{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 4087, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 4.999926141863722e-05, "loss": 1.0318, "step": 10 }, { "epoch": 0.0, "learning_rate": 4.999704571818908e-05, "loss": 0.8189, "step": 20 }, { "epoch": 0.01, "learning_rate": 4.999335302957357e-05, "loss": 0.6632, "step": 30 }, { "epoch": 0.01, "learning_rate": 4.9988183570978795e-05, "loss": 0.5815, "step": 40 }, { "epoch": 0.01, "learning_rate": 4.998153764784998e-05, "loss": 0.5221, "step": 50 }, { "epoch": 0.01, "learning_rate": 4.9973415652871555e-05, "loss": 0.4635, "step": 60 }, { "epoch": 0.02, "learning_rate": 4.9963818065943846e-05, "loss": 0.4266, "step": 70 }, { "epoch": 0.02, "learning_rate": 4.9952745454154736e-05, "loss": 0.4073, "step": 80 }, { "epoch": 0.02, "learning_rate": 4.9940198471746226e-05, "loss": 0.373, "step": 90 }, { "epoch": 0.02, "learning_rate": 4.9926177860075696e-05, "loss": 0.3172, "step": 100 }, { "epoch": 0.03, "learning_rate": 4.991068444757214e-05, "loss": 0.2985, "step": 110 }, { "epoch": 0.03, "learning_rate": 4.989371914968722e-05, "loss": 0.2965, "step": 120 }, { "epoch": 0.03, "learning_rate": 4.987528296884116e-05, "loss": 0.29, "step": 130 }, { "epoch": 0.03, "learning_rate": 4.985537699436353e-05, "loss": 0.2725, "step": 140 }, { "epoch": 0.04, "learning_rate": 4.983400240242887e-05, "loss": 0.236, "step": 150 }, { "epoch": 0.04, "learning_rate": 4.9811160455987196e-05, "loss": 0.2672, "step": 160 }, { "epoch": 0.04, "learning_rate": 4.978685250468939e-05, "loss": 0.2448, "step": 170 }, { "epoch": 0.04, "learning_rate": 4.9761079984807426e-05, "loss": 0.2553, "step": 180 }, { "epoch": 0.05, "learning_rate": 4.973384441914953e-05, "loss": 0.2133, "step": 190 }, { "epoch": 0.05, "learning_rate": 4.970514741697021e-05, "loss": 0.2351, "step": 200 }, { "epoch": 0.05, "learning_rate": 4.967499067387514e-05, "loss": 0.1805, "step": 210 }, { "epoch": 0.05, "learning_rate": 4.964337597172099e-05, "loss": 0.2253, "step": 220 }, { "epoch": 0.06, "learning_rate": 4.961030517851014e-05, "loss": 0.1853, "step": 230 }, { "epoch": 0.06, "learning_rate": 4.957578024828033e-05, "loss": 0.2271, "step": 240 }, { "epoch": 0.06, "learning_rate": 4.953980322098914e-05, "loss": 0.2112, "step": 250 }, { "epoch": 0.06, "learning_rate": 4.9502376222393525e-05, "loss": 0.22, "step": 260 }, { "epoch": 0.07, "learning_rate": 4.9463501463924175e-05, "loss": 0.1825, "step": 270 }, { "epoch": 0.07, "learning_rate": 4.942318124255487e-05, "loss": 0.242, "step": 280 }, { "epoch": 0.07, "learning_rate": 4.9381417940666715e-05, "loss": 0.2179, "step": 290 }, { "epoch": 0.07, "learning_rate": 4.9338214025907424e-05, "loss": 0.183, "step": 300 }, { "epoch": 0.08, "learning_rate": 4.9293572051045514e-05, "loss": 0.1886, "step": 310 }, { "epoch": 0.08, "learning_rate": 4.924749465381942e-05, "loss": 0.2062, "step": 320 }, { "epoch": 0.08, "learning_rate": 4.91999845567817e-05, "loss": 0.1647, "step": 330 }, { "epoch": 0.08, "learning_rate": 4.915104456713812e-05, "loss": 0.1828, "step": 340 }, { "epoch": 0.09, "learning_rate": 4.910067757658182e-05, "loss": 0.1781, "step": 350 }, { "epoch": 0.09, "learning_rate": 4.904888656112246e-05, "loss": 0.2016, "step": 360 }, { "epoch": 0.09, "learning_rate": 4.899567458091031e-05, "loss": 0.1761, "step": 370 }, { "epoch": 0.09, "learning_rate": 4.894104478005554e-05, "loss": 0.1729, "step": 380 }, { "epoch": 0.1, "learning_rate": 4.888500038644238e-05, "loss": 0.1894, "step": 390 }, { "epoch": 0.1, "learning_rate": 4.8827544711538386e-05, "loss": 0.2072, "step": 400 }, { "epoch": 0.1, "learning_rate": 4.876868115019881e-05, "loss": 0.1724, "step": 410 }, { "epoch": 0.1, "learning_rate": 4.8708413180466e-05, "loss": 0.2249, "step": 420 }, { "epoch": 0.11, "learning_rate": 4.8646744363363904e-05, "loss": 0.2007, "step": 430 }, { "epoch": 0.11, "learning_rate": 4.858367834268763e-05, "loss": 0.1877, "step": 440 }, { "epoch": 0.11, "learning_rate": 4.851921884478817e-05, "loss": 0.1827, "step": 450 }, { "epoch": 0.11, "learning_rate": 4.845336967835226e-05, "loss": 0.1819, "step": 460 }, { "epoch": 0.11, "learning_rate": 4.8386134734177225e-05, "loss": 0.1856, "step": 470 }, { "epoch": 0.12, "learning_rate": 4.831751798494123e-05, "loss": 0.1685, "step": 480 }, { "epoch": 0.12, "learning_rate": 4.824752348496844e-05, "loss": 0.1679, "step": 490 }, { "epoch": 0.12, "learning_rate": 4.8176155369989505e-05, "loss": 0.1788, "step": 500 }, { "epoch": 0.12, "learning_rate": 4.8103417856897206e-05, "loss": 0.1872, "step": 510 }, { "epoch": 0.13, "learning_rate": 4.8029315243497244e-05, "loss": 0.1697, "step": 520 }, { "epoch": 0.13, "learning_rate": 4.7953851908254376e-05, "loss": 0.162, "step": 530 }, { "epoch": 0.13, "learning_rate": 4.787703231003363e-05, "loss": 0.1718, "step": 540 }, { "epoch": 0.13, "learning_rate": 4.77988609878369e-05, "loss": 0.1866, "step": 550 }, { "epoch": 0.14, "learning_rate": 4.771934256053471e-05, "loss": 0.1609, "step": 560 }, { "epoch": 0.14, "learning_rate": 4.763848172659333e-05, "loss": 0.1629, "step": 570 }, { "epoch": 0.14, "learning_rate": 4.755628326379716e-05, "loss": 0.1947, "step": 580 }, { "epoch": 0.14, "learning_rate": 4.747275202896642e-05, "loss": 0.1813, "step": 590 }, { "epoch": 0.15, "learning_rate": 4.738789295767016e-05, "loss": 0.1672, "step": 600 }, { "epoch": 0.15, "learning_rate": 4.730171106393466e-05, "loss": 0.1597, "step": 610 }, { "epoch": 0.15, "learning_rate": 4.7214211439947174e-05, "loss": 0.1855, "step": 620 }, { "epoch": 0.15, "learning_rate": 4.712539925575502e-05, "loss": 0.1729, "step": 630 }, { "epoch": 0.16, "learning_rate": 4.7035279758960104e-05, "loss": 0.1771, "step": 640 }, { "epoch": 0.16, "learning_rate": 4.6943858274408915e-05, "loss": 0.1891, "step": 650 }, { "epoch": 0.16, "learning_rate": 4.68511402038778e-05, "loss": 0.1442, "step": 660 }, { "epoch": 0.16, "learning_rate": 4.6757131025753886e-05, "loss": 0.1666, "step": 670 }, { "epoch": 0.17, "learning_rate": 4.666183629471132e-05, "loss": 0.1672, "step": 680 }, { "epoch": 0.17, "learning_rate": 4.656526164138307e-05, "loss": 0.1525, "step": 690 }, { "epoch": 0.17, "learning_rate": 4.646741277202829e-05, "loss": 0.1758, "step": 700 }, { "epoch": 0.17, "learning_rate": 4.6368295468195064e-05, "loss": 0.1708, "step": 710 }, { "epoch": 0.18, "learning_rate": 4.626791558637887e-05, "loss": 0.1682, "step": 720 }, { "epoch": 0.18, "learning_rate": 4.6166279057676494e-05, "loss": 0.1515, "step": 730 }, { "epoch": 0.18, "learning_rate": 4.6063391887435605e-05, "loss": 0.1687, "step": 740 }, { "epoch": 0.18, "learning_rate": 4.595926015489991e-05, "loss": 0.1641, "step": 750 }, { "epoch": 0.19, "learning_rate": 4.585389001284999e-05, "loss": 0.1536, "step": 760 }, { "epoch": 0.19, "learning_rate": 4.574728768723965e-05, "loss": 0.1683, "step": 770 }, { "epoch": 0.19, "learning_rate": 4.563945947682821e-05, "loss": 0.1464, "step": 780 }, { "epoch": 0.19, "learning_rate": 4.5530411752808166e-05, "loss": 0.1764, "step": 790 }, { "epoch": 0.2, "learning_rate": 4.542015095842885e-05, "loss": 0.19, "step": 800 }, { "epoch": 0.2, "learning_rate": 4.5308683608615706e-05, "loss": 0.1819, "step": 810 }, { "epoch": 0.2, "learning_rate": 4.519601628958529e-05, "loss": 0.1512, "step": 820 }, { "epoch": 0.2, "learning_rate": 4.508215565845615e-05, "loss": 0.1689, "step": 830 }, { "epoch": 0.21, "learning_rate": 4.496710844285551e-05, "loss": 0.1739, "step": 840 }, { "epoch": 0.21, "learning_rate": 4.485088144052172e-05, "loss": 0.1535, "step": 850 }, { "epoch": 0.21, "learning_rate": 4.4733481518902584e-05, "loss": 0.142, "step": 860 }, { "epoch": 0.21, "learning_rate": 4.4614915614749645e-05, "loss": 0.1584, "step": 870 }, { "epoch": 0.22, "learning_rate": 4.449519073370827e-05, "loss": 0.1483, "step": 880 }, { "epoch": 0.22, "learning_rate": 4.43743139499037e-05, "loss": 0.1587, "step": 890 }, { "epoch": 0.22, "learning_rate": 4.425229240552313e-05, "loss": 0.1497, "step": 900 }, { "epoch": 0.22, "learning_rate": 4.4129133310393646e-05, "loss": 0.1368, "step": 910 }, { "epoch": 0.23, "learning_rate": 4.400484394155622e-05, "loss": 0.1565, "step": 920 }, { "epoch": 0.23, "learning_rate": 4.3879431642835784e-05, "loss": 0.1377, "step": 930 }, { "epoch": 0.23, "learning_rate": 4.3752903824407234e-05, "loss": 0.1755, "step": 940 }, { "epoch": 0.23, "learning_rate": 4.3625267962357676e-05, "loss": 0.165, "step": 950 }, { "epoch": 0.23, "learning_rate": 4.3496531598244616e-05, "loss": 0.1397, "step": 960 }, { "epoch": 0.24, "learning_rate": 4.33667023386504e-05, "loss": 0.1518, "step": 970 }, { "epoch": 0.24, "learning_rate": 4.323578785473272e-05, "loss": 0.1332, "step": 980 }, { "epoch": 0.24, "learning_rate": 4.3103795881771454e-05, "loss": 0.1541, "step": 990 }, { "epoch": 0.24, "learning_rate": 4.2970734218711474e-05, "loss": 0.1508, "step": 1000 }, { "epoch": 0.25, "learning_rate": 4.283661072770193e-05, "loss": 0.1465, "step": 1010 }, { "epoch": 0.25, "learning_rate": 4.270143333363171e-05, "loss": 0.1505, "step": 1020 }, { "epoch": 0.25, "learning_rate": 4.2565210023661104e-05, "loss": 0.1674, "step": 1030 }, { "epoch": 0.25, "learning_rate": 4.242794884674996e-05, "loss": 0.1352, "step": 1040 }, { "epoch": 0.26, "learning_rate": 4.2289657913182045e-05, "loss": 0.1667, "step": 1050 }, { "epoch": 0.26, "learning_rate": 4.2150345394085845e-05, "loss": 0.1495, "step": 1060 }, { "epoch": 0.26, "learning_rate": 4.201001952095178e-05, "loss": 0.1538, "step": 1070 }, { "epoch": 0.26, "learning_rate": 4.186868858514582e-05, "loss": 0.1425, "step": 1080 }, { "epoch": 0.27, "learning_rate": 4.172636093741959e-05, "loss": 0.1357, "step": 1090 }, { "epoch": 0.27, "learning_rate": 4.158304498741691e-05, "loss": 0.1493, "step": 1100 }, { "epoch": 0.27, "learning_rate": 4.143874920317696e-05, "loss": 0.1439, "step": 1110 }, { "epoch": 0.27, "learning_rate": 4.1293482110633905e-05, "loss": 0.1449, "step": 1120 }, { "epoch": 0.28, "learning_rate": 4.114725229311311e-05, "loss": 0.1451, "step": 1130 }, { "epoch": 0.28, "learning_rate": 4.100006839082402e-05, "loss": 0.1559, "step": 1140 }, { "epoch": 0.28, "learning_rate": 4.085193910034959e-05, "loss": 0.1452, "step": 1150 }, { "epoch": 0.28, "learning_rate": 4.0702873174132483e-05, "loss": 0.1584, "step": 1160 }, { "epoch": 0.29, "learning_rate": 4.0552879419957904e-05, "loss": 0.1341, "step": 1170 }, { "epoch": 0.29, "learning_rate": 4.0401966700433144e-05, "loss": 0.1433, "step": 1180 }, { "epoch": 0.29, "learning_rate": 4.025014393246399e-05, "loss": 0.1583, "step": 1190 }, { "epoch": 0.29, "learning_rate": 4.0097420086727766e-05, "loss": 0.1508, "step": 1200 }, { "epoch": 0.3, "learning_rate": 3.9943804187143384e-05, "loss": 0.1577, "step": 1210 }, { "epoch": 0.3, "learning_rate": 3.978930531033807e-05, "loss": 0.1431, "step": 1220 }, { "epoch": 0.3, "learning_rate": 3.963393258511109e-05, "loss": 0.161, "step": 1230 }, { "epoch": 0.3, "learning_rate": 3.9477695191894395e-05, "loss": 0.1396, "step": 1240 }, { "epoch": 0.31, "learning_rate": 3.9320602362210116e-05, "loss": 0.1477, "step": 1250 }, { "epoch": 0.31, "learning_rate": 3.9162663378125166e-05, "loss": 0.1461, "step": 1260 }, { "epoch": 0.31, "learning_rate": 3.900388757170274e-05, "loss": 0.152, "step": 1270 }, { "epoch": 0.31, "learning_rate": 3.8844284324450956e-05, "loss": 0.1563, "step": 1280 }, { "epoch": 0.32, "learning_rate": 3.8683863066768534e-05, "loss": 0.1441, "step": 1290 }, { "epoch": 0.32, "learning_rate": 3.8522633277387555e-05, "loss": 0.1448, "step": 1300 }, { "epoch": 0.32, "learning_rate": 3.836060448281342e-05, "loss": 0.1659, "step": 1310 }, { "epoch": 0.32, "learning_rate": 3.819778625676197e-05, "loss": 0.1464, "step": 1320 }, { "epoch": 0.33, "learning_rate": 3.803418821959379e-05, "loss": 0.147, "step": 1330 }, { "epoch": 0.33, "learning_rate": 3.7869820037745776e-05, "loss": 0.1479, "step": 1340 }, { "epoch": 0.33, "learning_rate": 3.770469142315997e-05, "loss": 0.1352, "step": 1350 }, { "epoch": 0.33, "learning_rate": 3.7538812132709775e-05, "loss": 0.1638, "step": 1360 }, { "epoch": 0.34, "learning_rate": 3.737219196762337e-05, "loss": 0.149, "step": 1370 }, { "epoch": 0.34, "learning_rate": 3.7204840772904645e-05, "loss": 0.155, "step": 1380 }, { "epoch": 0.34, "learning_rate": 3.703676843675147e-05, "loss": 0.159, "step": 1390 }, { "epoch": 0.34, "learning_rate": 3.686798488997146e-05, "loss": 0.1455, "step": 1400 }, { "epoch": 0.34, "learning_rate": 3.669850010539518e-05, "loss": 0.1439, "step": 1410 }, { "epoch": 0.35, "learning_rate": 3.652832409728686e-05, "loss": 0.1344, "step": 1420 }, { "epoch": 0.35, "learning_rate": 3.635746692075276e-05, "loss": 0.1519, "step": 1430 }, { "epoch": 0.35, "learning_rate": 3.6185938671146975e-05, "loss": 0.1469, "step": 1440 }, { "epoch": 0.35, "learning_rate": 3.601374948347498e-05, "loss": 0.1343, "step": 1450 }, { "epoch": 0.36, "learning_rate": 3.584090953179475e-05, "loss": 0.143, "step": 1460 }, { "epoch": 0.36, "learning_rate": 3.566742902861566e-05, "loss": 0.137, "step": 1470 }, { "epoch": 0.36, "learning_rate": 3.549331822429503e-05, "loss": 0.1485, "step": 1480 }, { "epoch": 0.36, "learning_rate": 3.531858740643247e-05, "loss": 0.1182, "step": 1490 }, { "epoch": 0.37, "learning_rate": 3.5143246899262014e-05, "loss": 0.1528, "step": 1500 }, { "epoch": 0.37, "learning_rate": 3.4967307063042134e-05, "loss": 0.1648, "step": 1510 }, { "epoch": 0.37, "learning_rate": 3.479077829344354e-05, "loss": 0.1213, "step": 1520 }, { "epoch": 0.37, "learning_rate": 3.461367102093498e-05, "loss": 0.1412, "step": 1530 }, { "epoch": 0.38, "learning_rate": 3.443599571016689e-05, "loss": 0.1397, "step": 1540 }, { "epoch": 0.38, "learning_rate": 3.425776285935315e-05, "loss": 0.1527, "step": 1550 }, { "epoch": 0.38, "learning_rate": 3.407898299965069e-05, "loss": 0.137, "step": 1560 }, { "epoch": 0.38, "learning_rate": 3.3899666694537307e-05, "loss": 0.121, "step": 1570 }, { "epoch": 0.39, "learning_rate": 3.3719824539187484e-05, "loss": 0.1401, "step": 1580 }, { "epoch": 0.39, "learning_rate": 3.353946715984635e-05, "loss": 0.1163, "step": 1590 }, { "epoch": 0.39, "learning_rate": 3.3358605213201835e-05, "loss": 0.1562, "step": 1600 }, { "epoch": 0.39, "learning_rate": 3.3177249385754964e-05, "loss": 0.1324, "step": 1610 }, { "epoch": 0.4, "learning_rate": 3.299541039318849e-05, "loss": 0.1586, "step": 1620 }, { "epoch": 0.4, "learning_rate": 3.281309897973368e-05, "loss": 0.1626, "step": 1630 }, { "epoch": 0.4, "learning_rate": 3.263032591753552e-05, "loss": 0.1366, "step": 1640 }, { "epoch": 0.4, "learning_rate": 3.2447102006016186e-05, "loss": 0.1336, "step": 1650 }, { "epoch": 0.41, "learning_rate": 3.226343807123697e-05, "loss": 0.1376, "step": 1660 }, { "epoch": 0.41, "learning_rate": 3.2079344965258645e-05, "loss": 0.132, "step": 1670 }, { "epoch": 0.41, "learning_rate": 3.189483356550015e-05, "loss": 0.1316, "step": 1680 }, { "epoch": 0.41, "learning_rate": 3.1709914774095975e-05, "loss": 0.1448, "step": 1690 }, { "epoch": 0.42, "learning_rate": 3.152459951725196e-05, "loss": 0.132, "step": 1700 }, { "epoch": 0.42, "learning_rate": 3.133889874459971e-05, "loss": 0.133, "step": 1710 }, { "epoch": 0.42, "learning_rate": 3.115282342854959e-05, "loss": 0.1558, "step": 1720 }, { "epoch": 0.42, "learning_rate": 3.096638456364245e-05, "loss": 0.152, "step": 1730 }, { "epoch": 0.43, "learning_rate": 3.0779593165899954e-05, "loss": 0.1293, "step": 1740 }, { "epoch": 0.43, "learning_rate": 3.059246027217372e-05, "loss": 0.1361, "step": 1750 }, { "epoch": 0.43, "learning_rate": 3.0404996939493156e-05, "loss": 0.1462, "step": 1760 }, { "epoch": 0.43, "learning_rate": 3.021721424441216e-05, "loss": 0.1671, "step": 1770 }, { "epoch": 0.44, "learning_rate": 3.0029123282354638e-05, "loss": 0.1269, "step": 1780 }, { "epoch": 0.44, "learning_rate": 2.984073516695892e-05, "loss": 0.1571, "step": 1790 }, { "epoch": 0.44, "learning_rate": 2.9652061029421087e-05, "loss": 0.139, "step": 1800 }, { "epoch": 0.44, "learning_rate": 2.9463112017837262e-05, "loss": 0.1234, "step": 1810 }, { "epoch": 0.45, "learning_rate": 2.927389929654494e-05, "loss": 0.1511, "step": 1820 }, { "epoch": 0.45, "learning_rate": 2.9084434045463255e-05, "loss": 0.1347, "step": 1830 }, { "epoch": 0.45, "learning_rate": 2.8894727459432504e-05, "loss": 0.1464, "step": 1840 }, { "epoch": 0.45, "learning_rate": 2.870479074755257e-05, "loss": 0.1446, "step": 1850 }, { "epoch": 0.46, "learning_rate": 2.8514635132520716e-05, "loss": 0.1371, "step": 1860 }, { "epoch": 0.46, "learning_rate": 2.8324271849968394e-05, "loss": 0.1345, "step": 1870 }, { "epoch": 0.46, "learning_rate": 2.8133712147797404e-05, "loss": 0.1396, "step": 1880 }, { "epoch": 0.46, "learning_rate": 2.794296728551533e-05, "loss": 0.1225, "step": 1890 }, { "epoch": 0.46, "learning_rate": 2.7752048533570186e-05, "loss": 0.1352, "step": 1900 }, { "epoch": 0.47, "learning_rate": 2.756096717268453e-05, "loss": 0.146, "step": 1910 }, { "epoch": 0.47, "learning_rate": 2.7369734493188924e-05, "loss": 0.1426, "step": 1920 }, { "epoch": 0.47, "learning_rate": 2.71783617943548e-05, "loss": 0.1494, "step": 1930 }, { "epoch": 0.47, "learning_rate": 2.6986860383726865e-05, "loss": 0.1098, "step": 1940 }, { "epoch": 0.48, "learning_rate": 2.679524157645494e-05, "loss": 0.1436, "step": 1950 }, { "epoch": 0.48, "learning_rate": 2.660351669462541e-05, "loss": 0.1269, "step": 1960 }, { "epoch": 0.48, "learning_rate": 2.6411697066592228e-05, "loss": 0.1283, "step": 1970 }, { "epoch": 0.48, "learning_rate": 2.62197940263076e-05, "loss": 0.1379, "step": 1980 }, { "epoch": 0.49, "learning_rate": 2.6027818912652218e-05, "loss": 0.135, "step": 1990 }, { "epoch": 0.49, "learning_rate": 2.5835783068765378e-05, "loss": 0.1365, "step": 2000 }, { "epoch": 0.49, "learning_rate": 2.564369784137472e-05, "loss": 0.1235, "step": 2010 }, { "epoch": 0.49, "learning_rate": 2.5451574580125738e-05, "loss": 0.1514, "step": 2020 }, { "epoch": 0.5, "learning_rate": 2.5259424636911256e-05, "loss": 0.1358, "step": 2030 }, { "epoch": 0.5, "learning_rate": 2.5067259365200624e-05, "loss": 0.1444, "step": 2040 }, { "epoch": 0.5, "learning_rate": 2.4875090119368905e-05, "loss": 0.1254, "step": 2050 }, { "epoch": 0.5, "learning_rate": 2.468292825402597e-05, "loss": 0.1247, "step": 2060 }, { "epoch": 0.51, "learning_rate": 2.449078512334562e-05, "loss": 0.1293, "step": 2070 }, { "epoch": 0.51, "learning_rate": 2.429867208039467e-05, "loss": 0.126, "step": 2080 }, { "epoch": 0.51, "learning_rate": 2.4106600476462173e-05, "loss": 0.1351, "step": 2090 }, { "epoch": 0.51, "learning_rate": 2.3914581660388683e-05, "loss": 0.1386, "step": 2100 }, { "epoch": 0.52, "learning_rate": 2.3722626977895706e-05, "loss": 0.1294, "step": 2110 }, { "epoch": 0.52, "learning_rate": 2.3530747770915328e-05, "loss": 0.1188, "step": 2120 }, { "epoch": 0.52, "learning_rate": 2.3338955376920034e-05, "loss": 0.1283, "step": 2130 }, { "epoch": 0.52, "learning_rate": 2.3147261128252858e-05, "loss": 0.1239, "step": 2140 }, { "epoch": 0.53, "learning_rate": 2.295567635145774e-05, "loss": 0.1258, "step": 2150 }, { "epoch": 0.53, "learning_rate": 2.276421236661032e-05, "loss": 0.145, "step": 2160 }, { "epoch": 0.53, "learning_rate": 2.2572880486649086e-05, "loss": 0.1232, "step": 2170 }, { "epoch": 0.53, "learning_rate": 2.2381692016706862e-05, "loss": 0.14, "step": 2180 }, { "epoch": 0.54, "learning_rate": 2.219065825344292e-05, "loss": 0.1594, "step": 2190 }, { "epoch": 0.54, "learning_rate": 2.1999790484375433e-05, "loss": 0.1475, "step": 2200 }, { "epoch": 0.54, "learning_rate": 2.1809099987214567e-05, "loss": 0.1176, "step": 2210 }, { "epoch": 0.54, "learning_rate": 2.1618598029196082e-05, "loss": 0.1317, "step": 2220 }, { "epoch": 0.55, "learning_rate": 2.142829586641566e-05, "loss": 0.1318, "step": 2230 }, { "epoch": 0.55, "learning_rate": 2.1238204743163764e-05, "loss": 0.1453, "step": 2240 }, { "epoch": 0.55, "learning_rate": 2.104833589126124e-05, "loss": 0.1297, "step": 2250 }, { "epoch": 0.55, "learning_rate": 2.085870052939572e-05, "loss": 0.1393, "step": 2260 }, { "epoch": 0.56, "learning_rate": 2.0669309862458756e-05, "loss": 0.1364, "step": 2270 }, { "epoch": 0.56, "learning_rate": 2.0480175080883663e-05, "loss": 0.1246, "step": 2280 }, { "epoch": 0.56, "learning_rate": 2.0291307359984435e-05, "loss": 0.1437, "step": 2290 }, { "epoch": 0.56, "learning_rate": 2.0102717859295365e-05, "loss": 0.1268, "step": 2300 }, { "epoch": 0.57, "learning_rate": 1.9914417721911687e-05, "loss": 0.1299, "step": 2310 }, { "epoch": 0.57, "learning_rate": 1.9726418073831167e-05, "loss": 0.1334, "step": 2320 }, { "epoch": 0.57, "learning_rate": 1.9538730023296708e-05, "loss": 0.138, "step": 2330 }, { "epoch": 0.57, "learning_rate": 1.9351364660140004e-05, "loss": 0.1247, "step": 2340 }, { "epoch": 0.57, "learning_rate": 1.9164333055126265e-05, "loss": 0.1218, "step": 2350 }, { "epoch": 0.58, "learning_rate": 1.897764625930012e-05, "loss": 0.1391, "step": 2360 }, { "epoch": 0.58, "learning_rate": 1.879131530333261e-05, "loss": 0.1405, "step": 2370 }, { "epoch": 0.58, "learning_rate": 1.8605351196869453e-05, "loss": 0.1454, "step": 2380 }, { "epoch": 0.58, "learning_rate": 1.8419764927880488e-05, "loss": 0.1404, "step": 2390 }, { "epoch": 0.59, "learning_rate": 1.8234567462010482e-05, "loss": 0.1277, "step": 2400 }, { "epoch": 0.59, "learning_rate": 1.8049769741931178e-05, "loss": 0.1298, "step": 2410 }, { "epoch": 0.59, "learning_rate": 1.786538268669472e-05, "loss": 0.1378, "step": 2420 }, { "epoch": 0.59, "learning_rate": 1.768141719108852e-05, "loss": 0.1336, "step": 2430 }, { "epoch": 0.6, "learning_rate": 1.749788412499149e-05, "loss": 0.131, "step": 2440 }, { "epoch": 0.6, "learning_rate": 1.7314794332731786e-05, "loss": 0.1187, "step": 2450 }, { "epoch": 0.6, "learning_rate": 1.7132158632446085e-05, "loss": 0.1291, "step": 2460 }, { "epoch": 0.6, "learning_rate": 1.6949987815440332e-05, "loss": 0.1277, "step": 2470 }, { "epoch": 0.61, "learning_rate": 1.676829264555216e-05, "loss": 0.1419, "step": 2480 }, { "epoch": 0.61, "learning_rate": 1.658708385851485e-05, "loss": 0.128, "step": 2490 }, { "epoch": 0.61, "learning_rate": 1.6406372161323035e-05, "loss": 0.1213, "step": 2500 }, { "epoch": 0.61, "learning_rate": 1.6226168231600058e-05, "loss": 0.1436, "step": 2510 }, { "epoch": 0.62, "learning_rate": 1.604648271696701e-05, "loss": 0.1297, "step": 2520 }, { "epoch": 0.62, "learning_rate": 1.5867326234413692e-05, "loss": 0.1481, "step": 2530 }, { "epoch": 0.62, "learning_rate": 1.5688709369671222e-05, "loss": 0.1237, "step": 2540 }, { "epoch": 0.62, "learning_rate": 1.5510642676586606e-05, "loss": 0.1342, "step": 2550 }, { "epoch": 0.63, "learning_rate": 1.5333136676499078e-05, "loss": 0.1265, "step": 2560 }, { "epoch": 0.63, "learning_rate": 1.5156201857618532e-05, "loss": 0.1294, "step": 2570 }, { "epoch": 0.63, "learning_rate": 1.4979848674405755e-05, "loss": 0.1218, "step": 2580 }, { "epoch": 0.63, "learning_rate": 1.480408754695467e-05, "loss": 0.1377, "step": 2590 }, { "epoch": 0.64, "learning_rate": 1.4628928860376733e-05, "loss": 0.1367, "step": 2600 }, { "epoch": 0.64, "learning_rate": 1.4454382964187263e-05, "loss": 0.1338, "step": 2610 }, { "epoch": 0.64, "learning_rate": 1.4280460171693924e-05, "loss": 0.1357, "step": 2620 }, { "epoch": 0.64, "learning_rate": 1.4107170759387364e-05, "loss": 0.1308, "step": 2630 }, { "epoch": 0.65, "learning_rate": 1.3934524966334012e-05, "loss": 0.1466, "step": 2640 }, { "epoch": 0.65, "learning_rate": 1.3762532993571079e-05, "loss": 0.1371, "step": 2650 }, { "epoch": 0.65, "learning_rate": 1.35912050035038e-05, "loss": 0.1357, "step": 2660 }, { "epoch": 0.65, "learning_rate": 1.342055111930502e-05, "loss": 0.1208, "step": 2670 }, { "epoch": 0.66, "learning_rate": 1.325058142431701e-05, "loss": 0.1168, "step": 2680 }, { "epoch": 0.66, "learning_rate": 1.3081305961455658e-05, "loss": 0.121, "step": 2690 }, { "epoch": 0.66, "learning_rate": 1.291273473261716e-05, "loss": 0.1215, "step": 2700 }, { "epoch": 0.66, "learning_rate": 1.2744877698086943e-05, "loss": 0.1302, "step": 2710 }, { "epoch": 0.67, "learning_rate": 1.2577744775951194e-05, "loss": 0.1381, "step": 2720 }, { "epoch": 0.67, "learning_rate": 1.2411345841510808e-05, "loss": 0.1398, "step": 2730 }, { "epoch": 0.67, "learning_rate": 1.2245690726697939e-05, "loss": 0.1279, "step": 2740 }, { "epoch": 0.67, "learning_rate": 1.2080789219495028e-05, "loss": 0.1419, "step": 2750 }, { "epoch": 0.68, "learning_rate": 1.1916651063356455e-05, "loss": 0.1466, "step": 2760 }, { "epoch": 0.68, "learning_rate": 1.1753285956632867e-05, "loss": 0.1237, "step": 2770 }, { "epoch": 0.68, "learning_rate": 1.1590703551998111e-05, "loss": 0.1248, "step": 2780 }, { "epoch": 0.68, "learning_rate": 1.1428913455878932e-05, "loss": 0.1256, "step": 2790 }, { "epoch": 0.69, "learning_rate": 1.1267925227887275e-05, "loss": 0.1272, "step": 2800 }, { "epoch": 0.69, "learning_rate": 1.1107748380255536e-05, "loss": 0.1325, "step": 2810 }, { "epoch": 0.69, "learning_rate": 1.0948392377274477e-05, "loss": 0.1099, "step": 2820 }, { "epoch": 0.69, "learning_rate": 1.0789866634733988e-05, "loss": 0.1418, "step": 2830 }, { "epoch": 0.69, "learning_rate": 1.0632180519366796e-05, "loss": 0.1202, "step": 2840 }, { "epoch": 0.7, "learning_rate": 1.0475343348294978e-05, "loss": 0.1304, "step": 2850 }, { "epoch": 0.7, "learning_rate": 1.0319364388479455e-05, "loss": 0.15, "step": 2860 }, { "epoch": 0.7, "learning_rate": 1.0164252856172445e-05, "loss": 0.1306, "step": 2870 }, { "epoch": 0.7, "learning_rate": 1.0010017916372908e-05, "loss": 0.1394, "step": 2880 }, { "epoch": 0.71, "learning_rate": 9.856668682285003e-06, "loss": 0.1545, "step": 2890 }, { "epoch": 0.71, "learning_rate": 9.704214214779625e-06, "loss": 0.1261, "step": 2900 }, { "epoch": 0.71, "learning_rate": 9.552663521859048e-06, "loss": 0.1222, "step": 2910 }, { "epoch": 0.71, "learning_rate": 9.402025558124658e-06, "loss": 0.1253, "step": 2920 }, { "epoch": 0.72, "learning_rate": 9.25230922424786e-06, "loss": 0.1418, "step": 2930 }, { "epoch": 0.72, "learning_rate": 9.103523366444164e-06, "loss": 0.1211, "step": 2940 }, { "epoch": 0.72, "learning_rate": 8.9556767759505e-06, "loss": 0.1251, "step": 2950 }, { "epoch": 0.72, "learning_rate": 8.808778188505779e-06, "loss": 0.1331, "step": 2960 }, { "epoch": 0.73, "learning_rate": 8.6628362838347e-06, "loss": 0.15, "step": 2970 }, { "epoch": 0.73, "learning_rate": 8.517859685134932e-06, "loss": 0.1152, "step": 2980 }, { "epoch": 0.73, "learning_rate": 8.373856958567589e-06, "loss": 0.1244, "step": 2990 }, { "epoch": 0.73, "learning_rate": 8.230836612751069e-06, "loss": 0.1491, "step": 3000 }, { "epoch": 0.74, "learning_rate": 8.088807098258328e-06, "loss": 0.1387, "step": 3010 }, { "epoch": 0.74, "learning_rate": 7.94777680711756e-06, "loss": 0.1497, "step": 3020 }, { "epoch": 0.74, "learning_rate": 7.807754072316323e-06, "loss": 0.1296, "step": 3030 }, { "epoch": 0.74, "learning_rate": 7.668747167309209e-06, "loss": 0.1278, "step": 3040 }, { "epoch": 0.75, "learning_rate": 7.530764305528959e-06, "loss": 0.1334, "step": 3050 }, { "epoch": 0.75, "learning_rate": 7.393813639901195e-06, "loss": 0.1314, "step": 3060 }, { "epoch": 0.75, "learning_rate": 7.257903262362637e-06, "loss": 0.1251, "step": 3070 }, { "epoch": 0.75, "learning_rate": 7.123041203383032e-06, "loss": 0.1127, "step": 3080 }, { "epoch": 0.76, "learning_rate": 6.989235431490676e-06, "loss": 0.1363, "step": 3090 }, { "epoch": 0.76, "learning_rate": 6.856493852801485e-06, "loss": 0.1184, "step": 3100 }, { "epoch": 0.76, "learning_rate": 6.724824310551961e-06, "loss": 0.1475, "step": 3110 }, { "epoch": 0.76, "learning_rate": 6.594234584635694e-06, "loss": 0.1336, "step": 3120 }, { "epoch": 0.77, "learning_rate": 6.464732391143713e-06, "loss": 0.127, "step": 3130 }, { "epoch": 0.77, "learning_rate": 6.336325381908529e-06, "loss": 0.1315, "step": 3140 }, { "epoch": 0.77, "learning_rate": 6.209021144052054e-06, "loss": 0.1257, "step": 3150 }, { "epoch": 0.77, "learning_rate": 6.082827199537295e-06, "loss": 0.1356, "step": 3160 }, { "epoch": 0.78, "learning_rate": 5.9577510047238825e-06, "loss": 0.1306, "step": 3170 }, { "epoch": 0.78, "learning_rate": 5.833799949927537e-06, "loss": 0.1163, "step": 3180 }, { "epoch": 0.78, "learning_rate": 5.710981358983384e-06, "loss": 0.1333, "step": 3190 }, { "epoch": 0.78, "learning_rate": 5.589302488813186e-06, "loss": 0.139, "step": 3200 }, { "epoch": 0.79, "learning_rate": 5.468770528996614e-06, "loss": 0.1261, "step": 3210 }, { "epoch": 0.79, "learning_rate": 5.349392601346398e-06, "loss": 0.1274, "step": 3220 }, { "epoch": 0.79, "learning_rate": 5.231175759487536e-06, "loss": 0.1136, "step": 3230 }, { "epoch": 0.79, "learning_rate": 5.114126988440523e-06, "loss": 0.1289, "step": 3240 }, { "epoch": 0.8, "learning_rate": 4.998253204208625e-06, "loss": 0.1462, "step": 3250 }, { "epoch": 0.8, "learning_rate": 4.8835612533692455e-06, "loss": 0.1209, "step": 3260 }, { "epoch": 0.8, "learning_rate": 4.770057912669362e-06, "loss": 0.1266, "step": 3270 }, { "epoch": 0.8, "learning_rate": 4.65774988862514e-06, "loss": 0.1487, "step": 3280 }, { "epoch": 0.8, "learning_rate": 4.546643817125662e-06, "loss": 0.1101, "step": 3290 }, { "epoch": 0.81, "learning_rate": 4.436746263040819e-06, "loss": 0.1183, "step": 3300 }, { "epoch": 0.81, "learning_rate": 4.328063719833433e-06, "loss": 0.1133, "step": 3310 }, { "epoch": 0.81, "learning_rate": 4.220602609175575e-06, "loss": 0.1159, "step": 3320 }, { "epoch": 0.81, "learning_rate": 4.114369280569136e-06, "loss": 0.1288, "step": 3330 }, { "epoch": 0.82, "learning_rate": 4.009370010970634e-06, "loss": 0.1222, "step": 3340 }, { "epoch": 0.82, "learning_rate": 3.90561100442036e-06, "loss": 0.1208, "step": 3350 }, { "epoch": 0.82, "learning_rate": 3.8030983916757966e-06, "loss": 0.1096, "step": 3360 }, { "epoch": 0.82, "learning_rate": 3.7018382298493635e-06, "loss": 0.1157, "step": 3370 }, { "epoch": 0.83, "learning_rate": 3.6018365020505124e-06, "loss": 0.1264, "step": 3380 }, { "epoch": 0.83, "learning_rate": 3.503099117032249e-06, "loss": 0.1315, "step": 3390 }, { "epoch": 0.83, "learning_rate": 3.405631908841966e-06, "loss": 0.1317, "step": 3400 }, { "epoch": 0.83, "learning_rate": 3.3094406364767267e-06, "loss": 0.1382, "step": 3410 }, { "epoch": 0.84, "learning_rate": 3.214530983543021e-06, "loss": 0.1366, "step": 3420 }, { "epoch": 0.84, "learning_rate": 3.120908557920918e-06, "loss": 0.1308, "step": 3430 }, { "epoch": 0.84, "learning_rate": 3.028578891432701e-06, "loss": 0.1448, "step": 3440 }, { "epoch": 0.84, "learning_rate": 2.9375474395160528e-06, "loss": 0.1292, "step": 3450 }, { "epoch": 0.85, "learning_rate": 2.8478195809016674e-06, "loss": 0.13, "step": 3460 }, { "epoch": 0.85, "learning_rate": 2.7594006172954906e-06, "loss": 0.1333, "step": 3470 }, { "epoch": 0.85, "learning_rate": 2.6722957730653973e-06, "loss": 0.1527, "step": 3480 }, { "epoch": 0.85, "learning_rate": 2.5865101949325567e-06, "loss": 0.1154, "step": 3490 }, { "epoch": 0.86, "learning_rate": 2.50204895166731e-06, "loss": 0.1247, "step": 3500 }, { "epoch": 0.86, "learning_rate": 2.418917033789661e-06, "loss": 0.1211, "step": 3510 }, { "epoch": 0.86, "learning_rate": 2.3371193532744305e-06, "loss": 0.1208, "step": 3520 }, { "epoch": 0.86, "learning_rate": 2.2566607432610047e-06, "loss": 0.1208, "step": 3530 }, { "epoch": 0.87, "learning_rate": 2.177545957767771e-06, "loss": 0.1251, "step": 3540 }, { "epoch": 0.87, "learning_rate": 2.0997796714112144e-06, "loss": 0.1175, "step": 3550 }, { "epoch": 0.87, "learning_rate": 2.023366479129718e-06, "loss": 0.1207, "step": 3560 }, { "epoch": 0.87, "learning_rate": 1.9483108959120587e-06, "loss": 0.1198, "step": 3570 }, { "epoch": 0.88, "learning_rate": 1.874617356530628e-06, "loss": 0.113, "step": 3580 }, { "epoch": 0.88, "learning_rate": 1.8022902152794036e-06, "loss": 0.1261, "step": 3590 }, { "epoch": 0.88, "learning_rate": 1.7313337457166707e-06, "loss": 0.1297, "step": 3600 }, { "epoch": 0.88, "learning_rate": 1.661752140412512e-06, "loss": 0.1197, "step": 3610 }, { "epoch": 0.89, "learning_rate": 1.5935495107010728e-06, "loss": 0.133, "step": 3620 }, { "epoch": 0.89, "learning_rate": 1.5267298864376527e-06, "loss": 0.1245, "step": 3630 }, { "epoch": 0.89, "learning_rate": 1.461297215760582e-06, "loss": 0.1337, "step": 3640 }, { "epoch": 0.89, "learning_rate": 1.3972553648579468e-06, "loss": 0.1437, "step": 3650 }, { "epoch": 0.9, "learning_rate": 1.3346081177391472e-06, "loss": 0.1467, "step": 3660 }, { "epoch": 0.9, "learning_rate": 1.2733591760113184e-06, "loss": 0.1298, "step": 3670 }, { "epoch": 0.9, "learning_rate": 1.2135121586606014e-06, "loss": 0.1239, "step": 3680 }, { "epoch": 0.9, "learning_rate": 1.1550706018383345e-06, "loss": 0.1252, "step": 3690 }, { "epoch": 0.91, "learning_rate": 1.09803795865209e-06, "loss": 0.1475, "step": 3700 }, { "epoch": 0.91, "learning_rate": 1.0424175989616563e-06, "loss": 0.1365, "step": 3710 }, { "epoch": 0.91, "learning_rate": 9.882128091799114e-07, "loss": 0.127, "step": 3720 }, { "epoch": 0.91, "learning_rate": 9.354267920786614e-07, "loss": 0.1281, "step": 3730 }, { "epoch": 0.92, "learning_rate": 8.840626665993807e-07, "loss": 0.1219, "step": 3740 }, { "epoch": 0.92, "learning_rate": 8.341234676689319e-07, "loss": 0.1227, "step": 3750 }, { "epoch": 0.92, "learning_rate": 7.856121460202398e-07, "loss": 0.1568, "step": 3760 }, { "epoch": 0.92, "learning_rate": 7.385315680179583e-07, "loss": 0.1202, "step": 3770 }, { "epoch": 0.92, "learning_rate": 6.928845154890757e-07, "loss": 0.1374, "step": 3780 }, { "epoch": 0.93, "learning_rate": 6.486736855585762e-07, "loss": 0.1235, "step": 3790 }, { "epoch": 0.93, "learning_rate": 6.059016904900628e-07, "loss": 0.1252, "step": 3800 }, { "epoch": 0.93, "learning_rate": 5.645710575314073e-07, "loss": 0.1261, "step": 3810 }, { "epoch": 0.93, "learning_rate": 5.246842287654235e-07, "loss": 0.1338, "step": 3820 }, { "epoch": 0.94, "learning_rate": 4.862435609655824e-07, "loss": 0.1345, "step": 3830 }, { "epoch": 0.94, "learning_rate": 4.492513254567482e-07, "loss": 0.1164, "step": 3840 }, { "epoch": 0.94, "learning_rate": 4.137097079809776e-07, "loss": 0.1212, "step": 3850 }, { "epoch": 0.94, "learning_rate": 3.796208085683733e-07, "loss": 0.1289, "step": 3860 }, { "epoch": 0.95, "learning_rate": 3.4698664141299686e-07, "loss": 0.1412, "step": 3870 }, { "epoch": 0.95, "learning_rate": 3.1580913475386107e-07, "loss": 0.1293, "step": 3880 }, { "epoch": 0.95, "learning_rate": 2.8609013076099365e-07, "loss": 0.1191, "step": 3890 }, { "epoch": 0.95, "learning_rate": 2.5783138542659345e-07, "loss": 0.1247, "step": 3900 }, { "epoch": 0.96, "learning_rate": 2.3103456846126913e-07, "loss": 0.1175, "step": 3910 }, { "epoch": 0.96, "learning_rate": 2.057012631953903e-07, "loss": 0.1317, "step": 3920 }, { "epoch": 0.96, "learning_rate": 1.8183296648552616e-07, "loss": 0.1421, "step": 3930 }, { "epoch": 0.96, "learning_rate": 1.5943108862600809e-07, "loss": 0.1365, "step": 3940 }, { "epoch": 0.97, "learning_rate": 1.384969532655933e-07, "loss": 0.1261, "step": 3950 }, { "epoch": 0.97, "learning_rate": 1.190317973292554e-07, "loss": 0.1054, "step": 3960 }, { "epoch": 0.97, "learning_rate": 1.0103677094510932e-07, "loss": 0.1251, "step": 3970 }, { "epoch": 0.97, "learning_rate": 8.451293737644638e-08, "loss": 0.1446, "step": 3980 }, { "epoch": 0.98, "learning_rate": 6.946127295890392e-08, "loss": 0.1271, "step": 3990 }, { "epoch": 0.98, "learning_rate": 5.5882667042789214e-08, "loss": 0.1372, "step": 4000 }, { "epoch": 0.98, "learning_rate": 4.377792194052432e-08, "loss": 0.1258, "step": 4010 }, { "epoch": 0.98, "learning_rate": 3.314775287923677e-08, "loss": 0.1268, "step": 4020 }, { "epoch": 0.99, "learning_rate": 2.399278795851001e-08, "loss": 0.1372, "step": 4030 }, { "epoch": 0.99, "learning_rate": 1.631356811325757e-08, "loss": 0.1274, "step": 4040 }, { "epoch": 0.99, "learning_rate": 1.0110547081776389e-08, "loss": 0.1361, "step": 4050 }, { "epoch": 0.99, "learning_rate": 5.384091378923817e-09, "loss": 0.1202, "step": 4060 }, { "epoch": 1.0, "learning_rate": 2.1344802744627246e-09, "loss": 0.1315, "step": 4070 }, { "epoch": 1.0, "learning_rate": 3.6190577657468914e-10, "loss": 0.1349, "step": 4080 }, { "epoch": 1.0, "step": 4087, "total_flos": 2.507754096179712e+18, "train_loss": 0.1560091979059696, "train_runtime": 21225.5051, "train_samples_per_second": 3.081, "train_steps_per_second": 0.193 } ], "logging_steps": 10, "max_steps": 4087, "num_train_epochs": 1, "save_steps": 1000, "total_flos": 2.507754096179712e+18, "trial_name": null, "trial_params": null }