoch": 4.957013574660634, + "grad_norm": 0.06116586229306116, + "learning_rate": 1.8010873761378224e-08, + "loss": 0.3755, + "step": 13146 + }, + { + "epoch": 4.957390648567119, + "grad_norm": 0.05975453996995025, + "learning_rate": 1.7696302596759142e-08, + "loss": 0.3805, + "step": 13147 + }, + { + "epoch": 4.957767722473605, + "grad_norm": 0.059876425036963075, + "learning_rate": 1.7384502170325613e-08, + "loss": 0.367, + "step": 13148 + }, + { + "epoch": 4.958144796380091, + "grad_norm": 0.0628348785498449, + "learning_rate": 1.7075472503678137e-08, + "loss": 0.3925, + "step": 13149 + }, + { + "epoch": 4.958521870286576, + "grad_norm": 0.061328120737419105, + "learning_rate": 1.6769213618239575e-08, + "loss": 0.3748, + "step": 13150 + }, + { + "epoch": 4.958898944193062, + "grad_norm": 0.05979045540427696, + "learning_rate": 1.6465725535224074e-08, + "loss": 0.3716, + "step": 13151 + }, + { + "epoch": 4.959276018099548, + "grad_norm": 0.059204406837663585, + "learning_rate": 1.616500827567258e-08, + "loss": 0.3694, + "step": 13152 + }, + { + "epoch": 4.959653092006033, + "grad_norm": 0.0622736186704046, + "learning_rate": 1.586706186042175e-08, + "loss": 0.3789, + "step": 13153 + }, + { + "epoch": 4.960030165912519, + "grad_norm": 0.060708752751297834, + "learning_rate": 1.557188631011286e-08, + "loss": 0.3885, + "step": 13154 + }, + { + "epoch": 4.960407239819005, + "grad_norm": 0.05926624784925409, + "learning_rate": 1.527948164520954e-08, + "loss": 0.3755, + "step": 13155 + }, + { + "epoch": 4.96078431372549, + "grad_norm": 0.060254925191246714, + "learning_rate": 1.4989847885971133e-08, + "loss": 0.3821, + "step": 13156 + }, + { + "epoch": 4.961161387631976, + "grad_norm": 0.05895092946156568, + "learning_rate": 1.4702985052470475e-08, + "loss": 0.3789, + "step": 13157 + }, + { + "epoch": 4.961538461538462, + "grad_norm": 0.06080263149142599, + "learning_rate": 1.4418893164585002e-08, + "loss": 0.3755, + "step": 13158 + }, + { + "epoch": 4.961915535444947, + "grad_norm": 0.06068550139852834, + "learning_rate": 1.4137572242001186e-08, + "loss": 0.3755, + "step": 13159 + }, + { + "epoch": 4.962292609351433, + "grad_norm": 0.0637507277569963, + "learning_rate": 1.3859022304218982e-08, + "loss": 0.3833, + "step": 13160 + }, + { + "epoch": 4.962669683257919, + "grad_norm": 0.06180903977582689, + "learning_rate": 1.3583243370538513e-08, + "loss": 0.3711, + "step": 13161 + }, + { + "epoch": 4.963046757164404, + "grad_norm": 0.06106886968886517, + "learning_rate": 1.3310235460068932e-08, + "loss": 0.3809, + "step": 13162 + }, + { + "epoch": 4.96342383107089, + "grad_norm": 0.0589654169491015, + "learning_rate": 1.3039998591732883e-08, + "loss": 0.3779, + "step": 13163 + }, + { + "epoch": 4.963800904977376, + "grad_norm": 0.06056830108840971, + "learning_rate": 1.2772532784257608e-08, + "loss": 0.3795, + "step": 13164 + }, + { + "epoch": 4.964177978883861, + "grad_norm": 0.06076513508000769, + "learning_rate": 1.2507838056174949e-08, + "loss": 0.3761, + "step": 13165 + }, + { + "epoch": 4.964555052790347, + "grad_norm": 0.06093532804056453, + "learning_rate": 1.224591442583467e-08, + "loss": 0.3991, + "step": 13166 + }, + { + "epoch": 4.964932126696833, + "grad_norm": 0.060551368653369596, + "learning_rate": 1.198676191137782e-08, + "loss": 0.3749, + "step": 13167 + }, + { + "epoch": 4.965309200603318, + "grad_norm": 0.061745902278241595, + "learning_rate": 1.1730380530772246e-08, + "loss": 0.3807, + "step": 13168 + }, + { + "epoch": 4.965686274509804, + "grad_norm": 0.058417039674130034, + "learning_rate": 1.1476770301781514e-08, + "loss": 0.3747, + "step": 13169 + }, + { + "epoch": 4.96606334841629, + "grad_norm": 0.062111116133189426, + "learning_rate": 1.1225931241982679e-08, + "loss": 0.3681, + "step": 13170 + }, + { + "epoch": 4.966440422322775, + "grad_norm": 0.05902460523412065, + "learning_rate": 1.097786336875295e-08, + "loss": 0.376, + "step": 13171 + }, + { + "epoch": 4.966817496229261, + "grad_norm": 0.06024515144771844, + "learning_rate": 1.0732566699291902e-08, + "loss": 0.3712, + "step": 13172 + }, + { + "epoch": 4.967194570135747, + "grad_norm": 0.06162022332979525, + "learning_rate": 1.0490041250590388e-08, + "loss": 0.3748, + "step": 13173 + }, + { + "epoch": 4.967571644042232, + "grad_norm": 0.060467057370733476, + "learning_rate": 1.0250287039457185e-08, + "loss": 0.3719, + "step": 13174 + }, + { + "epoch": 4.967948717948718, + "grad_norm": 0.05959329516610389, + "learning_rate": 1.001330408251011e-08, + "loss": 0.3932, + "step": 13175 + }, + { + "epoch": 4.968325791855204, + "grad_norm": 0.059469520432479085, + "learning_rate": 9.779092396171586e-09, + "loss": 0.385, + "step": 13176 + }, + { + "epoch": 4.968702865761689, + "grad_norm": 0.059308959890714205, + "learning_rate": 9.547651996668628e-09, + "loss": 0.3695, + "step": 13177 + }, + { + "epoch": 4.969079939668175, + "grad_norm": 0.06028909716839308, + "learning_rate": 9.31898290004174e-09, + "loss": 0.3916, + "step": 13178 + }, + { + "epoch": 4.969457013574661, + "grad_norm": 0.06042532028056355, + "learning_rate": 9.093085122136025e-09, + "loss": 0.3864, + "step": 13179 + }, + { + "epoch": 4.969834087481146, + "grad_norm": 0.06131159546664093, + "learning_rate": 8.869958678610068e-09, + "loss": 0.3841, + "step": 13180 + }, + { + "epoch": 4.970211161387632, + "grad_norm": 0.06096970970494818, + "learning_rate": 8.649603584922617e-09, + "loss": 0.3764, + "step": 13181 + }, + { + "epoch": 4.970588235294118, + "grad_norm": 0.06151794173276826, + "learning_rate": 8.432019856345896e-09, + "loss": 0.4001, + "step": 13182 + }, + { + "epoch": 4.970965309200603, + "grad_norm": 0.060784788001937415, + "learning_rate": 8.217207507961178e-09, + "loss": 0.3938, + "step": 13183 + }, + { + "epoch": 4.971342383107089, + "grad_norm": 0.06345724939838755, + "learning_rate": 8.00516655464545e-09, + "loss": 0.3862, + "step": 13184 + }, + { + "epoch": 4.971719457013575, + "grad_norm": 0.05982858801862618, + "learning_rate": 7.795897011102505e-09, + "loss": 0.3828, + "step": 13185 + }, + { + "epoch": 4.97209653092006, + "grad_norm": 0.061065106398701885, + "learning_rate": 7.589398891827415e-09, + "loss": 0.3755, + "step": 13186 + }, + { + "epoch": 4.972473604826546, + "grad_norm": 0.05857470888713415, + "learning_rate": 7.385672211137618e-09, + "loss": 0.3731, + "step": 13187 + }, + { + "epoch": 4.972850678733032, + "grad_norm": 0.060607681697858476, + "learning_rate": 7.184716983146267e-09, + "loss": 0.3761, + "step": 13188 + }, + { + "epoch": 4.973227752639517, + "grad_norm": 0.06171233000013234, + "learning_rate": 6.986533221780001e-09, + "loss": 0.3702, + "step": 13189 + }, + { + "epoch": 4.973604826546003, + "grad_norm": 0.05985521722618749, + "learning_rate": 6.791120940774498e-09, + "loss": 0.3753, + "step": 13190 + }, + { + "epoch": 4.973981900452489, + "grad_norm": 0.0610493678485976, + "learning_rate": 6.5984801536655986e-09, + "loss": 0.3783, + "step": 13191 + }, + { + "epoch": 4.9743589743589745, + "grad_norm": 0.058784153454236966, + "learning_rate": 6.4086108738115045e-09, + "loss": 0.3665, + "step": 13192 + }, + { + "epoch": 4.97473604826546, + "grad_norm": 0.060571890354658264, + "learning_rate": 6.22151311436614e-09, + "loss": 0.3794, + "step": 13193 + }, + { + "epoch": 4.975113122171946, + "grad_norm": 0.0607088385170689, + "learning_rate": 6.037186888296909e-09, + "loss": 0.3814, + "step": 13194 + }, + { + "epoch": 4.9754901960784315, + "grad_norm": 0.058545813517264865, + "learning_rate": 5.8556322083758165e-09, + "loss": 0.3595, + "step": 13195 + }, + { + "epoch": 4.975867269984917, + "grad_norm": 0.05881526563825401, + "learning_rate": 5.676849087183911e-09, + "loss": 0.3668, + "step": 13196 + }, + { + "epoch": 4.976244343891403, + "grad_norm": 0.0637425085179292, + "learning_rate": 5.500837537111281e-09, + "loss": 0.3888, + "step": 13197 + }, + { + "epoch": 4.9766214177978885, + "grad_norm": 0.059383193602452894, + "learning_rate": 5.3275975703526166e-09, + "loss": 0.3771, + "step": 13198 + }, + { + "epoch": 4.976998491704374, + "grad_norm": 0.06221727369782548, + "learning_rate": 5.157129198920529e-09, + "loss": 0.374, + "step": 13199 + }, + { + "epoch": 4.97737556561086, + "grad_norm": 0.06076453932154122, + "learning_rate": 4.989432434623353e-09, + "loss": 0.3709, + "step": 13200 + }, + { + "epoch": 4.9777526395173455, + "grad_norm": 0.06071347558521433, + "learning_rate": 4.824507289082903e-09, + "loss": 0.3771, + "step": 13201 + }, + { + "epoch": 4.978129713423831, + "grad_norm": 0.06024266859385876, + "learning_rate": 4.662353773725592e-09, + "loss": 0.3927, + "step": 13202 + }, + { + "epoch": 4.978506787330317, + "grad_norm": 0.0613620103642454, + "learning_rate": 4.50297189979132e-09, + "loss": 0.3708, + "step": 13203 + }, + { + "epoch": 4.9788838612368025, + "grad_norm": 0.0608420672253551, + "learning_rate": 4.346361678329025e-09, + "loss": 0.3778, + "step": 13204 + }, + { + "epoch": 4.979260935143288, + "grad_norm": 0.05940877152156493, + "learning_rate": 4.192523120187808e-09, + "loss": 0.3783, + "step": 13205 + }, + { + "epoch": 4.979638009049774, + "grad_norm": 0.060178450362061146, + "learning_rate": 4.0414562360258095e-09, + "loss": 0.3749, + "step": 13206 + }, + { + "epoch": 4.9800150829562595, + "grad_norm": 0.060456506513527804, + "learning_rate": 3.893161036314652e-09, + "loss": 0.3841, + "step": 13207 + }, + { + "epoch": 4.980392156862745, + "grad_norm": 0.061475710947213845, + "learning_rate": 3.747637531330561e-09, + "loss": 0.3668, + "step": 13208 + }, + { + "epoch": 4.980769230769231, + "grad_norm": 0.06162884516841655, + "learning_rate": 3.6048857311588024e-09, + "loss": 0.3783, + "step": 13209 + }, + { + "epoch": 4.9811463046757165, + "grad_norm": 0.05984006771462693, + "learning_rate": 3.4649056456936837e-09, + "loss": 0.3783, + "step": 13210 + }, + { + "epoch": 4.981523378582202, + "grad_norm": 0.061102217135047425, + "learning_rate": 3.3276972846341126e-09, + "loss": 0.3796, + "step": 13211 + }, + { + "epoch": 4.981900452488688, + "grad_norm": 0.06036059573231087, + "learning_rate": 3.193260657483599e-09, + "loss": 0.3714, + "step": 13212 + }, + { + "epoch": 4.9822775263951735, + "grad_norm": 0.06239420131782545, + "learning_rate": 3.0615957735680157e-09, + "loss": 0.3733, + "step": 13213 + }, + { + "epoch": 4.982654600301659, + "grad_norm": 0.06193513958317255, + "learning_rate": 2.9327026420045144e-09, + "loss": 0.38, + "step": 13214 + }, + { + "epoch": 4.983031674208145, + "grad_norm": 0.06120252569671836, + "learning_rate": 2.8065812717281703e-09, + "loss": 0.3773, + "step": 13215 + }, + { + "epoch": 4.9834087481146305, + "grad_norm": 0.06183938252119402, + "learning_rate": 2.683231671474218e-09, + "loss": 0.3777, + "step": 13216 + }, + { + "epoch": 4.983785822021116, + "grad_norm": 0.06104202031393948, + "learning_rate": 2.562653849800256e-09, + "loss": 0.3795, + "step": 13217 + }, + { + "epoch": 4.984162895927602, + "grad_norm": 0.06051680455692009, + "learning_rate": 2.4448478150551624e-09, + "loss": 0.3803, + "step": 13218 + }, + { + "epoch": 4.9845399698340875, + "grad_norm": 0.061257014618590684, + "learning_rate": 2.329813575401296e-09, + "loss": 0.382, + "step": 13219 + }, + { + "epoch": 4.984917043740573, + "grad_norm": 0.06298849531988936, + "learning_rate": 2.2175511388145e-09, + "loss": 0.3716, + "step": 13220 + }, + { + "epoch": 4.985294117647059, + "grad_norm": 0.060031573209312315, + "learning_rate": 2.1080605130752162e-09, + "loss": 0.3904, + "step": 13221 + }, + { + "epoch": 4.9856711915535445, + "grad_norm": 0.06188482472155701, + "learning_rate": 2.0013417057684893e-09, + "loss": 0.3844, + "step": 13222 + }, + { + "epoch": 4.98604826546003, + "grad_norm": 0.05920128404196606, + "learning_rate": 1.8973947242928446e-09, + "loss": 0.3697, + "step": 13223 + }, + { + "epoch": 4.986425339366516, + "grad_norm": 0.060737135573242874, + "learning_rate": 1.7962195758469691e-09, + "loss": 0.3753, + "step": 13224 + }, + { + "epoch": 4.9868024132730016, + "grad_norm": 0.060709120448153416, + "learning_rate": 1.697816267443031e-09, + "loss": 0.3733, + "step": 13225 + }, + { + "epoch": 4.987179487179487, + "grad_norm": 0.060374409321870465, + "learning_rate": 1.6021848059066813e-09, + "loss": 0.3747, + "step": 13226 + }, + { + "epoch": 4.987556561085973, + "grad_norm": 0.05860393825049868, + "learning_rate": 1.5093251978548495e-09, + "loss": 0.3609, + "step": 13227 + }, + { + "epoch": 4.987933634992459, + "grad_norm": 0.05898295280383007, + "learning_rate": 1.41923744973127e-09, + "loss": 0.3812, + "step": 13228 + }, + { + "epoch": 4.988310708898944, + "grad_norm": 0.061416707870507965, + "learning_rate": 1.331921567775396e-09, + "loss": 0.3729, + "step": 13229 + }, + { + "epoch": 4.98868778280543, + "grad_norm": 0.0585541051890743, + "learning_rate": 1.2473775580401636e-09, + "loss": 0.3744, + "step": 13230 + }, + { + "epoch": 4.989064856711916, + "grad_norm": 0.0609010927968747, + "learning_rate": 1.1656054263831095e-09, + "loss": 0.374, + "step": 13231 + }, + { + "epoch": 4.989441930618401, + "grad_norm": 0.06286002670842014, + "learning_rate": 1.0866051784663712e-09, + "loss": 0.3744, + "step": 13232 + }, + { + "epoch": 4.989819004524887, + "grad_norm": 0.059031406609269615, + "learning_rate": 1.01037681977445e-09, + "loss": 0.3702, + "step": 13233 + }, + { + "epoch": 4.990196078431373, + "grad_norm": 0.06006122205156335, + "learning_rate": 9.369203555786854e-10, + "loss": 0.3947, + "step": 13234 + }, + { + "epoch": 4.990573152337858, + "grad_norm": 0.05894179933807814, + "learning_rate": 8.662357909772212e-10, + "loss": 0.3706, + "step": 13235 + }, + { + "epoch": 4.990950226244344, + "grad_norm": 0.06159389184885852, + "learning_rate": 7.983231308683615e-10, + "loss": 0.372, + "step": 13236 + }, + { + "epoch": 4.99132730015083, + "grad_norm": 0.05965010013071588, + "learning_rate": 7.331823799550109e-10, + "loss": 0.3852, + "step": 13237 + }, + { + "epoch": 4.991704374057315, + "grad_norm": 0.05999111335466328, + "learning_rate": 6.708135427535567e-10, + "loss": 0.388, + "step": 13238 + }, + { + "epoch": 4.992081447963801, + "grad_norm": 0.0607207381020247, + "learning_rate": 6.11216623584987e-10, + "loss": 0.3736, + "step": 13239 + }, + { + "epoch": 4.992458521870287, + "grad_norm": 0.062975624110886, + "learning_rate": 5.543916265793315e-10, + "loss": 0.3792, + "step": 13240 + }, + { + "epoch": 4.992835595776772, + "grad_norm": 0.05846506261620183, + "learning_rate": 5.003385556756612e-10, + "loss": 0.3796, + "step": 13241 + }, + { + "epoch": 4.993212669683258, + "grad_norm": 0.06015321859006523, + "learning_rate": 4.490574146176485e-10, + "loss": 0.3786, + "step": 13242 + }, + { + "epoch": 4.993589743589744, + "grad_norm": 0.06088869166204602, + "learning_rate": 4.005482069624478e-10, + "loss": 0.3779, + "step": 13243 + }, + { + "epoch": 4.993966817496229, + "grad_norm": 0.06169728904747893, + "learning_rate": 3.548109360673735e-10, + "loss": 0.3853, + "step": 13244 + }, + { + "epoch": 4.994343891402715, + "grad_norm": 0.06426950435091726, + "learning_rate": 3.118456051076635e-10, + "loss": 0.3909, + "step": 13245 + }, + { + "epoch": 4.994720965309201, + "grad_norm": 0.061775452336764576, + "learning_rate": 2.716522170542746e-10, + "loss": 0.3932, + "step": 13246 + }, + { + "epoch": 4.995098039215686, + "grad_norm": 0.0622066194176199, + "learning_rate": 2.3423077469608703e-10, + "loss": 0.3781, + "step": 13247 + }, + { + "epoch": 4.995475113122172, + "grad_norm": 0.06052801329548249, + "learning_rate": 1.9958128062658178e-10, + "loss": 0.3789, + "step": 13248 + }, + { + "epoch": 4.995852187028658, + "grad_norm": 0.06086046866974991, + "learning_rate": 1.677037372438406e-10, + "loss": 0.3662, + "step": 13249 + }, + { + "epoch": 4.996229260935143, + "grad_norm": 0.06050339554394108, + "learning_rate": 1.3859814676386862e-10, + "loss": 0.3882, + "step": 13250 + }, + { + "epoch": 4.996606334841629, + "grad_norm": 0.06282854798474254, + "learning_rate": 1.122645111939491e-10, + "loss": 0.3932, + "step": 13251 + }, + { + "epoch": 4.996983408748115, + "grad_norm": 0.05851683249005294, + "learning_rate": 8.870283236817046e-11, + "loss": 0.3744, + "step": 13252 + }, + { + "epoch": 4.9973604826546, + "grad_norm": 0.06093870494936303, + "learning_rate": 6.791311191634009e-11, + "loss": 0.3786, + "step": 13253 + }, + { + "epoch": 4.997737556561086, + "grad_norm": 0.06348296256471041, + "learning_rate": 4.989535127730705e-11, + "loss": 0.3799, + "step": 13254 + }, + { + "epoch": 4.998114630467572, + "grad_norm": 0.05998594496529717, + "learning_rate": 3.4649551698962006e-11, + "loss": 0.3699, + "step": 13255 + }, + { + "epoch": 4.998491704374057, + "grad_norm": 0.0615795766826707, + "learning_rate": 2.217571423823728e-11, + "loss": 0.3684, + "step": 13256 + }, + { + "epoch": 4.998868778280543, + "grad_norm": 0.05861326258552221, + "learning_rate": 1.2473839765547724e-11, + "loss": 0.3747, + "step": 13257 + }, + { + "epoch": 4.999245852187029, + "grad_norm": 0.05942048985684221, + "learning_rate": 5.543928942586263e-12, + "loss": 0.3712, + "step": 13258 + }, + { + "epoch": 4.999622926093514, + "grad_norm": 0.0607255488048606, + "learning_rate": 1.3859822578510263e-12, + "loss": 0.3723, + "step": 13259 + }, + { + "epoch": 5.0, + "grad_norm": 0.0682200086214866, + "learning_rate": 0.0, + "loss": 0.3741, + "step": 13260 + }, + { + "epoch": 5.0, + "step": 13260, + "total_flos": 2.2245164634341376e+17, + "train_loss": 0.10862892258625764, + "train_runtime": 69777.7052, + "train_samples_per_second": 97.272, + "train_steps_per_second": 0.19 + } + ], + "logging_steps": 1, + "max_steps": 13260, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2.2245164634341376e+17, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}