| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9993227226549273, | |
| "eval_steps": 500, | |
| "global_step": 2952, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 11.120684732523612, | |
| "learning_rate": 6.756756756756757e-08, | |
| "loss": 2.4477, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 11.774652911409536, | |
| "learning_rate": 3.378378378378379e-07, | |
| "loss": 2.4743, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 12.419730474838534, | |
| "learning_rate": 6.756756756756758e-07, | |
| "loss": 2.4645, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.095378312665702, | |
| "learning_rate": 1.0135135135135136e-06, | |
| "loss": 2.4392, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.615296162867837, | |
| "learning_rate": 1.3513513513513515e-06, | |
| "loss": 2.3786, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.297332915695467, | |
| "learning_rate": 1.6891891891891894e-06, | |
| "loss": 2.3767, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.1552306936844987, | |
| "learning_rate": 2.0270270270270273e-06, | |
| "loss": 2.3153, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.9425651285523513, | |
| "learning_rate": 2.364864864864865e-06, | |
| "loss": 2.3097, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.7387554776828598, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 2.2923, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.6966595072053137, | |
| "learning_rate": 3.040540540540541e-06, | |
| "loss": 2.245, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.3994061705483904, | |
| "learning_rate": 3.3783783783783788e-06, | |
| "loss": 2.2321, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.354333109658359, | |
| "learning_rate": 3.7162162162162162e-06, | |
| "loss": 2.24, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.3337171978672469, | |
| "learning_rate": 4.0540540540540545e-06, | |
| "loss": 2.2361, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.2852075781932446, | |
| "learning_rate": 4.391891891891892e-06, | |
| "loss": 2.223, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.2694804239215547, | |
| "learning_rate": 4.72972972972973e-06, | |
| "loss": 2.2024, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.2950266734252336, | |
| "learning_rate": 5.067567567567568e-06, | |
| "loss": 2.2216, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.2180963016508284, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 2.2042, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.2706298695152691, | |
| "learning_rate": 5.743243243243244e-06, | |
| "loss": 2.1702, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.225095411191456, | |
| "learning_rate": 6.081081081081082e-06, | |
| "loss": 2.1981, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.243002613296369, | |
| "learning_rate": 6.41891891891892e-06, | |
| "loss": 2.1781, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.2560076766685662, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 2.1736, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.1891064395338329, | |
| "learning_rate": 7.0945945945945946e-06, | |
| "loss": 2.1686, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.226934274632558, | |
| "learning_rate": 7.4324324324324324e-06, | |
| "loss": 2.18, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.2138187714548274, | |
| "learning_rate": 7.77027027027027e-06, | |
| "loss": 2.184, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.2024066179815232, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 2.165, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.187417375853499, | |
| "learning_rate": 8.445945945945948e-06, | |
| "loss": 2.1521, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.2395735830069963, | |
| "learning_rate": 8.783783783783785e-06, | |
| "loss": 2.1917, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.2057670018158435, | |
| "learning_rate": 9.121621621621622e-06, | |
| "loss": 2.1914, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 1.188365088426718, | |
| "learning_rate": 9.45945945945946e-06, | |
| "loss": 2.1472, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1850606269765205, | |
| "learning_rate": 9.797297297297298e-06, | |
| "loss": 2.1478, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.1754912020641461, | |
| "learning_rate": 1.0135135135135136e-05, | |
| "loss": 2.1482, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.200054193355099, | |
| "learning_rate": 1.0472972972972975e-05, | |
| "loss": 2.16, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.2069624738319888, | |
| "learning_rate": 1.0810810810810812e-05, | |
| "loss": 2.179, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.1875211726858623, | |
| "learning_rate": 1.114864864864865e-05, | |
| "loss": 2.1442, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.1970442354954136, | |
| "learning_rate": 1.1486486486486488e-05, | |
| "loss": 2.1452, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2380085636256686, | |
| "learning_rate": 1.1824324324324325e-05, | |
| "loss": 2.1617, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 1.2186028086528804, | |
| "learning_rate": 1.2162162162162164e-05, | |
| "loss": 2.1608, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.1914285662360458, | |
| "learning_rate": 1.25e-05, | |
| "loss": 2.1583, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.2728483786556704, | |
| "learning_rate": 1.283783783783784e-05, | |
| "loss": 2.1461, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.2612669236589429, | |
| "learning_rate": 1.3175675675675676e-05, | |
| "loss": 2.1733, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.223268595309795, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 2.1515, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.1402100950657685, | |
| "learning_rate": 1.3851351351351352e-05, | |
| "loss": 2.1704, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 1.2042308858654536, | |
| "learning_rate": 1.4189189189189189e-05, | |
| "loss": 2.1558, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.1976327364781685, | |
| "learning_rate": 1.4527027027027028e-05, | |
| "loss": 2.1512, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.1648839908303132, | |
| "learning_rate": 1.4864864864864865e-05, | |
| "loss": 2.1433, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.2761560435776693, | |
| "learning_rate": 1.5202702702702704e-05, | |
| "loss": 2.1616, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.1936717423207983, | |
| "learning_rate": 1.554054054054054e-05, | |
| "loss": 2.1514, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.1866474451010465, | |
| "learning_rate": 1.587837837837838e-05, | |
| "loss": 2.1425, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.1804394482439402, | |
| "learning_rate": 1.6216216216216218e-05, | |
| "loss": 2.1487, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.19571568650813, | |
| "learning_rate": 1.6554054054054057e-05, | |
| "loss": 2.1504, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.189974666975512, | |
| "learning_rate": 1.6891891891891896e-05, | |
| "loss": 2.1563, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.2121020660874857, | |
| "learning_rate": 1.722972972972973e-05, | |
| "loss": 2.1599, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2213881126051926, | |
| "learning_rate": 1.756756756756757e-05, | |
| "loss": 2.1637, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2039085562610379, | |
| "learning_rate": 1.790540540540541e-05, | |
| "loss": 2.1658, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.2533570096566717, | |
| "learning_rate": 1.8243243243243244e-05, | |
| "loss": 2.1455, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.1967415340782568, | |
| "learning_rate": 1.8581081081081082e-05, | |
| "loss": 2.1641, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.2176261062565596, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 2.1393, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 1.2097988408193903, | |
| "learning_rate": 1.925675675675676e-05, | |
| "loss": 2.1597, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.2214887920434019, | |
| "learning_rate": 1.9594594594594595e-05, | |
| "loss": 2.1408, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.1995715557157443, | |
| "learning_rate": 1.9932432432432434e-05, | |
| "loss": 2.1289, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.2146596386103696, | |
| "learning_rate": 1.9999888073536733e-05, | |
| "loss": 2.1567, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.1871587471621745, | |
| "learning_rate": 1.9999433376573803e-05, | |
| "loss": 2.1256, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.1620165315368023, | |
| "learning_rate": 1.9998628929598956e-05, | |
| "loss": 2.1772, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.1905321330401686, | |
| "learning_rate": 1.9997474760749314e-05, | |
| "loss": 2.1272, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.1854631764766068, | |
| "learning_rate": 1.9995970910394228e-05, | |
| "loss": 2.1439, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.1706205848945217, | |
| "learning_rate": 1.9994117431133843e-05, | |
| "loss": 2.1388, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.183111751493484, | |
| "learning_rate": 1.9991914387797266e-05, | |
| "loss": 2.1396, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.1718173935731735, | |
| "learning_rate": 1.9989361857440318e-05, | |
| "loss": 2.1667, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.1553887899123578, | |
| "learning_rate": 1.99864599293428e-05, | |
| "loss": 2.1337, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.1591936171753476, | |
| "learning_rate": 1.998320870500541e-05, | |
| "loss": 2.1549, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.1876321771856482, | |
| "learning_rate": 1.997960829814616e-05, | |
| "loss": 2.1502, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.1496756947355502, | |
| "learning_rate": 1.997565883469643e-05, | |
| "loss": 2.1464, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.254225286118427, | |
| "learning_rate": 1.9971360452796523e-05, | |
| "loss": 2.1423, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.1829653120874923, | |
| "learning_rate": 1.9966713302790875e-05, | |
| "loss": 2.1594, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.2331026648285757, | |
| "learning_rate": 1.9961717547222775e-05, | |
| "loss": 2.138, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.1432519800917014, | |
| "learning_rate": 1.995637336082868e-05, | |
| "loss": 2.1319, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.1665043849652652, | |
| "learning_rate": 1.9950680930532107e-05, | |
| "loss": 2.16, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.2234802198858359, | |
| "learning_rate": 1.994464045543709e-05, | |
| "loss": 2.1389, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.1727647815503826, | |
| "learning_rate": 1.9938252146821236e-05, | |
| "loss": 2.1531, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.2173942744598583, | |
| "learning_rate": 1.9931516228128295e-05, | |
| "loss": 2.1215, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.1684578346995043, | |
| "learning_rate": 1.9924432934960384e-05, | |
| "loss": 2.1478, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.1866757835839474, | |
| "learning_rate": 1.9917002515069732e-05, | |
| "loss": 2.1367, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.1981986885472475, | |
| "learning_rate": 1.9909225228350007e-05, | |
| "loss": 2.1456, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.1573172050728477, | |
| "learning_rate": 1.9901101346827233e-05, | |
| "loss": 2.1253, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.1777082337932487, | |
| "learning_rate": 1.989263115465028e-05, | |
| "loss": 2.1392, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.1238340239895352, | |
| "learning_rate": 1.9883814948080918e-05, | |
| "loss": 2.1402, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.1325433734718393, | |
| "learning_rate": 1.987465303548345e-05, | |
| "loss": 2.1237, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.1992543472262565, | |
| "learning_rate": 1.9865145737313937e-05, | |
| "loss": 2.1396, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.1631062343254268, | |
| "learning_rate": 1.9855293386108995e-05, | |
| "loss": 2.1344, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.1680414315123386, | |
| "learning_rate": 1.9845096326474135e-05, | |
| "loss": 2.1263, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.135418556953998, | |
| "learning_rate": 1.9834554915071745e-05, | |
| "loss": 2.1348, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.1764551162297905, | |
| "learning_rate": 1.982366952060859e-05, | |
| "loss": 2.1129, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.1306792793513711, | |
| "learning_rate": 1.981244052382293e-05, | |
| "loss": 2.1228, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.1819737279135798, | |
| "learning_rate": 1.9800868317471196e-05, | |
| "loss": 2.1332, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.1634001658406754, | |
| "learning_rate": 1.978895330631425e-05, | |
| "loss": 2.136, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.2145877501446412, | |
| "learning_rate": 1.977669590710324e-05, | |
| "loss": 2.1453, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.1335463914960855, | |
| "learning_rate": 1.976409654856501e-05, | |
| "loss": 2.1403, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.1689667725563104, | |
| "learning_rate": 1.975115567138711e-05, | |
| "loss": 2.1325, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 1.1250476400769953, | |
| "learning_rate": 1.9737873728202376e-05, | |
| "loss": 2.1362, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.1070265136556228, | |
| "learning_rate": 1.972425118357312e-05, | |
| "loss": 2.1507, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.1524405517321128, | |
| "learning_rate": 1.9710288513974846e-05, | |
| "loss": 2.1221, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.148194020672431, | |
| "learning_rate": 1.969598620777962e-05, | |
| "loss": 2.1174, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.1797355397893157, | |
| "learning_rate": 1.9681344765238958e-05, | |
| "loss": 2.1252, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.2271289828620666, | |
| "learning_rate": 1.966636469846635e-05, | |
| "loss": 2.1286, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.1868659451800048, | |
| "learning_rate": 1.9651046531419335e-05, | |
| "loss": 2.1054, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.130037832731905, | |
| "learning_rate": 1.9635390799881186e-05, | |
| "loss": 2.1078, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.1266023724170184, | |
| "learning_rate": 1.961939805144216e-05, | |
| "loss": 2.1334, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.1148557064465363, | |
| "learning_rate": 1.9603068845480347e-05, | |
| "loss": 2.1342, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.1347560168290718, | |
| "learning_rate": 1.9586403753142104e-05, | |
| "loss": 2.1322, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.1462877739298016, | |
| "learning_rate": 1.956940335732209e-05, | |
| "loss": 2.1584, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.1124505887293903, | |
| "learning_rate": 1.9552068252642858e-05, | |
| "loss": 2.1331, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.113156082275959, | |
| "learning_rate": 1.9534399045434073e-05, | |
| "loss": 2.1376, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.1386903791281329, | |
| "learning_rate": 1.9516396353711297e-05, | |
| "loss": 2.1427, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.1329464952286028, | |
| "learning_rate": 1.9498060807154368e-05, | |
| "loss": 2.1303, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.1251111696000566, | |
| "learning_rate": 1.9479393047085392e-05, | |
| "loss": 2.1375, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.1237720236234878, | |
| "learning_rate": 1.94603937264463e-05, | |
| "loss": 2.1389, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.1610343006739035, | |
| "learning_rate": 1.9441063509776003e-05, | |
| "loss": 2.123, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.146790657360117, | |
| "learning_rate": 1.9421403073187162e-05, | |
| "loss": 2.1352, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.1200470095088182, | |
| "learning_rate": 1.9401413104342535e-05, | |
| "loss": 2.1138, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.1169776305442638, | |
| "learning_rate": 1.938109430243093e-05, | |
| "loss": 2.1236, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.1777999758928328, | |
| "learning_rate": 1.936044737814273e-05, | |
| "loss": 2.1243, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.1481847604301374, | |
| "learning_rate": 1.9339473053645056e-05, | |
| "loss": 2.1342, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 1.1191204695363768, | |
| "learning_rate": 1.931817206255651e-05, | |
| "loss": 2.1147, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.113720888666362, | |
| "learning_rate": 1.9296545149921488e-05, | |
| "loss": 2.1135, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.1473063574306148, | |
| "learning_rate": 1.9274593072184152e-05, | |
| "loss": 2.1241, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.1320738105413926, | |
| "learning_rate": 1.9252316597161947e-05, | |
| "loss": 2.1344, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.124582310854143, | |
| "learning_rate": 1.922971650401877e-05, | |
| "loss": 2.1296, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.1240431751971176, | |
| "learning_rate": 1.920679358323769e-05, | |
| "loss": 2.1306, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.1675648906192655, | |
| "learning_rate": 1.9183548636593322e-05, | |
| "loss": 2.1397, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.1183247099705995, | |
| "learning_rate": 1.9159982477123776e-05, | |
| "loss": 2.1287, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.1995227425757071, | |
| "learning_rate": 1.9136095929102204e-05, | |
| "loss": 2.1107, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.1496316241314066, | |
| "learning_rate": 1.9111889828007997e-05, | |
| "loss": 2.1392, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.1379803405847972, | |
| "learning_rate": 1.908736502049754e-05, | |
| "loss": 2.1169, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.1185357090534018, | |
| "learning_rate": 1.9062522364374617e-05, | |
| "loss": 2.117, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.1493081478070881, | |
| "learning_rate": 1.903736272856038e-05, | |
| "loss": 2.1057, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.1406631430361929, | |
| "learning_rate": 1.9011886993062994e-05, | |
| "loss": 2.1338, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.236921611965705, | |
| "learning_rate": 1.8986096048946826e-05, | |
| "loss": 2.1112, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 1.0973004897827627, | |
| "learning_rate": 1.8959990798301286e-05, | |
| "loss": 2.1123, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.1362850942192249, | |
| "learning_rate": 1.893357215420929e-05, | |
| "loss": 2.1299, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.143824677944863, | |
| "learning_rate": 1.8906841040715304e-05, | |
| "loss": 2.1498, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.1142169829634871, | |
| "learning_rate": 1.8879798392793033e-05, | |
| "loss": 2.1145, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1234810265193804, | |
| "learning_rate": 1.8852445156312713e-05, | |
| "loss": 2.1446, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1512490139550935, | |
| "learning_rate": 1.8824782288008038e-05, | |
| "loss": 2.1125, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.141728562719004, | |
| "learning_rate": 1.8796810755442675e-05, | |
| "loss": 2.1209, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.107105608845359, | |
| "learning_rate": 1.8768531536976452e-05, | |
| "loss": 2.111, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.1536543446807266, | |
| "learning_rate": 1.8739945621731114e-05, | |
| "loss": 2.1279, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.1243046575233615, | |
| "learning_rate": 1.8711054009555736e-05, | |
| "loss": 2.1161, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.1697848400011563, | |
| "learning_rate": 1.8681857710991745e-05, | |
| "loss": 2.1182, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.1431398557689867, | |
| "learning_rate": 1.865235774723758e-05, | |
| "loss": 2.1321, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.1954741616430697, | |
| "learning_rate": 1.862255515011298e-05, | |
| "loss": 2.129, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.1395616282832628, | |
| "learning_rate": 1.8592450962022872e-05, | |
| "loss": 2.1495, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.069366088459952, | |
| "learning_rate": 1.8562046235920938e-05, | |
| "loss": 2.0934, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.0953604901901421, | |
| "learning_rate": 1.8531342035272768e-05, | |
| "loss": 2.1186, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.141014164895258, | |
| "learning_rate": 1.850033943401867e-05, | |
| "loss": 2.1199, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.130487869190755, | |
| "learning_rate": 1.84690395165361e-05, | |
| "loss": 2.1311, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.1270531234995143, | |
| "learning_rate": 1.8437443377601736e-05, | |
| "loss": 2.1092, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.1011174255230582, | |
| "learning_rate": 1.8405552122353212e-05, | |
| "loss": 2.1274, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.105020692646057, | |
| "learning_rate": 1.8373366866250407e-05, | |
| "loss": 2.1296, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.090030313602243, | |
| "learning_rate": 1.8340888735036485e-05, | |
| "loss": 2.1193, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.1078722193307318, | |
| "learning_rate": 1.830811886469849e-05, | |
| "loss": 2.107, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.1358149656604917, | |
| "learning_rate": 1.8275058401427622e-05, | |
| "loss": 2.1213, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.1112469947883081, | |
| "learning_rate": 1.8241708501579146e-05, | |
| "loss": 2.1172, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.1188463095778778, | |
| "learning_rate": 1.8208070331631943e-05, | |
| "loss": 2.1182, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.1347354181246418, | |
| "learning_rate": 1.817414506814772e-05, | |
| "loss": 2.1293, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.1443165084836204, | |
| "learning_rate": 1.8139933897729833e-05, | |
| "loss": 2.1341, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.1590217288330331, | |
| "learning_rate": 1.8105438016981816e-05, | |
| "loss": 2.1089, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.1507315477006572, | |
| "learning_rate": 1.807065863246551e-05, | |
| "loss": 2.1168, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.1297637027488059, | |
| "learning_rate": 1.8035596960658856e-05, | |
| "loss": 2.1222, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.1030216110989035, | |
| "learning_rate": 1.8000254227913346e-05, | |
| "loss": 2.1093, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.0995954035038586, | |
| "learning_rate": 1.7964631670411154e-05, | |
| "loss": 2.1131, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.0780386854335375, | |
| "learning_rate": 1.7928730534121872e-05, | |
| "loss": 2.1167, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.0789206146363102, | |
| "learning_rate": 1.7892552074758932e-05, | |
| "loss": 2.0905, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.2118385014350053, | |
| "learning_rate": 1.7856097557735697e-05, | |
| "loss": 2.1171, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1058902525153855, | |
| "learning_rate": 1.7819368258121188e-05, | |
| "loss": 2.1106, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1479602850406405, | |
| "learning_rate": 1.77823654605955e-05, | |
| "loss": 2.122, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1276232191680498, | |
| "learning_rate": 1.7745090459404853e-05, | |
| "loss": 2.118, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.0918830389365917, | |
| "learning_rate": 1.7707544558316332e-05, | |
| "loss": 2.1093, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.1081061914662866, | |
| "learning_rate": 1.766972907057229e-05, | |
| "loss": 2.1052, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.0753316019333576, | |
| "learning_rate": 1.763164531884439e-05, | |
| "loss": 2.0933, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.139253596755427, | |
| "learning_rate": 1.7593294635187385e-05, | |
| "loss": 2.1454, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.1006108083060764, | |
| "learning_rate": 1.7554678360992475e-05, | |
| "loss": 2.091, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.1147375647006343, | |
| "learning_rate": 1.7515797846940435e-05, | |
| "loss": 2.1091, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.08902530557411, | |
| "learning_rate": 1.7476654452954354e-05, | |
| "loss": 2.1197, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.0637262113474155, | |
| "learning_rate": 1.7437249548152046e-05, | |
| "loss": 2.1228, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.091282896685236, | |
| "learning_rate": 1.7397584510798208e-05, | |
| "loss": 2.1101, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.0634146000550102, | |
| "learning_rate": 1.735766072825618e-05, | |
| "loss": 2.0997, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.086009489299668, | |
| "learning_rate": 1.7317479596939426e-05, | |
| "loss": 2.1059, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 1.091394126233442, | |
| "learning_rate": 1.72770425222627e-05, | |
| "loss": 2.1139, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.1511755690868908, | |
| "learning_rate": 1.723635091859287e-05, | |
| "loss": 2.1127, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0836319436950377, | |
| "learning_rate": 1.7195406209199473e-05, | |
| "loss": 2.1175, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0898105340267281, | |
| "learning_rate": 1.7154209826204916e-05, | |
| "loss": 2.1129, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.138471848511822, | |
| "learning_rate": 1.7112763210534395e-05, | |
| "loss": 2.0964, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.1206740818407968, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 2.111, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.131153459996651, | |
| "learning_rate": 1.7029125088577418e-05, | |
| "loss": 2.0983, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.0856246507797631, | |
| "learning_rate": 1.698693650770014e-05, | |
| "loss": 2.1011, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.1292122488308165, | |
| "learning_rate": 1.6944503544862924e-05, | |
| "loss": 2.1236, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.1346297426666019, | |
| "learning_rate": 1.690182768424279e-05, | |
| "loss": 2.0895, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.1034630210612935, | |
| "learning_rate": 1.6858910418512592e-05, | |
| "loss": 2.082, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.069627012568006, | |
| "learning_rate": 1.6815753248788814e-05, | |
| "loss": 2.1211, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.079271874383566, | |
| "learning_rate": 1.677235768457904e-05, | |
| "loss": 2.1109, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.1501866458790928, | |
| "learning_rate": 1.672872524372919e-05, | |
| "loss": 2.1135, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.111220910287794, | |
| "learning_rate": 1.668485745237041e-05, | |
| "loss": 2.1261, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.1067290913022252, | |
| "learning_rate": 1.6640755844865693e-05, | |
| "loss": 2.116, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.1023368782366267, | |
| "learning_rate": 1.6596421963756225e-05, | |
| "loss": 2.0944, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.0929201382514886, | |
| "learning_rate": 1.6551857359707407e-05, | |
| "loss": 2.0872, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.151997458775109, | |
| "learning_rate": 1.650706359145464e-05, | |
| "loss": 2.1191, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.0954291665406306, | |
| "learning_rate": 1.6462042225748803e-05, | |
| "loss": 2.1104, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.1137869292323443, | |
| "learning_rate": 1.641679483730144e-05, | |
| "loss": 2.1157, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.1209205183746023, | |
| "learning_rate": 1.637132300872969e-05, | |
| "loss": 2.1129, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.096510571575026, | |
| "learning_rate": 1.6325628330500928e-05, | |
| "loss": 2.0991, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.0812996975184992, | |
| "learning_rate": 1.627971240087715e-05, | |
| "loss": 2.1009, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.1185402689619315, | |
| "learning_rate": 1.6233576825859042e-05, | |
| "loss": 2.1182, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.078472600530305, | |
| "learning_rate": 1.6187223219129825e-05, | |
| "loss": 2.0959, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.0780740671282019, | |
| "learning_rate": 1.6140653201998833e-05, | |
| "loss": 2.1099, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.1153437142857043, | |
| "learning_rate": 1.609386840334475e-05, | |
| "loss": 2.1184, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.0870094438447029, | |
| "learning_rate": 1.60468704595587e-05, | |
| "loss": 2.1094, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.1081080494728075, | |
| "learning_rate": 1.5999661014486956e-05, | |
| "loss": 2.1015, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.1136424168229873, | |
| "learning_rate": 1.595224171937349e-05, | |
| "loss": 2.1401, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.1992802658998873, | |
| "learning_rate": 1.5904614232802173e-05, | |
| "loss": 2.1286, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.1068566272428728, | |
| "learning_rate": 1.58567802206388e-05, | |
| "loss": 2.0977, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.1133775714723446, | |
| "learning_rate": 1.5808741355972808e-05, | |
| "loss": 2.0996, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.0990574825146329, | |
| "learning_rate": 1.5760499319058754e-05, | |
| "loss": 2.0869, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.0789115119651662, | |
| "learning_rate": 1.5712055797257542e-05, | |
| "loss": 2.0829, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.1478272565935816, | |
| "learning_rate": 1.5663412484977417e-05, | |
| "loss": 2.1113, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.1191023412010976, | |
| "learning_rate": 1.5614571083614683e-05, | |
| "loss": 2.1147, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.1633940359115014, | |
| "learning_rate": 1.5565533301494217e-05, | |
| "loss": 2.1064, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.1004272946600961, | |
| "learning_rate": 1.5516300853809688e-05, | |
| "loss": 2.1297, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.1053014229400129, | |
| "learning_rate": 1.5466875462563592e-05, | |
| "loss": 2.0982, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.0538002603359922, | |
| "learning_rate": 1.5417258856506995e-05, | |
| "loss": 2.1049, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.1154282573123684, | |
| "learning_rate": 1.536745277107909e-05, | |
| "loss": 2.0871, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.1564226545846201, | |
| "learning_rate": 1.5317458948346482e-05, | |
| "loss": 2.1232, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.086490249464435, | |
| "learning_rate": 1.526727913694226e-05, | |
| "loss": 2.1161, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.0991487658843775, | |
| "learning_rate": 1.5216915092004847e-05, | |
| "loss": 2.0915, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.1260740272011343, | |
| "learning_rate": 1.516636857511658e-05, | |
| "loss": 2.1062, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.1437815072207562, | |
| "learning_rate": 1.5115641354242135e-05, | |
| "loss": 2.1031, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.116238517298388, | |
| "learning_rate": 1.5064735203666644e-05, | |
| "loss": 2.1021, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.0957743132196522, | |
| "learning_rate": 1.5013651903933686e-05, | |
| "loss": 2.1289, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.083118453369977, | |
| "learning_rate": 1.4962393241782966e-05, | |
| "loss": 2.1104, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.0916183452086836, | |
| "learning_rate": 1.491096101008784e-05, | |
| "loss": 2.1012, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.1226757963072318, | |
| "learning_rate": 1.485935700779261e-05, | |
| "loss": 2.0931, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.0779842189539774, | |
| "learning_rate": 1.4807583039849589e-05, | |
| "loss": 2.1156, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.055621607141943, | |
| "learning_rate": 1.4755640917155975e-05, | |
| "loss": 2.1107, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.1181305935952355, | |
| "learning_rate": 1.4703532456490518e-05, | |
| "loss": 2.1119, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.1393150890595864, | |
| "learning_rate": 1.465125948044996e-05, | |
| "loss": 2.1125, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.045965344032516, | |
| "learning_rate": 1.45988238173853e-05, | |
| "loss": 2.098, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.0879905688965106, | |
| "learning_rate": 1.4546227301337838e-05, | |
| "loss": 2.1113, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.0888667998390096, | |
| "learning_rate": 1.449347177197503e-05, | |
| "loss": 2.0972, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.0748269885773398, | |
| "learning_rate": 1.4440559074526126e-05, | |
| "loss": 2.1156, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.0871004289217108, | |
| "learning_rate": 1.4387491059717653e-05, | |
| "loss": 2.0987, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.0783349855333741, | |
| "learning_rate": 1.4334269583708664e-05, | |
| "loss": 2.0922, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.0841025974661314, | |
| "learning_rate": 1.4280896508025825e-05, | |
| "loss": 2.0778, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.0672100583853599, | |
| "learning_rate": 1.4227373699498305e-05, | |
| "loss": 2.1158, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.1057161227495689, | |
| "learning_rate": 1.4173703030192467e-05, | |
| "loss": 2.0995, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.0659900967881801, | |
| "learning_rate": 1.4119886377346402e-05, | |
| "loss": 2.1055, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.084877612752922, | |
| "learning_rate": 1.4065925623304267e-05, | |
| "loss": 2.1061, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.0796554409194477, | |
| "learning_rate": 1.4011822655450444e-05, | |
| "loss": 2.0776, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.0872310869838189, | |
| "learning_rate": 1.3957579366143521e-05, | |
| "loss": 2.0849, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.0553350688423493, | |
| "learning_rate": 1.3903197652650114e-05, | |
| "loss": 2.1038, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.148898225686392, | |
| "learning_rate": 1.3848679417078487e-05, | |
| "loss": 2.0989, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.0979778758372292, | |
| "learning_rate": 1.379402656631205e-05, | |
| "loss": 2.0957, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.0953035530756736, | |
| "learning_rate": 1.373924101194264e-05, | |
| "loss": 2.0935, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.0940463875072963, | |
| "learning_rate": 1.3684324670203658e-05, | |
| "loss": 2.1008, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.066991634591455, | |
| "learning_rate": 1.3629279461903064e-05, | |
| "loss": 2.109, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.0805973199508876, | |
| "learning_rate": 1.3574107312356174e-05, | |
| "loss": 2.0996, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.0473176890904303, | |
| "learning_rate": 1.351881015131833e-05, | |
| "loss": 2.0985, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.032511298259745, | |
| "learning_rate": 1.3463389912917401e-05, | |
| "loss": 2.0987, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.1216427751794493, | |
| "learning_rate": 1.3407848535586122e-05, | |
| "loss": 2.1124, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.0882088993618682, | |
| "learning_rate": 1.3352187961994309e-05, | |
| "loss": 2.0898, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.1053224310721825, | |
| "learning_rate": 1.3296410138980903e-05, | |
| "loss": 2.1081, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.0450067885627594, | |
| "learning_rate": 1.3240517017485872e-05, | |
| "loss": 2.0944, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.0899253739402426, | |
| "learning_rate": 1.3184510552481988e-05, | |
| "loss": 2.0985, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.0894862395136626, | |
| "learning_rate": 1.3128392702906418e-05, | |
| "loss": 2.0959, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.1068341669858077, | |
| "learning_rate": 1.307216543159225e-05, | |
| "loss": 2.11, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.0983702188982325, | |
| "learning_rate": 1.301583070519979e-05, | |
| "loss": 2.1102, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.0723770536726198, | |
| "learning_rate": 1.2959390494147814e-05, | |
| "loss": 2.1049, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.0944372562195301, | |
| "learning_rate": 1.2902846772544625e-05, | |
| "loss": 2.0977, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.0733243605919334, | |
| "learning_rate": 1.2846201518119019e-05, | |
| "loss": 2.0889, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.1061864768462715, | |
| "learning_rate": 1.2789456712151102e-05, | |
| "loss": 2.1192, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.098280857435917, | |
| "learning_rate": 1.273261433940299e-05, | |
| "loss": 2.0994, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.0645547922822787, | |
| "learning_rate": 1.2675676388049394e-05, | |
| "loss": 2.1013, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.0964522797866272, | |
| "learning_rate": 1.2618644849608068e-05, | |
| "loss": 2.1049, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.0810004605641046, | |
| "learning_rate": 1.2561521718870175e-05, | |
| "loss": 2.0941, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.0517685506211454, | |
| "learning_rate": 1.2504308993830488e-05, | |
| "loss": 2.0985, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.0586739276001895, | |
| "learning_rate": 1.2447008675617522e-05, | |
| "loss": 2.0915, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.0678975728007227, | |
| "learning_rate": 1.2389622768423537e-05, | |
| "loss": 2.108, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.0711219567905255, | |
| "learning_rate": 1.2332153279434438e-05, | |
| "loss": 2.0799, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.1065207631743585, | |
| "learning_rate": 1.2274602218759569e-05, | |
| "loss": 2.0824, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.2566524370061822, | |
| "learning_rate": 1.2216971599361402e-05, | |
| "loss": 2.1037, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.1011087188220579, | |
| "learning_rate": 1.2159263436985139e-05, | |
| "loss": 2.081, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.073927753354677, | |
| "learning_rate": 1.2101479750088197e-05, | |
| "loss": 2.0945, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.1180642368339424, | |
| "learning_rate": 1.2043622559769614e-05, | |
| "loss": 2.1009, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.0550378741908701, | |
| "learning_rate": 1.1985693889699359e-05, | |
| "loss": 2.0882, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.096018855144859, | |
| "learning_rate": 1.192769576604754e-05, | |
| "loss": 2.1055, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.0885957016186576, | |
| "learning_rate": 1.186963021741355e-05, | |
| "loss": 2.0898, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0969394787827396, | |
| "learning_rate": 1.1811499274755099e-05, | |
| "loss": 2.1031, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0511933672341713, | |
| "learning_rate": 1.1753304971317181e-05, | |
| "loss": 2.0885, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.0929336547851562, | |
| "eval_runtime": 85.3403, | |
| "eval_samples_per_second": 15.292, | |
| "eval_steps_per_second": 1.922, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.1101314242434477, | |
| "learning_rate": 1.1695049342560969e-05, | |
| "loss": 1.9882, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.146795174152589, | |
| "learning_rate": 1.1636734426092597e-05, | |
| "loss": 1.9718, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.1259132281295647, | |
| "learning_rate": 1.1578362261591916e-05, | |
| "loss": 1.9801, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.1749987325350009, | |
| "learning_rate": 1.1519934890741129e-05, | |
| "loss": 1.9758, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.1790386310837542, | |
| "learning_rate": 1.1461454357153407e-05, | |
| "loss": 1.9518, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.1797255451152497, | |
| "learning_rate": 1.1402922706301373e-05, | |
| "loss": 1.974, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.136660714984583, | |
| "learning_rate": 1.1344341985445594e-05, | |
| "loss": 1.9669, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.1169494634592092, | |
| "learning_rate": 1.1285714243562951e-05, | |
| "loss": 1.9532, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.1497963948136107, | |
| "learning_rate": 1.1227041531274978e-05, | |
| "loss": 1.9578, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.151486059445285, | |
| "learning_rate": 1.116832590077614e-05, | |
| "loss": 1.9782, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.149672171470685, | |
| "learning_rate": 1.1109569405762051e-05, | |
| "loss": 1.9914, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.1429059174539071, | |
| "learning_rate": 1.1050774101357638e-05, | |
| "loss": 1.9748, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.1790345851550625, | |
| "learning_rate": 1.0991942044045277e-05, | |
| "loss": 1.9658, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 1.1651391272301412, | |
| "learning_rate": 1.0933075291592832e-05, | |
| "loss": 1.962, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 1.1714870616571722, | |
| "learning_rate": 1.0874175902981711e-05, | |
| "loss": 1.9703, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 1.1632503164724157, | |
| "learning_rate": 1.0815245938334833e-05, | |
| "loss": 1.9628, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.1897086315120196, | |
| "learning_rate": 1.075628745884457e-05, | |
| "loss": 1.9725, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.1785590576959835, | |
| "learning_rate": 1.069730252670066e-05, | |
| "loss": 1.9417, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.1964045578322557, | |
| "learning_rate": 1.0638293205018084e-05, | |
| "loss": 1.9611, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.1815338539316877, | |
| "learning_rate": 1.0579261557764884e-05, | |
| "loss": 1.9682, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.1539070785331453, | |
| "learning_rate": 1.0520209649689979e-05, | |
| "loss": 1.9724, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.2002661304956936, | |
| "learning_rate": 1.0461139546250969e-05, | |
| "loss": 1.9992, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.2030184707417146, | |
| "learning_rate": 1.0402053313541854e-05, | |
| "loss": 1.976, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.174542114214098, | |
| "learning_rate": 1.03429530182208e-05, | |
| "loss": 1.9854, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.1626381884444785, | |
| "learning_rate": 1.0283840727437832e-05, | |
| "loss": 1.9577, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.2089610074412056, | |
| "learning_rate": 1.0224718508762549e-05, | |
| "loss": 1.9679, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.1587189836673264, | |
| "learning_rate": 1.0165588430111787e-05, | |
| "loss": 1.9593, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.1785455588782179, | |
| "learning_rate": 1.0106452559677309e-05, | |
| "loss": 1.9675, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.2098618060981559, | |
| "learning_rate": 1.0047312965853456e-05, | |
| "loss": 1.9609, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.1597512100566287, | |
| "learning_rate": 9.988171717164795e-06, | |
| "loss": 1.9881, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.2114808985923757, | |
| "learning_rate": 9.92903088219378e-06, | |
| "loss": 1.9669, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.1555571334890251, | |
| "learning_rate": 9.8698925295084e-06, | |
| "loss": 1.9601, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.200387042685194, | |
| "learning_rate": 9.810758727589814e-06, | |
| "loss": 1.9837, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.2024261695413923, | |
| "learning_rate": 9.751631544760004e-06, | |
| "loss": 1.9656, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.1570615707384821, | |
| "learning_rate": 9.69251304910945e-06, | |
| "loss": 1.9658, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.214720721568559, | |
| "learning_rate": 9.633405308424774e-06, | |
| "loss": 1.9712, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.1594367826830254, | |
| "learning_rate": 9.574310390116421e-06, | |
| "loss": 1.9656, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.209225990369828, | |
| "learning_rate": 9.51523036114634e-06, | |
| "loss": 1.9717, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.2523073282069046, | |
| "learning_rate": 9.45616728795571e-06, | |
| "loss": 1.9599, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.2102566135159636, | |
| "learning_rate": 9.397123236392647e-06, | |
| "loss": 1.9739, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 1.2216048036726168, | |
| "learning_rate": 9.338100271639932e-06, | |
| "loss": 1.97, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 1.166750451424913, | |
| "learning_rate": 9.279100458142815e-06, | |
| "loss": 1.9728, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 1.1608933836574207, | |
| "learning_rate": 9.220125859536766e-06, | |
| "loss": 1.9789, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.2157585537183477, | |
| "learning_rate": 9.161178538575321e-06, | |
| "loss": 1.973, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.1865919569086423, | |
| "learning_rate": 9.102260557057936e-06, | |
| "loss": 1.9654, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.2039575961205946, | |
| "learning_rate": 9.04337397575785e-06, | |
| "loss": 1.9733, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.2133138937608146, | |
| "learning_rate": 8.984520854350014e-06, | |
| "loss": 1.9646, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.185209016818463, | |
| "learning_rate": 8.925703251339061e-06, | |
| "loss": 1.9612, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.1439083359284015, | |
| "learning_rate": 8.866923223987303e-06, | |
| "loss": 1.9538, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.222566521665365, | |
| "learning_rate": 8.808182828242755e-06, | |
| "loss": 1.9648, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.247624939080409, | |
| "learning_rate": 8.749484118667242e-06, | |
| "loss": 1.9612, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.2530178888599242, | |
| "learning_rate": 8.690829148364544e-06, | |
| "loss": 1.9871, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.2108185860480414, | |
| "learning_rate": 8.632219968908556e-06, | |
| "loss": 1.9679, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.23598528854516, | |
| "learning_rate": 8.573658630271562e-06, | |
| "loss": 1.997, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.189859809726367, | |
| "learning_rate": 8.515147180752504e-06, | |
| "loss": 1.9536, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.1845678493113165, | |
| "learning_rate": 8.45668766690537e-06, | |
| "loss": 1.9598, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.2061467072734369, | |
| "learning_rate": 8.398282133467579e-06, | |
| "loss": 1.9426, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.1927014872549593, | |
| "learning_rate": 8.339932623288487e-06, | |
| "loss": 1.9656, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.227454731430014, | |
| "learning_rate": 8.281641177257929e-06, | |
| "loss": 1.98, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.1913450740502733, | |
| "learning_rate": 8.223409834234821e-06, | |
| "loss": 1.9727, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.2324504759141486, | |
| "learning_rate": 8.165240630975863e-06, | |
| "loss": 1.9663, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.2349713187916738, | |
| "learning_rate": 8.107135602064288e-06, | |
| "loss": 1.9683, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.1670802314687547, | |
| "learning_rate": 8.04909677983872e-06, | |
| "loss": 1.9705, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.1683607058295353, | |
| "learning_rate": 7.991126194322052e-06, | |
| "loss": 1.9877, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.201067059502362, | |
| "learning_rate": 7.93322587315047e-06, | |
| "loss": 1.9596, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 1.2339345747888217, | |
| "learning_rate": 7.875397841502536e-06, | |
| "loss": 1.9606, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.2187115438160399, | |
| "learning_rate": 7.81764412202833e-06, | |
| "loss": 1.9866, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.2034775846049928, | |
| "learning_rate": 7.75996673477872e-06, | |
| "loss": 1.9752, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.2936572681086704, | |
| "learning_rate": 7.702367697134701e-06, | |
| "loss": 1.9755, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.2201257056722639, | |
| "learning_rate": 7.644849023736842e-06, | |
| "loss": 1.9681, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.2209293093898896, | |
| "learning_rate": 7.5874127264148075e-06, | |
| "loss": 1.959, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.1645380938161174, | |
| "learning_rate": 7.530060814116992e-06, | |
| "loss": 1.981, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.1899566875269345, | |
| "learning_rate": 7.4727952928402695e-06, | |
| "loss": 1.9893, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.2154234991513635, | |
| "learning_rate": 7.415618165559799e-06, | |
| "loss": 1.959, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.216421504538409, | |
| "learning_rate": 7.358531432159002e-06, | |
| "loss": 1.9758, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.216796726948752, | |
| "learning_rate": 7.301537089359576e-06, | |
| "loss": 1.9497, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.2792638434318382, | |
| "learning_rate": 7.2446371306516935e-06, | |
| "loss": 1.9796, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 1.2091264739061314, | |
| "learning_rate": 7.187833546224245e-06, | |
| "loss": 1.9551, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.2392847383939514, | |
| "learning_rate": 7.1311283228952345e-06, | |
| "loss": 1.9491, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.1833236702350767, | |
| "learning_rate": 7.074523444042305e-06, | |
| "loss": 1.9838, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 1.2004605276986497, | |
| "learning_rate": 7.018020889533348e-06, | |
| "loss": 1.9829, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.205521789935667, | |
| "learning_rate": 6.961622635657251e-06, | |
| "loss": 1.9635, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.195881358789121, | |
| "learning_rate": 6.905330655054782e-06, | |
| "loss": 1.959, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.2074977318496216, | |
| "learning_rate": 6.8491469166496e-06, | |
| "loss": 1.9857, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.2161563440959358, | |
| "learning_rate": 6.793073385579373e-06, | |
| "loss": 1.9603, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.183673034033136, | |
| "learning_rate": 6.737112023127039e-06, | |
| "loss": 1.9701, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.228061251430918, | |
| "learning_rate": 6.681264786652236e-06, | |
| "loss": 1.9716, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.1709004416957736, | |
| "learning_rate": 6.6255336295228e-06, | |
| "loss": 1.9794, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.1751434527858315, | |
| "learning_rate": 6.569920501046474e-06, | |
| "loss": 1.9543, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.1690362024391214, | |
| "learning_rate": 6.514427346402712e-06, | |
| "loss": 1.9662, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 1.165267884553653, | |
| "learning_rate": 6.459056106574646e-06, | |
| "loss": 1.9849, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 1.205654513940418, | |
| "learning_rate": 6.403808718281188e-06, | |
| "loss": 1.9583, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 1.1948226708166962, | |
| "learning_rate": 6.348687113909304e-06, | |
| "loss": 1.9642, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.1986883734367584, | |
| "learning_rate": 6.293693221446424e-06, | |
| "loss": 1.987, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.2040471446109975, | |
| "learning_rate": 6.238828964412989e-06, | |
| "loss": 1.9825, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.2189778738931885, | |
| "learning_rate": 6.184096261795185e-06, | |
| "loss": 1.9714, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.2449157729698268, | |
| "learning_rate": 6.129497027977829e-06, | |
| "loss": 1.9663, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.208670388915054, | |
| "learning_rate": 6.075033172677398e-06, | |
| "loss": 1.9678, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.198125546182056, | |
| "learning_rate": 6.020706600875237e-06, | |
| "loss": 1.9483, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 1.173321161075678, | |
| "learning_rate": 5.966519212750923e-06, | |
| "loss": 1.9537, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 1.2174367958841394, | |
| "learning_rate": 5.912472903615821e-06, | |
| "loss": 1.9765, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 1.2142082544285866, | |
| "learning_rate": 5.858569563846764e-06, | |
| "loss": 1.9506, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.210910628942189, | |
| "learning_rate": 5.804811078819956e-06, | |
| "loss": 1.9706, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.182159914084399, | |
| "learning_rate": 5.75119932884502e-06, | |
| "loss": 1.9548, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.229433719183134, | |
| "learning_rate": 5.69773618909923e-06, | |
| "loss": 1.9575, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.172186632136802, | |
| "learning_rate": 5.644423529561911e-06, | |
| "loss": 1.9632, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.223273709243437, | |
| "learning_rate": 5.591263214949059e-06, | |
| "loss": 1.9616, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.2273846674644888, | |
| "learning_rate": 5.538257104648095e-06, | |
| "loss": 1.9755, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.1926932816399, | |
| "learning_rate": 5.4854070526528445e-06, | |
| "loss": 1.9615, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.2285290958363317, | |
| "learning_rate": 5.432714907498682e-06, | |
| "loss": 1.9821, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.2179688883081763, | |
| "learning_rate": 5.380182512197873e-06, | |
| "loss": 1.9667, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.258350454919681, | |
| "learning_rate": 5.327811704175123e-06, | |
| "loss": 1.9489, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 1.2225500456502807, | |
| "learning_rate": 5.2756043152032934e-06, | |
| "loss": 1.9742, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.2199895887278935, | |
| "learning_rate": 5.22356217133934e-06, | |
| "loss": 1.9723, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.1923102063541555, | |
| "learning_rate": 5.171687092860464e-06, | |
| "loss": 1.9664, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 1.2238483216379505, | |
| "learning_rate": 5.119980894200391e-06, | |
| "loss": 1.9656, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.2216032156031935, | |
| "learning_rate": 5.068445383885961e-06, | |
| "loss": 1.9682, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.1925371814493226, | |
| "learning_rate": 5.017082364473849e-06, | |
| "loss": 1.9729, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.2349996235847593, | |
| "learning_rate": 4.965893632487511e-06, | |
| "loss": 1.9739, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.2256513043025108, | |
| "learning_rate": 4.914880978354365e-06, | |
| "loss": 1.9703, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.1870736830630215, | |
| "learning_rate": 4.86404618634314e-06, | |
| "loss": 1.9419, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.189455313499183, | |
| "learning_rate": 4.813391034501505e-06, | |
| "loss": 1.9659, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.1922850605560722, | |
| "learning_rate": 4.762917294593851e-06, | |
| "loss": 1.9665, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.1790899419010261, | |
| "learning_rate": 4.712626732039311e-06, | |
| "loss": 1.9664, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.1761424320225644, | |
| "learning_rate": 4.662521105850056e-06, | |
| "loss": 1.9579, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.1939587817974457, | |
| "learning_rate": 4.6126021685697134e-06, | |
| "loss": 1.9507, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.2073541128422212, | |
| "learning_rate": 4.562871666212107e-06, | |
| "loss": 1.9657, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.2081527749730767, | |
| "learning_rate": 4.5133313382001865e-06, | |
| "loss": 1.9722, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.2184462455867358, | |
| "learning_rate": 4.463982917305155e-06, | |
| "loss": 1.9711, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.2158339715721698, | |
| "learning_rate": 4.414828129585901e-06, | |
| "loss": 1.953, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.177280144065104, | |
| "learning_rate": 4.365868694328599e-06, | |
| "loss": 1.9594, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.223747330887482, | |
| "learning_rate": 4.317106323986588e-06, | |
| "loss": 1.956, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.1858766619432295, | |
| "learning_rate": 4.2685427241204755e-06, | |
| "loss": 1.9596, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.197303530908462, | |
| "learning_rate": 4.220179593338459e-06, | |
| "loss": 1.9546, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.2299146675065473, | |
| "learning_rate": 4.1720186232369595e-06, | |
| "loss": 1.9584, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.240163069324589, | |
| "learning_rate": 4.124061498341403e-06, | |
| "loss": 1.9705, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.2396371274858464, | |
| "learning_rate": 4.076309896047337e-06, | |
| "loss": 1.9648, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.2012016328696529, | |
| "learning_rate": 4.028765486561749e-06, | |
| "loss": 1.9658, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.235953143109444, | |
| "learning_rate": 3.981429932844641e-06, | |
| "loss": 1.9793, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.19033811293556, | |
| "learning_rate": 3.934304890550878e-06, | |
| "loss": 1.9612, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.219886705989132, | |
| "learning_rate": 3.887392007972265e-06, | |
| "loss": 1.9629, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.2217119896905162, | |
| "learning_rate": 3.840692925979901e-06, | |
| "loss": 1.9598, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.2001629929744646, | |
| "learning_rate": 3.7942092779667904e-06, | |
| "loss": 1.9634, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.206387239403751, | |
| "learning_rate": 3.7479426897906946e-06, | |
| "loss": 1.9545, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.203411383413321, | |
| "learning_rate": 3.7018947797172864e-06, | |
| "loss": 1.9788, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.2082093372659748, | |
| "learning_rate": 3.6560671583635467e-06, | |
| "loss": 1.9667, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.1429855352087663, | |
| "learning_rate": 3.610461428641402e-06, | |
| "loss": 1.9801, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.2203395846743075, | |
| "learning_rate": 3.56507918570169e-06, | |
| "loss": 1.9624, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.2079078874905969, | |
| "learning_rate": 3.5199220168783564e-06, | |
| "loss": 1.9525, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 1.211711753662139, | |
| "learning_rate": 3.474991501632928e-06, | |
| "loss": 1.9706, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 1.2156495737616455, | |
| "learning_rate": 3.4302892114992814e-06, | |
| "loss": 1.9592, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 1.2054019636474047, | |
| "learning_rate": 3.385816710028651e-06, | |
| "loss": 1.9528, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.2196370662361549, | |
| "learning_rate": 3.341575552734978e-06, | |
| "loss": 1.9546, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.1981424496084816, | |
| "learning_rate": 3.2975672870404717e-06, | |
| "loss": 1.9719, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.1873573410362823, | |
| "learning_rate": 3.253793452221485e-06, | |
| "loss": 1.943, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.1817597577710646, | |
| "learning_rate": 3.210255579354713e-06, | |
| "loss": 1.9611, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.1976251823218045, | |
| "learning_rate": 3.1669551912635874e-06, | |
| "loss": 1.9293, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 1.2010547018207023, | |
| "learning_rate": 3.123893802465051e-06, | |
| "loss": 1.9744, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.2081281906690435, | |
| "learning_rate": 3.0810729191165722e-06, | |
| "loss": 1.9742, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.2004388915217468, | |
| "learning_rate": 3.0384940389634623e-06, | |
| "loss": 1.963, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.1815960741958818, | |
| "learning_rate": 2.9961586512864947e-06, | |
| "loss": 1.945, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.2253233973991853, | |
| "learning_rate": 2.9540682368497996e-06, | |
| "loss": 1.9723, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.2229213167923887, | |
| "learning_rate": 2.9122242678490975e-06, | |
| "loss": 1.9613, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.2080846010735247, | |
| "learning_rate": 2.8706282078601844e-06, | |
| "loss": 1.9697, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.2050808183555228, | |
| "learning_rate": 2.829281511787739e-06, | |
| "loss": 1.974, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.1966791127110614, | |
| "learning_rate": 2.7881856258144557e-06, | |
| "loss": 1.9488, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.2122528809626594, | |
| "learning_rate": 2.747341987350439e-06, | |
| "loss": 1.9672, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.1941495738129242, | |
| "learning_rate": 2.7067520249829425e-06, | |
| "loss": 1.955, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.1766932671802308, | |
| "learning_rate": 2.666417158426393e-06, | |
| "loss": 1.9633, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.2119445107062896, | |
| "learning_rate": 2.626338798472733e-06, | |
| "loss": 1.9678, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.196743410982704, | |
| "learning_rate": 2.586518346942082e-06, | |
| "loss": 1.9639, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.2025850507638536, | |
| "learning_rate": 2.546957196633697e-06, | |
| "loss": 1.9641, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.1921161816320012, | |
| "learning_rate": 2.5076567312772635e-06, | |
| "loss": 1.9723, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.1895102870427157, | |
| "learning_rate": 2.4686183254844976e-06, | |
| "loss": 1.9444, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.1972245662619143, | |
| "learning_rate": 2.4298433447010495e-06, | |
| "loss": 1.9585, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.25737022642945, | |
| "learning_rate": 2.3913331451587663e-06, | |
| "loss": 1.9668, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.2125664488413237, | |
| "learning_rate": 2.353089073828255e-06, | |
| "loss": 1.9663, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.2113319709812511, | |
| "learning_rate": 2.315112468371742e-06, | |
| "loss": 1.9405, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 1.1763917837443918, | |
| "learning_rate": 2.2774046570963147e-06, | |
| "loss": 1.9511, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 1.2175853109302095, | |
| "learning_rate": 2.2399669589074526e-06, | |
| "loss": 1.944, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 1.1869388674362804, | |
| "learning_rate": 2.202800683262888e-06, | |
| "loss": 1.9692, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 1.1710404114311668, | |
| "learning_rate": 2.165907130126821e-06, | |
| "loss": 1.9663, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 1.2220979463610013, | |
| "learning_rate": 2.1292875899244237e-06, | |
| "loss": 1.9787, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 1.1853896026325352, | |
| "learning_rate": 2.0929433434967472e-06, | |
| "loss": 1.9842, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.1809684031767618, | |
| "learning_rate": 2.056875662055874e-06, | |
| "loss": 1.943, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.2235497697137312, | |
| "learning_rate": 2.0210858071404883e-06, | |
| "loss": 1.9818, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.1991510683517859, | |
| "learning_rate": 1.985575030571746e-06, | |
| "loss": 1.9557, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 1.2133326198159218, | |
| "learning_rate": 1.950344574409471e-06, | |
| "loss": 1.9704, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 1.2226937532359674, | |
| "learning_rate": 1.915395670908734e-06, | |
| "loss": 1.9728, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 1.171909392967019, | |
| "learning_rate": 1.8807295424767424e-06, | |
| "loss": 1.9678, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.2065336174822567, | |
| "learning_rate": 1.8463474016300831e-06, | |
| "loss": 1.9641, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.1900688990418653, | |
| "learning_rate": 1.8122504509523154e-06, | |
| "loss": 1.9438, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.2117044669481427, | |
| "learning_rate": 1.7784398830519002e-06, | |
| "loss": 1.9687, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.2145575550121779, | |
| "learning_rate": 1.7449168805205041e-06, | |
| "loss": 1.9671, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.1583053162801586, | |
| "learning_rate": 1.7116826158916212e-06, | |
| "loss": 1.9741, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.2285636610045698, | |
| "learning_rate": 1.6787382515995565e-06, | |
| "loss": 1.9676, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.170026556939685, | |
| "learning_rate": 1.6460849399387845e-06, | |
| "loss": 1.968, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.2117305333864605, | |
| "learning_rate": 1.6137238230236352e-06, | |
| "loss": 1.9593, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.1912426781505197, | |
| "learning_rate": 1.5816560327483466e-06, | |
| "loss": 1.965, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 1.2241393308488153, | |
| "learning_rate": 1.549882690747475e-06, | |
| "loss": 1.965, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 1.2449286278537521, | |
| "learning_rate": 1.5184049083566688e-06, | |
| "loss": 1.9532, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 1.215075361187514, | |
| "learning_rate": 1.487223786573786e-06, | |
| "loss": 1.9483, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.2266617250923817, | |
| "learning_rate": 1.4563404160204008e-06, | |
| "loss": 1.9894, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.1796604610198975, | |
| "learning_rate": 1.4257558769036327e-06, | |
| "loss": 1.9502, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.1804223433145569, | |
| "learning_rate": 1.3954712389783998e-06, | |
| "loss": 1.9547, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.1933182406748242, | |
| "learning_rate": 1.3654875615099627e-06, | |
| "loss": 1.9769, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.178151100389304, | |
| "learning_rate": 1.3358058932369012e-06, | |
| "loss": 1.9714, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.224264951402397, | |
| "learning_rate": 1.3064272723344263e-06, | |
| "loss": 1.9671, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 1.617760635498734, | |
| "learning_rate": 1.2773527263780626e-06, | |
| "loss": 1.9851, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 1.246370986700957, | |
| "learning_rate": 1.248583272307712e-06, | |
| "loss": 1.9552, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 1.2213517682161756, | |
| "learning_rate": 1.2201199163920829e-06, | |
| "loss": 1.9519, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.174912927149578, | |
| "learning_rate": 1.191963654193493e-06, | |
| "loss": 1.9578, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.1971894827684488, | |
| "learning_rate": 1.1641154705330503e-06, | |
| "loss": 1.9563, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.2097678641776448, | |
| "learning_rate": 1.1365763394561968e-06, | |
| "loss": 1.9644, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.2219276384147493, | |
| "learning_rate": 1.1093472241986636e-06, | |
| "loss": 1.9494, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.2388673205475542, | |
| "learning_rate": 1.0824290771527458e-06, | |
| "loss": 1.9418, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.198911525869777, | |
| "learning_rate": 1.0558228398340188e-06, | |
| "loss": 1.9469, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 1.1864891366051398, | |
| "learning_rate": 1.0295294428483993e-06, | |
| "loss": 1.9781, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 1.221891735671, | |
| "learning_rate": 1.0035498058595838e-06, | |
| "loss": 1.9743, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 1.1671483732649974, | |
| "learning_rate": 9.778848375568972e-07, | |
| "loss": 1.9529, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.2201967318370808, | |
| "learning_rate": 9.525354356235006e-07, | |
| "loss": 1.9623, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.200619098782056, | |
| "learning_rate": 9.275024867049975e-07, | |
| "loss": 1.9645, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 1.1820963955186015, | |
| "learning_rate": 9.027868663784189e-07, | |
| "loss": 1.9554, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.1712410306613592, | |
| "learning_rate": 8.78389439121593e-07, | |
| "loss": 1.9873, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.1829036729003595, | |
| "learning_rate": 8.543110582829272e-07, | |
| "loss": 1.9641, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.1757107072465847, | |
| "learning_rate": 8.305525660515379e-07, | |
| "loss": 1.9518, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.191069329452327, | |
| "learning_rate": 8.071147934278045e-07, | |
| "loss": 1.9624, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.1598933470978174, | |
| "learning_rate": 7.839985601943034e-07, | |
| "loss": 1.9569, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.273967581044274, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 1.9645, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.1980715381259817, | |
| "learning_rate": 7.387339347676348e-07, | |
| "loss": 1.9452, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.2053075331410248, | |
| "learning_rate": 7.165871257945046e-07, | |
| "loss": 1.9648, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.2235580080377466, | |
| "learning_rate": 6.947650225963054e-07, | |
| "loss": 1.9623, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 1.247754572439595, | |
| "learning_rate": 6.732683884443736e-07, | |
| "loss": 1.983, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 1.1879678922220196, | |
| "learning_rate": 6.520979752261136e-07, | |
| "loss": 1.9631, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 1.191619967088088, | |
| "learning_rate": 6.312545234187106e-07, | |
| "loss": 1.9683, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.1979401454618652, | |
| "learning_rate": 6.107387620632299e-07, | |
| "loss": 1.9523, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.1878206203692598, | |
| "learning_rate": 5.905514087391051e-07, | |
| "loss": 1.9545, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.196733209060229, | |
| "learning_rate": 5.706931695390527e-07, | |
| "loss": 1.9754, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.2505354159099629, | |
| "learning_rate": 5.511647390443697e-07, | |
| "loss": 1.9888, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.228757992372372, | |
| "learning_rate": 5.319668003006373e-07, | |
| "loss": 1.9814, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.2516332685879819, | |
| "learning_rate": 5.131000247938367e-07, | |
| "loss": 1.9744, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.202006466653939, | |
| "learning_rate": 4.945650724268514e-07, | |
| "loss": 1.9684, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.2126601707093525, | |
| "learning_rate": 4.7636259149639983e-07, | |
| "loss": 1.9612, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 1.182582667447835, | |
| "learning_rate": 4.5849321867034944e-07, | |
| "loss": 1.9714, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 1.2236920761413304, | |
| "learning_rate": 4.4095757896544747e-07, | |
| "loss": 1.9753, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 1.1605426111405623, | |
| "learning_rate": 4.2375628572547287e-07, | |
| "loss": 1.9443, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.220871345737379, | |
| "learning_rate": 4.0688994059976175e-07, | |
| "loss": 1.9553, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.1964590943514957, | |
| "learning_rate": 3.903591335221857e-07, | |
| "loss": 1.9407, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.182268560053183, | |
| "learning_rate": 3.7416444269050335e-07, | |
| "loss": 1.9598, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.1697285321134465, | |
| "learning_rate": 3.5830643454614177e-07, | |
| "loss": 1.9542, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.2176539946158462, | |
| "learning_rate": 3.427856637543825e-07, | |
| "loss": 1.9726, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.1990481410686826, | |
| "learning_rate": 3.276026731849635e-07, | |
| "loss": 1.9568, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.1952585081608356, | |
| "learning_rate": 3.127579938930891e-07, | |
| "loss": 1.9358, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.2512418821041569, | |
| "learning_rate": 2.9825214510085575e-07, | |
| "loss": 1.966, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.172951820001413, | |
| "learning_rate": 2.8408563417908875e-07, | |
| "loss": 1.9358, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 1.2064447728611092, | |
| "learning_rate": 2.702589566296043e-07, | |
| "loss": 1.9455, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 1.1930180093439473, | |
| "learning_rate": 2.5677259606786686e-07, | |
| "loss": 1.9624, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 1.2365417188887344, | |
| "learning_rate": 2.436270242060812e-07, | |
| "loss": 1.9605, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.2170814732884094, | |
| "learning_rate": 2.3082270083669255e-07, | |
| "loss": 1.9697, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.195282860234351, | |
| "learning_rate": 2.1836007381630276e-07, | |
| "loss": 1.9622, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.2044986380826657, | |
| "learning_rate": 2.0623957905000603e-07, | |
| "loss": 1.9643, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.2036075432865145, | |
| "learning_rate": 1.9446164047614234e-07, | |
| "loss": 1.9678, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.188102229066059, | |
| "learning_rate": 1.8302667005146936e-07, | |
| "loss": 1.975, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.1885751008087175, | |
| "learning_rate": 1.7193506773675594e-07, | |
| "loss": 1.9607, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.1809108277033225, | |
| "learning_rate": 1.6118722148278586e-07, | |
| "loss": 1.9501, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.2042395457191926, | |
| "learning_rate": 1.5078350721679625e-07, | |
| "loss": 1.9372, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.1973298325941055, | |
| "learning_rate": 1.40724288829327e-07, | |
| "loss": 1.9681, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.1614975148633706, | |
| "learning_rate": 1.3100991816148657e-07, | |
| "loss": 1.9641, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.2049090434491252, | |
| "learning_rate": 1.2164073499265404e-07, | |
| "loss": 1.9683, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.1827811263674493, | |
| "learning_rate": 1.1261706702859198e-07, | |
| "loss": 1.9487, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.2119012714479263, | |
| "learning_rate": 1.0393922988997995e-07, | |
| "loss": 1.9717, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.1859531246908646, | |
| "learning_rate": 9.560752710138343e-08, | |
| "loss": 1.9787, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.2164147951415898, | |
| "learning_rate": 8.762225008062675e-08, | |
| "loss": 1.9657, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.2125607822440714, | |
| "learning_rate": 7.998367812861007e-08, | |
| "loss": 1.9614, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.2433209891346064, | |
| "learning_rate": 7.269207841953618e-08, | |
| "loss": 1.9559, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.2162270807881994, | |
| "learning_rate": 6.574770599156344e-08, | |
| "loss": 1.9562, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 1.206596012063816, | |
| "learning_rate": 5.915080373788962e-08, | |
| "loss": 1.9542, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 1.2090103241460801, | |
| "learning_rate": 5.290160239825315e-08, | |
| "loss": 1.9631, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 1.1776912988876318, | |
| "learning_rate": 4.700032055086179e-08, | |
| "loss": 1.9591, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.2080138102613605, | |
| "learning_rate": 4.1447164604753175e-08, | |
| "loss": 1.9722, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.18758236255762, | |
| "learning_rate": 3.6242328792567286e-08, | |
| "loss": 1.9401, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.1634742807000598, | |
| "learning_rate": 3.138599516375962e-08, | |
| "loss": 1.9507, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.165982746047221, | |
| "learning_rate": 2.6878333578227446e-08, | |
| "loss": 1.9654, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.1901335328038745, | |
| "learning_rate": 2.2719501700374514e-08, | |
| "loss": 1.9526, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.1878772105958921, | |
| "learning_rate": 1.890964499359327e-08, | |
| "loss": 1.9522, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.1911857730595692, | |
| "learning_rate": 1.5448896715177798e-08, | |
| "loss": 1.9653, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.2013895455868808, | |
| "learning_rate": 1.2337377911660897e-08, | |
| "loss": 1.9434, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.1933098943249791, | |
| "learning_rate": 9.575197414583015e-09, | |
| "loss": 1.9655, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.2205615809925663, | |
| "learning_rate": 7.162451836685291e-09, | |
| "loss": 1.9677, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.2135911892018199, | |
| "learning_rate": 5.0992255685289315e-09, | |
| "loss": 1.9629, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.1988764433311674, | |
| "learning_rate": 3.385590775544234e-09, | |
| "loss": 1.9525, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 1.1874798996817988, | |
| "learning_rate": 2.0216073955070527e-09, | |
| "loss": 1.9622, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 1.2006175349314985, | |
| "learning_rate": 1.007323136438254e-09, | |
| "loss": 1.9648, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 1.223178074142642, | |
| "learning_rate": 3.42773474943936e-10, | |
| "loss": 1.9608, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1671199103833645, | |
| "learning_rate": 2.7981654966424686e-11, | |
| "loss": 1.9596, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.086709499359131, | |
| "eval_runtime": 84.9602, | |
| "eval_samples_per_second": 15.36, | |
| "eval_steps_per_second": 1.93, | |
| "step": 2952 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 2952, | |
| "total_flos": 53473726955520.0, | |
| "train_loss": 2.0487296481119586, | |
| "train_runtime": 14639.4312, | |
| "train_samples_per_second": 3.227, | |
| "train_steps_per_second": 0.202 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2952, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "total_flos": 53473726955520.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |