{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 412, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0048543689320388345, "grad_norm": 13.700736745706756, "learning_rate": 9.999854640567861e-06, "loss": 0.9326, "step": 1 }, { "epoch": 0.009708737864077669, "grad_norm": 7.653530875325563, "learning_rate": 9.999418570723189e-06, "loss": 0.6856, "step": 2 }, { "epoch": 0.014563106796116505, "grad_norm": 9.577989544673871, "learning_rate": 9.998691815820732e-06, "loss": 0.8421, "step": 3 }, { "epoch": 0.019417475728155338, "grad_norm": 7.140812268728353, "learning_rate": 9.997674418116759e-06, "loss": 0.749, "step": 4 }, { "epoch": 0.024271844660194174, "grad_norm": 7.6218338777610475, "learning_rate": 9.996366436766612e-06, "loss": 0.6104, "step": 5 }, { "epoch": 0.02912621359223301, "grad_norm": 5.967564365749561, "learning_rate": 9.994767947821261e-06, "loss": 0.6626, "step": 6 }, { "epoch": 0.03398058252427184, "grad_norm": 5.502705416941371, "learning_rate": 9.992879044222887e-06, "loss": 0.6681, "step": 7 }, { "epoch": 0.038834951456310676, "grad_norm": 5.4548993216652475, "learning_rate": 9.99069983579947e-06, "loss": 0.6568, "step": 8 }, { "epoch": 0.043689320388349516, "grad_norm": 5.92613191204192, "learning_rate": 9.988230449258409e-06, "loss": 0.7622, "step": 9 }, { "epoch": 0.04854368932038835, "grad_norm": 4.444825817835861, "learning_rate": 9.985471028179155e-06, "loss": 0.4877, "step": 10 }, { "epoch": 0.05339805825242718, "grad_norm": 5.8973316051643, "learning_rate": 9.982421733004857e-06, "loss": 0.7607, "step": 11 }, { "epoch": 0.05825242718446602, "grad_norm": 4.284149653974444, "learning_rate": 9.979082741033047e-06, "loss": 0.5874, "step": 12 }, { "epoch": 0.06310679611650485, "grad_norm": 4.4685297094322625, "learning_rate": 9.975454246405312e-06, "loss": 0.5163, "step": 13 }, { "epoch": 0.06796116504854369, "grad_norm": 6.070020033852334, "learning_rate": 9.971536460096021e-06, "loss": 0.8964, "step": 14 }, { "epoch": 0.07281553398058252, "grad_norm": 5.042066769396814, "learning_rate": 9.96732960990005e-06, "loss": 0.7094, "step": 15 }, { "epoch": 0.07766990291262135, "grad_norm": 4.761120484405765, "learning_rate": 9.96283394041954e-06, "loss": 0.6713, "step": 16 }, { "epoch": 0.0825242718446602, "grad_norm": 5.264846009576267, "learning_rate": 9.95804971304968e-06, "loss": 0.6306, "step": 17 }, { "epoch": 0.08737864077669903, "grad_norm": 3.9581925980742634, "learning_rate": 9.952977205963496e-06, "loss": 0.4716, "step": 18 }, { "epoch": 0.09223300970873786, "grad_norm": 4.302772918081245, "learning_rate": 9.94761671409569e-06, "loss": 0.6351, "step": 19 }, { "epoch": 0.0970873786407767, "grad_norm": 5.701003291841117, "learning_rate": 9.941968549125481e-06, "loss": 0.7364, "step": 20 }, { "epoch": 0.10194174757281553, "grad_norm": 5.0702795999876304, "learning_rate": 9.936033039458494e-06, "loss": 0.5886, "step": 21 }, { "epoch": 0.10679611650485436, "grad_norm": 5.152834692977829, "learning_rate": 9.929810530207651e-06, "loss": 0.6578, "step": 22 }, { "epoch": 0.11165048543689321, "grad_norm": 4.822469046990982, "learning_rate": 9.923301383173119e-06, "loss": 0.6027, "step": 23 }, { "epoch": 0.11650485436893204, "grad_norm": 4.6175566380863176, "learning_rate": 9.916505976821262e-06, "loss": 0.6046, "step": 24 }, { "epoch": 0.12135922330097088, "grad_norm": 4.38348505825168, "learning_rate": 9.909424706262647e-06, "loss": 0.5412, "step": 25 }, { "epoch": 0.1262135922330097, "grad_norm": 4.816398668030298, "learning_rate": 9.902057983229059e-06, "loss": 0.6874, "step": 26 }, { "epoch": 0.13106796116504854, "grad_norm": 5.5776981022172185, "learning_rate": 9.894406236049569e-06, "loss": 0.6001, "step": 27 }, { "epoch": 0.13592233009708737, "grad_norm": 4.780954878516279, "learning_rate": 9.886469909625624e-06, "loss": 0.6741, "step": 28 }, { "epoch": 0.1407766990291262, "grad_norm": 4.9801591936535266, "learning_rate": 9.87824946540519e-06, "loss": 0.6434, "step": 29 }, { "epoch": 0.14563106796116504, "grad_norm": 5.291290855277338, "learning_rate": 9.869745381355906e-06, "loss": 0.6493, "step": 30 }, { "epoch": 0.15048543689320387, "grad_norm": 4.80793570944734, "learning_rate": 9.860958151937303e-06, "loss": 0.6887, "step": 31 }, { "epoch": 0.1553398058252427, "grad_norm": 4.673937175953779, "learning_rate": 9.851888288072053e-06, "loss": 0.5977, "step": 32 }, { "epoch": 0.16019417475728157, "grad_norm": 4.807903875618923, "learning_rate": 9.842536317116262e-06, "loss": 0.6596, "step": 33 }, { "epoch": 0.1650485436893204, "grad_norm": 4.4033172139270125, "learning_rate": 9.832902782828801e-06, "loss": 0.5933, "step": 34 }, { "epoch": 0.16990291262135923, "grad_norm": 4.306109664471615, "learning_rate": 9.822988245339701e-06, "loss": 0.7037, "step": 35 }, { "epoch": 0.17475728155339806, "grad_norm": 4.940968620361455, "learning_rate": 9.81279328111758e-06, "loss": 0.6237, "step": 36 }, { "epoch": 0.1796116504854369, "grad_norm": 4.584175575943769, "learning_rate": 9.802318482936121e-06, "loss": 0.6387, "step": 37 }, { "epoch": 0.18446601941747573, "grad_norm": 4.663687766674046, "learning_rate": 9.791564459839609e-06, "loss": 0.6093, "step": 38 }, { "epoch": 0.18932038834951456, "grad_norm": 5.498597967202127, "learning_rate": 9.780531837107519e-06, "loss": 0.7233, "step": 39 }, { "epoch": 0.1941747572815534, "grad_norm": 5.3466847312213766, "learning_rate": 9.769221256218165e-06, "loss": 0.7023, "step": 40 }, { "epoch": 0.19902912621359223, "grad_norm": 5.3956768882738695, "learning_rate": 9.75763337481139e-06, "loss": 0.7399, "step": 41 }, { "epoch": 0.20388349514563106, "grad_norm": 4.916004098589441, "learning_rate": 9.745768866650339e-06, "loss": 0.7587, "step": 42 }, { "epoch": 0.2087378640776699, "grad_norm": 3.846742353174306, "learning_rate": 9.73362842158228e-06, "loss": 0.5237, "step": 43 }, { "epoch": 0.21359223300970873, "grad_norm": 4.29794561649259, "learning_rate": 9.721212745498493e-06, "loss": 0.655, "step": 44 }, { "epoch": 0.21844660194174756, "grad_norm": 5.109488219103711, "learning_rate": 9.70852256029323e-06, "loss": 0.6921, "step": 45 }, { "epoch": 0.22330097087378642, "grad_norm": 4.982670313966775, "learning_rate": 9.695558603821735e-06, "loss": 0.6938, "step": 46 }, { "epoch": 0.22815533980582525, "grad_norm": 5.583121475226154, "learning_rate": 9.682321629857348e-06, "loss": 0.6347, "step": 47 }, { "epoch": 0.23300970873786409, "grad_norm": 4.708017381486753, "learning_rate": 9.66881240804768e-06, "loss": 0.6695, "step": 48 }, { "epoch": 0.23786407766990292, "grad_norm": 5.139844952024056, "learning_rate": 9.655031723869848e-06, "loss": 0.6346, "step": 49 }, { "epoch": 0.24271844660194175, "grad_norm": 5.861883019001733, "learning_rate": 9.64098037858483e-06, "loss": 0.6977, "step": 50 }, { "epoch": 0.24757281553398058, "grad_norm": 4.6475114203513925, "learning_rate": 9.626659189190852e-06, "loss": 0.645, "step": 51 }, { "epoch": 0.2524271844660194, "grad_norm": 4.2994059145086245, "learning_rate": 9.612068988375898e-06, "loss": 0.6203, "step": 52 }, { "epoch": 0.25728155339805825, "grad_norm": 5.790605752150191, "learning_rate": 9.597210624469288e-06, "loss": 0.5849, "step": 53 }, { "epoch": 0.2621359223300971, "grad_norm": 5.361231666151757, "learning_rate": 9.582084961392358e-06, "loss": 0.6884, "step": 54 }, { "epoch": 0.2669902912621359, "grad_norm": 4.760024225174705, "learning_rate": 9.566692878608229e-06, "loss": 0.5778, "step": 55 }, { "epoch": 0.27184466019417475, "grad_norm": 5.123081621403656, "learning_rate": 9.551035271070665e-06, "loss": 0.5888, "step": 56 }, { "epoch": 0.2766990291262136, "grad_norm": 5.466643332389697, "learning_rate": 9.53511304917204e-06, "loss": 0.6971, "step": 57 }, { "epoch": 0.2815533980582524, "grad_norm": 5.0930073227574395, "learning_rate": 9.51892713869041e-06, "loss": 0.5525, "step": 58 }, { "epoch": 0.28640776699029125, "grad_norm": 5.125644166962777, "learning_rate": 9.502478480735678e-06, "loss": 0.6693, "step": 59 }, { "epoch": 0.2912621359223301, "grad_norm": 4.93420686835216, "learning_rate": 9.485768031694872e-06, "loss": 0.6365, "step": 60 }, { "epoch": 0.2961165048543689, "grad_norm": 4.836642913482464, "learning_rate": 9.468796763176549e-06, "loss": 0.7249, "step": 61 }, { "epoch": 0.30097087378640774, "grad_norm": 4.839544300303768, "learning_rate": 9.45156566195429e-06, "loss": 0.5816, "step": 62 }, { "epoch": 0.3058252427184466, "grad_norm": 6.0522421709064576, "learning_rate": 9.43407572990933e-06, "loss": 0.6257, "step": 63 }, { "epoch": 0.3106796116504854, "grad_norm": 4.9936992820066255, "learning_rate": 9.416327983972304e-06, "loss": 0.6462, "step": 64 }, { "epoch": 0.3155339805825243, "grad_norm": 4.5334446974452325, "learning_rate": 9.398323456064124e-06, "loss": 0.616, "step": 65 }, { "epoch": 0.32038834951456313, "grad_norm": 5.301359089885595, "learning_rate": 9.380063193035968e-06, "loss": 0.6011, "step": 66 }, { "epoch": 0.32524271844660196, "grad_norm": 4.872404630511316, "learning_rate": 9.361548256608421e-06, "loss": 0.6468, "step": 67 }, { "epoch": 0.3300970873786408, "grad_norm": 4.16002503203696, "learning_rate": 9.342779723309746e-06, "loss": 0.6813, "step": 68 }, { "epoch": 0.33495145631067963, "grad_norm": 4.390658439818213, "learning_rate": 9.323758684413272e-06, "loss": 0.5254, "step": 69 }, { "epoch": 0.33980582524271846, "grad_norm": 4.781526541263613, "learning_rate": 9.304486245873973e-06, "loss": 0.6176, "step": 70 }, { "epoch": 0.3446601941747573, "grad_norm": 4.724986433021162, "learning_rate": 9.284963528264133e-06, "loss": 0.5654, "step": 71 }, { "epoch": 0.34951456310679613, "grad_norm": 4.8972023611957045, "learning_rate": 9.26519166670821e-06, "loss": 0.6679, "step": 72 }, { "epoch": 0.35436893203883496, "grad_norm": 5.913767391821486, "learning_rate": 9.24517181081683e-06, "loss": 0.7364, "step": 73 }, { "epoch": 0.3592233009708738, "grad_norm": 4.766317066174581, "learning_rate": 9.22490512461995e-06, "loss": 0.5499, "step": 74 }, { "epoch": 0.3640776699029126, "grad_norm": 5.076729732381392, "learning_rate": 9.204392786499168e-06, "loss": 0.6739, "step": 75 }, { "epoch": 0.36893203883495146, "grad_norm": 4.02828502445423, "learning_rate": 9.183635989119211e-06, "loss": 0.5146, "step": 76 }, { "epoch": 0.3737864077669903, "grad_norm": 4.9456154568543935, "learning_rate": 9.162635939358593e-06, "loss": 0.5799, "step": 77 }, { "epoch": 0.3786407766990291, "grad_norm": 4.706444978491252, "learning_rate": 9.141393858239435e-06, "loss": 0.5559, "step": 78 }, { "epoch": 0.38349514563106796, "grad_norm": 5.636385287039295, "learning_rate": 9.119910980856477e-06, "loss": 0.6384, "step": 79 }, { "epoch": 0.3883495145631068, "grad_norm": 4.151399863186262, "learning_rate": 9.098188556305262e-06, "loss": 0.5744, "step": 80 }, { "epoch": 0.3932038834951456, "grad_norm": 4.529177940065554, "learning_rate": 9.076227847609513e-06, "loss": 0.5626, "step": 81 }, { "epoch": 0.39805825242718446, "grad_norm": 3.9279601380476628, "learning_rate": 9.054030131647682e-06, "loss": 0.5655, "step": 82 }, { "epoch": 0.4029126213592233, "grad_norm": 4.7672881807629555, "learning_rate": 9.031596699078727e-06, "loss": 0.7735, "step": 83 }, { "epoch": 0.4077669902912621, "grad_norm": 4.80824322834049, "learning_rate": 9.008928854267054e-06, "loss": 0.6445, "step": 84 }, { "epoch": 0.41262135922330095, "grad_norm": 5.260290861656673, "learning_rate": 8.986027915206686e-06, "loss": 0.529, "step": 85 }, { "epoch": 0.4174757281553398, "grad_norm": 4.897688668301753, "learning_rate": 8.962895213444618e-06, "loss": 0.6177, "step": 86 }, { "epoch": 0.4223300970873786, "grad_norm": 3.728537289995296, "learning_rate": 8.939532094003409e-06, "loss": 0.499, "step": 87 }, { "epoch": 0.42718446601941745, "grad_norm": 4.355224288889637, "learning_rate": 8.91593991530297e-06, "loss": 0.5745, "step": 88 }, { "epoch": 0.4320388349514563, "grad_norm": 5.354244078351413, "learning_rate": 8.892120049081577e-06, "loss": 0.6856, "step": 89 }, { "epoch": 0.4368932038834951, "grad_norm": 4.552981386977979, "learning_rate": 8.868073880316125e-06, "loss": 0.6618, "step": 90 }, { "epoch": 0.441747572815534, "grad_norm": 4.718124149535382, "learning_rate": 8.843802807141584e-06, "loss": 0.5998, "step": 91 }, { "epoch": 0.44660194174757284, "grad_norm": 5.617438256496521, "learning_rate": 8.819308240769726e-06, "loss": 0.6578, "step": 92 }, { "epoch": 0.45145631067961167, "grad_norm": 4.430519863035246, "learning_rate": 8.794591605407047e-06, "loss": 0.7492, "step": 93 }, { "epoch": 0.4563106796116505, "grad_norm": 4.986123337744732, "learning_rate": 8.769654338171986e-06, "loss": 0.7528, "step": 94 }, { "epoch": 0.46116504854368934, "grad_norm": 5.369401522372315, "learning_rate": 8.744497889011344e-06, "loss": 0.6214, "step": 95 }, { "epoch": 0.46601941747572817, "grad_norm": 4.654104574921331, "learning_rate": 8.71912372061598e-06, "loss": 0.7001, "step": 96 }, { "epoch": 0.470873786407767, "grad_norm": 4.843623532539826, "learning_rate": 8.693533308335786e-06, "loss": 0.684, "step": 97 }, { "epoch": 0.47572815533980584, "grad_norm": 4.8416624230835, "learning_rate": 8.667728140093876e-06, "loss": 0.7585, "step": 98 }, { "epoch": 0.48058252427184467, "grad_norm": 5.813328813425333, "learning_rate": 8.641709716300092e-06, "loss": 0.7049, "step": 99 }, { "epoch": 0.4854368932038835, "grad_norm": 5.44925011361989, "learning_rate": 8.615479549763756e-06, "loss": 0.7286, "step": 100 }, { "epoch": 0.49029126213592233, "grad_norm": 4.267160815318575, "learning_rate": 8.589039165605716e-06, "loss": 0.6111, "step": 101 }, { "epoch": 0.49514563106796117, "grad_norm": 5.197518058605588, "learning_rate": 8.56239010116966e-06, "loss": 0.7136, "step": 102 }, { "epoch": 0.5, "grad_norm": 4.22548367094451, "learning_rate": 8.535533905932739e-06, "loss": 0.6056, "step": 103 }, { "epoch": 0.5048543689320388, "grad_norm": 4.130512584878891, "learning_rate": 8.508472141415468e-06, "loss": 0.5206, "step": 104 }, { "epoch": 0.5097087378640777, "grad_norm": 4.863452461576489, "learning_rate": 8.481206381090934e-06, "loss": 0.6708, "step": 105 }, { "epoch": 0.5145631067961165, "grad_norm": 4.381330827874045, "learning_rate": 8.453738210293316e-06, "loss": 0.5166, "step": 106 }, { "epoch": 0.5194174757281553, "grad_norm": 4.64450353123871, "learning_rate": 8.426069226125695e-06, "loss": 0.5254, "step": 107 }, { "epoch": 0.5242718446601942, "grad_norm": 4.789784427108796, "learning_rate": 8.398201037367202e-06, "loss": 0.5922, "step": 108 }, { "epoch": 0.529126213592233, "grad_norm": 4.9374440323093465, "learning_rate": 8.370135264379475e-06, "loss": 0.6798, "step": 109 }, { "epoch": 0.5339805825242718, "grad_norm": 4.449074918325947, "learning_rate": 8.341873539012443e-06, "loss": 0.5926, "step": 110 }, { "epoch": 0.5388349514563107, "grad_norm": 4.9976782967365185, "learning_rate": 8.313417504509446e-06, "loss": 0.5774, "step": 111 }, { "epoch": 0.5436893203883495, "grad_norm": 4.205266082241357, "learning_rate": 8.284768815411693e-06, "loss": 0.5607, "step": 112 }, { "epoch": 0.5485436893203883, "grad_norm": 5.394379865691564, "learning_rate": 8.255929137462049e-06, "loss": 0.768, "step": 113 }, { "epoch": 0.5533980582524272, "grad_norm": 5.446027769445993, "learning_rate": 8.226900147508205e-06, "loss": 0.6709, "step": 114 }, { "epoch": 0.558252427184466, "grad_norm": 4.3442092653936975, "learning_rate": 8.197683533405156e-06, "loss": 0.5935, "step": 115 }, { "epoch": 0.5631067961165048, "grad_norm": 5.04306598769716, "learning_rate": 8.168280993917078e-06, "loss": 0.6847, "step": 116 }, { "epoch": 0.5679611650485437, "grad_norm": 5.046597270632832, "learning_rate": 8.138694238618543e-06, "loss": 0.6968, "step": 117 }, { "epoch": 0.5728155339805825, "grad_norm": 4.612488835149653, "learning_rate": 8.108924987795137e-06, "loss": 0.591, "step": 118 }, { "epoch": 0.5776699029126213, "grad_norm": 4.452427393299512, "learning_rate": 8.078974972343414e-06, "loss": 0.6167, "step": 119 }, { "epoch": 0.5825242718446602, "grad_norm": 4.810855216412306, "learning_rate": 8.048845933670274e-06, "loss": 0.641, "step": 120 }, { "epoch": 0.587378640776699, "grad_norm": 4.259329203974203, "learning_rate": 8.01853962359169e-06, "loss": 0.5174, "step": 121 }, { "epoch": 0.5922330097087378, "grad_norm": 4.525018864641845, "learning_rate": 7.988057804230878e-06, "loss": 0.5923, "step": 122 }, { "epoch": 0.5970873786407767, "grad_norm": 4.287171824496176, "learning_rate": 7.957402247915817e-06, "loss": 0.5201, "step": 123 }, { "epoch": 0.6019417475728155, "grad_norm": 4.568547315330446, "learning_rate": 7.92657473707621e-06, "loss": 0.5593, "step": 124 }, { "epoch": 0.6067961165048543, "grad_norm": 5.619787834257479, "learning_rate": 7.895577064139847e-06, "loss": 0.755, "step": 125 }, { "epoch": 0.6116504854368932, "grad_norm": 5.038045288458346, "learning_rate": 7.864411031428379e-06, "loss": 0.5929, "step": 126 }, { "epoch": 0.616504854368932, "grad_norm": 4.308789083354014, "learning_rate": 7.833078451052537e-06, "loss": 0.5325, "step": 127 }, { "epoch": 0.6213592233009708, "grad_norm": 4.876457980237497, "learning_rate": 7.801581144806752e-06, "loss": 0.6042, "step": 128 }, { "epoch": 0.6262135922330098, "grad_norm": 4.715115120896625, "learning_rate": 7.769920944063244e-06, "loss": 0.5506, "step": 129 }, { "epoch": 0.6310679611650486, "grad_norm": 4.093929899155603, "learning_rate": 7.73809968966554e-06, "loss": 0.5067, "step": 130 }, { "epoch": 0.6359223300970874, "grad_norm": 5.204971809306947, "learning_rate": 7.706119231821423e-06, "loss": 0.678, "step": 131 }, { "epoch": 0.6407766990291263, "grad_norm": 4.598319690119657, "learning_rate": 7.673981429995372e-06, "loss": 0.5404, "step": 132 }, { "epoch": 0.6456310679611651, "grad_norm": 4.411590644201943, "learning_rate": 7.641688152800433e-06, "loss": 0.5449, "step": 133 }, { "epoch": 0.6504854368932039, "grad_norm": 4.567080038328939, "learning_rate": 7.609241277889583e-06, "loss": 0.5207, "step": 134 }, { "epoch": 0.6553398058252428, "grad_norm": 5.874481895781281, "learning_rate": 7.5766426918465455e-06, "loss": 0.739, "step": 135 }, { "epoch": 0.6601941747572816, "grad_norm": 4.691099974012714, "learning_rate": 7.5438942900761035e-06, "loss": 0.6806, "step": 136 }, { "epoch": 0.6650485436893204, "grad_norm": 4.790000632253766, "learning_rate": 7.51099797669389e-06, "loss": 0.5998, "step": 137 }, { "epoch": 0.6699029126213593, "grad_norm": 5.358966015470333, "learning_rate": 7.477955664415678e-06, "loss": 0.7577, "step": 138 }, { "epoch": 0.6747572815533981, "grad_norm": 4.752903028752795, "learning_rate": 7.444769274446168e-06, "loss": 0.6375, "step": 139 }, { "epoch": 0.6796116504854369, "grad_norm": 4.518059343323462, "learning_rate": 7.411440736367281e-06, "loss": 0.5483, "step": 140 }, { "epoch": 0.6844660194174758, "grad_norm": 4.260389601318289, "learning_rate": 7.377971988025964e-06, "loss": 0.5732, "step": 141 }, { "epoch": 0.6893203883495146, "grad_norm": 4.585815720669657, "learning_rate": 7.3443649754215175e-06, "loss": 0.5305, "step": 142 }, { "epoch": 0.6941747572815534, "grad_norm": 5.858287351875903, "learning_rate": 7.310621652592449e-06, "loss": 0.7093, "step": 143 }, { "epoch": 0.6990291262135923, "grad_norm": 4.7194455615920905, "learning_rate": 7.276743981502856e-06, "loss": 0.5611, "step": 144 }, { "epoch": 0.7038834951456311, "grad_norm": 4.479691664973672, "learning_rate": 7.242733931928352e-06, "loss": 0.5848, "step": 145 }, { "epoch": 0.7087378640776699, "grad_norm": 3.7387563494373026, "learning_rate": 7.208593481341536e-06, "loss": 0.5149, "step": 146 }, { "epoch": 0.7135922330097088, "grad_norm": 3.960569599785761, "learning_rate": 7.1743246147970095e-06, "loss": 0.5172, "step": 147 }, { "epoch": 0.7184466019417476, "grad_norm": 4.964800551732467, "learning_rate": 7.139929324815965e-06, "loss": 0.6144, "step": 148 }, { "epoch": 0.7233009708737864, "grad_norm": 4.285834062715167, "learning_rate": 7.105409611270332e-06, "loss": 0.5468, "step": 149 }, { "epoch": 0.7281553398058253, "grad_norm": 4.640435777755612, "learning_rate": 7.070767481266493e-06, "loss": 0.7393, "step": 150 }, { "epoch": 0.7330097087378641, "grad_norm": 6.160564423668876, "learning_rate": 7.036004949028587e-06, "loss": 0.596, "step": 151 }, { "epoch": 0.7378640776699029, "grad_norm": 5.045193112548703, "learning_rate": 7.00112403578139e-06, "loss": 0.7241, "step": 152 }, { "epoch": 0.7427184466019418, "grad_norm": 4.6726231717702165, "learning_rate": 6.9661267696328015e-06, "loss": 0.6641, "step": 153 }, { "epoch": 0.7475728155339806, "grad_norm": 4.498579507506946, "learning_rate": 6.931015185455915e-06, "loss": 0.6593, "step": 154 }, { "epoch": 0.7524271844660194, "grad_norm": 4.418286446544229, "learning_rate": 6.895791324770702e-06, "loss": 0.5237, "step": 155 }, { "epoch": 0.7572815533980582, "grad_norm": 4.324444133479052, "learning_rate": 6.860457235625322e-06, "loss": 0.5238, "step": 156 }, { "epoch": 0.7621359223300971, "grad_norm": 4.623653724808323, "learning_rate": 6.825014972477024e-06, "loss": 0.627, "step": 157 }, { "epoch": 0.7669902912621359, "grad_norm": 4.225434812151843, "learning_rate": 6.7894665960727105e-06, "loss": 0.5557, "step": 158 }, { "epoch": 0.7718446601941747, "grad_norm": 4.326323911723923, "learning_rate": 6.7538141733291e-06, "loss": 0.5044, "step": 159 }, { "epoch": 0.7766990291262136, "grad_norm": 4.736446978361883, "learning_rate": 6.7180597772125665e-06, "loss": 0.5791, "step": 160 }, { "epoch": 0.7815533980582524, "grad_norm": 5.53164688155515, "learning_rate": 6.682205486618592e-06, "loss": 0.7067, "step": 161 }, { "epoch": 0.7864077669902912, "grad_norm": 4.35445538897878, "learning_rate": 6.646253386250909e-06, "loss": 0.5252, "step": 162 }, { "epoch": 0.7912621359223301, "grad_norm": 4.029242537177906, "learning_rate": 6.610205566500272e-06, "loss": 0.5451, "step": 163 }, { "epoch": 0.7961165048543689, "grad_norm": 4.989028966521606, "learning_rate": 6.574064123322925e-06, "loss": 0.7189, "step": 164 }, { "epoch": 0.8009708737864077, "grad_norm": 4.113230209748707, "learning_rate": 6.537831158118733e-06, "loss": 0.5319, "step": 165 }, { "epoch": 0.8058252427184466, "grad_norm": 4.463636912689356, "learning_rate": 6.50150877760899e-06, "loss": 0.6048, "step": 166 }, { "epoch": 0.8106796116504854, "grad_norm": 4.9148662074020715, "learning_rate": 6.465099093713944e-06, "loss": 0.7315, "step": 167 }, { "epoch": 0.8155339805825242, "grad_norm": 4.226811763646021, "learning_rate": 6.42860422342998e-06, "loss": 0.6551, "step": 168 }, { "epoch": 0.8203883495145631, "grad_norm": 4.1577606585165725, "learning_rate": 6.392026288706549e-06, "loss": 0.562, "step": 169 }, { "epoch": 0.8252427184466019, "grad_norm": 3.7034304723831575, "learning_rate": 6.3553674163227786e-06, "loss": 0.4522, "step": 170 }, { "epoch": 0.8300970873786407, "grad_norm": 4.300090221607775, "learning_rate": 6.318629737763818e-06, "loss": 0.5555, "step": 171 }, { "epoch": 0.8349514563106796, "grad_norm": 5.949444522959021, "learning_rate": 6.281815389096903e-06, "loss": 0.7808, "step": 172 }, { "epoch": 0.8398058252427184, "grad_norm": 5.031092073483118, "learning_rate": 6.244926510847162e-06, "loss": 0.6009, "step": 173 }, { "epoch": 0.8446601941747572, "grad_norm": 5.187101736478666, "learning_rate": 6.207965247873151e-06, "loss": 0.6132, "step": 174 }, { "epoch": 0.8495145631067961, "grad_norm": 5.24456318411293, "learning_rate": 6.1709337492421515e-06, "loss": 0.5247, "step": 175 }, { "epoch": 0.8543689320388349, "grad_norm": 4.371333738660568, "learning_rate": 6.133834168105206e-06, "loss": 0.6485, "step": 176 }, { "epoch": 0.8592233009708737, "grad_norm": 4.684330980217502, "learning_rate": 6.096668661571934e-06, "loss": 0.6254, "step": 177 }, { "epoch": 0.8640776699029126, "grad_norm": 4.955262238546974, "learning_rate": 6.0594393905851065e-06, "loss": 0.6422, "step": 178 }, { "epoch": 0.8689320388349514, "grad_norm": 4.462259052477423, "learning_rate": 6.0221485197949995e-06, "loss": 0.6756, "step": 179 }, { "epoch": 0.8737864077669902, "grad_norm": 4.878770988965721, "learning_rate": 5.9847982174335314e-06, "loss": 0.7392, "step": 180 }, { "epoch": 0.8786407766990292, "grad_norm": 4.543627472198425, "learning_rate": 5.9473906551881985e-06, "loss": 0.6347, "step": 181 }, { "epoch": 0.883495145631068, "grad_norm": 5.202468191483863, "learning_rate": 5.9099280080758085e-06, "loss": 0.5598, "step": 182 }, { "epoch": 0.8883495145631068, "grad_norm": 4.315780849113243, "learning_rate": 5.872412454315999e-06, "loss": 0.6663, "step": 183 }, { "epoch": 0.8932038834951457, "grad_norm": 5.372325357350075, "learning_rate": 5.834846175204612e-06, "loss": 0.6421, "step": 184 }, { "epoch": 0.8980582524271845, "grad_norm": 4.349486246810395, "learning_rate": 5.797231354986842e-06, "loss": 0.5004, "step": 185 }, { "epoch": 0.9029126213592233, "grad_norm": 3.930987322352072, "learning_rate": 5.759570180730255e-06, "loss": 0.6901, "step": 186 }, { "epoch": 0.9077669902912622, "grad_norm": 4.161806833595799, "learning_rate": 5.721864842197612e-06, "loss": 0.5021, "step": 187 }, { "epoch": 0.912621359223301, "grad_norm": 4.826643850863039, "learning_rate": 5.684117531719552e-06, "loss": 0.8215, "step": 188 }, { "epoch": 0.9174757281553398, "grad_norm": 5.661081361235644, "learning_rate": 5.646330444067121e-06, "loss": 0.6133, "step": 189 }, { "epoch": 0.9223300970873787, "grad_norm": 3.70874900259711, "learning_rate": 5.608505776324158e-06, "loss": 0.48, "step": 190 }, { "epoch": 0.9271844660194175, "grad_norm": 4.580224834317882, "learning_rate": 5.570645727759558e-06, "loss": 0.6633, "step": 191 }, { "epoch": 0.9320388349514563, "grad_norm": 4.3084843720006685, "learning_rate": 5.532752499699381e-06, "loss": 0.5887, "step": 192 }, { "epoch": 0.9368932038834952, "grad_norm": 4.436869902088287, "learning_rate": 5.494828295398874e-06, "loss": 0.594, "step": 193 }, { "epoch": 0.941747572815534, "grad_norm": 4.3282846742588434, "learning_rate": 5.456875319914355e-06, "loss": 0.5275, "step": 194 }, { "epoch": 0.9466019417475728, "grad_norm": 4.589521089894314, "learning_rate": 5.4188957799750145e-06, "loss": 0.5684, "step": 195 }, { "epoch": 0.9514563106796117, "grad_norm": 3.9061087486112798, "learning_rate": 5.380891883854591e-06, "loss": 0.5671, "step": 196 }, { "epoch": 0.9563106796116505, "grad_norm": 4.390976409109075, "learning_rate": 5.34286584124299e-06, "loss": 0.6359, "step": 197 }, { "epoch": 0.9611650485436893, "grad_norm": 3.9715583054517305, "learning_rate": 5.304819863117796e-06, "loss": 0.5545, "step": 198 }, { "epoch": 0.9660194174757282, "grad_norm": 4.28227466317342, "learning_rate": 5.266756161615719e-06, "loss": 0.5643, "step": 199 }, { "epoch": 0.970873786407767, "grad_norm": 4.857073395531744, "learning_rate": 5.228676949903974e-06, "loss": 0.6413, "step": 200 }, { "epoch": 0.9757281553398058, "grad_norm": 4.699879680047218, "learning_rate": 5.190584442051594e-06, "loss": 0.583, "step": 201 }, { "epoch": 0.9805825242718447, "grad_norm": 4.785423263456382, "learning_rate": 5.1524808529007075e-06, "loss": 0.6833, "step": 202 }, { "epoch": 0.9854368932038835, "grad_norm": 5.299382767400942, "learning_rate": 5.114368397937744e-06, "loss": 0.7587, "step": 203 }, { "epoch": 0.9902912621359223, "grad_norm": 4.220195153555258, "learning_rate": 5.07624929316463e-06, "loss": 0.5928, "step": 204 }, { "epoch": 0.9951456310679612, "grad_norm": 4.885376452095594, "learning_rate": 5.038125754969933e-06, "loss": 0.536, "step": 205 }, { "epoch": 1.0, "grad_norm": 4.953486790656896, "learning_rate": 5e-06, "loss": 0.4953, "step": 206 }, { "epoch": 1.0048543689320388, "grad_norm": 4.2998226477752475, "learning_rate": 4.9618742450300675e-06, "loss": 0.4054, "step": 207 }, { "epoch": 1.0097087378640777, "grad_norm": 3.9199439325425876, "learning_rate": 4.923750706835371e-06, "loss": 0.4169, "step": 208 }, { "epoch": 1.0145631067961165, "grad_norm": 3.974696778634173, "learning_rate": 4.8856316020622564e-06, "loss": 0.462, "step": 209 }, { "epoch": 1.0194174757281553, "grad_norm": 3.2484168037321197, "learning_rate": 4.847519147099294e-06, "loss": 0.3426, "step": 210 }, { "epoch": 1.0242718446601942, "grad_norm": 4.207830949487246, "learning_rate": 4.809415557948407e-06, "loss": 0.476, "step": 211 }, { "epoch": 1.029126213592233, "grad_norm": 3.4000088344982986, "learning_rate": 4.771323050096028e-06, "loss": 0.376, "step": 212 }, { "epoch": 1.0339805825242718, "grad_norm": 4.546136704529267, "learning_rate": 4.733243838384282e-06, "loss": 0.4709, "step": 213 }, { "epoch": 1.0388349514563107, "grad_norm": 4.518392025037244, "learning_rate": 4.6951801368822055e-06, "loss": 0.4269, "step": 214 }, { "epoch": 1.0436893203883495, "grad_norm": 4.443482492740436, "learning_rate": 4.6571341587570114e-06, "loss": 0.4226, "step": 215 }, { "epoch": 1.0485436893203883, "grad_norm": 3.9767242121603483, "learning_rate": 4.619108116145411e-06, "loss": 0.4995, "step": 216 }, { "epoch": 1.0533980582524272, "grad_norm": 4.335780318473907, "learning_rate": 4.581104220024988e-06, "loss": 0.4465, "step": 217 }, { "epoch": 1.058252427184466, "grad_norm": 4.387439414333554, "learning_rate": 4.5431246800856455e-06, "loss": 0.4979, "step": 218 }, { "epoch": 1.0631067961165048, "grad_norm": 4.189311933354572, "learning_rate": 4.505171704601128e-06, "loss": 0.3814, "step": 219 }, { "epoch": 1.0679611650485437, "grad_norm": 3.997478332212308, "learning_rate": 4.467247500300621e-06, "loss": 0.3909, "step": 220 }, { "epoch": 1.0728155339805825, "grad_norm": 4.77703496689477, "learning_rate": 4.4293542722404435e-06, "loss": 0.458, "step": 221 }, { "epoch": 1.0776699029126213, "grad_norm": 5.902657681692468, "learning_rate": 4.391494223675843e-06, "loss": 0.415, "step": 222 }, { "epoch": 1.0825242718446602, "grad_norm": 4.564960979293292, "learning_rate": 4.3536695559328816e-06, "loss": 0.4116, "step": 223 }, { "epoch": 1.087378640776699, "grad_norm": 4.474800558361843, "learning_rate": 4.31588246828045e-06, "loss": 0.3916, "step": 224 }, { "epoch": 1.0922330097087378, "grad_norm": 4.705977443335298, "learning_rate": 4.278135157802389e-06, "loss": 0.3803, "step": 225 }, { "epoch": 1.0970873786407767, "grad_norm": 4.462404772400245, "learning_rate": 4.240429819269746e-06, "loss": 0.4053, "step": 226 }, { "epoch": 1.1019417475728155, "grad_norm": 3.9527750733792884, "learning_rate": 4.20276864501316e-06, "loss": 0.3213, "step": 227 }, { "epoch": 1.1067961165048543, "grad_norm": 5.230759530090539, "learning_rate": 4.165153824795391e-06, "loss": 0.4666, "step": 228 }, { "epoch": 1.1116504854368932, "grad_norm": 4.263994053644524, "learning_rate": 4.127587545684002e-06, "loss": 0.4391, "step": 229 }, { "epoch": 1.116504854368932, "grad_norm": 4.406086901097569, "learning_rate": 4.090071991924194e-06, "loss": 0.4033, "step": 230 }, { "epoch": 1.1213592233009708, "grad_norm": 3.9140973464405286, "learning_rate": 4.052609344811802e-06, "loss": 0.3706, "step": 231 }, { "epoch": 1.1262135922330097, "grad_norm": 4.397586852739592, "learning_rate": 4.015201782566471e-06, "loss": 0.4412, "step": 232 }, { "epoch": 1.1310679611650485, "grad_norm": 4.246548217308422, "learning_rate": 3.977851480205003e-06, "loss": 0.3776, "step": 233 }, { "epoch": 1.1359223300970873, "grad_norm": 3.8964754483880095, "learning_rate": 3.940560609414894e-06, "loss": 0.3252, "step": 234 }, { "epoch": 1.1407766990291262, "grad_norm": 4.489750540716874, "learning_rate": 3.903331338428067e-06, "loss": 0.4068, "step": 235 }, { "epoch": 1.145631067961165, "grad_norm": 4.095950745293965, "learning_rate": 3.866165831894796e-06, "loss": 0.4483, "step": 236 }, { "epoch": 1.1504854368932038, "grad_norm": 4.005574368490938, "learning_rate": 3.829066250757851e-06, "loss": 0.3712, "step": 237 }, { "epoch": 1.1553398058252426, "grad_norm": 5.2140906656942825, "learning_rate": 3.7920347521268514e-06, "loss": 0.4676, "step": 238 }, { "epoch": 1.1601941747572815, "grad_norm": 4.118605527869571, "learning_rate": 3.7550734891528413e-06, "loss": 0.399, "step": 239 }, { "epoch": 1.1650485436893203, "grad_norm": 4.535521891874924, "learning_rate": 3.7181846109031007e-06, "loss": 0.4607, "step": 240 }, { "epoch": 1.1699029126213591, "grad_norm": 3.523227265648954, "learning_rate": 3.6813702622361858e-06, "loss": 0.3345, "step": 241 }, { "epoch": 1.174757281553398, "grad_norm": 4.098225629427337, "learning_rate": 3.6446325836772244e-06, "loss": 0.3472, "step": 242 }, { "epoch": 1.1796116504854368, "grad_norm": 4.467051517827206, "learning_rate": 3.6079737112934533e-06, "loss": 0.4033, "step": 243 }, { "epoch": 1.1844660194174756, "grad_norm": 3.637313655779216, "learning_rate": 3.5713957765700224e-06, "loss": 0.3446, "step": 244 }, { "epoch": 1.1893203883495145, "grad_norm": 4.056284852584977, "learning_rate": 3.5349009062860586e-06, "loss": 0.4759, "step": 245 }, { "epoch": 1.1941747572815533, "grad_norm": 4.705435349219051, "learning_rate": 3.4984912223910105e-06, "loss": 0.4082, "step": 246 }, { "epoch": 1.1990291262135921, "grad_norm": 4.107890455577559, "learning_rate": 3.46216884188127e-06, "loss": 0.3445, "step": 247 }, { "epoch": 1.203883495145631, "grad_norm": 3.9854145467687245, "learning_rate": 3.425935876677077e-06, "loss": 0.3844, "step": 248 }, { "epoch": 1.2087378640776698, "grad_norm": 4.606912943038989, "learning_rate": 3.38979443349973e-06, "loss": 0.4149, "step": 249 }, { "epoch": 1.2135922330097086, "grad_norm": 4.318034154528551, "learning_rate": 3.3537466137490937e-06, "loss": 0.3787, "step": 250 }, { "epoch": 1.2184466019417475, "grad_norm": 3.8453064971033273, "learning_rate": 3.3177945133814093e-06, "loss": 0.4277, "step": 251 }, { "epoch": 1.2233009708737863, "grad_norm": 4.585295333838764, "learning_rate": 3.2819402227874364e-06, "loss": 0.4946, "step": 252 }, { "epoch": 1.2281553398058254, "grad_norm": 4.315991751752694, "learning_rate": 3.2461858266709017e-06, "loss": 0.3499, "step": 253 }, { "epoch": 1.233009708737864, "grad_norm": 3.683056037012155, "learning_rate": 3.2105334039272924e-06, "loss": 0.3793, "step": 254 }, { "epoch": 1.237864077669903, "grad_norm": 3.611086757154011, "learning_rate": 3.1749850275229777e-06, "loss": 0.3663, "step": 255 }, { "epoch": 1.2427184466019416, "grad_norm": 4.256844964809122, "learning_rate": 3.1395427643746802e-06, "loss": 0.424, "step": 256 }, { "epoch": 1.2475728155339807, "grad_norm": 4.6632407000047635, "learning_rate": 3.1042086752292995e-06, "loss": 0.3705, "step": 257 }, { "epoch": 1.2524271844660193, "grad_norm": 4.062276851228071, "learning_rate": 3.068984814544087e-06, "loss": 0.4258, "step": 258 }, { "epoch": 1.2572815533980584, "grad_norm": 5.695523440771089, "learning_rate": 3.0338732303671993e-06, "loss": 0.3971, "step": 259 }, { "epoch": 1.262135922330097, "grad_norm": 4.148731993978135, "learning_rate": 2.99887596421861e-06, "loss": 0.3462, "step": 260 }, { "epoch": 1.266990291262136, "grad_norm": 5.884360430854941, "learning_rate": 2.9639950509714138e-06, "loss": 0.452, "step": 261 }, { "epoch": 1.2718446601941746, "grad_norm": 4.777875804249764, "learning_rate": 2.929232518733507e-06, "loss": 0.528, "step": 262 }, { "epoch": 1.2766990291262137, "grad_norm": 4.753463999442962, "learning_rate": 2.8945903887296686e-06, "loss": 0.3473, "step": 263 }, { "epoch": 1.2815533980582523, "grad_norm": 4.159019900404467, "learning_rate": 2.860070675184036e-06, "loss": 0.3765, "step": 264 }, { "epoch": 1.2864077669902914, "grad_norm": 4.0950310456747765, "learning_rate": 2.8256753852029917e-06, "loss": 0.4084, "step": 265 }, { "epoch": 1.29126213592233, "grad_norm": 4.132024264410155, "learning_rate": 2.7914065186584637e-06, "loss": 0.415, "step": 266 }, { "epoch": 1.296116504854369, "grad_norm": 5.6427278620163035, "learning_rate": 2.757266068071648e-06, "loss": 0.4325, "step": 267 }, { "epoch": 1.3009708737864076, "grad_norm": 4.969560750168923, "learning_rate": 2.7232560184971437e-06, "loss": 0.4409, "step": 268 }, { "epoch": 1.3058252427184467, "grad_norm": 4.5520470651696945, "learning_rate": 2.689378347407553e-06, "loss": 0.4372, "step": 269 }, { "epoch": 1.3106796116504853, "grad_norm": 4.401842760321957, "learning_rate": 2.6556350245784833e-06, "loss": 0.4392, "step": 270 }, { "epoch": 1.3155339805825244, "grad_norm": 4.362124241022186, "learning_rate": 2.6220280119740376e-06, "loss": 0.3732, "step": 271 }, { "epoch": 1.3203883495145632, "grad_norm": 6.7958206995705925, "learning_rate": 2.588559263632719e-06, "loss": 0.63, "step": 272 }, { "epoch": 1.325242718446602, "grad_norm": 4.228733543063519, "learning_rate": 2.555230725553832e-06, "loss": 0.2853, "step": 273 }, { "epoch": 1.3300970873786409, "grad_norm": 4.287441420272118, "learning_rate": 2.522044335584322e-06, "loss": 0.423, "step": 274 }, { "epoch": 1.3349514563106797, "grad_norm": 4.594999132471548, "learning_rate": 2.489002023306112e-06, "loss": 0.439, "step": 275 }, { "epoch": 1.3398058252427185, "grad_norm": 5.685734323847062, "learning_rate": 2.4561057099238973e-06, "loss": 0.5615, "step": 276 }, { "epoch": 1.3446601941747574, "grad_norm": 4.894857125218613, "learning_rate": 2.423357308153454e-06, "loss": 0.3219, "step": 277 }, { "epoch": 1.3495145631067962, "grad_norm": 3.906056695907201, "learning_rate": 2.390758722110418e-06, "loss": 0.395, "step": 278 }, { "epoch": 1.354368932038835, "grad_norm": 4.292758454406484, "learning_rate": 2.358311847199567e-06, "loss": 0.4565, "step": 279 }, { "epoch": 1.3592233009708738, "grad_norm": 4.1580017163845415, "learning_rate": 2.3260185700046295e-06, "loss": 0.4092, "step": 280 }, { "epoch": 1.3640776699029127, "grad_norm": 3.9655592928181598, "learning_rate": 2.2938807681785764e-06, "loss": 0.4131, "step": 281 }, { "epoch": 1.3689320388349515, "grad_norm": 4.374260902231569, "learning_rate": 2.2619003103344607e-06, "loss": 0.4369, "step": 282 }, { "epoch": 1.3737864077669903, "grad_norm": 4.383382260794347, "learning_rate": 2.2300790559367553e-06, "loss": 0.4091, "step": 283 }, { "epoch": 1.3786407766990292, "grad_norm": 4.248560762101138, "learning_rate": 2.1984188551932513e-06, "loss": 0.407, "step": 284 }, { "epoch": 1.383495145631068, "grad_norm": 4.613689374125919, "learning_rate": 2.166921548947466e-06, "loss": 0.3831, "step": 285 }, { "epoch": 1.3883495145631068, "grad_norm": 4.981809011827919, "learning_rate": 2.1355889685716225e-06, "loss": 0.4303, "step": 286 }, { "epoch": 1.3932038834951457, "grad_norm": 4.592480510311169, "learning_rate": 2.1044229358601543e-06, "loss": 0.4503, "step": 287 }, { "epoch": 1.3980582524271845, "grad_norm": 4.173987949053065, "learning_rate": 2.0734252629237892e-06, "loss": 0.3728, "step": 288 }, { "epoch": 1.4029126213592233, "grad_norm": 4.22006738427565, "learning_rate": 2.0425977520841837e-06, "loss": 0.3797, "step": 289 }, { "epoch": 1.4077669902912622, "grad_norm": 3.8510122954562043, "learning_rate": 2.011942195769122e-06, "loss": 0.4101, "step": 290 }, { "epoch": 1.412621359223301, "grad_norm": 4.372990559839991, "learning_rate": 1.9814603764083112e-06, "loss": 0.4158, "step": 291 }, { "epoch": 1.4174757281553398, "grad_norm": 5.575649245558948, "learning_rate": 1.9511540663297284e-06, "loss": 0.477, "step": 292 }, { "epoch": 1.4223300970873787, "grad_norm": 4.141046008086547, "learning_rate": 1.921025027656587e-06, "loss": 0.4314, "step": 293 }, { "epoch": 1.4271844660194175, "grad_norm": 4.237180409344308, "learning_rate": 1.8910750122048638e-06, "loss": 0.3802, "step": 294 }, { "epoch": 1.4320388349514563, "grad_norm": 4.122454194672351, "learning_rate": 1.8613057613814584e-06, "loss": 0.4027, "step": 295 }, { "epoch": 1.4368932038834952, "grad_norm": 4.848299331536871, "learning_rate": 1.8317190060829242e-06, "loss": 0.4636, "step": 296 }, { "epoch": 1.441747572815534, "grad_norm": 4.529457597978418, "learning_rate": 1.8023164665948455e-06, "loss": 0.4457, "step": 297 }, { "epoch": 1.4466019417475728, "grad_norm": 3.729095413494349, "learning_rate": 1.773099852491796e-06, "loss": 0.3206, "step": 298 }, { "epoch": 1.4514563106796117, "grad_norm": 4.122736134286334, "learning_rate": 1.7440708625379503e-06, "loss": 0.4391, "step": 299 }, { "epoch": 1.4563106796116505, "grad_norm": 6.190186805722854, "learning_rate": 1.7152311845883096e-06, "loss": 0.5189, "step": 300 }, { "epoch": 1.4611650485436893, "grad_norm": 4.789287657761732, "learning_rate": 1.686582495490554e-06, "loss": 0.4275, "step": 301 }, { "epoch": 1.4660194174757282, "grad_norm": 5.182489368234093, "learning_rate": 1.658126460987558e-06, "loss": 0.3923, "step": 302 }, { "epoch": 1.470873786407767, "grad_norm": 5.367009189045723, "learning_rate": 1.6298647356205255e-06, "loss": 0.4636, "step": 303 }, { "epoch": 1.4757281553398058, "grad_norm": 4.917083767169665, "learning_rate": 1.601798962632799e-06, "loss": 0.3767, "step": 304 }, { "epoch": 1.4805825242718447, "grad_norm": 5.338254289658298, "learning_rate": 1.573930773874306e-06, "loss": 0.4162, "step": 305 }, { "epoch": 1.4854368932038835, "grad_norm": 4.05097071015025, "learning_rate": 1.5462617897066863e-06, "loss": 0.4038, "step": 306 }, { "epoch": 1.4902912621359223, "grad_norm": 4.378749092818973, "learning_rate": 1.5187936189090668e-06, "loss": 0.4558, "step": 307 }, { "epoch": 1.4951456310679612, "grad_norm": 4.07659258923228, "learning_rate": 1.491527858584535e-06, "loss": 0.4543, "step": 308 }, { "epoch": 1.5, "grad_norm": 4.13730586703872, "learning_rate": 1.4644660940672628e-06, "loss": 0.4169, "step": 309 }, { "epoch": 1.5048543689320388, "grad_norm": 4.508381029166123, "learning_rate": 1.4376098988303406e-06, "loss": 0.4213, "step": 310 }, { "epoch": 1.5097087378640777, "grad_norm": 4.338275903526633, "learning_rate": 1.4109608343942855e-06, "loss": 0.4262, "step": 311 }, { "epoch": 1.5145631067961165, "grad_norm": 4.772965787816261, "learning_rate": 1.3845204502362442e-06, "loss": 0.449, "step": 312 }, { "epoch": 1.5194174757281553, "grad_norm": 4.71513436120805, "learning_rate": 1.35829028369991e-06, "loss": 0.4622, "step": 313 }, { "epoch": 1.5242718446601942, "grad_norm": 3.9957568067652467, "learning_rate": 1.3322718599061252e-06, "loss": 0.4202, "step": 314 }, { "epoch": 1.529126213592233, "grad_norm": 4.630497432290958, "learning_rate": 1.306466691664216e-06, "loss": 0.4441, "step": 315 }, { "epoch": 1.5339805825242718, "grad_norm": 4.2467017569500145, "learning_rate": 1.28087627938402e-06, "loss": 0.398, "step": 316 }, { "epoch": 1.5388349514563107, "grad_norm": 4.231264847945665, "learning_rate": 1.2555021109886589e-06, "loss": 0.4212, "step": 317 }, { "epoch": 1.5436893203883495, "grad_norm": 4.291061419787607, "learning_rate": 1.2303456618280141e-06, "loss": 0.421, "step": 318 }, { "epoch": 1.5485436893203883, "grad_norm": 4.46374928202354, "learning_rate": 1.2054083945929534e-06, "loss": 0.4476, "step": 319 }, { "epoch": 1.5533980582524272, "grad_norm": 3.6982923322876515, "learning_rate": 1.1806917592302763e-06, "loss": 0.4248, "step": 320 }, { "epoch": 1.558252427184466, "grad_norm": 4.418373857413791, "learning_rate": 1.1561971928584158e-06, "loss": 0.4163, "step": 321 }, { "epoch": 1.5631067961165048, "grad_norm": 3.9837995844743594, "learning_rate": 1.1319261196838782e-06, "loss": 0.3893, "step": 322 }, { "epoch": 1.5679611650485437, "grad_norm": 5.3287624087130885, "learning_rate": 1.1078799509184246e-06, "loss": 0.4268, "step": 323 }, { "epoch": 1.5728155339805825, "grad_norm": 5.138257361829326, "learning_rate": 1.0840600846970333e-06, "loss": 0.3857, "step": 324 }, { "epoch": 1.5776699029126213, "grad_norm": 4.704200975668369, "learning_rate": 1.0604679059965923e-06, "loss": 0.5596, "step": 325 }, { "epoch": 1.5825242718446602, "grad_norm": 4.339874274644684, "learning_rate": 1.0371047865553847e-06, "loss": 0.451, "step": 326 }, { "epoch": 1.587378640776699, "grad_norm": 4.148887853101413, "learning_rate": 1.0139720847933166e-06, "loss": 0.3548, "step": 327 }, { "epoch": 1.5922330097087378, "grad_norm": 4.753262194818213, "learning_rate": 9.91071145732948e-07, "loss": 0.4236, "step": 328 }, { "epoch": 1.5970873786407767, "grad_norm": 4.540220779697058, "learning_rate": 9.684033009212752e-07, "loss": 0.4209, "step": 329 }, { "epoch": 1.6019417475728155, "grad_norm": 5.224847165003191, "learning_rate": 9.459698683523205e-07, "loss": 0.4181, "step": 330 }, { "epoch": 1.6067961165048543, "grad_norm": 5.132607637774081, "learning_rate": 9.237721523904891e-07, "loss": 0.5417, "step": 331 }, { "epoch": 1.6116504854368932, "grad_norm": 3.184625685148546, "learning_rate": 9.018114436947373e-07, "loss": 0.3499, "step": 332 }, { "epoch": 1.616504854368932, "grad_norm": 4.554258438073612, "learning_rate": 8.80089019143524e-07, "loss": 0.4828, "step": 333 }, { "epoch": 1.6213592233009708, "grad_norm": 4.1770814878829, "learning_rate": 8.586061417605668e-07, "loss": 0.3553, "step": 334 }, { "epoch": 1.6262135922330097, "grad_norm": 3.8196121690200404, "learning_rate": 8.373640606414097e-07, "loss": 0.3288, "step": 335 }, { "epoch": 1.6310679611650487, "grad_norm": 4.496781666333166, "learning_rate": 8.163640108807897e-07, "loss": 0.4416, "step": 336 }, { "epoch": 1.6359223300970873, "grad_norm": 4.35546151645318, "learning_rate": 7.956072135008336e-07, "loss": 0.4127, "step": 337 }, { "epoch": 1.6407766990291264, "grad_norm": 4.795587023864251, "learning_rate": 7.750948753800508e-07, "loss": 0.4885, "step": 338 }, { "epoch": 1.645631067961165, "grad_norm": 6.008422227883028, "learning_rate": 7.548281891831715e-07, "loss": 0.5137, "step": 339 }, { "epoch": 1.650485436893204, "grad_norm": 4.253999972499131, "learning_rate": 7.348083332917927e-07, "loss": 0.3728, "step": 340 }, { "epoch": 1.6553398058252426, "grad_norm": 4.676580266713877, "learning_rate": 7.150364717358699e-07, "loss": 0.4033, "step": 341 }, { "epoch": 1.6601941747572817, "grad_norm": 4.098918930298827, "learning_rate": 6.955137541260287e-07, "loss": 0.4282, "step": 342 }, { "epoch": 1.6650485436893203, "grad_norm": 5.528552237471654, "learning_rate": 6.762413155867276e-07, "loss": 0.4606, "step": 343 }, { "epoch": 1.6699029126213594, "grad_norm": 5.283918840269048, "learning_rate": 6.572202766902569e-07, "loss": 0.478, "step": 344 }, { "epoch": 1.674757281553398, "grad_norm": 3.8318985373037147, "learning_rate": 6.384517433915794e-07, "loss": 0.3176, "step": 345 }, { "epoch": 1.679611650485437, "grad_norm": 5.0678793505848105, "learning_rate": 6.199368069640343e-07, "loss": 0.4649, "step": 346 }, { "epoch": 1.6844660194174756, "grad_norm": 4.295352351273493, "learning_rate": 6.016765439358774e-07, "loss": 0.4594, "step": 347 }, { "epoch": 1.6893203883495147, "grad_norm": 4.826675553955059, "learning_rate": 5.836720160276971e-07, "loss": 0.4858, "step": 348 }, { "epoch": 1.6941747572815533, "grad_norm": 4.345920465381718, "learning_rate": 5.659242700906719e-07, "loss": 0.4368, "step": 349 }, { "epoch": 1.6990291262135924, "grad_norm": 4.622703067279455, "learning_rate": 5.484343380457124e-07, "loss": 0.4953, "step": 350 }, { "epoch": 1.703883495145631, "grad_norm": 5.109183875190284, "learning_rate": 5.312032368234527e-07, "loss": 0.4819, "step": 351 }, { "epoch": 1.70873786407767, "grad_norm": 4.2112710044910155, "learning_rate": 5.1423196830513e-07, "loss": 0.4029, "step": 352 }, { "epoch": 1.7135922330097086, "grad_norm": 4.701355983600365, "learning_rate": 4.975215192643246e-07, "loss": 0.4028, "step": 353 }, { "epoch": 1.7184466019417477, "grad_norm": 4.167059878812549, "learning_rate": 4.81072861309591e-07, "loss": 0.3343, "step": 354 }, { "epoch": 1.7233009708737863, "grad_norm": 3.9901597887845837, "learning_rate": 4.648869508279613e-07, "loss": 0.3471, "step": 355 }, { "epoch": 1.7281553398058254, "grad_norm": 3.9583463325118307, "learning_rate": 4.4896472892933693e-07, "loss": 0.3412, "step": 356 }, { "epoch": 1.733009708737864, "grad_norm": 4.126806703784011, "learning_rate": 4.333071213917722e-07, "loss": 0.3931, "step": 357 }, { "epoch": 1.737864077669903, "grad_norm": 4.4340363361603545, "learning_rate": 4.179150386076425e-07, "loss": 0.4409, "step": 358 }, { "epoch": 1.7427184466019416, "grad_norm": 4.510979728312916, "learning_rate": 4.027893755307144e-07, "loss": 0.432, "step": 359 }, { "epoch": 1.7475728155339807, "grad_norm": 4.378809231799341, "learning_rate": 3.8793101162410417e-07, "loss": 0.4176, "step": 360 }, { "epoch": 1.7524271844660193, "grad_norm": 3.943131476031349, "learning_rate": 3.733408108091485e-07, "loss": 0.3474, "step": 361 }, { "epoch": 1.7572815533980584, "grad_norm": 4.9135907839975195, "learning_rate": 3.5901962141516975e-07, "loss": 0.4312, "step": 362 }, { "epoch": 1.762135922330097, "grad_norm": 4.657329300668235, "learning_rate": 3.4496827613015206e-07, "loss": 0.454, "step": 363 }, { "epoch": 1.766990291262136, "grad_norm": 4.204345187477797, "learning_rate": 3.3118759195232273e-07, "loss": 0.3929, "step": 364 }, { "epoch": 1.7718446601941746, "grad_norm": 4.37280312690859, "learning_rate": 3.176783701426528e-07, "loss": 0.4313, "step": 365 }, { "epoch": 1.7766990291262137, "grad_norm": 5.889047588141843, "learning_rate": 3.0444139617826605e-07, "loss": 0.5399, "step": 366 }, { "epoch": 1.7815533980582523, "grad_norm": 4.584247874736123, "learning_rate": 2.91477439706771e-07, "loss": 0.486, "step": 367 }, { "epoch": 1.7864077669902914, "grad_norm": 4.172917625522165, "learning_rate": 2.787872545015069e-07, "loss": 0.4059, "step": 368 }, { "epoch": 1.79126213592233, "grad_norm": 4.979803673311964, "learning_rate": 2.663715784177201e-07, "loss": 0.4701, "step": 369 }, { "epoch": 1.796116504854369, "grad_norm": 5.111662276722985, "learning_rate": 2.542311333496622e-07, "loss": 0.4564, "step": 370 }, { "epoch": 1.8009708737864076, "grad_norm": 4.525496837095561, "learning_rate": 2.423666251886114e-07, "loss": 0.4331, "step": 371 }, { "epoch": 1.8058252427184467, "grad_norm": 7.678846935783622, "learning_rate": 2.307787437818365e-07, "loss": 0.3864, "step": 372 }, { "epoch": 1.8106796116504853, "grad_norm": 4.478717479008666, "learning_rate": 2.1946816289248163e-07, "loss": 0.4552, "step": 373 }, { "epoch": 1.8155339805825244, "grad_norm": 4.504614128426268, "learning_rate": 2.0843554016039326e-07, "loss": 0.4273, "step": 374 }, { "epoch": 1.820388349514563, "grad_norm": 3.962970116983484, "learning_rate": 1.9768151706388016e-07, "loss": 0.3642, "step": 375 }, { "epoch": 1.825242718446602, "grad_norm": 5.2177132289814905, "learning_rate": 1.8720671888242058e-07, "loss": 0.397, "step": 376 }, { "epoch": 1.8300970873786406, "grad_norm": 4.914409178275186, "learning_rate": 1.7701175466029895e-07, "loss": 0.4609, "step": 377 }, { "epoch": 1.8349514563106797, "grad_norm": 4.724898560128005, "learning_rate": 1.6709721717120042e-07, "loss": 0.3956, "step": 378 }, { "epoch": 1.8398058252427183, "grad_norm": 4.3589930725548856, "learning_rate": 1.574636828837395e-07, "loss": 0.4422, "step": 379 }, { "epoch": 1.8446601941747574, "grad_norm": 4.004566262385127, "learning_rate": 1.4811171192794628e-07, "loss": 0.3568, "step": 380 }, { "epoch": 1.849514563106796, "grad_norm": 4.137120731244176, "learning_rate": 1.3904184806269705e-07, "loss": 0.3777, "step": 381 }, { "epoch": 1.854368932038835, "grad_norm": 3.9475318613646198, "learning_rate": 1.3025461864409395e-07, "loss": 0.3617, "step": 382 }, { "epoch": 1.8592233009708736, "grad_norm": 4.1095123692638555, "learning_rate": 1.2175053459481e-07, "loss": 0.4421, "step": 383 }, { "epoch": 1.8640776699029127, "grad_norm": 4.501880113149294, "learning_rate": 1.1353009037437523e-07, "loss": 0.3947, "step": 384 }, { "epoch": 1.8689320388349513, "grad_norm": 4.593770979573446, "learning_rate": 1.0559376395043285e-07, "loss": 0.491, "step": 385 }, { "epoch": 1.8737864077669903, "grad_norm": 3.798511279483271, "learning_rate": 9.794201677094162e-08, "loss": 0.38, "step": 386 }, { "epoch": 1.8786407766990292, "grad_norm": 4.803454088515766, "learning_rate": 9.05752937373533e-08, "loss": 0.4515, "step": 387 }, { "epoch": 1.883495145631068, "grad_norm": 4.661442569573281, "learning_rate": 8.34940231787379e-08, "loss": 0.4853, "step": 388 }, { "epoch": 1.8883495145631068, "grad_norm": 4.320440519414222, "learning_rate": 7.66986168268824e-08, "loss": 0.3878, "step": 389 }, { "epoch": 1.8932038834951457, "grad_norm": 4.557288928938388, "learning_rate": 7.018946979234997e-08, "loss": 0.4072, "step": 390 }, { "epoch": 1.8980582524271845, "grad_norm": 4.968514839079913, "learning_rate": 6.396696054150719e-08, "loss": 0.4359, "step": 391 }, { "epoch": 1.9029126213592233, "grad_norm": 4.394357400381863, "learning_rate": 5.803145087451945e-08, "loss": 0.375, "step": 392 }, { "epoch": 1.9077669902912622, "grad_norm": 4.860351986655913, "learning_rate": 5.238328590431163e-08, "loss": 0.4817, "step": 393 }, { "epoch": 1.912621359223301, "grad_norm": 3.789287914838989, "learning_rate": 4.702279403650534e-08, "loss": 0.3856, "step": 394 }, { "epoch": 1.9174757281553398, "grad_norm": 4.674754754897535, "learning_rate": 4.195028695032133e-08, "loss": 0.4212, "step": 395 }, { "epoch": 1.9223300970873787, "grad_norm": 5.579755683545321, "learning_rate": 3.716605958046071e-08, "loss": 0.4225, "step": 396 }, { "epoch": 1.9271844660194175, "grad_norm": 4.4346523529104696, "learning_rate": 3.2670390099951985e-08, "loss": 0.4321, "step": 397 }, { "epoch": 1.9320388349514563, "grad_norm": 4.485415478621165, "learning_rate": 2.846353990398065e-08, "loss": 0.4617, "step": 398 }, { "epoch": 1.9368932038834952, "grad_norm": 5.509857885947148, "learning_rate": 2.4545753594688582e-08, "loss": 0.4323, "step": 399 }, { "epoch": 1.941747572815534, "grad_norm": 5.782992007265136, "learning_rate": 2.0917258966953735e-08, "loss": 0.4941, "step": 400 }, { "epoch": 1.9466019417475728, "grad_norm": 4.575382012942858, "learning_rate": 1.757826699514298e-08, "loss": 0.4305, "step": 401 }, { "epoch": 1.9514563106796117, "grad_norm": 4.884196576797662, "learning_rate": 1.4528971820846894e-08, "loss": 0.4338, "step": 402 }, { "epoch": 1.9563106796116505, "grad_norm": 3.913224417483387, "learning_rate": 1.176955074159214e-08, "loss": 0.324, "step": 403 }, { "epoch": 1.9611650485436893, "grad_norm": 4.847779458303026, "learning_rate": 9.300164200530815e-09, "loss": 0.434, "step": 404 }, { "epoch": 1.9660194174757282, "grad_norm": 4.77264465584264, "learning_rate": 7.120955777112915e-09, "loss": 0.3824, "step": 405 }, { "epoch": 1.970873786407767, "grad_norm": 4.096993724726267, "learning_rate": 5.232052178738567e-09, "loss": 0.3225, "step": 406 }, { "epoch": 1.9757281553398058, "grad_norm": 4.667643921754874, "learning_rate": 3.633563233388926e-09, "loss": 0.3843, "step": 407 }, { "epoch": 1.9805825242718447, "grad_norm": 4.541822373694423, "learning_rate": 2.3255818832423894e-09, "loss": 0.5116, "step": 408 }, { "epoch": 1.9854368932038835, "grad_norm": 5.070880322038374, "learning_rate": 1.3081841792694783e-09, "loss": 0.4627, "step": 409 }, { "epoch": 1.9902912621359223, "grad_norm": 4.09781487009922, "learning_rate": 5.814292768108187e-10, "loss": 0.4259, "step": 410 }, { "epoch": 1.9951456310679612, "grad_norm": 4.536175330246984, "learning_rate": 1.453594321393359e-10, "loss": 0.4569, "step": 411 }, { "epoch": 2.0, "grad_norm": 3.695576408194871, "learning_rate": 0.0, "loss": 0.307, "step": 412 }, { "epoch": 2.0, "step": 412, "total_flos": 871050006528.0, "train_loss": 0.523200351274708, "train_runtime": 256.3511, "train_samples_per_second": 12.803, "train_steps_per_second": 1.607 } ], "logging_steps": 1, "max_steps": 412, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 50000000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 871050006528.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }