{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8589220528237063, "eval_steps": 500, "global_step": 1500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0005726147018824708, "grad_norm": 7.761023998260498, "learning_rate": 0.0, "loss": 6.0592, "step": 1 }, { "epoch": 0.0011452294037649416, "grad_norm": 7.8541951179504395, "learning_rate": 5.714285714285715e-07, "loss": 6.0156, "step": 2 }, { "epoch": 0.0022904588075298832, "grad_norm": 7.347611904144287, "learning_rate": 1.7142857142857145e-06, "loss": 6.0103, "step": 4 }, { "epoch": 0.003435688211294825, "grad_norm": 5.382428169250488, "learning_rate": 2.8571428571428573e-06, "loss": 5.9221, "step": 6 }, { "epoch": 0.0045809176150597665, "grad_norm": 5.063406467437744, "learning_rate": 4.000000000000001e-06, "loss": 6.0365, "step": 8 }, { "epoch": 0.005726147018824708, "grad_norm": 9.779157638549805, "learning_rate": 5.142857142857143e-06, "loss": 6.0336, "step": 10 }, { "epoch": 0.00687137642258965, "grad_norm": 7.555446147918701, "learning_rate": 6.285714285714287e-06, "loss": 6.0328, "step": 12 }, { "epoch": 0.008016605826354592, "grad_norm": 6.790043354034424, "learning_rate": 7.428571428571429e-06, "loss": 5.7848, "step": 14 }, { "epoch": 0.009161835230119533, "grad_norm": 4.4132208824157715, "learning_rate": 8.571428571428573e-06, "loss": 5.8207, "step": 16 }, { "epoch": 0.010307064633884476, "grad_norm": 4.064995765686035, "learning_rate": 9.714285714285715e-06, "loss": 5.6497, "step": 18 }, { "epoch": 0.011452294037649417, "grad_norm": 3.357184410095215, "learning_rate": 1.0857142857142858e-05, "loss": 5.7758, "step": 20 }, { "epoch": 0.012597523441414358, "grad_norm": 2.742230176925659, "learning_rate": 1.2e-05, "loss": 5.6173, "step": 22 }, { "epoch": 0.0137427528451793, "grad_norm": 2.491459369659424, "learning_rate": 1.3142857142857143e-05, "loss": 5.6681, "step": 24 }, { "epoch": 0.014887982248944241, "grad_norm": 2.7569029331207275, "learning_rate": 1.4285714285714285e-05, "loss": 5.6393, "step": 26 }, { "epoch": 0.016033211652709184, "grad_norm": 2.208378791809082, "learning_rate": 1.5428571428571428e-05, "loss": 5.5768, "step": 28 }, { "epoch": 0.017178441056474127, "grad_norm": 3.2770133018493652, "learning_rate": 1.657142857142857e-05, "loss": 5.484, "step": 30 }, { "epoch": 0.018323670460239066, "grad_norm": 3.177299976348877, "learning_rate": 1.7714285714285713e-05, "loss": 5.528, "step": 32 }, { "epoch": 0.01946889986400401, "grad_norm": 2.1981537342071533, "learning_rate": 1.885714285714286e-05, "loss": 5.6327, "step": 34 }, { "epoch": 0.02061412926776895, "grad_norm": 3.265881061553955, "learning_rate": 2e-05, "loss": 5.6288, "step": 36 }, { "epoch": 0.02175935867153389, "grad_norm": 3.6059298515319824, "learning_rate": 2.1142857142857144e-05, "loss": 5.4789, "step": 38 }, { "epoch": 0.022904588075298833, "grad_norm": 2.4080026149749756, "learning_rate": 2.2285714285714287e-05, "loss": 5.4046, "step": 40 }, { "epoch": 0.024049817479063776, "grad_norm": 2.142902135848999, "learning_rate": 2.342857142857143e-05, "loss": 5.4738, "step": 42 }, { "epoch": 0.025195046882828715, "grad_norm": 2.4021224975585938, "learning_rate": 2.4571428571428572e-05, "loss": 5.4649, "step": 44 }, { "epoch": 0.026340276286593658, "grad_norm": 2.172009229660034, "learning_rate": 2.5714285714285714e-05, "loss": 5.4302, "step": 46 }, { "epoch": 0.0274855056903586, "grad_norm": 2.9737730026245117, "learning_rate": 2.6857142857142857e-05, "loss": 5.3045, "step": 48 }, { "epoch": 0.028630735094123543, "grad_norm": 3.0378615856170654, "learning_rate": 2.8000000000000003e-05, "loss": 5.2185, "step": 50 }, { "epoch": 0.029775964497888482, "grad_norm": 3.4448676109313965, "learning_rate": 2.9142857142857146e-05, "loss": 5.1838, "step": 52 }, { "epoch": 0.030921193901653425, "grad_norm": 2.469245672225952, "learning_rate": 3.0285714285714288e-05, "loss": 5.1637, "step": 54 }, { "epoch": 0.03206642330541837, "grad_norm": 3.58486008644104, "learning_rate": 3.142857142857143e-05, "loss": 5.2063, "step": 56 }, { "epoch": 0.03321165270918331, "grad_norm": 3.0815446376800537, "learning_rate": 3.257142857142857e-05, "loss": 5.2317, "step": 58 }, { "epoch": 0.03435688211294825, "grad_norm": 3.6842119693756104, "learning_rate": 3.3714285714285716e-05, "loss": 5.2695, "step": 60 }, { "epoch": 0.03550211151671319, "grad_norm": 2.9440791606903076, "learning_rate": 3.485714285714286e-05, "loss": 5.2686, "step": 62 }, { "epoch": 0.03664734092047813, "grad_norm": 3.9632568359375, "learning_rate": 3.6e-05, "loss": 5.1262, "step": 64 }, { "epoch": 0.037792570324243074, "grad_norm": 4.045065402984619, "learning_rate": 3.7142857142857143e-05, "loss": 5.1546, "step": 66 }, { "epoch": 0.03893779972800802, "grad_norm": 3.5707085132598877, "learning_rate": 3.8285714285714286e-05, "loss": 5.0036, "step": 68 }, { "epoch": 0.04008302913177296, "grad_norm": 3.014404535293579, "learning_rate": 3.942857142857143e-05, "loss": 5.026, "step": 70 }, { "epoch": 0.0412282585355379, "grad_norm": 2.708796977996826, "learning_rate": 4.057142857142857e-05, "loss": 4.9442, "step": 72 }, { "epoch": 0.04237348793930284, "grad_norm": 2.5384011268615723, "learning_rate": 4.1714285714285714e-05, "loss": 5.0223, "step": 74 }, { "epoch": 0.04351871734306778, "grad_norm": 3.006281852722168, "learning_rate": 4.2857142857142856e-05, "loss": 4.9827, "step": 76 }, { "epoch": 0.044663946746832724, "grad_norm": 2.5772130489349365, "learning_rate": 4.4000000000000006e-05, "loss": 4.9675, "step": 78 }, { "epoch": 0.045809176150597666, "grad_norm": 3.456017255783081, "learning_rate": 4.514285714285714e-05, "loss": 5.0341, "step": 80 }, { "epoch": 0.04695440555436261, "grad_norm": 3.3163113594055176, "learning_rate": 4.628571428571429e-05, "loss": 4.9867, "step": 82 }, { "epoch": 0.04809963495812755, "grad_norm": 3.7568469047546387, "learning_rate": 4.742857142857143e-05, "loss": 4.8652, "step": 84 }, { "epoch": 0.049244864361892494, "grad_norm": 4.19318151473999, "learning_rate": 4.8571428571428576e-05, "loss": 5.0602, "step": 86 }, { "epoch": 0.05039009376565743, "grad_norm": 5.1034064292907715, "learning_rate": 4.971428571428572e-05, "loss": 4.9757, "step": 88 }, { "epoch": 0.05153532316942237, "grad_norm": 4.0827484130859375, "learning_rate": 5.085714285714286e-05, "loss": 4.8486, "step": 90 }, { "epoch": 0.052680552573187316, "grad_norm": 4.6189446449279785, "learning_rate": 5.2000000000000004e-05, "loss": 4.9595, "step": 92 }, { "epoch": 0.05382578197695226, "grad_norm": 3.988513469696045, "learning_rate": 5.314285714285715e-05, "loss": 4.9035, "step": 94 }, { "epoch": 0.0549710113807172, "grad_norm": 3.857276678085327, "learning_rate": 5.428571428571428e-05, "loss": 4.8277, "step": 96 }, { "epoch": 0.056116240784482144, "grad_norm": 3.5372354984283447, "learning_rate": 5.542857142857143e-05, "loss": 4.7718, "step": 98 }, { "epoch": 0.057261470188247086, "grad_norm": 3.3853676319122314, "learning_rate": 5.6571428571428574e-05, "loss": 4.8098, "step": 100 }, { "epoch": 0.05840669959201202, "grad_norm": 2.1142077445983887, "learning_rate": 5.771428571428572e-05, "loss": 4.7975, "step": 102 }, { "epoch": 0.059551928995776965, "grad_norm": 3.2275538444519043, "learning_rate": 5.885714285714285e-05, "loss": 4.8509, "step": 104 }, { "epoch": 0.06069715839954191, "grad_norm": 3.5413126945495605, "learning_rate": 6e-05, "loss": 4.6069, "step": 106 }, { "epoch": 0.06184238780330685, "grad_norm": 2.755648374557495, "learning_rate": 6.114285714285714e-05, "loss": 4.6951, "step": 108 }, { "epoch": 0.06298761720707179, "grad_norm": 2.980039596557617, "learning_rate": 6.22857142857143e-05, "loss": 4.7012, "step": 110 }, { "epoch": 0.06413284661083674, "grad_norm": 4.890020370483398, "learning_rate": 6.342857142857143e-05, "loss": 4.8008, "step": 112 }, { "epoch": 0.06527807601460167, "grad_norm": 4.35846471786499, "learning_rate": 6.457142857142856e-05, "loss": 4.8587, "step": 114 }, { "epoch": 0.06642330541836662, "grad_norm": 3.6171813011169434, "learning_rate": 6.571428571428571e-05, "loss": 4.7473, "step": 116 }, { "epoch": 0.06756853482213156, "grad_norm": 2.4927010536193848, "learning_rate": 6.685714285714286e-05, "loss": 4.7113, "step": 118 }, { "epoch": 0.0687137642258965, "grad_norm": 3.3327009677886963, "learning_rate": 6.800000000000001e-05, "loss": 4.6105, "step": 120 }, { "epoch": 0.06985899362966144, "grad_norm": 3.1123206615448, "learning_rate": 6.914285714285715e-05, "loss": 4.5968, "step": 122 }, { "epoch": 0.07100422303342638, "grad_norm": 2.6985421180725098, "learning_rate": 7.028571428571428e-05, "loss": 4.6323, "step": 124 }, { "epoch": 0.07214945243719133, "grad_norm": 2.058084011077881, "learning_rate": 7.142857142857143e-05, "loss": 4.5721, "step": 126 }, { "epoch": 0.07329468184095626, "grad_norm": 2.144658327102661, "learning_rate": 7.257142857142858e-05, "loss": 4.6125, "step": 128 }, { "epoch": 0.07443991124472121, "grad_norm": 2.477219820022583, "learning_rate": 7.371428571428572e-05, "loss": 4.4727, "step": 130 }, { "epoch": 0.07558514064848615, "grad_norm": 3.8517298698425293, "learning_rate": 7.485714285714285e-05, "loss": 4.5696, "step": 132 }, { "epoch": 0.0767303700522511, "grad_norm": 3.0253565311431885, "learning_rate": 7.6e-05, "loss": 4.4838, "step": 134 }, { "epoch": 0.07787559945601603, "grad_norm": 3.397569179534912, "learning_rate": 7.714285714285715e-05, "loss": 4.6431, "step": 136 }, { "epoch": 0.07902082885978097, "grad_norm": 2.435197114944458, "learning_rate": 7.828571428571429e-05, "loss": 4.4681, "step": 138 }, { "epoch": 0.08016605826354592, "grad_norm": 2.6476476192474365, "learning_rate": 7.942857142857143e-05, "loss": 4.4462, "step": 140 }, { "epoch": 0.08131128766731086, "grad_norm": 2.1929690837860107, "learning_rate": 8.057142857142857e-05, "loss": 4.5136, "step": 142 }, { "epoch": 0.0824565170710758, "grad_norm": 2.4533395767211914, "learning_rate": 8.171428571428572e-05, "loss": 4.5572, "step": 144 }, { "epoch": 0.08360174647484074, "grad_norm": 2.601806879043579, "learning_rate": 8.285714285714287e-05, "loss": 4.4121, "step": 146 }, { "epoch": 0.08474697587860568, "grad_norm": 3.233973741531372, "learning_rate": 8.4e-05, "loss": 4.4599, "step": 148 }, { "epoch": 0.08589220528237063, "grad_norm": 2.6353538036346436, "learning_rate": 8.514285714285714e-05, "loss": 4.4533, "step": 150 }, { "epoch": 0.08703743468613556, "grad_norm": 2.8465511798858643, "learning_rate": 8.62857142857143e-05, "loss": 4.5246, "step": 152 }, { "epoch": 0.08818266408990051, "grad_norm": 2.8642711639404297, "learning_rate": 8.742857142857144e-05, "loss": 4.4659, "step": 154 }, { "epoch": 0.08932789349366545, "grad_norm": 2.793112277984619, "learning_rate": 8.857142857142857e-05, "loss": 4.5107, "step": 156 }, { "epoch": 0.0904731228974304, "grad_norm": 3.43472957611084, "learning_rate": 8.971428571428571e-05, "loss": 4.4079, "step": 158 }, { "epoch": 0.09161835230119533, "grad_norm": 2.9260294437408447, "learning_rate": 9.085714285714286e-05, "loss": 4.4047, "step": 160 }, { "epoch": 0.09276358170496027, "grad_norm": 2.6336724758148193, "learning_rate": 9.200000000000001e-05, "loss": 4.4777, "step": 162 }, { "epoch": 0.09390881110872522, "grad_norm": 2.8348231315612793, "learning_rate": 9.314285714285715e-05, "loss": 4.3445, "step": 164 }, { "epoch": 0.09505404051249015, "grad_norm": 4.271595478057861, "learning_rate": 9.428571428571429e-05, "loss": 4.4234, "step": 166 }, { "epoch": 0.0961992699162551, "grad_norm": 3.4789109230041504, "learning_rate": 9.542857142857143e-05, "loss": 4.2872, "step": 168 }, { "epoch": 0.09734449932002004, "grad_norm": 2.57273530960083, "learning_rate": 9.657142857142858e-05, "loss": 4.4177, "step": 170 }, { "epoch": 0.09848972872378499, "grad_norm": 2.185086250305176, "learning_rate": 9.771428571428572e-05, "loss": 4.3568, "step": 172 }, { "epoch": 0.09963495812754992, "grad_norm": 2.771744966506958, "learning_rate": 9.885714285714286e-05, "loss": 4.3392, "step": 174 }, { "epoch": 0.10078018753131486, "grad_norm": 1.950353741645813, "learning_rate": 0.0001, "loss": 4.1931, "step": 176 }, { "epoch": 0.10192541693507981, "grad_norm": 2.4709694385528564, "learning_rate": 9.999991040472416e-05, "loss": 4.2936, "step": 178 }, { "epoch": 0.10307064633884475, "grad_norm": 2.140997886657715, "learning_rate": 9.999964161921776e-05, "loss": 4.1653, "step": 180 }, { "epoch": 0.1042158757426097, "grad_norm": 2.491321563720703, "learning_rate": 9.999919364444403e-05, "loss": 4.3202, "step": 182 }, { "epoch": 0.10536110514637463, "grad_norm": 2.5410189628601074, "learning_rate": 9.999856648200845e-05, "loss": 4.2657, "step": 184 }, { "epoch": 0.10650633455013958, "grad_norm": 2.1820590496063232, "learning_rate": 9.999776013415866e-05, "loss": 4.2282, "step": 186 }, { "epoch": 0.10765156395390452, "grad_norm": 1.7251808643341064, "learning_rate": 9.999677460378444e-05, "loss": 4.3421, "step": 188 }, { "epoch": 0.10879679335766945, "grad_norm": 2.002145290374756, "learning_rate": 9.999560989441779e-05, "loss": 4.1361, "step": 190 }, { "epoch": 0.1099420227614344, "grad_norm": 1.9663431644439697, "learning_rate": 9.999426601023274e-05, "loss": 4.201, "step": 192 }, { "epoch": 0.11108725216519934, "grad_norm": 2.1406776905059814, "learning_rate": 9.999274295604558e-05, "loss": 4.1086, "step": 194 }, { "epoch": 0.11223248156896429, "grad_norm": 3.3888607025146484, "learning_rate": 9.999104073731458e-05, "loss": 4.2723, "step": 196 }, { "epoch": 0.11337771097272922, "grad_norm": 2.371840715408325, "learning_rate": 9.998915936014024e-05, "loss": 4.1893, "step": 198 }, { "epoch": 0.11452294037649417, "grad_norm": 2.0502302646636963, "learning_rate": 9.998709883126502e-05, "loss": 4.1395, "step": 200 }, { "epoch": 0.11566816978025911, "grad_norm": 1.6674678325653076, "learning_rate": 9.998485915807347e-05, "loss": 4.071, "step": 202 }, { "epoch": 0.11681339918402404, "grad_norm": 1.7829004526138306, "learning_rate": 9.998244034859219e-05, "loss": 4.1107, "step": 204 }, { "epoch": 0.117958628587789, "grad_norm": 1.763493299484253, "learning_rate": 9.997984241148967e-05, "loss": 4.1142, "step": 206 }, { "epoch": 0.11910385799155393, "grad_norm": 2.069258213043213, "learning_rate": 9.997706535607649e-05, "loss": 4.047, "step": 208 }, { "epoch": 0.12024908739531888, "grad_norm": 2.4262139797210693, "learning_rate": 9.997410919230505e-05, "loss": 4.0396, "step": 210 }, { "epoch": 0.12139431679908382, "grad_norm": 1.820494532585144, "learning_rate": 9.997097393076971e-05, "loss": 4.1548, "step": 212 }, { "epoch": 0.12253954620284876, "grad_norm": 2.1332643032073975, "learning_rate": 9.996765958270664e-05, "loss": 4.1384, "step": 214 }, { "epoch": 0.1236847756066137, "grad_norm": 2.1329920291900635, "learning_rate": 9.996416615999384e-05, "loss": 4.0315, "step": 216 }, { "epoch": 0.12483000501037864, "grad_norm": 2.29955792427063, "learning_rate": 9.996049367515108e-05, "loss": 4.0963, "step": 218 }, { "epoch": 0.12597523441414357, "grad_norm": 2.225827693939209, "learning_rate": 9.995664214133983e-05, "loss": 4.1247, "step": 220 }, { "epoch": 0.12712046381790854, "grad_norm": 1.794838786125183, "learning_rate": 9.99526115723633e-05, "loss": 4.0449, "step": 222 }, { "epoch": 0.12826569322167347, "grad_norm": 1.7548491954803467, "learning_rate": 9.994840198266626e-05, "loss": 3.927, "step": 224 }, { "epoch": 0.1294109226254384, "grad_norm": 1.487001895904541, "learning_rate": 9.994401338733508e-05, "loss": 3.9714, "step": 226 }, { "epoch": 0.13055615202920334, "grad_norm": 1.9811242818832397, "learning_rate": 9.993944580209768e-05, "loss": 4.0094, "step": 228 }, { "epoch": 0.13170138143296828, "grad_norm": 1.4257248640060425, "learning_rate": 9.99346992433234e-05, "loss": 4.0213, "step": 230 }, { "epoch": 0.13284661083673324, "grad_norm": 1.545812726020813, "learning_rate": 9.992977372802302e-05, "loss": 4.0076, "step": 232 }, { "epoch": 0.13399184024049818, "grad_norm": 1.8193179368972778, "learning_rate": 9.992466927384865e-05, "loss": 4.0536, "step": 234 }, { "epoch": 0.1351370696442631, "grad_norm": 2.329951763153076, "learning_rate": 9.991938589909369e-05, "loss": 3.9284, "step": 236 }, { "epoch": 0.13628229904802805, "grad_norm": 1.928336501121521, "learning_rate": 9.991392362269276e-05, "loss": 3.9462, "step": 238 }, { "epoch": 0.137427528451793, "grad_norm": 1.4073456525802612, "learning_rate": 9.990828246422164e-05, "loss": 3.9525, "step": 240 }, { "epoch": 0.13857275785555795, "grad_norm": 1.6663973331451416, "learning_rate": 9.990246244389713e-05, "loss": 3.9685, "step": 242 }, { "epoch": 0.13971798725932288, "grad_norm": 1.8091737031936646, "learning_rate": 9.989646358257715e-05, "loss": 3.9284, "step": 244 }, { "epoch": 0.14086321666308782, "grad_norm": 1.5511283874511719, "learning_rate": 9.989028590176044e-05, "loss": 3.9289, "step": 246 }, { "epoch": 0.14200844606685276, "grad_norm": 1.5394625663757324, "learning_rate": 9.988392942358664e-05, "loss": 3.9849, "step": 248 }, { "epoch": 0.14315367547061772, "grad_norm": 1.680882453918457, "learning_rate": 9.98773941708362e-05, "loss": 3.9452, "step": 250 }, { "epoch": 0.14429890487438266, "grad_norm": 1.6341670751571655, "learning_rate": 9.98706801669302e-05, "loss": 3.8317, "step": 252 }, { "epoch": 0.1454441342781476, "grad_norm": 1.9933757781982422, "learning_rate": 9.986378743593036e-05, "loss": 3.9665, "step": 254 }, { "epoch": 0.14658936368191253, "grad_norm": 2.2253994941711426, "learning_rate": 9.985671600253894e-05, "loss": 3.9239, "step": 256 }, { "epoch": 0.14773459308567746, "grad_norm": 2.2543365955352783, "learning_rate": 9.984946589209862e-05, "loss": 3.8639, "step": 258 }, { "epoch": 0.14887982248944243, "grad_norm": 1.8106629848480225, "learning_rate": 9.984203713059241e-05, "loss": 3.9178, "step": 260 }, { "epoch": 0.15002505189320736, "grad_norm": 1.638542652130127, "learning_rate": 9.983442974464362e-05, "loss": 3.9169, "step": 262 }, { "epoch": 0.1511702812969723, "grad_norm": 1.3521384000778198, "learning_rate": 9.982664376151564e-05, "loss": 3.8682, "step": 264 }, { "epoch": 0.15231551070073723, "grad_norm": 1.6458699703216553, "learning_rate": 9.981867920911201e-05, "loss": 3.9566, "step": 266 }, { "epoch": 0.1534607401045022, "grad_norm": 1.7851066589355469, "learning_rate": 9.981053611597615e-05, "loss": 3.9085, "step": 268 }, { "epoch": 0.15460596950826713, "grad_norm": 1.6740517616271973, "learning_rate": 9.980221451129137e-05, "loss": 3.8899, "step": 270 }, { "epoch": 0.15575119891203207, "grad_norm": 1.117129921913147, "learning_rate": 9.979371442488073e-05, "loss": 3.7544, "step": 272 }, { "epoch": 0.156896428315797, "grad_norm": 1.5676058530807495, "learning_rate": 9.978503588720694e-05, "loss": 3.7753, "step": 274 }, { "epoch": 0.15804165771956194, "grad_norm": 1.6609163284301758, "learning_rate": 9.977617892937223e-05, "loss": 3.8463, "step": 276 }, { "epoch": 0.1591868871233269, "grad_norm": 1.7229987382888794, "learning_rate": 9.976714358311828e-05, "loss": 3.8446, "step": 278 }, { "epoch": 0.16033211652709184, "grad_norm": 1.6770962476730347, "learning_rate": 9.975792988082603e-05, "loss": 3.8684, "step": 280 }, { "epoch": 0.16147734593085677, "grad_norm": 1.215281367301941, "learning_rate": 9.974853785551568e-05, "loss": 3.7788, "step": 282 }, { "epoch": 0.1626225753346217, "grad_norm": 1.208257794380188, "learning_rate": 9.973896754084646e-05, "loss": 3.8338, "step": 284 }, { "epoch": 0.16376780473838665, "grad_norm": 1.4068255424499512, "learning_rate": 9.972921897111658e-05, "loss": 3.8583, "step": 286 }, { "epoch": 0.1649130341421516, "grad_norm": 1.4898021221160889, "learning_rate": 9.971929218126306e-05, "loss": 3.8051, "step": 288 }, { "epoch": 0.16605826354591655, "grad_norm": 1.6303211450576782, "learning_rate": 9.970918720686164e-05, "loss": 3.8598, "step": 290 }, { "epoch": 0.16720349294968148, "grad_norm": 1.6599496603012085, "learning_rate": 9.969890408412665e-05, "loss": 3.7214, "step": 292 }, { "epoch": 0.16834872235344642, "grad_norm": 1.1958950757980347, "learning_rate": 9.968844284991086e-05, "loss": 3.7042, "step": 294 }, { "epoch": 0.16949395175721135, "grad_norm": 1.3099420070648193, "learning_rate": 9.967780354170533e-05, "loss": 3.7405, "step": 296 }, { "epoch": 0.17063918116097632, "grad_norm": 1.5054072141647339, "learning_rate": 9.966698619763936e-05, "loss": 3.7827, "step": 298 }, { "epoch": 0.17178441056474125, "grad_norm": 1.444757103919983, "learning_rate": 9.965599085648025e-05, "loss": 3.7361, "step": 300 }, { "epoch": 0.1729296399685062, "grad_norm": 0.9423370361328125, "learning_rate": 9.964481755763322e-05, "loss": 3.7063, "step": 302 }, { "epoch": 0.17407486937227112, "grad_norm": 1.044169306755066, "learning_rate": 9.963346634114128e-05, "loss": 3.7999, "step": 304 }, { "epoch": 0.1752200987760361, "grad_norm": 1.578296184539795, "learning_rate": 9.962193724768503e-05, "loss": 3.7448, "step": 306 }, { "epoch": 0.17636532817980102, "grad_norm": 1.4953491687774658, "learning_rate": 9.961023031858258e-05, "loss": 3.7625, "step": 308 }, { "epoch": 0.17751055758356596, "grad_norm": 1.295817494392395, "learning_rate": 9.959834559578934e-05, "loss": 3.7042, "step": 310 }, { "epoch": 0.1786557869873309, "grad_norm": 1.4001609086990356, "learning_rate": 9.95862831218979e-05, "loss": 3.7272, "step": 312 }, { "epoch": 0.17980101639109583, "grad_norm": 1.8881722688674927, "learning_rate": 9.95740429401379e-05, "loss": 3.6904, "step": 314 }, { "epoch": 0.1809462457948608, "grad_norm": 1.919791340827942, "learning_rate": 9.956162509437584e-05, "loss": 3.7071, "step": 316 }, { "epoch": 0.18209147519862573, "grad_norm": 1.758253574371338, "learning_rate": 9.954902962911494e-05, "loss": 3.7906, "step": 318 }, { "epoch": 0.18323670460239067, "grad_norm": 1.480323314666748, "learning_rate": 9.953625658949494e-05, "loss": 3.7697, "step": 320 }, { "epoch": 0.1843819340061556, "grad_norm": 1.5573948621749878, "learning_rate": 9.952330602129202e-05, "loss": 3.752, "step": 322 }, { "epoch": 0.18552716340992054, "grad_norm": 1.3204878568649292, "learning_rate": 9.951017797091858e-05, "loss": 3.6479, "step": 324 }, { "epoch": 0.1866723928136855, "grad_norm": 1.5514147281646729, "learning_rate": 9.949687248542303e-05, "loss": 3.7199, "step": 326 }, { "epoch": 0.18781762221745044, "grad_norm": 1.2910770177841187, "learning_rate": 9.948338961248977e-05, "loss": 3.7427, "step": 328 }, { "epoch": 0.18896285162121537, "grad_norm": 1.1663178205490112, "learning_rate": 9.946972940043882e-05, "loss": 3.6616, "step": 330 }, { "epoch": 0.1901080810249803, "grad_norm": 1.3439650535583496, "learning_rate": 9.945589189822584e-05, "loss": 3.7385, "step": 332 }, { "epoch": 0.19125331042874527, "grad_norm": 1.1256877183914185, "learning_rate": 9.94418771554418e-05, "loss": 3.6056, "step": 334 }, { "epoch": 0.1923985398325102, "grad_norm": 1.1813896894454956, "learning_rate": 9.942768522231289e-05, "loss": 3.5544, "step": 336 }, { "epoch": 0.19354376923627514, "grad_norm": 1.2541157007217407, "learning_rate": 9.941331614970031e-05, "loss": 3.6401, "step": 338 }, { "epoch": 0.19468899864004008, "grad_norm": 1.237069010734558, "learning_rate": 9.939876998910012e-05, "loss": 3.7564, "step": 340 }, { "epoch": 0.19583422804380501, "grad_norm": 1.1157530546188354, "learning_rate": 9.938404679264301e-05, "loss": 3.6164, "step": 342 }, { "epoch": 0.19697945744756998, "grad_norm": 1.149465560913086, "learning_rate": 9.936914661309412e-05, "loss": 3.6968, "step": 344 }, { "epoch": 0.1981246868513349, "grad_norm": 0.9530683755874634, "learning_rate": 9.93540695038529e-05, "loss": 3.6194, "step": 346 }, { "epoch": 0.19926991625509985, "grad_norm": 1.1686296463012695, "learning_rate": 9.933881551895281e-05, "loss": 3.7604, "step": 348 }, { "epoch": 0.20041514565886479, "grad_norm": 1.2699095010757446, "learning_rate": 9.93233847130613e-05, "loss": 3.6371, "step": 350 }, { "epoch": 0.20156037506262972, "grad_norm": 1.1345208883285522, "learning_rate": 9.930777714147945e-05, "loss": 3.6146, "step": 352 }, { "epoch": 0.20270560446639468, "grad_norm": 1.3319895267486572, "learning_rate": 9.929199286014185e-05, "loss": 3.6443, "step": 354 }, { "epoch": 0.20385083387015962, "grad_norm": 1.6053088903427124, "learning_rate": 9.927603192561637e-05, "loss": 3.6277, "step": 356 }, { "epoch": 0.20499606327392456, "grad_norm": 1.2149386405944824, "learning_rate": 9.925989439510398e-05, "loss": 3.5555, "step": 358 }, { "epoch": 0.2061412926776895, "grad_norm": 1.0859287977218628, "learning_rate": 9.924358032643855e-05, "loss": 3.6253, "step": 360 }, { "epoch": 0.20728652208145446, "grad_norm": 0.9613994359970093, "learning_rate": 9.922708977808663e-05, "loss": 3.5826, "step": 362 }, { "epoch": 0.2084317514852194, "grad_norm": 1.0509222745895386, "learning_rate": 9.921042280914721e-05, "loss": 3.6263, "step": 364 }, { "epoch": 0.20957698088898433, "grad_norm": 1.3777049779891968, "learning_rate": 9.919357947935156e-05, "loss": 3.6187, "step": 366 }, { "epoch": 0.21072221029274926, "grad_norm": 1.3364644050598145, "learning_rate": 9.9176559849063e-05, "loss": 3.5946, "step": 368 }, { "epoch": 0.2118674396965142, "grad_norm": 1.4562104940414429, "learning_rate": 9.915936397927665e-05, "loss": 3.6099, "step": 370 }, { "epoch": 0.21301266910027916, "grad_norm": 1.066383719444275, "learning_rate": 9.91419919316193e-05, "loss": 3.5395, "step": 372 }, { "epoch": 0.2141578985040441, "grad_norm": 1.6498733758926392, "learning_rate": 9.912444376834903e-05, "loss": 3.6083, "step": 374 }, { "epoch": 0.21530312790780903, "grad_norm": 0.9828553795814514, "learning_rate": 9.910671955235518e-05, "loss": 3.5409, "step": 376 }, { "epoch": 0.21644835731157397, "grad_norm": 1.178269624710083, "learning_rate": 9.908881934715798e-05, "loss": 3.6018, "step": 378 }, { "epoch": 0.2175935867153389, "grad_norm": 1.3328818082809448, "learning_rate": 9.907074321690838e-05, "loss": 3.5718, "step": 380 }, { "epoch": 0.21873881611910387, "grad_norm": 1.1077896356582642, "learning_rate": 9.905249122638783e-05, "loss": 3.581, "step": 382 }, { "epoch": 0.2198840455228688, "grad_norm": 1.220638394355774, "learning_rate": 9.903406344100798e-05, "loss": 3.5813, "step": 384 }, { "epoch": 0.22102927492663374, "grad_norm": 1.5574766397476196, "learning_rate": 9.901545992681057e-05, "loss": 3.5785, "step": 386 }, { "epoch": 0.22217450433039868, "grad_norm": 1.013902187347412, "learning_rate": 9.899668075046706e-05, "loss": 3.6156, "step": 388 }, { "epoch": 0.2233197337341636, "grad_norm": 1.197936773300171, "learning_rate": 9.897772597927848e-05, "loss": 3.5428, "step": 390 }, { "epoch": 0.22446496313792857, "grad_norm": 0.9838180541992188, "learning_rate": 9.895859568117512e-05, "loss": 3.534, "step": 392 }, { "epoch": 0.2256101925416935, "grad_norm": 1.0316840410232544, "learning_rate": 9.893928992471639e-05, "loss": 3.5691, "step": 394 }, { "epoch": 0.22675542194545845, "grad_norm": 0.9378739595413208, "learning_rate": 9.891980877909045e-05, "loss": 3.5368, "step": 396 }, { "epoch": 0.22790065134922338, "grad_norm": 1.4947346448898315, "learning_rate": 9.890015231411404e-05, "loss": 3.5709, "step": 398 }, { "epoch": 0.22904588075298835, "grad_norm": 0.9118148684501648, "learning_rate": 9.888032060023225e-05, "loss": 3.527, "step": 400 }, { "epoch": 0.23019111015675328, "grad_norm": 1.2407753467559814, "learning_rate": 9.886031370851816e-05, "loss": 3.5301, "step": 402 }, { "epoch": 0.23133633956051822, "grad_norm": 1.7163093090057373, "learning_rate": 9.88401317106727e-05, "loss": 3.5828, "step": 404 }, { "epoch": 0.23248156896428315, "grad_norm": 1.0757009983062744, "learning_rate": 9.881977467902434e-05, "loss": 3.4831, "step": 406 }, { "epoch": 0.2336267983680481, "grad_norm": 0.9473862648010254, "learning_rate": 9.879924268652885e-05, "loss": 3.5196, "step": 408 }, { "epoch": 0.23477202777181305, "grad_norm": 1.199771761894226, "learning_rate": 9.877853580676897e-05, "loss": 3.574, "step": 410 }, { "epoch": 0.235917257175578, "grad_norm": 0.9006698131561279, "learning_rate": 9.875765411395428e-05, "loss": 3.5348, "step": 412 }, { "epoch": 0.23706248657934292, "grad_norm": 1.1242282390594482, "learning_rate": 9.873659768292081e-05, "loss": 3.5249, "step": 414 }, { "epoch": 0.23820771598310786, "grad_norm": 1.0675747394561768, "learning_rate": 9.871536658913082e-05, "loss": 3.5086, "step": 416 }, { "epoch": 0.2393529453868728, "grad_norm": 0.8544116616249084, "learning_rate": 9.869396090867255e-05, "loss": 3.546, "step": 418 }, { "epoch": 0.24049817479063776, "grad_norm": 1.3136742115020752, "learning_rate": 9.867238071825992e-05, "loss": 3.4937, "step": 420 }, { "epoch": 0.2416434041944027, "grad_norm": 1.3740772008895874, "learning_rate": 9.865062609523223e-05, "loss": 3.4303, "step": 422 }, { "epoch": 0.24278863359816763, "grad_norm": 1.342213749885559, "learning_rate": 9.862869711755397e-05, "loss": 3.4982, "step": 424 }, { "epoch": 0.24393386300193257, "grad_norm": 1.0677942037582397, "learning_rate": 9.860659386381443e-05, "loss": 3.4288, "step": 426 }, { "epoch": 0.24507909240569753, "grad_norm": 0.9615838527679443, "learning_rate": 9.858431641322749e-05, "loss": 3.4787, "step": 428 }, { "epoch": 0.24622432180946247, "grad_norm": 1.0572890043258667, "learning_rate": 9.856186484563134e-05, "loss": 3.5314, "step": 430 }, { "epoch": 0.2473695512132274, "grad_norm": 1.158275842666626, "learning_rate": 9.853923924148815e-05, "loss": 3.5504, "step": 432 }, { "epoch": 0.24851478061699234, "grad_norm": 1.171581745147705, "learning_rate": 9.851643968188383e-05, "loss": 3.5478, "step": 434 }, { "epoch": 0.24966001002075727, "grad_norm": 1.0333714485168457, "learning_rate": 9.849346624852764e-05, "loss": 3.5497, "step": 436 }, { "epoch": 0.2508052394245222, "grad_norm": 0.9459155797958374, "learning_rate": 9.847031902375207e-05, "loss": 3.5074, "step": 438 }, { "epoch": 0.25195046882828714, "grad_norm": 1.0424790382385254, "learning_rate": 9.84469980905124e-05, "loss": 3.4961, "step": 440 }, { "epoch": 0.25309569823205214, "grad_norm": 1.0463571548461914, "learning_rate": 9.842350353238642e-05, "loss": 3.4405, "step": 442 }, { "epoch": 0.25424092763581707, "grad_norm": 1.000319242477417, "learning_rate": 9.839983543357421e-05, "loss": 3.4595, "step": 444 }, { "epoch": 0.255386157039582, "grad_norm": 1.2526150941848755, "learning_rate": 9.837599387889773e-05, "loss": 3.5012, "step": 446 }, { "epoch": 0.25653138644334694, "grad_norm": 1.3148843050003052, "learning_rate": 9.835197895380065e-05, "loss": 3.4767, "step": 448 }, { "epoch": 0.2576766158471119, "grad_norm": 1.3939634561538696, "learning_rate": 9.83277907443479e-05, "loss": 3.3783, "step": 450 }, { "epoch": 0.2588218452508768, "grad_norm": 1.0367929935455322, "learning_rate": 9.830342933722545e-05, "loss": 3.4289, "step": 452 }, { "epoch": 0.25996707465464175, "grad_norm": 0.9439120888710022, "learning_rate": 9.827889481974e-05, "loss": 3.4728, "step": 454 }, { "epoch": 0.2611123040584067, "grad_norm": 1.2146074771881104, "learning_rate": 9.82541872798186e-05, "loss": 3.4257, "step": 456 }, { "epoch": 0.2622575334621716, "grad_norm": 1.0530729293823242, "learning_rate": 9.822930680600841e-05, "loss": 3.4681, "step": 458 }, { "epoch": 0.26340276286593656, "grad_norm": 1.1026678085327148, "learning_rate": 9.820425348747637e-05, "loss": 3.4298, "step": 460 }, { "epoch": 0.26454799226970155, "grad_norm": 1.2520779371261597, "learning_rate": 9.817902741400879e-05, "loss": 3.4191, "step": 462 }, { "epoch": 0.2656932216734665, "grad_norm": 1.1041593551635742, "learning_rate": 9.815362867601121e-05, "loss": 3.466, "step": 464 }, { "epoch": 0.2668384510772314, "grad_norm": 0.881693422794342, "learning_rate": 9.812805736450786e-05, "loss": 3.4929, "step": 466 }, { "epoch": 0.26798368048099636, "grad_norm": 1.3125033378601074, "learning_rate": 9.810231357114152e-05, "loss": 3.4592, "step": 468 }, { "epoch": 0.2691289098847613, "grad_norm": 1.2968268394470215, "learning_rate": 9.807639738817307e-05, "loss": 3.4851, "step": 470 }, { "epoch": 0.2702741392885262, "grad_norm": 0.9855544567108154, "learning_rate": 9.805030890848119e-05, "loss": 3.4487, "step": 472 }, { "epoch": 0.27141936869229116, "grad_norm": 1.3063323497772217, "learning_rate": 9.802404822556209e-05, "loss": 3.4961, "step": 474 }, { "epoch": 0.2725645980960561, "grad_norm": 1.0567957162857056, "learning_rate": 9.79976154335291e-05, "loss": 3.3975, "step": 476 }, { "epoch": 0.27370982749982103, "grad_norm": 0.9473979473114014, "learning_rate": 9.797101062711231e-05, "loss": 3.4573, "step": 478 }, { "epoch": 0.274855056903586, "grad_norm": 1.2931294441223145, "learning_rate": 9.794423390165837e-05, "loss": 3.3732, "step": 480 }, { "epoch": 0.27600028630735096, "grad_norm": 1.233302116394043, "learning_rate": 9.791728535312998e-05, "loss": 3.419, "step": 482 }, { "epoch": 0.2771455157111159, "grad_norm": 0.9638918042182922, "learning_rate": 9.789016507810564e-05, "loss": 3.4119, "step": 484 }, { "epoch": 0.27829074511488083, "grad_norm": 1.105643391609192, "learning_rate": 9.786287317377929e-05, "loss": 3.3909, "step": 486 }, { "epoch": 0.27943597451864577, "grad_norm": 0.9666796922683716, "learning_rate": 9.783540973795998e-05, "loss": 3.4194, "step": 488 }, { "epoch": 0.2805812039224107, "grad_norm": 1.3533586263656616, "learning_rate": 9.780777486907146e-05, "loss": 3.3789, "step": 490 }, { "epoch": 0.28172643332617564, "grad_norm": 1.1253416538238525, "learning_rate": 9.777996866615186e-05, "loss": 3.4385, "step": 492 }, { "epoch": 0.2828716627299406, "grad_norm": 0.7198868989944458, "learning_rate": 9.775199122885339e-05, "loss": 3.4038, "step": 494 }, { "epoch": 0.2840168921337055, "grad_norm": 0.9696770310401917, "learning_rate": 9.772384265744188e-05, "loss": 3.4576, "step": 496 }, { "epoch": 0.28516212153747045, "grad_norm": 1.321269154548645, "learning_rate": 9.76955230527965e-05, "loss": 3.4348, "step": 498 }, { "epoch": 0.28630735094123544, "grad_norm": 1.3119802474975586, "learning_rate": 9.766703251640934e-05, "loss": 3.3848, "step": 500 }, { "epoch": 0.2874525803450004, "grad_norm": 1.0199967622756958, "learning_rate": 9.763837115038513e-05, "loss": 3.4108, "step": 502 }, { "epoch": 0.2885978097487653, "grad_norm": 0.9925194382667542, "learning_rate": 9.760953905744075e-05, "loss": 3.31, "step": 504 }, { "epoch": 0.28974303915253025, "grad_norm": 0.9447107315063477, "learning_rate": 9.758053634090502e-05, "loss": 3.3598, "step": 506 }, { "epoch": 0.2908882685562952, "grad_norm": 1.052873134613037, "learning_rate": 9.755136310471817e-05, "loss": 3.3704, "step": 508 }, { "epoch": 0.2920334979600601, "grad_norm": 1.061514139175415, "learning_rate": 9.752201945343156e-05, "loss": 3.3642, "step": 510 }, { "epoch": 0.29317872736382505, "grad_norm": 0.8627074956893921, "learning_rate": 9.74925054922073e-05, "loss": 3.367, "step": 512 }, { "epoch": 0.29432395676759, "grad_norm": 1.0214530229568481, "learning_rate": 9.746282132681785e-05, "loss": 3.3266, "step": 514 }, { "epoch": 0.2954691861713549, "grad_norm": 1.1223275661468506, "learning_rate": 9.743296706364565e-05, "loss": 3.4194, "step": 516 }, { "epoch": 0.2966144155751199, "grad_norm": 0.9849138259887695, "learning_rate": 9.740294280968273e-05, "loss": 3.3664, "step": 518 }, { "epoch": 0.29775964497888485, "grad_norm": 0.7025099396705627, "learning_rate": 9.737274867253034e-05, "loss": 3.3772, "step": 520 }, { "epoch": 0.2989048743826498, "grad_norm": 0.936536967754364, "learning_rate": 9.734238476039858e-05, "loss": 3.3196, "step": 522 }, { "epoch": 0.3000501037864147, "grad_norm": 1.113277792930603, "learning_rate": 9.731185118210598e-05, "loss": 3.4606, "step": 524 }, { "epoch": 0.30119533319017966, "grad_norm": 1.0153186321258545, "learning_rate": 9.728114804707909e-05, "loss": 3.4079, "step": 526 }, { "epoch": 0.3023405625939446, "grad_norm": 1.1675206422805786, "learning_rate": 9.725027546535215e-05, "loss": 3.4111, "step": 528 }, { "epoch": 0.30348579199770953, "grad_norm": 0.9518959522247314, "learning_rate": 9.721923354756665e-05, "loss": 3.3905, "step": 530 }, { "epoch": 0.30463102140147447, "grad_norm": 0.9693425297737122, "learning_rate": 9.718802240497098e-05, "loss": 3.4364, "step": 532 }, { "epoch": 0.3057762508052394, "grad_norm": 1.1249076128005981, "learning_rate": 9.715664214941997e-05, "loss": 3.3373, "step": 534 }, { "epoch": 0.3069214802090044, "grad_norm": 0.8406875133514404, "learning_rate": 9.712509289337453e-05, "loss": 3.321, "step": 536 }, { "epoch": 0.30806670961276933, "grad_norm": 0.9538395404815674, "learning_rate": 9.709337474990121e-05, "loss": 3.4007, "step": 538 }, { "epoch": 0.30921193901653427, "grad_norm": 0.8003599047660828, "learning_rate": 9.706148783267187e-05, "loss": 3.3798, "step": 540 }, { "epoch": 0.3103571684202992, "grad_norm": 0.8605026602745056, "learning_rate": 9.702943225596316e-05, "loss": 3.2908, "step": 542 }, { "epoch": 0.31150239782406414, "grad_norm": 0.7349815964698792, "learning_rate": 9.699720813465625e-05, "loss": 3.408, "step": 544 }, { "epoch": 0.3126476272278291, "grad_norm": 1.1622780561447144, "learning_rate": 9.696481558423628e-05, "loss": 3.3212, "step": 546 }, { "epoch": 0.313792856631594, "grad_norm": 0.9829496145248413, "learning_rate": 9.693225472079204e-05, "loss": 3.4067, "step": 548 }, { "epoch": 0.31493808603535894, "grad_norm": 1.1378313302993774, "learning_rate": 9.689952566101548e-05, "loss": 3.3556, "step": 550 }, { "epoch": 0.3160833154391239, "grad_norm": 0.9355561137199402, "learning_rate": 9.686662852220142e-05, "loss": 3.3281, "step": 552 }, { "epoch": 0.3172285448428888, "grad_norm": 0.9328277111053467, "learning_rate": 9.683356342224694e-05, "loss": 3.313, "step": 554 }, { "epoch": 0.3183737742466538, "grad_norm": 1.277377724647522, "learning_rate": 9.680033047965114e-05, "loss": 3.3499, "step": 556 }, { "epoch": 0.31951900365041874, "grad_norm": 1.0239235162734985, "learning_rate": 9.67669298135146e-05, "loss": 3.3936, "step": 558 }, { "epoch": 0.3206642330541837, "grad_norm": 0.6908963322639465, "learning_rate": 9.673336154353899e-05, "loss": 3.3584, "step": 560 }, { "epoch": 0.3218094624579486, "grad_norm": 0.8835290670394897, "learning_rate": 9.669962579002664e-05, "loss": 3.3728, "step": 562 }, { "epoch": 0.32295469186171355, "grad_norm": 1.0561710596084595, "learning_rate": 9.666572267388013e-05, "loss": 3.3579, "step": 564 }, { "epoch": 0.3240999212654785, "grad_norm": 0.8400120735168457, "learning_rate": 9.663165231660181e-05, "loss": 3.3224, "step": 566 }, { "epoch": 0.3252451506692434, "grad_norm": 0.8960584998130798, "learning_rate": 9.659741484029341e-05, "loss": 3.3434, "step": 568 }, { "epoch": 0.32639038007300836, "grad_norm": 0.9615944027900696, "learning_rate": 9.656301036765558e-05, "loss": 3.2587, "step": 570 }, { "epoch": 0.3275356094767733, "grad_norm": 0.983391523361206, "learning_rate": 9.652843902198743e-05, "loss": 3.2396, "step": 572 }, { "epoch": 0.3286808388805383, "grad_norm": 0.7758197784423828, "learning_rate": 9.649370092718615e-05, "loss": 3.2948, "step": 574 }, { "epoch": 0.3298260682843032, "grad_norm": 0.9714862704277039, "learning_rate": 9.64587962077465e-05, "loss": 3.3381, "step": 576 }, { "epoch": 0.33097129768806816, "grad_norm": 0.8628116846084595, "learning_rate": 9.64237249887604e-05, "loss": 3.294, "step": 578 }, { "epoch": 0.3321165270918331, "grad_norm": 0.9794777035713196, "learning_rate": 9.638848739591646e-05, "loss": 3.3119, "step": 580 }, { "epoch": 0.333261756495598, "grad_norm": 0.8179820775985718, "learning_rate": 9.635308355549957e-05, "loss": 3.3009, "step": 582 }, { "epoch": 0.33440698589936296, "grad_norm": 0.8732323050498962, "learning_rate": 9.63175135943904e-05, "loss": 3.3207, "step": 584 }, { "epoch": 0.3355522153031279, "grad_norm": 1.0355788469314575, "learning_rate": 9.628177764006497e-05, "loss": 3.2889, "step": 586 }, { "epoch": 0.33669744470689283, "grad_norm": 0.8974720239639282, "learning_rate": 9.624587582059417e-05, "loss": 3.3089, "step": 588 }, { "epoch": 0.33784267411065777, "grad_norm": 0.7800531387329102, "learning_rate": 9.620980826464335e-05, "loss": 3.2999, "step": 590 }, { "epoch": 0.3389879035144227, "grad_norm": 0.7294676899909973, "learning_rate": 9.617357510147182e-05, "loss": 3.3634, "step": 592 }, { "epoch": 0.3401331329181877, "grad_norm": 0.7799131274223328, "learning_rate": 9.613717646093239e-05, "loss": 3.308, "step": 594 }, { "epoch": 0.34127836232195263, "grad_norm": 0.9899328947067261, "learning_rate": 9.610061247347091e-05, "loss": 3.3191, "step": 596 }, { "epoch": 0.34242359172571757, "grad_norm": 1.0520347356796265, "learning_rate": 9.606388327012579e-05, "loss": 3.389, "step": 598 }, { "epoch": 0.3435688211294825, "grad_norm": 0.9768466353416443, "learning_rate": 9.602698898252756e-05, "loss": 3.2905, "step": 600 }, { "epoch": 0.34471405053324744, "grad_norm": 0.9359555244445801, "learning_rate": 9.598992974289837e-05, "loss": 3.3022, "step": 602 }, { "epoch": 0.3458592799370124, "grad_norm": 0.7487738728523254, "learning_rate": 9.595270568405156e-05, "loss": 3.2234, "step": 604 }, { "epoch": 0.3470045093407773, "grad_norm": 0.8295655846595764, "learning_rate": 9.591531693939109e-05, "loss": 3.3506, "step": 606 }, { "epoch": 0.34814973874454225, "grad_norm": 0.9020605683326721, "learning_rate": 9.587776364291117e-05, "loss": 3.3026, "step": 608 }, { "epoch": 0.3492949681483072, "grad_norm": 0.7868961095809937, "learning_rate": 9.58400459291957e-05, "loss": 3.2393, "step": 610 }, { "epoch": 0.3504401975520722, "grad_norm": 0.9779835939407349, "learning_rate": 9.580216393341785e-05, "loss": 3.3254, "step": 612 }, { "epoch": 0.3515854269558371, "grad_norm": 0.8962246179580688, "learning_rate": 9.576411779133956e-05, "loss": 3.2486, "step": 614 }, { "epoch": 0.35273065635960205, "grad_norm": 0.9166551828384399, "learning_rate": 9.572590763931097e-05, "loss": 3.2193, "step": 616 }, { "epoch": 0.353875885763367, "grad_norm": 0.7779364585876465, "learning_rate": 9.568753361427009e-05, "loss": 3.2469, "step": 618 }, { "epoch": 0.3550211151671319, "grad_norm": 0.750092089176178, "learning_rate": 9.564899585374214e-05, "loss": 3.2532, "step": 620 }, { "epoch": 0.35616634457089685, "grad_norm": 1.0269392728805542, "learning_rate": 9.561029449583919e-05, "loss": 3.3331, "step": 622 }, { "epoch": 0.3573115739746618, "grad_norm": 0.7937965989112854, "learning_rate": 9.557142967925956e-05, "loss": 3.314, "step": 624 }, { "epoch": 0.3584568033784267, "grad_norm": 1.1338940858840942, "learning_rate": 9.553240154328744e-05, "loss": 3.3375, "step": 626 }, { "epoch": 0.35960203278219166, "grad_norm": 0.7937076091766357, "learning_rate": 9.549321022779229e-05, "loss": 3.2691, "step": 628 }, { "epoch": 0.36074726218595665, "grad_norm": 0.8552340865135193, "learning_rate": 9.545385587322839e-05, "loss": 3.3107, "step": 630 }, { "epoch": 0.3618924915897216, "grad_norm": 1.0279617309570312, "learning_rate": 9.541433862063429e-05, "loss": 3.2552, "step": 632 }, { "epoch": 0.3630377209934865, "grad_norm": 0.9652466177940369, "learning_rate": 9.537465861163237e-05, "loss": 3.242, "step": 634 }, { "epoch": 0.36418295039725146, "grad_norm": 0.9129723310470581, "learning_rate": 9.533481598842827e-05, "loss": 3.3131, "step": 636 }, { "epoch": 0.3653281798010164, "grad_norm": 0.9316424131393433, "learning_rate": 9.529481089381042e-05, "loss": 3.3288, "step": 638 }, { "epoch": 0.36647340920478133, "grad_norm": 0.7987300753593445, "learning_rate": 9.525464347114953e-05, "loss": 3.2832, "step": 640 }, { "epoch": 0.36761863860854627, "grad_norm": 0.7103368043899536, "learning_rate": 9.521431386439807e-05, "loss": 3.2339, "step": 642 }, { "epoch": 0.3687638680123112, "grad_norm": 0.7420955896377563, "learning_rate": 9.517382221808969e-05, "loss": 3.1662, "step": 644 }, { "epoch": 0.36990909741607614, "grad_norm": 0.8201749324798584, "learning_rate": 9.513316867733883e-05, "loss": 3.2837, "step": 646 }, { "epoch": 0.3710543268198411, "grad_norm": 0.8581364154815674, "learning_rate": 9.509235338784009e-05, "loss": 3.2949, "step": 648 }, { "epoch": 0.37219955622360607, "grad_norm": 0.956118643283844, "learning_rate": 9.505137649586775e-05, "loss": 3.316, "step": 650 }, { "epoch": 0.373344785627371, "grad_norm": 0.708759069442749, "learning_rate": 9.501023814827524e-05, "loss": 3.1951, "step": 652 }, { "epoch": 0.37449001503113594, "grad_norm": 0.8143038153648376, "learning_rate": 9.496893849249464e-05, "loss": 3.2738, "step": 654 }, { "epoch": 0.3756352444349009, "grad_norm": 0.6578754782676697, "learning_rate": 9.492747767653611e-05, "loss": 3.2809, "step": 656 }, { "epoch": 0.3767804738386658, "grad_norm": 0.8550508618354797, "learning_rate": 9.488585584898738e-05, "loss": 3.2668, "step": 658 }, { "epoch": 0.37792570324243074, "grad_norm": 0.795080304145813, "learning_rate": 9.48440731590132e-05, "loss": 3.28, "step": 660 }, { "epoch": 0.3790709326461957, "grad_norm": 0.9933105707168579, "learning_rate": 9.480212975635486e-05, "loss": 3.3104, "step": 662 }, { "epoch": 0.3802161620499606, "grad_norm": 1.224338412284851, "learning_rate": 9.476002579132957e-05, "loss": 3.29, "step": 664 }, { "epoch": 0.38136139145372555, "grad_norm": 0.8564585447311401, "learning_rate": 9.471776141483e-05, "loss": 3.2, "step": 666 }, { "epoch": 0.38250662085749054, "grad_norm": 1.160684585571289, "learning_rate": 9.467533677832365e-05, "loss": 3.2226, "step": 668 }, { "epoch": 0.3836518502612555, "grad_norm": 0.8671857714653015, "learning_rate": 9.463275203385244e-05, "loss": 3.2453, "step": 670 }, { "epoch": 0.3847970796650204, "grad_norm": 1.0225045680999756, "learning_rate": 9.459000733403205e-05, "loss": 3.2283, "step": 672 }, { "epoch": 0.38594230906878535, "grad_norm": 0.8350477814674377, "learning_rate": 9.454710283205139e-05, "loss": 3.2584, "step": 674 }, { "epoch": 0.3870875384725503, "grad_norm": 0.8098021745681763, "learning_rate": 9.450403868167208e-05, "loss": 3.2836, "step": 676 }, { "epoch": 0.3882327678763152, "grad_norm": 0.8174638748168945, "learning_rate": 9.446081503722792e-05, "loss": 3.1896, "step": 678 }, { "epoch": 0.38937799728008016, "grad_norm": 0.6904940009117126, "learning_rate": 9.441743205362426e-05, "loss": 3.2464, "step": 680 }, { "epoch": 0.3905232266838451, "grad_norm": 0.692864716053009, "learning_rate": 9.437388988633752e-05, "loss": 3.2277, "step": 682 }, { "epoch": 0.39166845608761003, "grad_norm": 0.7014842629432678, "learning_rate": 9.433018869141464e-05, "loss": 3.2372, "step": 684 }, { "epoch": 0.39281368549137496, "grad_norm": 0.6166806817054749, "learning_rate": 9.428632862547237e-05, "loss": 3.2501, "step": 686 }, { "epoch": 0.39395891489513996, "grad_norm": 0.7060846090316772, "learning_rate": 9.424230984569696e-05, "loss": 3.2881, "step": 688 }, { "epoch": 0.3951041442989049, "grad_norm": 0.7771391272544861, "learning_rate": 9.419813250984337e-05, "loss": 3.2149, "step": 690 }, { "epoch": 0.3962493737026698, "grad_norm": 0.6290923953056335, "learning_rate": 9.415379677623485e-05, "loss": 3.1555, "step": 692 }, { "epoch": 0.39739460310643476, "grad_norm": 0.7270971536636353, "learning_rate": 9.410930280376225e-05, "loss": 3.2554, "step": 694 }, { "epoch": 0.3985398325101997, "grad_norm": 0.681962788105011, "learning_rate": 9.40646507518836e-05, "loss": 3.1671, "step": 696 }, { "epoch": 0.39968506191396463, "grad_norm": 0.5727997422218323, "learning_rate": 9.40198407806234e-05, "loss": 3.237, "step": 698 }, { "epoch": 0.40083029131772957, "grad_norm": 0.7687988877296448, "learning_rate": 9.39748730505721e-05, "loss": 3.2357, "step": 700 }, { "epoch": 0.4019755207214945, "grad_norm": 0.7813317179679871, "learning_rate": 9.392974772288558e-05, "loss": 3.2101, "step": 702 }, { "epoch": 0.40312075012525944, "grad_norm": 0.8766132593154907, "learning_rate": 9.388446495928446e-05, "loss": 3.2852, "step": 704 }, { "epoch": 0.40426597952902443, "grad_norm": 0.7857736349105835, "learning_rate": 9.383902492205363e-05, "loss": 3.2113, "step": 706 }, { "epoch": 0.40541120893278937, "grad_norm": 0.9073331356048584, "learning_rate": 9.379342777404159e-05, "loss": 3.2478, "step": 708 }, { "epoch": 0.4065564383365543, "grad_norm": 0.8033682107925415, "learning_rate": 9.374767367865989e-05, "loss": 3.3159, "step": 710 }, { "epoch": 0.40770166774031924, "grad_norm": 0.7821508646011353, "learning_rate": 9.370176279988256e-05, "loss": 3.2362, "step": 712 }, { "epoch": 0.4088468971440842, "grad_norm": 0.8257923126220703, "learning_rate": 9.365569530224554e-05, "loss": 3.1832, "step": 714 }, { "epoch": 0.4099921265478491, "grad_norm": 0.8349987864494324, "learning_rate": 9.360947135084603e-05, "loss": 3.1995, "step": 716 }, { "epoch": 0.41113735595161405, "grad_norm": 0.8590210676193237, "learning_rate": 9.356309111134191e-05, "loss": 3.2119, "step": 718 }, { "epoch": 0.412282585355379, "grad_norm": 0.8512969017028809, "learning_rate": 9.351655474995122e-05, "loss": 3.2323, "step": 720 }, { "epoch": 0.4134278147591439, "grad_norm": 0.6388457417488098, "learning_rate": 9.346986243345149e-05, "loss": 3.1677, "step": 722 }, { "epoch": 0.4145730441629089, "grad_norm": 0.8811210989952087, "learning_rate": 9.342301432917912e-05, "loss": 3.2307, "step": 724 }, { "epoch": 0.41571827356667385, "grad_norm": 0.9297654628753662, "learning_rate": 9.337601060502891e-05, "loss": 3.1838, "step": 726 }, { "epoch": 0.4168635029704388, "grad_norm": 0.750491201877594, "learning_rate": 9.332885142945329e-05, "loss": 3.23, "step": 728 }, { "epoch": 0.4180087323742037, "grad_norm": 0.8282638192176819, "learning_rate": 9.328153697146186e-05, "loss": 3.1789, "step": 730 }, { "epoch": 0.41915396177796865, "grad_norm": 0.7395208477973938, "learning_rate": 9.323406740062068e-05, "loss": 3.2881, "step": 732 }, { "epoch": 0.4202991911817336, "grad_norm": 0.5959879755973816, "learning_rate": 9.318644288705172e-05, "loss": 3.1879, "step": 734 }, { "epoch": 0.4214444205854985, "grad_norm": 0.6063298583030701, "learning_rate": 9.313866360143227e-05, "loss": 3.273, "step": 736 }, { "epoch": 0.42258964998926346, "grad_norm": 0.6868070960044861, "learning_rate": 9.309072971499422e-05, "loss": 3.2145, "step": 738 }, { "epoch": 0.4237348793930284, "grad_norm": 0.6153081655502319, "learning_rate": 9.304264139952356e-05, "loss": 3.0791, "step": 740 }, { "epoch": 0.42488010879679333, "grad_norm": 0.6345932483673096, "learning_rate": 9.299439882735977e-05, "loss": 3.1991, "step": 742 }, { "epoch": 0.4260253382005583, "grad_norm": 0.7605310082435608, "learning_rate": 9.294600217139506e-05, "loss": 3.1272, "step": 744 }, { "epoch": 0.42717056760432326, "grad_norm": 0.6695173382759094, "learning_rate": 9.289745160507395e-05, "loss": 3.1482, "step": 746 }, { "epoch": 0.4283157970080882, "grad_norm": 0.8121134638786316, "learning_rate": 9.284874730239244e-05, "loss": 3.2122, "step": 748 }, { "epoch": 0.42946102641185313, "grad_norm": 0.8771198391914368, "learning_rate": 9.279988943789759e-05, "loss": 3.1768, "step": 750 }, { "epoch": 0.43060625581561807, "grad_norm": 0.7993550300598145, "learning_rate": 9.275087818668675e-05, "loss": 3.1944, "step": 752 }, { "epoch": 0.431751485219383, "grad_norm": 0.6639721393585205, "learning_rate": 9.270171372440697e-05, "loss": 3.1418, "step": 754 }, { "epoch": 0.43289671462314794, "grad_norm": 0.7494943737983704, "learning_rate": 9.265239622725438e-05, "loss": 3.1956, "step": 756 }, { "epoch": 0.4340419440269129, "grad_norm": 0.7307000160217285, "learning_rate": 9.26029258719736e-05, "loss": 3.133, "step": 758 }, { "epoch": 0.4351871734306778, "grad_norm": 0.7357375621795654, "learning_rate": 9.255330283585701e-05, "loss": 3.1898, "step": 760 }, { "epoch": 0.4363324028344428, "grad_norm": 0.6649693250656128, "learning_rate": 9.250352729674422e-05, "loss": 3.2147, "step": 762 }, { "epoch": 0.43747763223820774, "grad_norm": 0.6873495578765869, "learning_rate": 9.245359943302133e-05, "loss": 3.2341, "step": 764 }, { "epoch": 0.43862286164197267, "grad_norm": 0.7320956587791443, "learning_rate": 9.240351942362038e-05, "loss": 3.1241, "step": 766 }, { "epoch": 0.4397680910457376, "grad_norm": 0.6137463450431824, "learning_rate": 9.235328744801868e-05, "loss": 3.1529, "step": 768 }, { "epoch": 0.44091332044950254, "grad_norm": 0.8658304214477539, "learning_rate": 9.230290368623809e-05, "loss": 3.2168, "step": 770 }, { "epoch": 0.4420585498532675, "grad_norm": 0.7436694502830505, "learning_rate": 9.225236831884454e-05, "loss": 3.1798, "step": 772 }, { "epoch": 0.4432037792570324, "grad_norm": 0.9040384888648987, "learning_rate": 9.220168152694722e-05, "loss": 3.2241, "step": 774 }, { "epoch": 0.44434900866079735, "grad_norm": 0.7236924171447754, "learning_rate": 9.215084349219801e-05, "loss": 3.183, "step": 776 }, { "epoch": 0.4454942380645623, "grad_norm": 0.8633347153663635, "learning_rate": 9.209985439679081e-05, "loss": 3.1776, "step": 778 }, { "epoch": 0.4466394674683272, "grad_norm": 0.730910062789917, "learning_rate": 9.204871442346091e-05, "loss": 3.1633, "step": 780 }, { "epoch": 0.4477846968720922, "grad_norm": 0.809923529624939, "learning_rate": 9.199742375548432e-05, "loss": 3.1736, "step": 782 }, { "epoch": 0.44892992627585715, "grad_norm": 0.7229586839675903, "learning_rate": 9.194598257667711e-05, "loss": 3.1813, "step": 784 }, { "epoch": 0.4500751556796221, "grad_norm": 0.6999960541725159, "learning_rate": 9.189439107139472e-05, "loss": 3.1125, "step": 786 }, { "epoch": 0.451220385083387, "grad_norm": 0.7234693169593811, "learning_rate": 9.184264942453138e-05, "loss": 3.137, "step": 788 }, { "epoch": 0.45236561448715196, "grad_norm": 0.7283908724784851, "learning_rate": 9.179075782151936e-05, "loss": 3.1672, "step": 790 }, { "epoch": 0.4535108438909169, "grad_norm": 0.793543815612793, "learning_rate": 9.173871644832834e-05, "loss": 3.1925, "step": 792 }, { "epoch": 0.45465607329468183, "grad_norm": 0.7263696789741516, "learning_rate": 9.168652549146481e-05, "loss": 3.1609, "step": 794 }, { "epoch": 0.45580130269844676, "grad_norm": 0.7698031663894653, "learning_rate": 9.163418513797126e-05, "loss": 3.2547, "step": 796 }, { "epoch": 0.4569465321022117, "grad_norm": 0.908698320388794, "learning_rate": 9.158169557542566e-05, "loss": 3.2165, "step": 798 }, { "epoch": 0.4580917615059767, "grad_norm": 0.9588857293128967, "learning_rate": 9.152905699194065e-05, "loss": 3.1743, "step": 800 }, { "epoch": 0.4592369909097416, "grad_norm": 0.7442302107810974, "learning_rate": 9.1476269576163e-05, "loss": 3.1088, "step": 802 }, { "epoch": 0.46038222031350656, "grad_norm": 0.7421006560325623, "learning_rate": 9.14233335172728e-05, "loss": 3.1497, "step": 804 }, { "epoch": 0.4615274497172715, "grad_norm": 0.8878415822982788, "learning_rate": 9.13702490049829e-05, "loss": 3.1924, "step": 806 }, { "epoch": 0.46267267912103643, "grad_norm": 0.6878317594528198, "learning_rate": 9.131701622953816e-05, "loss": 3.1366, "step": 808 }, { "epoch": 0.46381790852480137, "grad_norm": 0.7945599555969238, "learning_rate": 9.126363538171478e-05, "loss": 3.1926, "step": 810 }, { "epoch": 0.4649631379285663, "grad_norm": 0.7997886538505554, "learning_rate": 9.121010665281964e-05, "loss": 3.1521, "step": 812 }, { "epoch": 0.46610836733233124, "grad_norm": 0.715614378452301, "learning_rate": 9.115643023468958e-05, "loss": 3.1904, "step": 814 }, { "epoch": 0.4672535967360962, "grad_norm": 0.7846017479896545, "learning_rate": 9.110260631969077e-05, "loss": 3.1338, "step": 816 }, { "epoch": 0.46839882613986117, "grad_norm": 0.6939677596092224, "learning_rate": 9.10486351007179e-05, "loss": 3.1635, "step": 818 }, { "epoch": 0.4695440555436261, "grad_norm": 0.7764283418655396, "learning_rate": 9.099451677119366e-05, "loss": 3.1922, "step": 820 }, { "epoch": 0.47068928494739104, "grad_norm": 0.753666877746582, "learning_rate": 9.094025152506788e-05, "loss": 3.0827, "step": 822 }, { "epoch": 0.471834514351156, "grad_norm": 0.6793937683105469, "learning_rate": 9.088583955681699e-05, "loss": 3.1235, "step": 824 }, { "epoch": 0.4729797437549209, "grad_norm": 0.645055890083313, "learning_rate": 9.08312810614432e-05, "loss": 3.1758, "step": 826 }, { "epoch": 0.47412497315868585, "grad_norm": 0.7241025567054749, "learning_rate": 9.077657623447379e-05, "loss": 3.1636, "step": 828 }, { "epoch": 0.4752702025624508, "grad_norm": 0.762117862701416, "learning_rate": 9.07217252719606e-05, "loss": 3.1423, "step": 830 }, { "epoch": 0.4764154319662157, "grad_norm": 0.7575943470001221, "learning_rate": 9.066672837047907e-05, "loss": 3.1304, "step": 832 }, { "epoch": 0.47756066136998065, "grad_norm": 0.8326764106750488, "learning_rate": 9.061158572712769e-05, "loss": 3.1807, "step": 834 }, { "epoch": 0.4787058907737456, "grad_norm": 0.7815741300582886, "learning_rate": 9.055629753952731e-05, "loss": 3.2113, "step": 836 }, { "epoch": 0.4798511201775106, "grad_norm": 0.7716583609580994, "learning_rate": 9.050086400582033e-05, "loss": 3.1791, "step": 838 }, { "epoch": 0.4809963495812755, "grad_norm": 0.6160004734992981, "learning_rate": 9.044528532467006e-05, "loss": 3.1696, "step": 840 }, { "epoch": 0.48214157898504045, "grad_norm": 0.8025004267692566, "learning_rate": 9.038956169525998e-05, "loss": 3.2002, "step": 842 }, { "epoch": 0.4832868083888054, "grad_norm": 0.733741819858551, "learning_rate": 9.033369331729307e-05, "loss": 3.1661, "step": 844 }, { "epoch": 0.4844320377925703, "grad_norm": 0.7210118770599365, "learning_rate": 9.027768039099103e-05, "loss": 3.1492, "step": 846 }, { "epoch": 0.48557726719633526, "grad_norm": 0.6915583610534668, "learning_rate": 9.02215231170936e-05, "loss": 3.1892, "step": 848 }, { "epoch": 0.4867224966001002, "grad_norm": 0.6812649965286255, "learning_rate": 9.016522169685783e-05, "loss": 3.1404, "step": 850 }, { "epoch": 0.48786772600386513, "grad_norm": 0.7272056341171265, "learning_rate": 9.010877633205738e-05, "loss": 3.1935, "step": 852 }, { "epoch": 0.48901295540763007, "grad_norm": 0.7162798643112183, "learning_rate": 9.005218722498177e-05, "loss": 3.1949, "step": 854 }, { "epoch": 0.49015818481139506, "grad_norm": 0.6110600829124451, "learning_rate": 8.999545457843568e-05, "loss": 3.1217, "step": 856 }, { "epoch": 0.49130341421516, "grad_norm": 0.657370924949646, "learning_rate": 8.993857859573818e-05, "loss": 3.1381, "step": 858 }, { "epoch": 0.49244864361892493, "grad_norm": 0.8181600570678711, "learning_rate": 8.988155948072203e-05, "loss": 3.1527, "step": 860 }, { "epoch": 0.49359387302268987, "grad_norm": 0.586644172668457, "learning_rate": 8.9824397437733e-05, "loss": 3.1328, "step": 862 }, { "epoch": 0.4947391024264548, "grad_norm": 0.8710150718688965, "learning_rate": 8.976709267162903e-05, "loss": 3.1509, "step": 864 }, { "epoch": 0.49588433183021974, "grad_norm": 0.7185545563697815, "learning_rate": 8.970964538777957e-05, "loss": 3.0628, "step": 866 }, { "epoch": 0.4970295612339847, "grad_norm": 0.7242484092712402, "learning_rate": 8.965205579206483e-05, "loss": 3.0603, "step": 868 }, { "epoch": 0.4981747906377496, "grad_norm": 0.7996972799301147, "learning_rate": 8.959432409087504e-05, "loss": 3.2346, "step": 870 }, { "epoch": 0.49932002004151455, "grad_norm": 0.6038782000541687, "learning_rate": 8.953645049110971e-05, "loss": 3.0751, "step": 872 }, { "epoch": 0.5004652494452795, "grad_norm": 0.7712786197662354, "learning_rate": 8.94784352001769e-05, "loss": 3.1086, "step": 874 }, { "epoch": 0.5016104788490444, "grad_norm": 0.6952617168426514, "learning_rate": 8.94202784259924e-05, "loss": 3.13, "step": 876 }, { "epoch": 0.5027557082528094, "grad_norm": 0.7420851588249207, "learning_rate": 8.936198037697916e-05, "loss": 3.1094, "step": 878 }, { "epoch": 0.5039009376565743, "grad_norm": 0.6883806586265564, "learning_rate": 8.930354126206634e-05, "loss": 3.0722, "step": 880 }, { "epoch": 0.5050461670603392, "grad_norm": 0.7546491026878357, "learning_rate": 8.92449612906887e-05, "loss": 3.1571, "step": 882 }, { "epoch": 0.5061913964641043, "grad_norm": 0.7471094727516174, "learning_rate": 8.918624067278576e-05, "loss": 3.1842, "step": 884 }, { "epoch": 0.5073366258678692, "grad_norm": 0.8344042897224426, "learning_rate": 8.912737961880116e-05, "loss": 3.1709, "step": 886 }, { "epoch": 0.5084818552716341, "grad_norm": 0.6555135250091553, "learning_rate": 8.906837833968174e-05, "loss": 3.1777, "step": 888 }, { "epoch": 0.5096270846753991, "grad_norm": 0.799281120300293, "learning_rate": 8.900923704687697e-05, "loss": 3.176, "step": 890 }, { "epoch": 0.510772314079164, "grad_norm": 0.8266319632530212, "learning_rate": 8.894995595233809e-05, "loss": 3.1353, "step": 892 }, { "epoch": 0.511917543482929, "grad_norm": 0.7263309955596924, "learning_rate": 8.889053526851729e-05, "loss": 3.0824, "step": 894 }, { "epoch": 0.5130627728866939, "grad_norm": 0.7665941119194031, "learning_rate": 8.88309752083671e-05, "loss": 3.1808, "step": 896 }, { "epoch": 0.5142080022904588, "grad_norm": 0.7014003396034241, "learning_rate": 8.877127598533952e-05, "loss": 3.1158, "step": 898 }, { "epoch": 0.5153532316942238, "grad_norm": 0.6320556998252869, "learning_rate": 8.871143781338529e-05, "loss": 3.1276, "step": 900 }, { "epoch": 0.5164984610979887, "grad_norm": 0.8376429677009583, "learning_rate": 8.865146090695308e-05, "loss": 3.1422, "step": 902 }, { "epoch": 0.5176436905017536, "grad_norm": 0.6639658212661743, "learning_rate": 8.859134548098883e-05, "loss": 3.0622, "step": 904 }, { "epoch": 0.5187889199055186, "grad_norm": 0.6442060470581055, "learning_rate": 8.853109175093486e-05, "loss": 3.1206, "step": 906 }, { "epoch": 0.5199341493092835, "grad_norm": 0.6882277131080627, "learning_rate": 8.847069993272912e-05, "loss": 3.1315, "step": 908 }, { "epoch": 0.5210793787130484, "grad_norm": 0.8141956329345703, "learning_rate": 8.841017024280449e-05, "loss": 3.1498, "step": 910 }, { "epoch": 0.5222246081168134, "grad_norm": 0.6133621335029602, "learning_rate": 8.834950289808796e-05, "loss": 3.0971, "step": 912 }, { "epoch": 0.5233698375205783, "grad_norm": 0.6844592690467834, "learning_rate": 8.828869811599982e-05, "loss": 3.1408, "step": 914 }, { "epoch": 0.5245150669243432, "grad_norm": 0.7407364845275879, "learning_rate": 8.822775611445289e-05, "loss": 3.1356, "step": 916 }, { "epoch": 0.5256602963281082, "grad_norm": 0.7962344884872437, "learning_rate": 8.816667711185183e-05, "loss": 3.037, "step": 918 }, { "epoch": 0.5268055257318731, "grad_norm": 0.6615867018699646, "learning_rate": 8.81054613270922e-05, "loss": 3.119, "step": 920 }, { "epoch": 0.5279507551356382, "grad_norm": 0.6886764168739319, "learning_rate": 8.804410897955986e-05, "loss": 3.1686, "step": 922 }, { "epoch": 0.5290959845394031, "grad_norm": 0.6714747548103333, "learning_rate": 8.798262028913e-05, "loss": 3.0539, "step": 924 }, { "epoch": 0.530241213943168, "grad_norm": 0.630648672580719, "learning_rate": 8.792099547616646e-05, "loss": 3.03, "step": 926 }, { "epoch": 0.531386443346933, "grad_norm": 0.6129744648933411, "learning_rate": 8.785923476152092e-05, "loss": 3.112, "step": 928 }, { "epoch": 0.5325316727506979, "grad_norm": 0.6656561493873596, "learning_rate": 8.779733836653213e-05, "loss": 3.0675, "step": 930 }, { "epoch": 0.5336769021544628, "grad_norm": 0.6855784058570862, "learning_rate": 8.773530651302506e-05, "loss": 3.0567, "step": 932 }, { "epoch": 0.5348221315582278, "grad_norm": 0.6233646869659424, "learning_rate": 8.767313942331016e-05, "loss": 3.1316, "step": 934 }, { "epoch": 0.5359673609619927, "grad_norm": 0.5950207710266113, "learning_rate": 8.761083732018253e-05, "loss": 3.0838, "step": 936 }, { "epoch": 0.5371125903657576, "grad_norm": 0.6548320651054382, "learning_rate": 8.754840042692114e-05, "loss": 3.128, "step": 938 }, { "epoch": 0.5382578197695226, "grad_norm": 0.5914682149887085, "learning_rate": 8.748582896728801e-05, "loss": 3.0995, "step": 940 }, { "epoch": 0.5394030491732875, "grad_norm": 0.8005509376525879, "learning_rate": 8.742312316552741e-05, "loss": 3.1194, "step": 942 }, { "epoch": 0.5405482785770525, "grad_norm": 0.6313744783401489, "learning_rate": 8.736028324636511e-05, "loss": 3.1332, "step": 944 }, { "epoch": 0.5416935079808174, "grad_norm": 0.6614211797714233, "learning_rate": 8.729730943500751e-05, "loss": 3.1808, "step": 946 }, { "epoch": 0.5428387373845823, "grad_norm": 0.6834341883659363, "learning_rate": 8.723420195714083e-05, "loss": 3.146, "step": 948 }, { "epoch": 0.5439839667883473, "grad_norm": 0.6552104353904724, "learning_rate": 8.717096103893034e-05, "loss": 3.1339, "step": 950 }, { "epoch": 0.5451291961921122, "grad_norm": 0.6134440302848816, "learning_rate": 8.710758690701957e-05, "loss": 3.1073, "step": 952 }, { "epoch": 0.5462744255958771, "grad_norm": 0.6617953181266785, "learning_rate": 8.704407978852941e-05, "loss": 3.0803, "step": 954 }, { "epoch": 0.5474196549996421, "grad_norm": 0.6200254559516907, "learning_rate": 8.698043991105738e-05, "loss": 3.0902, "step": 956 }, { "epoch": 0.548564884403407, "grad_norm": 0.766797661781311, "learning_rate": 8.691666750267677e-05, "loss": 3.0533, "step": 958 }, { "epoch": 0.549710113807172, "grad_norm": 0.8062739372253418, "learning_rate": 8.685276279193583e-05, "loss": 3.074, "step": 960 }, { "epoch": 0.550855343210937, "grad_norm": 0.7662980556488037, "learning_rate": 8.678872600785702e-05, "loss": 3.0567, "step": 962 }, { "epoch": 0.5520005726147019, "grad_norm": 0.610495388507843, "learning_rate": 8.672455737993601e-05, "loss": 3.0505, "step": 964 }, { "epoch": 0.5531458020184669, "grad_norm": 0.7125016450881958, "learning_rate": 8.666025713814106e-05, "loss": 3.1392, "step": 966 }, { "epoch": 0.5542910314222318, "grad_norm": 0.6135743260383606, "learning_rate": 8.65958255129121e-05, "loss": 3.0789, "step": 968 }, { "epoch": 0.5554362608259967, "grad_norm": 0.5998417735099792, "learning_rate": 8.653126273515988e-05, "loss": 3.0702, "step": 970 }, { "epoch": 0.5565814902297617, "grad_norm": 0.7088379859924316, "learning_rate": 8.64665690362652e-05, "loss": 3.084, "step": 972 }, { "epoch": 0.5577267196335266, "grad_norm": 0.6542948484420776, "learning_rate": 8.640174464807805e-05, "loss": 3.1164, "step": 974 }, { "epoch": 0.5588719490372915, "grad_norm": 0.5674989819526672, "learning_rate": 8.63367898029168e-05, "loss": 3.1048, "step": 976 }, { "epoch": 0.5600171784410565, "grad_norm": 0.6628077626228333, "learning_rate": 8.627170473356733e-05, "loss": 3.0492, "step": 978 }, { "epoch": 0.5611624078448214, "grad_norm": 0.6939430832862854, "learning_rate": 8.620648967328224e-05, "loss": 3.1041, "step": 980 }, { "epoch": 0.5623076372485863, "grad_norm": 0.7876750230789185, "learning_rate": 8.614114485577996e-05, "loss": 3.1109, "step": 982 }, { "epoch": 0.5634528666523513, "grad_norm": 0.7786777019500732, "learning_rate": 8.607567051524399e-05, "loss": 3.0289, "step": 984 }, { "epoch": 0.5645980960561162, "grad_norm": 0.6663212180137634, "learning_rate": 8.601006688632199e-05, "loss": 3.0807, "step": 986 }, { "epoch": 0.5657433254598812, "grad_norm": 0.7165863513946533, "learning_rate": 8.594433420412496e-05, "loss": 3.0755, "step": 988 }, { "epoch": 0.5668885548636461, "grad_norm": 0.7116391658782959, "learning_rate": 8.587847270422642e-05, "loss": 3.064, "step": 990 }, { "epoch": 0.568033784267411, "grad_norm": 0.7216659188270569, "learning_rate": 8.581248262266155e-05, "loss": 3.0844, "step": 992 }, { "epoch": 0.569179013671176, "grad_norm": 0.600975751876831, "learning_rate": 8.57463641959263e-05, "loss": 2.9771, "step": 994 }, { "epoch": 0.5703242430749409, "grad_norm": 0.6743506789207458, "learning_rate": 8.568011766097666e-05, "loss": 3.1177, "step": 996 }, { "epoch": 0.5714694724787059, "grad_norm": 0.6986669301986694, "learning_rate": 8.561374325522764e-05, "loss": 3.0838, "step": 998 }, { "epoch": 0.5726147018824709, "grad_norm": 0.8114129900932312, "learning_rate": 8.554724121655262e-05, "loss": 3.1444, "step": 1000 }, { "epoch": 0.5737599312862358, "grad_norm": 0.7919934988021851, "learning_rate": 8.548061178328233e-05, "loss": 3.0166, "step": 1002 }, { "epoch": 0.5749051606900007, "grad_norm": 0.6979469656944275, "learning_rate": 8.541385519420403e-05, "loss": 3.0737, "step": 1004 }, { "epoch": 0.5760503900937657, "grad_norm": 0.6499598026275635, "learning_rate": 8.534697168856076e-05, "loss": 3.0649, "step": 1006 }, { "epoch": 0.5771956194975306, "grad_norm": 0.7335128784179688, "learning_rate": 8.527996150605034e-05, "loss": 3.0403, "step": 1008 }, { "epoch": 0.5783408489012956, "grad_norm": 0.7250447869300842, "learning_rate": 8.521282488682463e-05, "loss": 3.0069, "step": 1010 }, { "epoch": 0.5794860783050605, "grad_norm": 0.6807704567909241, "learning_rate": 8.514556207148857e-05, "loss": 3.119, "step": 1012 }, { "epoch": 0.5806313077088254, "grad_norm": 0.7379552721977234, "learning_rate": 8.507817330109936e-05, "loss": 3.0773, "step": 1014 }, { "epoch": 0.5817765371125904, "grad_norm": 0.6113300919532776, "learning_rate": 8.501065881716566e-05, "loss": 3.0768, "step": 1016 }, { "epoch": 0.5829217665163553, "grad_norm": 0.6463739275932312, "learning_rate": 8.494301886164658e-05, "loss": 3.0759, "step": 1018 }, { "epoch": 0.5840669959201202, "grad_norm": 0.6680572032928467, "learning_rate": 8.487525367695098e-05, "loss": 3.032, "step": 1020 }, { "epoch": 0.5852122253238852, "grad_norm": 0.7283656597137451, "learning_rate": 8.480736350593644e-05, "loss": 3.0986, "step": 1022 }, { "epoch": 0.5863574547276501, "grad_norm": 0.6844098567962646, "learning_rate": 8.473934859190853e-05, "loss": 3.0703, "step": 1024 }, { "epoch": 0.587502684131415, "grad_norm": 0.6737761497497559, "learning_rate": 8.467120917861984e-05, "loss": 3.0775, "step": 1026 }, { "epoch": 0.58864791353518, "grad_norm": 0.8135201930999756, "learning_rate": 8.460294551026916e-05, "loss": 3.0802, "step": 1028 }, { "epoch": 0.5897931429389449, "grad_norm": 0.6999467015266418, "learning_rate": 8.453455783150054e-05, "loss": 3.047, "step": 1030 }, { "epoch": 0.5909383723427099, "grad_norm": 0.7999339699745178, "learning_rate": 8.446604638740256e-05, "loss": 3.1247, "step": 1032 }, { "epoch": 0.5920836017464748, "grad_norm": 0.7229709029197693, "learning_rate": 8.439741142350725e-05, "loss": 3.1009, "step": 1034 }, { "epoch": 0.5932288311502398, "grad_norm": 0.7535393238067627, "learning_rate": 8.432865318578935e-05, "loss": 3.0566, "step": 1036 }, { "epoch": 0.5943740605540048, "grad_norm": 0.7364835143089294, "learning_rate": 8.425977192066539e-05, "loss": 3.0751, "step": 1038 }, { "epoch": 0.5955192899577697, "grad_norm": 0.7996159791946411, "learning_rate": 8.419076787499283e-05, "loss": 3.1277, "step": 1040 }, { "epoch": 0.5966645193615346, "grad_norm": 0.6993304491043091, "learning_rate": 8.412164129606911e-05, "loss": 3.0713, "step": 1042 }, { "epoch": 0.5978097487652996, "grad_norm": 0.6996495127677917, "learning_rate": 8.405239243163084e-05, "loss": 3.0595, "step": 1044 }, { "epoch": 0.5989549781690645, "grad_norm": 0.6525830626487732, "learning_rate": 8.398302152985285e-05, "loss": 3.0666, "step": 1046 }, { "epoch": 0.6001002075728294, "grad_norm": 0.6092258095741272, "learning_rate": 8.391352883934733e-05, "loss": 3.0453, "step": 1048 }, { "epoch": 0.6012454369765944, "grad_norm": 0.7509777545928955, "learning_rate": 8.3843914609163e-05, "loss": 3.016, "step": 1050 }, { "epoch": 0.6023906663803593, "grad_norm": 0.6374807953834534, "learning_rate": 8.377417908878406e-05, "loss": 2.9986, "step": 1052 }, { "epoch": 0.6035358957841243, "grad_norm": 0.6541762351989746, "learning_rate": 8.370432252812946e-05, "loss": 3.0299, "step": 1054 }, { "epoch": 0.6046811251878892, "grad_norm": 0.5995933413505554, "learning_rate": 8.363434517755191e-05, "loss": 3.0853, "step": 1056 }, { "epoch": 0.6058263545916541, "grad_norm": 0.7466599941253662, "learning_rate": 8.356424728783702e-05, "loss": 3.1482, "step": 1058 }, { "epoch": 0.6069715839954191, "grad_norm": 0.7076915502548218, "learning_rate": 8.349402911020234e-05, "loss": 3.0681, "step": 1060 }, { "epoch": 0.608116813399184, "grad_norm": 0.6121742725372314, "learning_rate": 8.34236908962966e-05, "loss": 3.0771, "step": 1062 }, { "epoch": 0.6092620428029489, "grad_norm": 0.7337540984153748, "learning_rate": 8.335323289819865e-05, "loss": 3.1927, "step": 1064 }, { "epoch": 0.6104072722067139, "grad_norm": 0.6119634509086609, "learning_rate": 8.328265536841662e-05, "loss": 3.0988, "step": 1066 }, { "epoch": 0.6115525016104788, "grad_norm": 0.6291252970695496, "learning_rate": 8.321195855988706e-05, "loss": 3.0667, "step": 1068 }, { "epoch": 0.6126977310142437, "grad_norm": 0.5802082419395447, "learning_rate": 8.314114272597398e-05, "loss": 3.0118, "step": 1070 }, { "epoch": 0.6138429604180088, "grad_norm": 0.6166985630989075, "learning_rate": 8.307020812046792e-05, "loss": 3.1762, "step": 1072 }, { "epoch": 0.6149881898217737, "grad_norm": 0.5751842260360718, "learning_rate": 8.299915499758514e-05, "loss": 3.0107, "step": 1074 }, { "epoch": 0.6161334192255387, "grad_norm": 0.6163948178291321, "learning_rate": 8.292798361196658e-05, "loss": 3.0617, "step": 1076 }, { "epoch": 0.6172786486293036, "grad_norm": 0.6665089130401611, "learning_rate": 8.285669421867703e-05, "loss": 3.0729, "step": 1078 }, { "epoch": 0.6184238780330685, "grad_norm": 0.630814254283905, "learning_rate": 8.278528707320421e-05, "loss": 2.9811, "step": 1080 }, { "epoch": 0.6195691074368335, "grad_norm": 0.785892903804779, "learning_rate": 8.271376243145786e-05, "loss": 3.0561, "step": 1082 }, { "epoch": 0.6207143368405984, "grad_norm": 0.6047619581222534, "learning_rate": 8.264212054976875e-05, "loss": 3.0595, "step": 1084 }, { "epoch": 0.6218595662443633, "grad_norm": 0.6675294041633606, "learning_rate": 8.257036168488785e-05, "loss": 3.1725, "step": 1086 }, { "epoch": 0.6230047956481283, "grad_norm": 0.6342408657073975, "learning_rate": 8.24984860939854e-05, "loss": 2.9766, "step": 1088 }, { "epoch": 0.6241500250518932, "grad_norm": 0.5901287794113159, "learning_rate": 8.242649403464989e-05, "loss": 3.1021, "step": 1090 }, { "epoch": 0.6252952544556581, "grad_norm": 0.5638805627822876, "learning_rate": 8.23543857648873e-05, "loss": 2.9866, "step": 1092 }, { "epoch": 0.6264404838594231, "grad_norm": 0.5482515692710876, "learning_rate": 8.228216154312001e-05, "loss": 3.0344, "step": 1094 }, { "epoch": 0.627585713263188, "grad_norm": 0.7258690595626831, "learning_rate": 8.2209821628186e-05, "loss": 3.0339, "step": 1096 }, { "epoch": 0.628730942666953, "grad_norm": 0.6626359820365906, "learning_rate": 8.213736627933786e-05, "loss": 3.1191, "step": 1098 }, { "epoch": 0.6298761720707179, "grad_norm": 0.5897409319877625, "learning_rate": 8.206479575624186e-05, "loss": 2.9604, "step": 1100 }, { "epoch": 0.6310214014744828, "grad_norm": 0.661314070224762, "learning_rate": 8.199211031897704e-05, "loss": 3.0568, "step": 1102 }, { "epoch": 0.6321666308782478, "grad_norm": 0.7244003415107727, "learning_rate": 8.191931022803427e-05, "loss": 3.0202, "step": 1104 }, { "epoch": 0.6333118602820127, "grad_norm": 0.6676930785179138, "learning_rate": 8.184639574431532e-05, "loss": 3.0692, "step": 1106 }, { "epoch": 0.6344570896857776, "grad_norm": 0.8161568641662598, "learning_rate": 8.177336712913194e-05, "loss": 3.0835, "step": 1108 }, { "epoch": 0.6356023190895427, "grad_norm": 0.9007164239883423, "learning_rate": 8.170022464420486e-05, "loss": 3.0665, "step": 1110 }, { "epoch": 0.6367475484933076, "grad_norm": 0.7467122673988342, "learning_rate": 8.162696855166294e-05, "loss": 3.1098, "step": 1112 }, { "epoch": 0.6378927778970725, "grad_norm": 0.5917842388153076, "learning_rate": 8.155359911404217e-05, "loss": 3.0047, "step": 1114 }, { "epoch": 0.6390380073008375, "grad_norm": 0.7426056861877441, "learning_rate": 8.148011659428474e-05, "loss": 3.1037, "step": 1116 }, { "epoch": 0.6401832367046024, "grad_norm": 0.8367446660995483, "learning_rate": 8.140652125573813e-05, "loss": 2.9628, "step": 1118 }, { "epoch": 0.6413284661083674, "grad_norm": 0.6199979186058044, "learning_rate": 8.133281336215412e-05, "loss": 3.0239, "step": 1120 }, { "epoch": 0.6424736955121323, "grad_norm": 0.7761691808700562, "learning_rate": 8.125899317768786e-05, "loss": 3.0609, "step": 1122 }, { "epoch": 0.6436189249158972, "grad_norm": 0.6743906140327454, "learning_rate": 8.118506096689698e-05, "loss": 3.0696, "step": 1124 }, { "epoch": 0.6447641543196622, "grad_norm": 0.6743597388267517, "learning_rate": 8.111101699474051e-05, "loss": 3.045, "step": 1126 }, { "epoch": 0.6459093837234271, "grad_norm": 0.6069556474685669, "learning_rate": 8.103686152657808e-05, "loss": 3.0171, "step": 1128 }, { "epoch": 0.647054613127192, "grad_norm": 0.5756711959838867, "learning_rate": 8.096259482816886e-05, "loss": 3.0161, "step": 1130 }, { "epoch": 0.648199842530957, "grad_norm": 0.6239808797836304, "learning_rate": 8.088821716567066e-05, "loss": 3.0887, "step": 1132 }, { "epoch": 0.6493450719347219, "grad_norm": 0.5236758589744568, "learning_rate": 8.081372880563898e-05, "loss": 2.9743, "step": 1134 }, { "epoch": 0.6504903013384868, "grad_norm": 0.6389586925506592, "learning_rate": 8.073913001502605e-05, "loss": 2.9972, "step": 1136 }, { "epoch": 0.6516355307422518, "grad_norm": 0.5799978971481323, "learning_rate": 8.066442106117978e-05, "loss": 3.0043, "step": 1138 }, { "epoch": 0.6527807601460167, "grad_norm": 0.6182774901390076, "learning_rate": 8.058960221184298e-05, "loss": 3.065, "step": 1140 }, { "epoch": 0.6539259895497816, "grad_norm": 0.5762799382209778, "learning_rate": 8.051467373515228e-05, "loss": 3.0374, "step": 1142 }, { "epoch": 0.6550712189535466, "grad_norm": 0.6707761287689209, "learning_rate": 8.043963589963714e-05, "loss": 3.0056, "step": 1144 }, { "epoch": 0.6562164483573115, "grad_norm": 0.6148689985275269, "learning_rate": 8.036448897421903e-05, "loss": 3.0222, "step": 1146 }, { "epoch": 0.6573616777610766, "grad_norm": 0.6503751277923584, "learning_rate": 8.028923322821031e-05, "loss": 3.0186, "step": 1148 }, { "epoch": 0.6585069071648415, "grad_norm": 0.7160323858261108, "learning_rate": 8.021386893131334e-05, "loss": 3.0785, "step": 1150 }, { "epoch": 0.6596521365686064, "grad_norm": 0.5561687350273132, "learning_rate": 8.013839635361953e-05, "loss": 3.0425, "step": 1152 }, { "epoch": 0.6607973659723714, "grad_norm": 0.7166488170623779, "learning_rate": 8.006281576560834e-05, "loss": 2.9722, "step": 1154 }, { "epoch": 0.6619425953761363, "grad_norm": 0.5913854241371155, "learning_rate": 7.99871274381463e-05, "loss": 2.9645, "step": 1156 }, { "epoch": 0.6630878247799012, "grad_norm": 0.5721243619918823, "learning_rate": 7.99113316424861e-05, "loss": 3.0326, "step": 1158 }, { "epoch": 0.6642330541836662, "grad_norm": 0.6886599659919739, "learning_rate": 7.983542865026552e-05, "loss": 2.9934, "step": 1160 }, { "epoch": 0.6653782835874311, "grad_norm": 0.6505165100097656, "learning_rate": 7.975941873350656e-05, "loss": 3.0275, "step": 1162 }, { "epoch": 0.666523512991196, "grad_norm": 0.5963652729988098, "learning_rate": 7.968330216461439e-05, "loss": 3.0581, "step": 1164 }, { "epoch": 0.667668742394961, "grad_norm": 0.7020843029022217, "learning_rate": 7.960707921637642e-05, "loss": 3.0214, "step": 1166 }, { "epoch": 0.6688139717987259, "grad_norm": 0.5729818344116211, "learning_rate": 7.953075016196128e-05, "loss": 3.0928, "step": 1168 }, { "epoch": 0.6699592012024909, "grad_norm": 0.6135843992233276, "learning_rate": 7.945431527491788e-05, "loss": 3.0281, "step": 1170 }, { "epoch": 0.6711044306062558, "grad_norm": 0.844972550868988, "learning_rate": 7.937777482917441e-05, "loss": 3.0451, "step": 1172 }, { "epoch": 0.6722496600100207, "grad_norm": 0.6200757026672363, "learning_rate": 7.930112909903737e-05, "loss": 2.9982, "step": 1174 }, { "epoch": 0.6733948894137857, "grad_norm": 0.6621441841125488, "learning_rate": 7.922437835919059e-05, "loss": 3.0133, "step": 1176 }, { "epoch": 0.6745401188175506, "grad_norm": 0.6099239587783813, "learning_rate": 7.914752288469418e-05, "loss": 3.0359, "step": 1178 }, { "epoch": 0.6756853482213155, "grad_norm": 0.6415863633155823, "learning_rate": 7.907056295098367e-05, "loss": 3.0456, "step": 1180 }, { "epoch": 0.6768305776250805, "grad_norm": 0.5651492476463318, "learning_rate": 7.89934988338689e-05, "loss": 3.0138, "step": 1182 }, { "epoch": 0.6779758070288454, "grad_norm": 0.6899843215942383, "learning_rate": 7.891633080953309e-05, "loss": 3.1091, "step": 1184 }, { "epoch": 0.6791210364326105, "grad_norm": 0.6236230134963989, "learning_rate": 7.883905915453191e-05, "loss": 3.0477, "step": 1186 }, { "epoch": 0.6802662658363754, "grad_norm": 0.6632122993469238, "learning_rate": 7.876168414579232e-05, "loss": 3.0023, "step": 1188 }, { "epoch": 0.6814114952401403, "grad_norm": 0.5697975158691406, "learning_rate": 7.868420606061174e-05, "loss": 3.0046, "step": 1190 }, { "epoch": 0.6825567246439053, "grad_norm": 0.6847530603408813, "learning_rate": 7.8606625176657e-05, "loss": 3.0155, "step": 1192 }, { "epoch": 0.6837019540476702, "grad_norm": 0.6649438738822937, "learning_rate": 7.852894177196333e-05, "loss": 3.0616, "step": 1194 }, { "epoch": 0.6848471834514351, "grad_norm": 0.6214346289634705, "learning_rate": 7.845115612493335e-05, "loss": 2.963, "step": 1196 }, { "epoch": 0.6859924128552001, "grad_norm": 0.6587514877319336, "learning_rate": 7.837326851433614e-05, "loss": 3.0344, "step": 1198 }, { "epoch": 0.687137642258965, "grad_norm": 0.6454896330833435, "learning_rate": 7.829527921930617e-05, "loss": 3.0191, "step": 1200 }, { "epoch": 0.68828287166273, "grad_norm": 0.7053160071372986, "learning_rate": 7.821718851934235e-05, "loss": 2.9888, "step": 1202 }, { "epoch": 0.6894281010664949, "grad_norm": 0.6341421604156494, "learning_rate": 7.813899669430695e-05, "loss": 3.0688, "step": 1204 }, { "epoch": 0.6905733304702598, "grad_norm": 0.5567854642868042, "learning_rate": 7.806070402442476e-05, "loss": 3.0409, "step": 1206 }, { "epoch": 0.6917185598740248, "grad_norm": 0.5970193147659302, "learning_rate": 7.798231079028186e-05, "loss": 3.0569, "step": 1208 }, { "epoch": 0.6928637892777897, "grad_norm": 0.6818400621414185, "learning_rate": 7.79038172728248e-05, "loss": 2.9987, "step": 1210 }, { "epoch": 0.6940090186815546, "grad_norm": 0.7994418740272522, "learning_rate": 7.782522375335954e-05, "loss": 2.9835, "step": 1212 }, { "epoch": 0.6951542480853196, "grad_norm": 0.7498995065689087, "learning_rate": 7.774653051355039e-05, "loss": 3.0296, "step": 1214 }, { "epoch": 0.6962994774890845, "grad_norm": 0.6866205930709839, "learning_rate": 7.766773783541902e-05, "loss": 3.0301, "step": 1216 }, { "epoch": 0.6974447068928494, "grad_norm": 0.6896367073059082, "learning_rate": 7.758884600134354e-05, "loss": 3.0219, "step": 1218 }, { "epoch": 0.6985899362966144, "grad_norm": 0.668050229549408, "learning_rate": 7.750985529405736e-05, "loss": 3.0341, "step": 1220 }, { "epoch": 0.6997351657003793, "grad_norm": 0.666374146938324, "learning_rate": 7.743076599664824e-05, "loss": 3.0484, "step": 1222 }, { "epoch": 0.7008803951041443, "grad_norm": 0.7941027283668518, "learning_rate": 7.735157839255728e-05, "loss": 3.0032, "step": 1224 }, { "epoch": 0.7020256245079093, "grad_norm": 0.6927073001861572, "learning_rate": 7.727229276557791e-05, "loss": 3.0483, "step": 1226 }, { "epoch": 0.7031708539116742, "grad_norm": 0.6417405009269714, "learning_rate": 7.71929093998548e-05, "loss": 3.0354, "step": 1228 }, { "epoch": 0.7043160833154392, "grad_norm": 0.6400049328804016, "learning_rate": 7.711342857988295e-05, "loss": 3.013, "step": 1230 }, { "epoch": 0.7054613127192041, "grad_norm": 0.6526817083358765, "learning_rate": 7.703385059050662e-05, "loss": 3.1439, "step": 1232 }, { "epoch": 0.706606542122969, "grad_norm": 0.5762894749641418, "learning_rate": 7.695417571691825e-05, "loss": 2.9981, "step": 1234 }, { "epoch": 0.707751771526734, "grad_norm": 0.6319015622138977, "learning_rate": 7.687440424465755e-05, "loss": 3.0592, "step": 1236 }, { "epoch": 0.7088970009304989, "grad_norm": 0.6258828043937683, "learning_rate": 7.679453645961039e-05, "loss": 3.0415, "step": 1238 }, { "epoch": 0.7100422303342638, "grad_norm": 0.6156604290008545, "learning_rate": 7.671457264800784e-05, "loss": 3.0128, "step": 1240 }, { "epoch": 0.7111874597380288, "grad_norm": 0.6246117353439331, "learning_rate": 7.663451309642509e-05, "loss": 3.0324, "step": 1242 }, { "epoch": 0.7123326891417937, "grad_norm": 0.5683282017707825, "learning_rate": 7.65543580917804e-05, "loss": 2.9937, "step": 1244 }, { "epoch": 0.7134779185455586, "grad_norm": 0.7062090635299683, "learning_rate": 7.647410792133422e-05, "loss": 3.0414, "step": 1246 }, { "epoch": 0.7146231479493236, "grad_norm": 0.6515636444091797, "learning_rate": 7.639376287268798e-05, "loss": 3.029, "step": 1248 }, { "epoch": 0.7157683773530885, "grad_norm": 0.6117521524429321, "learning_rate": 7.631332323378314e-05, "loss": 3.0049, "step": 1250 }, { "epoch": 0.7169136067568534, "grad_norm": 0.7275700569152832, "learning_rate": 7.623278929290013e-05, "loss": 3.0346, "step": 1252 }, { "epoch": 0.7180588361606184, "grad_norm": 0.6181052923202515, "learning_rate": 7.615216133865744e-05, "loss": 2.9997, "step": 1254 }, { "epoch": 0.7192040655643833, "grad_norm": 0.5979108810424805, "learning_rate": 7.607143966001041e-05, "loss": 3.0351, "step": 1256 }, { "epoch": 0.7203492949681483, "grad_norm": 0.6783589720726013, "learning_rate": 7.599062454625023e-05, "loss": 2.9636, "step": 1258 }, { "epoch": 0.7214945243719133, "grad_norm": 0.6468908190727234, "learning_rate": 7.590971628700305e-05, "loss": 3.0285, "step": 1260 }, { "epoch": 0.7226397537756782, "grad_norm": 0.6515734195709229, "learning_rate": 7.582871517222876e-05, "loss": 3.0494, "step": 1262 }, { "epoch": 0.7237849831794432, "grad_norm": 0.6727483868598938, "learning_rate": 7.574762149222007e-05, "loss": 3.08, "step": 1264 }, { "epoch": 0.7249302125832081, "grad_norm": 0.70179283618927, "learning_rate": 7.566643553760138e-05, "loss": 2.9622, "step": 1266 }, { "epoch": 0.726075441986973, "grad_norm": 0.7478881478309631, "learning_rate": 7.558515759932782e-05, "loss": 3.0535, "step": 1268 }, { "epoch": 0.727220671390738, "grad_norm": 0.7690526247024536, "learning_rate": 7.550378796868417e-05, "loss": 3.0019, "step": 1270 }, { "epoch": 0.7283659007945029, "grad_norm": 0.6407628059387207, "learning_rate": 7.542232693728379e-05, "loss": 2.9642, "step": 1272 }, { "epoch": 0.7295111301982679, "grad_norm": 0.5795389413833618, "learning_rate": 7.534077479706764e-05, "loss": 3.0628, "step": 1274 }, { "epoch": 0.7306563596020328, "grad_norm": 0.6217789053916931, "learning_rate": 7.525913184030319e-05, "loss": 3.0264, "step": 1276 }, { "epoch": 0.7318015890057977, "grad_norm": 0.6992785930633545, "learning_rate": 7.517739835958335e-05, "loss": 3.0133, "step": 1278 }, { "epoch": 0.7329468184095627, "grad_norm": 0.6179572343826294, "learning_rate": 7.509557464782546e-05, "loss": 2.9636, "step": 1280 }, { "epoch": 0.7340920478133276, "grad_norm": 0.5889167189598083, "learning_rate": 7.501366099827025e-05, "loss": 3.0598, "step": 1282 }, { "epoch": 0.7352372772170925, "grad_norm": 0.6687774062156677, "learning_rate": 7.493165770448078e-05, "loss": 3.0248, "step": 1284 }, { "epoch": 0.7363825066208575, "grad_norm": 0.6092299818992615, "learning_rate": 7.484956506034136e-05, "loss": 3.0009, "step": 1286 }, { "epoch": 0.7375277360246224, "grad_norm": 0.5790627598762512, "learning_rate": 7.476738336005647e-05, "loss": 2.9792, "step": 1288 }, { "epoch": 0.7386729654283873, "grad_norm": 0.5962207913398743, "learning_rate": 7.468511289814983e-05, "loss": 2.9984, "step": 1290 }, { "epoch": 0.7398181948321523, "grad_norm": 0.6334844827651978, "learning_rate": 7.460275396946323e-05, "loss": 3.041, "step": 1292 }, { "epoch": 0.7409634242359172, "grad_norm": 0.6076721549034119, "learning_rate": 7.45203068691555e-05, "loss": 3.0255, "step": 1294 }, { "epoch": 0.7421086536396821, "grad_norm": 0.7529584169387817, "learning_rate": 7.443777189270147e-05, "loss": 3.0043, "step": 1296 }, { "epoch": 0.7432538830434472, "grad_norm": 0.6646864414215088, "learning_rate": 7.435514933589089e-05, "loss": 3.0181, "step": 1298 }, { "epoch": 0.7443991124472121, "grad_norm": 0.6356727480888367, "learning_rate": 7.427243949482741e-05, "loss": 3.0157, "step": 1300 }, { "epoch": 0.7455443418509771, "grad_norm": 0.5824810266494751, "learning_rate": 7.418964266592744e-05, "loss": 2.9286, "step": 1302 }, { "epoch": 0.746689571254742, "grad_norm": 0.6385902166366577, "learning_rate": 7.410675914591921e-05, "loss": 2.9807, "step": 1304 }, { "epoch": 0.7478348006585069, "grad_norm": 0.7312979102134705, "learning_rate": 7.402378923184156e-05, "loss": 3.0521, "step": 1306 }, { "epoch": 0.7489800300622719, "grad_norm": 0.5312052965164185, "learning_rate": 7.394073322104298e-05, "loss": 2.8997, "step": 1308 }, { "epoch": 0.7501252594660368, "grad_norm": 0.6032156348228455, "learning_rate": 7.385759141118055e-05, "loss": 2.9969, "step": 1310 }, { "epoch": 0.7512704888698017, "grad_norm": 0.6151460409164429, "learning_rate": 7.377436410021878e-05, "loss": 2.9889, "step": 1312 }, { "epoch": 0.7524157182735667, "grad_norm": 0.6497369408607483, "learning_rate": 7.369105158642863e-05, "loss": 3.0514, "step": 1314 }, { "epoch": 0.7535609476773316, "grad_norm": 0.5319530367851257, "learning_rate": 7.360765416838643e-05, "loss": 2.9673, "step": 1316 }, { "epoch": 0.7547061770810966, "grad_norm": 0.6640217900276184, "learning_rate": 7.352417214497272e-05, "loss": 2.9538, "step": 1318 }, { "epoch": 0.7558514064848615, "grad_norm": 0.5952315926551819, "learning_rate": 7.344060581537134e-05, "loss": 3.0166, "step": 1320 }, { "epoch": 0.7569966358886264, "grad_norm": 0.6607936024665833, "learning_rate": 7.335695547906821e-05, "loss": 3.053, "step": 1322 }, { "epoch": 0.7581418652923914, "grad_norm": 0.5945944786071777, "learning_rate": 7.327322143585033e-05, "loss": 2.9579, "step": 1324 }, { "epoch": 0.7592870946961563, "grad_norm": 0.5982474088668823, "learning_rate": 7.318940398580467e-05, "loss": 3.0381, "step": 1326 }, { "epoch": 0.7604323240999212, "grad_norm": 0.557156503200531, "learning_rate": 7.310550342931714e-05, "loss": 2.9517, "step": 1328 }, { "epoch": 0.7615775535036862, "grad_norm": 0.519133985042572, "learning_rate": 7.30215200670715e-05, "loss": 2.95, "step": 1330 }, { "epoch": 0.7627227829074511, "grad_norm": 0.6734493970870972, "learning_rate": 7.293745420004823e-05, "loss": 3.0136, "step": 1332 }, { "epoch": 0.763868012311216, "grad_norm": 0.6518685817718506, "learning_rate": 7.28533061295235e-05, "loss": 2.9696, "step": 1334 }, { "epoch": 0.7650132417149811, "grad_norm": 0.5936789512634277, "learning_rate": 7.276907615706814e-05, "loss": 2.9877, "step": 1336 }, { "epoch": 0.766158471118746, "grad_norm": 0.5671294331550598, "learning_rate": 7.268476458454642e-05, "loss": 2.9858, "step": 1338 }, { "epoch": 0.767303700522511, "grad_norm": 0.5697634816169739, "learning_rate": 7.260037171411508e-05, "loss": 2.9434, "step": 1340 }, { "epoch": 0.7684489299262759, "grad_norm": 0.6007529497146606, "learning_rate": 7.251589784822224e-05, "loss": 2.9387, "step": 1342 }, { "epoch": 0.7695941593300408, "grad_norm": 0.5928565263748169, "learning_rate": 7.243134328960625e-05, "loss": 2.9275, "step": 1344 }, { "epoch": 0.7707393887338058, "grad_norm": 0.6670402884483337, "learning_rate": 7.234670834129469e-05, "loss": 2.9812, "step": 1346 }, { "epoch": 0.7718846181375707, "grad_norm": 0.6268984079360962, "learning_rate": 7.226199330660322e-05, "loss": 2.9359, "step": 1348 }, { "epoch": 0.7730298475413356, "grad_norm": 0.5983556509017944, "learning_rate": 7.217719848913451e-05, "loss": 2.9524, "step": 1350 }, { "epoch": 0.7741750769451006, "grad_norm": 0.5522739291191101, "learning_rate": 7.209232419277714e-05, "loss": 2.9182, "step": 1352 }, { "epoch": 0.7753203063488655, "grad_norm": 0.5227957367897034, "learning_rate": 7.20073707217046e-05, "loss": 2.9703, "step": 1354 }, { "epoch": 0.7764655357526304, "grad_norm": 0.484824538230896, "learning_rate": 7.192233838037403e-05, "loss": 2.985, "step": 1356 }, { "epoch": 0.7776107651563954, "grad_norm": 0.5953473448753357, "learning_rate": 7.183722747352531e-05, "loss": 2.9942, "step": 1358 }, { "epoch": 0.7787559945601603, "grad_norm": 0.550989031791687, "learning_rate": 7.175203830617983e-05, "loss": 2.9782, "step": 1360 }, { "epoch": 0.7799012239639252, "grad_norm": 0.5300703048706055, "learning_rate": 7.166677118363945e-05, "loss": 2.9471, "step": 1362 }, { "epoch": 0.7810464533676902, "grad_norm": 0.6735565066337585, "learning_rate": 7.158142641148546e-05, "loss": 2.9581, "step": 1364 }, { "epoch": 0.7821916827714551, "grad_norm": 0.6232147216796875, "learning_rate": 7.149600429557734e-05, "loss": 2.9921, "step": 1366 }, { "epoch": 0.7833369121752201, "grad_norm": 0.563504159450531, "learning_rate": 7.14105051420519e-05, "loss": 2.9923, "step": 1368 }, { "epoch": 0.784482141578985, "grad_norm": 0.5748158097267151, "learning_rate": 7.132492925732187e-05, "loss": 3.0145, "step": 1370 }, { "epoch": 0.7856273709827499, "grad_norm": 0.5430135130882263, "learning_rate": 7.12392769480751e-05, "loss": 2.9535, "step": 1372 }, { "epoch": 0.786772600386515, "grad_norm": 0.6344389319419861, "learning_rate": 7.115354852127324e-05, "loss": 3.0599, "step": 1374 }, { "epoch": 0.7879178297902799, "grad_norm": 0.6093223094940186, "learning_rate": 7.106774428415079e-05, "loss": 3.0051, "step": 1376 }, { "epoch": 0.7890630591940448, "grad_norm": 0.5969236493110657, "learning_rate": 7.098186454421393e-05, "loss": 2.9106, "step": 1378 }, { "epoch": 0.7902082885978098, "grad_norm": 0.5839152336120605, "learning_rate": 7.089590960923943e-05, "loss": 2.9917, "step": 1380 }, { "epoch": 0.7913535180015747, "grad_norm": 0.6248955130577087, "learning_rate": 7.080987978727349e-05, "loss": 3.0072, "step": 1382 }, { "epoch": 0.7924987474053397, "grad_norm": 0.6621485352516174, "learning_rate": 7.072377538663079e-05, "loss": 2.9616, "step": 1384 }, { "epoch": 0.7936439768091046, "grad_norm": 0.5360151529312134, "learning_rate": 7.063759671589319e-05, "loss": 3.0115, "step": 1386 }, { "epoch": 0.7947892062128695, "grad_norm": 0.6625118851661682, "learning_rate": 7.055134408390877e-05, "loss": 3.0551, "step": 1388 }, { "epoch": 0.7959344356166345, "grad_norm": 0.5611973404884338, "learning_rate": 7.046501779979068e-05, "loss": 2.9427, "step": 1390 }, { "epoch": 0.7970796650203994, "grad_norm": 0.6025654673576355, "learning_rate": 7.037861817291598e-05, "loss": 2.985, "step": 1392 }, { "epoch": 0.7982248944241643, "grad_norm": 0.7166057229042053, "learning_rate": 7.029214551292465e-05, "loss": 3.006, "step": 1394 }, { "epoch": 0.7993701238279293, "grad_norm": 0.586137056350708, "learning_rate": 7.020560012971832e-05, "loss": 2.9835, "step": 1396 }, { "epoch": 0.8005153532316942, "grad_norm": 0.7084041833877563, "learning_rate": 7.011898233345931e-05, "loss": 2.9931, "step": 1398 }, { "epoch": 0.8016605826354591, "grad_norm": 0.6771812438964844, "learning_rate": 7.003229243456944e-05, "loss": 3.0061, "step": 1400 }, { "epoch": 0.8028058120392241, "grad_norm": 0.6903461813926697, "learning_rate": 6.994553074372891e-05, "loss": 2.977, "step": 1402 }, { "epoch": 0.803951041442989, "grad_norm": 0.6033274531364441, "learning_rate": 6.985869757187523e-05, "loss": 2.935, "step": 1404 }, { "epoch": 0.805096270846754, "grad_norm": 0.6533209085464478, "learning_rate": 6.977179323020207e-05, "loss": 3.0233, "step": 1406 }, { "epoch": 0.8062415002505189, "grad_norm": 0.5902218222618103, "learning_rate": 6.96848180301582e-05, "loss": 3.0012, "step": 1408 }, { "epoch": 0.8073867296542838, "grad_norm": 0.6266187429428101, "learning_rate": 6.959777228344628e-05, "loss": 2.9518, "step": 1410 }, { "epoch": 0.8085319590580489, "grad_norm": 0.5876622200012207, "learning_rate": 6.95106563020218e-05, "loss": 2.9377, "step": 1412 }, { "epoch": 0.8096771884618138, "grad_norm": 0.5977484583854675, "learning_rate": 6.942347039809201e-05, "loss": 2.9814, "step": 1414 }, { "epoch": 0.8108224178655787, "grad_norm": 0.64298015832901, "learning_rate": 6.933621488411468e-05, "loss": 3.0269, "step": 1416 }, { "epoch": 0.8119676472693437, "grad_norm": 0.5676386952400208, "learning_rate": 6.924889007279712e-05, "loss": 2.9237, "step": 1418 }, { "epoch": 0.8131128766731086, "grad_norm": 0.5827841758728027, "learning_rate": 6.916149627709494e-05, "loss": 2.867, "step": 1420 }, { "epoch": 0.8142581060768735, "grad_norm": 0.6391154527664185, "learning_rate": 6.907403381021097e-05, "loss": 2.9802, "step": 1422 }, { "epoch": 0.8154033354806385, "grad_norm": 0.5791281461715698, "learning_rate": 6.89865029855942e-05, "loss": 2.9945, "step": 1424 }, { "epoch": 0.8165485648844034, "grad_norm": 0.5360828638076782, "learning_rate": 6.88989041169385e-05, "loss": 2.9882, "step": 1426 }, { "epoch": 0.8176937942881684, "grad_norm": 0.5865568518638611, "learning_rate": 6.881123751818175e-05, "loss": 2.9096, "step": 1428 }, { "epoch": 0.8188390236919333, "grad_norm": 0.5686046481132507, "learning_rate": 6.87235035035044e-05, "loss": 3.0026, "step": 1430 }, { "epoch": 0.8199842530956982, "grad_norm": 0.6230041980743408, "learning_rate": 6.863570238732862e-05, "loss": 2.9601, "step": 1432 }, { "epoch": 0.8211294824994632, "grad_norm": 0.61517733335495, "learning_rate": 6.854783448431702e-05, "loss": 2.9286, "step": 1434 }, { "epoch": 0.8222747119032281, "grad_norm": 0.5126785635948181, "learning_rate": 6.845990010937152e-05, "loss": 2.9896, "step": 1436 }, { "epoch": 0.823419941306993, "grad_norm": 0.5101936459541321, "learning_rate": 6.837189957763234e-05, "loss": 2.9612, "step": 1438 }, { "epoch": 0.824565170710758, "grad_norm": 0.6595968008041382, "learning_rate": 6.828383320447675e-05, "loss": 3.0712, "step": 1440 }, { "epoch": 0.8257104001145229, "grad_norm": 0.5970346927642822, "learning_rate": 6.8195701305518e-05, "loss": 2.9534, "step": 1442 }, { "epoch": 0.8268556295182878, "grad_norm": 0.645318865776062, "learning_rate": 6.810750419660415e-05, "loss": 2.9673, "step": 1444 }, { "epoch": 0.8280008589220528, "grad_norm": 0.6072613596916199, "learning_rate": 6.801924219381695e-05, "loss": 2.9525, "step": 1446 }, { "epoch": 0.8291460883258178, "grad_norm": 0.5918760299682617, "learning_rate": 6.793091561347078e-05, "loss": 2.961, "step": 1448 }, { "epoch": 0.8302913177295828, "grad_norm": 0.6208258271217346, "learning_rate": 6.784252477211138e-05, "loss": 2.9874, "step": 1450 }, { "epoch": 0.8314365471333477, "grad_norm": 0.601787805557251, "learning_rate": 6.775406998651484e-05, "loss": 2.9412, "step": 1452 }, { "epoch": 0.8325817765371126, "grad_norm": 0.5628416538238525, "learning_rate": 6.76655515736864e-05, "loss": 2.9098, "step": 1454 }, { "epoch": 0.8337270059408776, "grad_norm": 0.559907078742981, "learning_rate": 6.757696985085931e-05, "loss": 3.0108, "step": 1456 }, { "epoch": 0.8348722353446425, "grad_norm": 0.5879938006401062, "learning_rate": 6.748832513549373e-05, "loss": 2.9376, "step": 1458 }, { "epoch": 0.8360174647484074, "grad_norm": 0.581619143486023, "learning_rate": 6.739961774527557e-05, "loss": 2.9491, "step": 1460 }, { "epoch": 0.8371626941521724, "grad_norm": 0.686090350151062, "learning_rate": 6.731084799811536e-05, "loss": 2.9458, "step": 1462 }, { "epoch": 0.8383079235559373, "grad_norm": 0.6862696409225464, "learning_rate": 6.72220162121471e-05, "loss": 3.0272, "step": 1464 }, { "epoch": 0.8394531529597022, "grad_norm": 0.5614147782325745, "learning_rate": 6.713312270572711e-05, "loss": 2.9941, "step": 1466 }, { "epoch": 0.8405983823634672, "grad_norm": 0.6119958162307739, "learning_rate": 6.704416779743295e-05, "loss": 2.9241, "step": 1468 }, { "epoch": 0.8417436117672321, "grad_norm": 0.5873327851295471, "learning_rate": 6.695515180606217e-05, "loss": 2.9539, "step": 1470 }, { "epoch": 0.842888841170997, "grad_norm": 0.5834409594535828, "learning_rate": 6.686607505063127e-05, "loss": 2.981, "step": 1472 }, { "epoch": 0.844034070574762, "grad_norm": 0.6550679206848145, "learning_rate": 6.677693785037452e-05, "loss": 2.9575, "step": 1474 }, { "epoch": 0.8451792999785269, "grad_norm": 0.7193329930305481, "learning_rate": 6.668774052474278e-05, "loss": 2.9882, "step": 1476 }, { "epoch": 0.8463245293822919, "grad_norm": 0.741956353187561, "learning_rate": 6.659848339340243e-05, "loss": 2.9814, "step": 1478 }, { "epoch": 0.8474697587860568, "grad_norm": 0.7412140369415283, "learning_rate": 6.650916677623415e-05, "loss": 2.9989, "step": 1480 }, { "epoch": 0.8486149881898217, "grad_norm": 0.709784984588623, "learning_rate": 6.641979099333183e-05, "loss": 2.9654, "step": 1482 }, { "epoch": 0.8497602175935867, "grad_norm": 0.6807482838630676, "learning_rate": 6.633035636500137e-05, "loss": 2.9919, "step": 1484 }, { "epoch": 0.8509054469973517, "grad_norm": 0.6146615147590637, "learning_rate": 6.62408632117596e-05, "loss": 2.9711, "step": 1486 }, { "epoch": 0.8520506764011166, "grad_norm": 0.4998982846736908, "learning_rate": 6.615131185433306e-05, "loss": 2.9859, "step": 1488 }, { "epoch": 0.8531959058048816, "grad_norm": 0.577151358127594, "learning_rate": 6.606170261365689e-05, "loss": 3.0008, "step": 1490 }, { "epoch": 0.8543411352086465, "grad_norm": 0.5823954939842224, "learning_rate": 6.597203581087367e-05, "loss": 2.9355, "step": 1492 }, { "epoch": 0.8554863646124115, "grad_norm": 0.5608075857162476, "learning_rate": 6.588231176733228e-05, "loss": 2.9701, "step": 1494 }, { "epoch": 0.8566315940161764, "grad_norm": 0.5515905618667603, "learning_rate": 6.579253080458676e-05, "loss": 2.9437, "step": 1496 }, { "epoch": 0.8577768234199413, "grad_norm": 0.5510479211807251, "learning_rate": 6.570269324439509e-05, "loss": 2.889, "step": 1498 }, { "epoch": 0.8589220528237063, "grad_norm": 0.5867220163345337, "learning_rate": 6.561279940871809e-05, "loss": 2.9781, "step": 1500 } ], "logging_steps": 2, "max_steps": 3494, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 300, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.688775168098304e+18, "train_batch_size": 16, "trial_name": null, "trial_params": null }