diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,7413 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.8846153846153846, + "eval_steps": 500, + "global_step": 12000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "learning_rate": 4.776868449810209e-06, + "loss": 5.5366, + "step": 10 + }, + { + "epoch": 0.0, + "learning_rate": 6.214849138543984e-06, + "loss": 5.8115, + "step": 20 + }, + { + "epoch": 0.01, + "learning_rate": 7.056013918214425e-06, + "loss": 4.483, + "step": 30 + }, + { + "epoch": 0.01, + "learning_rate": 7.65282982727776e-06, + "loss": 3.6979, + "step": 40 + }, + { + "epoch": 0.01, + "learning_rate": 8.115756210886642e-06, + "loss": 2.1631, + "step": 50 + }, + { + "epoch": 0.01, + "learning_rate": 8.4939946069482e-06, + "loss": 1.081, + "step": 60 + }, + { + "epoch": 0.02, + "learning_rate": 8.813790614150759e-06, + "loss": 0.6225, + "step": 70 + }, + { + "epoch": 0.02, + "learning_rate": 9.090810516011535e-06, + "loss": 0.5162, + "step": 80 + }, + { + "epoch": 0.02, + "learning_rate": 9.335159386618639e-06, + "loss": 0.5221, + "step": 90 + }, + { + "epoch": 0.02, + "learning_rate": 9.553736899620418e-06, + "loss": 0.3936, + "step": 100 + }, + { + "epoch": 0.03, + "learning_rate": 9.75146431140567e-06, + "loss": 0.3957, + "step": 110 + }, + { + "epoch": 0.03, + "learning_rate": 9.931975295681976e-06, + "loss": 0.3244, + "step": 120 + }, + { + "epoch": 0.03, + "learning_rate": 9.995953382971836e-06, + "loss": 0.2308, + "step": 130 + }, + { + "epoch": 0.03, + "learning_rate": 9.987860148915507e-06, + "loss": 0.341, + "step": 140 + }, + { + "epoch": 0.04, + "learning_rate": 9.979766914859177e-06, + "loss": 0.2614, + "step": 150 + }, + { + "epoch": 0.04, + "learning_rate": 9.97167368080285e-06, + "loss": 0.2419, + "step": 160 + }, + { + "epoch": 0.04, + "learning_rate": 9.96358044674652e-06, + "loss": 0.2571, + "step": 170 + }, + { + "epoch": 0.04, + "learning_rate": 9.95548721269019e-06, + "loss": 0.2388, + "step": 180 + }, + { + "epoch": 0.05, + "learning_rate": 9.947393978633863e-06, + "loss": 0.2855, + "step": 190 + }, + { + "epoch": 0.05, + "learning_rate": 9.939300744577534e-06, + "loss": 0.2295, + "step": 200 + }, + { + "epoch": 0.05, + "learning_rate": 9.931207510521204e-06, + "loss": 0.1594, + "step": 210 + }, + { + "epoch": 0.05, + "learning_rate": 9.923114276464876e-06, + "loss": 0.1693, + "step": 220 + }, + { + "epoch": 0.06, + "learning_rate": 9.915021042408548e-06, + "loss": 0.2279, + "step": 230 + }, + { + "epoch": 0.06, + "learning_rate": 9.906927808352218e-06, + "loss": 0.2483, + "step": 240 + }, + { + "epoch": 0.06, + "learning_rate": 9.89883457429589e-06, + "loss": 0.2755, + "step": 250 + }, + { + "epoch": 0.06, + "learning_rate": 9.890741340239561e-06, + "loss": 0.1696, + "step": 260 + }, + { + "epoch": 0.06, + "learning_rate": 9.882648106183231e-06, + "loss": 0.1762, + "step": 270 + }, + { + "epoch": 0.07, + "learning_rate": 9.874554872126903e-06, + "loss": 0.1674, + "step": 280 + }, + { + "epoch": 0.07, + "learning_rate": 9.866461638070574e-06, + "loss": 0.2658, + "step": 290 + }, + { + "epoch": 0.07, + "learning_rate": 9.858368404014244e-06, + "loss": 0.147, + "step": 300 + }, + { + "epoch": 0.07, + "learning_rate": 9.850275169957916e-06, + "loss": 0.1612, + "step": 310 + }, + { + "epoch": 0.08, + "learning_rate": 9.842181935901588e-06, + "loss": 0.1837, + "step": 320 + }, + { + "epoch": 0.08, + "learning_rate": 9.834088701845258e-06, + "loss": 0.1751, + "step": 330 + }, + { + "epoch": 0.08, + "learning_rate": 9.82599546778893e-06, + "loss": 0.1923, + "step": 340 + }, + { + "epoch": 0.08, + "learning_rate": 9.817902233732601e-06, + "loss": 0.1124, + "step": 350 + }, + { + "epoch": 0.09, + "learning_rate": 9.809808999676271e-06, + "loss": 0.1667, + "step": 360 + }, + { + "epoch": 0.09, + "learning_rate": 9.801715765619943e-06, + "loss": 0.235, + "step": 370 + }, + { + "epoch": 0.09, + "learning_rate": 9.793622531563614e-06, + "loss": 0.2465, + "step": 380 + }, + { + "epoch": 0.09, + "learning_rate": 9.785529297507284e-06, + "loss": 0.1298, + "step": 390 + }, + { + "epoch": 0.1, + "learning_rate": 9.777436063450956e-06, + "loss": 0.1935, + "step": 400 + }, + { + "epoch": 0.1, + "learning_rate": 9.769342829394628e-06, + "loss": 0.1941, + "step": 410 + }, + { + "epoch": 0.1, + "learning_rate": 9.761249595338298e-06, + "loss": 0.1175, + "step": 420 + }, + { + "epoch": 0.1, + "learning_rate": 9.75315636128197e-06, + "loss": 0.1604, + "step": 430 + }, + { + "epoch": 0.11, + "learning_rate": 9.745063127225641e-06, + "loss": 0.1223, + "step": 440 + }, + { + "epoch": 0.11, + "learning_rate": 9.736969893169311e-06, + "loss": 0.1323, + "step": 450 + }, + { + "epoch": 0.11, + "learning_rate": 9.728876659112983e-06, + "loss": 0.1742, + "step": 460 + }, + { + "epoch": 0.11, + "learning_rate": 9.720783425056654e-06, + "loss": 0.1031, + "step": 470 + }, + { + "epoch": 0.12, + "learning_rate": 9.712690191000324e-06, + "loss": 0.2592, + "step": 480 + }, + { + "epoch": 0.12, + "learning_rate": 9.704596956943996e-06, + "loss": 0.1171, + "step": 490 + }, + { + "epoch": 0.12, + "learning_rate": 9.696503722887668e-06, + "loss": 0.2295, + "step": 500 + }, + { + "epoch": 0.12, + "eval_loss": 0.14800795912742615, + "eval_runtime": 420.4385, + "eval_samples_per_second": 14.839, + "eval_steps_per_second": 2.474, + "step": 500 + }, + { + "epoch": 0.12, + "learning_rate": 9.688410488831338e-06, + "loss": 0.1971, + "step": 510 + }, + { + "epoch": 0.12, + "learning_rate": 9.68031725477501e-06, + "loss": 0.117, + "step": 520 + }, + { + "epoch": 0.13, + "learning_rate": 9.672224020718681e-06, + "loss": 0.116, + "step": 530 + }, + { + "epoch": 0.13, + "learning_rate": 9.664130786662351e-06, + "loss": 0.1839, + "step": 540 + }, + { + "epoch": 0.13, + "learning_rate": 9.656037552606023e-06, + "loss": 0.1317, + "step": 550 + }, + { + "epoch": 0.13, + "learning_rate": 9.647944318549693e-06, + "loss": 0.094, + "step": 560 + }, + { + "epoch": 0.14, + "learning_rate": 9.639851084493364e-06, + "loss": 0.0871, + "step": 570 + }, + { + "epoch": 0.14, + "learning_rate": 9.631757850437036e-06, + "loss": 0.2401, + "step": 580 + }, + { + "epoch": 0.14, + "learning_rate": 9.623664616380706e-06, + "loss": 0.1376, + "step": 590 + }, + { + "epoch": 0.14, + "learning_rate": 9.615571382324378e-06, + "loss": 0.1962, + "step": 600 + }, + { + "epoch": 0.15, + "learning_rate": 9.60747814826805e-06, + "loss": 0.1894, + "step": 610 + }, + { + "epoch": 0.15, + "learning_rate": 9.59938491421172e-06, + "loss": 0.122, + "step": 620 + }, + { + "epoch": 0.15, + "learning_rate": 9.591291680155391e-06, + "loss": 0.1447, + "step": 630 + }, + { + "epoch": 0.15, + "learning_rate": 9.583198446099063e-06, + "loss": 0.1256, + "step": 640 + }, + { + "epoch": 0.16, + "learning_rate": 9.575105212042733e-06, + "loss": 0.1813, + "step": 650 + }, + { + "epoch": 0.16, + "learning_rate": 9.567011977986404e-06, + "loss": 0.1816, + "step": 660 + }, + { + "epoch": 0.16, + "learning_rate": 9.558918743930076e-06, + "loss": 0.1417, + "step": 670 + }, + { + "epoch": 0.16, + "learning_rate": 9.550825509873746e-06, + "loss": 0.1504, + "step": 680 + }, + { + "epoch": 0.17, + "learning_rate": 9.542732275817418e-06, + "loss": 0.1417, + "step": 690 + }, + { + "epoch": 0.17, + "learning_rate": 9.53463904176109e-06, + "loss": 0.2278, + "step": 700 + }, + { + "epoch": 0.17, + "learning_rate": 9.52654580770476e-06, + "loss": 0.0975, + "step": 710 + }, + { + "epoch": 0.17, + "learning_rate": 9.518452573648431e-06, + "loss": 0.1511, + "step": 720 + }, + { + "epoch": 0.18, + "learning_rate": 9.510359339592103e-06, + "loss": 0.1863, + "step": 730 + }, + { + "epoch": 0.18, + "learning_rate": 9.502266105535773e-06, + "loss": 0.1307, + "step": 740 + }, + { + "epoch": 0.18, + "learning_rate": 9.494172871479443e-06, + "loss": 0.1832, + "step": 750 + }, + { + "epoch": 0.18, + "learning_rate": 9.486079637423114e-06, + "loss": 0.0811, + "step": 760 + }, + { + "epoch": 0.19, + "learning_rate": 9.477986403366786e-06, + "loss": 0.1348, + "step": 770 + }, + { + "epoch": 0.19, + "learning_rate": 9.469893169310456e-06, + "loss": 0.1243, + "step": 780 + }, + { + "epoch": 0.19, + "learning_rate": 9.461799935254128e-06, + "loss": 0.0965, + "step": 790 + }, + { + "epoch": 0.19, + "learning_rate": 9.4537067011978e-06, + "loss": 0.1227, + "step": 800 + }, + { + "epoch": 0.19, + "learning_rate": 9.44561346714147e-06, + "loss": 0.0882, + "step": 810 + }, + { + "epoch": 0.2, + "learning_rate": 9.437520233085141e-06, + "loss": 0.1245, + "step": 820 + }, + { + "epoch": 0.2, + "learning_rate": 9.429426999028813e-06, + "loss": 0.0908, + "step": 830 + }, + { + "epoch": 0.2, + "learning_rate": 9.421333764972483e-06, + "loss": 0.1084, + "step": 840 + }, + { + "epoch": 0.2, + "learning_rate": 9.413240530916154e-06, + "loss": 0.1101, + "step": 850 + }, + { + "epoch": 0.21, + "learning_rate": 9.405147296859826e-06, + "loss": 0.1388, + "step": 860 + }, + { + "epoch": 0.21, + "learning_rate": 9.397054062803496e-06, + "loss": 0.1372, + "step": 870 + }, + { + "epoch": 0.21, + "learning_rate": 9.388960828747168e-06, + "loss": 0.1032, + "step": 880 + }, + { + "epoch": 0.21, + "learning_rate": 9.38086759469084e-06, + "loss": 0.1065, + "step": 890 + }, + { + "epoch": 0.22, + "learning_rate": 9.37277436063451e-06, + "loss": 0.0605, + "step": 900 + }, + { + "epoch": 0.22, + "learning_rate": 9.364681126578181e-06, + "loss": 0.0967, + "step": 910 + }, + { + "epoch": 0.22, + "learning_rate": 9.356587892521853e-06, + "loss": 0.1294, + "step": 920 + }, + { + "epoch": 0.22, + "learning_rate": 9.348494658465523e-06, + "loss": 0.0591, + "step": 930 + }, + { + "epoch": 0.23, + "learning_rate": 9.340401424409194e-06, + "loss": 0.0643, + "step": 940 + }, + { + "epoch": 0.23, + "learning_rate": 9.332308190352866e-06, + "loss": 0.1745, + "step": 950 + }, + { + "epoch": 0.23, + "learning_rate": 9.324214956296536e-06, + "loss": 0.1285, + "step": 960 + }, + { + "epoch": 0.23, + "learning_rate": 9.316121722240208e-06, + "loss": 0.0945, + "step": 970 + }, + { + "epoch": 0.24, + "learning_rate": 9.30802848818388e-06, + "loss": 0.1411, + "step": 980 + }, + { + "epoch": 0.24, + "learning_rate": 9.29993525412755e-06, + "loss": 0.0819, + "step": 990 + }, + { + "epoch": 0.24, + "learning_rate": 9.291842020071221e-06, + "loss": 0.1516, + "step": 1000 + }, + { + "epoch": 0.24, + "eval_loss": 0.10547837615013123, + "eval_runtime": 420.9793, + "eval_samples_per_second": 14.82, + "eval_steps_per_second": 2.47, + "step": 1000 + }, + { + "epoch": 0.24, + "learning_rate": 9.283748786014893e-06, + "loss": 0.1138, + "step": 1010 + }, + { + "epoch": 0.25, + "learning_rate": 9.275655551958563e-06, + "loss": 0.1164, + "step": 1020 + }, + { + "epoch": 0.25, + "learning_rate": 9.267562317902235e-06, + "loss": 0.0835, + "step": 1030 + }, + { + "epoch": 0.25, + "learning_rate": 9.259469083845906e-06, + "loss": 0.136, + "step": 1040 + }, + { + "epoch": 0.25, + "learning_rate": 9.251375849789576e-06, + "loss": 0.1011, + "step": 1050 + }, + { + "epoch": 0.25, + "learning_rate": 9.243282615733248e-06, + "loss": 0.1471, + "step": 1060 + }, + { + "epoch": 0.26, + "learning_rate": 9.23518938167692e-06, + "loss": 0.1014, + "step": 1070 + }, + { + "epoch": 0.26, + "learning_rate": 9.22709614762059e-06, + "loss": 0.0704, + "step": 1080 + }, + { + "epoch": 0.26, + "learning_rate": 9.219002913564261e-06, + "loss": 0.0775, + "step": 1090 + }, + { + "epoch": 0.26, + "learning_rate": 9.210909679507933e-06, + "loss": 0.1406, + "step": 1100 + }, + { + "epoch": 0.27, + "learning_rate": 9.202816445451603e-06, + "loss": 0.1182, + "step": 1110 + }, + { + "epoch": 0.27, + "learning_rate": 9.194723211395275e-06, + "loss": 0.0706, + "step": 1120 + }, + { + "epoch": 0.27, + "learning_rate": 9.186629977338946e-06, + "loss": 0.0891, + "step": 1130 + }, + { + "epoch": 0.27, + "learning_rate": 9.178536743282616e-06, + "loss": 0.0515, + "step": 1140 + }, + { + "epoch": 0.28, + "learning_rate": 9.170443509226288e-06, + "loss": 0.1169, + "step": 1150 + }, + { + "epoch": 0.28, + "learning_rate": 9.16235027516996e-06, + "loss": 0.1162, + "step": 1160 + }, + { + "epoch": 0.28, + "learning_rate": 9.15425704111363e-06, + "loss": 0.1404, + "step": 1170 + }, + { + "epoch": 0.28, + "learning_rate": 9.146163807057301e-06, + "loss": 0.1248, + "step": 1180 + }, + { + "epoch": 0.29, + "learning_rate": 9.138070573000973e-06, + "loss": 0.1024, + "step": 1190 + }, + { + "epoch": 0.29, + "learning_rate": 9.129977338944643e-06, + "loss": 0.0956, + "step": 1200 + }, + { + "epoch": 0.29, + "learning_rate": 9.121884104888315e-06, + "loss": 0.0477, + "step": 1210 + }, + { + "epoch": 0.29, + "learning_rate": 9.113790870831986e-06, + "loss": 0.109, + "step": 1220 + }, + { + "epoch": 0.3, + "learning_rate": 9.105697636775656e-06, + "loss": 0.1411, + "step": 1230 + }, + { + "epoch": 0.3, + "learning_rate": 9.097604402719328e-06, + "loss": 0.0602, + "step": 1240 + }, + { + "epoch": 0.3, + "learning_rate": 9.089511168663e-06, + "loss": 0.068, + "step": 1250 + }, + { + "epoch": 0.3, + "learning_rate": 9.08141793460667e-06, + "loss": 0.0803, + "step": 1260 + }, + { + "epoch": 0.31, + "learning_rate": 9.073324700550341e-06, + "loss": 0.0626, + "step": 1270 + }, + { + "epoch": 0.31, + "learning_rate": 9.065231466494013e-06, + "loss": 0.0749, + "step": 1280 + }, + { + "epoch": 0.31, + "learning_rate": 9.057138232437683e-06, + "loss": 0.132, + "step": 1290 + }, + { + "epoch": 0.31, + "learning_rate": 9.049044998381355e-06, + "loss": 0.1092, + "step": 1300 + }, + { + "epoch": 0.31, + "learning_rate": 9.040951764325026e-06, + "loss": 0.0968, + "step": 1310 + }, + { + "epoch": 0.32, + "learning_rate": 9.032858530268696e-06, + "loss": 0.1175, + "step": 1320 + }, + { + "epoch": 0.32, + "learning_rate": 9.024765296212368e-06, + "loss": 0.0756, + "step": 1330 + }, + { + "epoch": 0.32, + "learning_rate": 9.016672062156038e-06, + "loss": 0.1387, + "step": 1340 + }, + { + "epoch": 0.32, + "learning_rate": 9.00857882809971e-06, + "loss": 0.0971, + "step": 1350 + }, + { + "epoch": 0.33, + "learning_rate": 9.00048559404338e-06, + "loss": 0.0696, + "step": 1360 + }, + { + "epoch": 0.33, + "learning_rate": 8.992392359987051e-06, + "loss": 0.0892, + "step": 1370 + }, + { + "epoch": 0.33, + "learning_rate": 8.984299125930723e-06, + "loss": 0.0675, + "step": 1380 + }, + { + "epoch": 0.33, + "learning_rate": 8.976205891874393e-06, + "loss": 0.0848, + "step": 1390 + }, + { + "epoch": 0.34, + "learning_rate": 8.968112657818065e-06, + "loss": 0.1111, + "step": 1400 + }, + { + "epoch": 0.34, + "learning_rate": 8.960019423761736e-06, + "loss": 0.0698, + "step": 1410 + }, + { + "epoch": 0.34, + "learning_rate": 8.951926189705406e-06, + "loss": 0.1462, + "step": 1420 + }, + { + "epoch": 0.34, + "learning_rate": 8.943832955649078e-06, + "loss": 0.0836, + "step": 1430 + }, + { + "epoch": 0.35, + "learning_rate": 8.935739721592748e-06, + "loss": 0.0846, + "step": 1440 + }, + { + "epoch": 0.35, + "learning_rate": 8.92764648753642e-06, + "loss": 0.097, + "step": 1450 + }, + { + "epoch": 0.35, + "learning_rate": 8.919553253480091e-06, + "loss": 0.0931, + "step": 1460 + }, + { + "epoch": 0.35, + "learning_rate": 8.911460019423761e-06, + "loss": 0.0297, + "step": 1470 + }, + { + "epoch": 0.36, + "learning_rate": 8.903366785367433e-06, + "loss": 0.1094, + "step": 1480 + }, + { + "epoch": 0.36, + "learning_rate": 8.895273551311105e-06, + "loss": 0.095, + "step": 1490 + }, + { + "epoch": 0.36, + "learning_rate": 8.887180317254775e-06, + "loss": 0.0768, + "step": 1500 + }, + { + "epoch": 0.36, + "eval_loss": 0.10153712332248688, + "eval_runtime": 417.8721, + "eval_samples_per_second": 14.93, + "eval_steps_per_second": 2.489, + "step": 1500 + }, + { + "epoch": 0.36, + "learning_rate": 8.879087083198446e-06, + "loss": 0.1861, + "step": 1510 + }, + { + "epoch": 0.37, + "learning_rate": 8.870993849142118e-06, + "loss": 0.0467, + "step": 1520 + }, + { + "epoch": 0.37, + "learning_rate": 8.862900615085788e-06, + "loss": 0.0914, + "step": 1530 + }, + { + "epoch": 0.37, + "learning_rate": 8.85480738102946e-06, + "loss": 0.0958, + "step": 1540 + }, + { + "epoch": 0.37, + "learning_rate": 8.846714146973131e-06, + "loss": 0.099, + "step": 1550 + }, + { + "epoch": 0.38, + "learning_rate": 8.838620912916801e-06, + "loss": 0.0949, + "step": 1560 + }, + { + "epoch": 0.38, + "learning_rate": 8.830527678860473e-06, + "loss": 0.0593, + "step": 1570 + }, + { + "epoch": 0.38, + "learning_rate": 8.822434444804145e-06, + "loss": 0.0861, + "step": 1580 + }, + { + "epoch": 0.38, + "learning_rate": 8.814341210747815e-06, + "loss": 0.0659, + "step": 1590 + }, + { + "epoch": 0.38, + "learning_rate": 8.806247976691486e-06, + "loss": 0.0929, + "step": 1600 + }, + { + "epoch": 0.39, + "learning_rate": 8.798154742635158e-06, + "loss": 0.112, + "step": 1610 + }, + { + "epoch": 0.39, + "learning_rate": 8.790061508578828e-06, + "loss": 0.1193, + "step": 1620 + }, + { + "epoch": 0.39, + "learning_rate": 8.7819682745225e-06, + "loss": 0.0739, + "step": 1630 + }, + { + "epoch": 0.39, + "learning_rate": 8.773875040466171e-06, + "loss": 0.066, + "step": 1640 + }, + { + "epoch": 0.4, + "learning_rate": 8.765781806409841e-06, + "loss": 0.1185, + "step": 1650 + }, + { + "epoch": 0.4, + "learning_rate": 8.757688572353513e-06, + "loss": 0.1405, + "step": 1660 + }, + { + "epoch": 0.4, + "learning_rate": 8.749595338297185e-06, + "loss": 0.0414, + "step": 1670 + }, + { + "epoch": 0.4, + "learning_rate": 8.741502104240855e-06, + "loss": 0.0982, + "step": 1680 + }, + { + "epoch": 0.41, + "learning_rate": 8.733408870184526e-06, + "loss": 0.0526, + "step": 1690 + }, + { + "epoch": 0.41, + "learning_rate": 8.725315636128198e-06, + "loss": 0.1257, + "step": 1700 + }, + { + "epoch": 0.41, + "learning_rate": 8.717222402071868e-06, + "loss": 0.0607, + "step": 1710 + }, + { + "epoch": 0.41, + "learning_rate": 8.70912916801554e-06, + "loss": 0.0923, + "step": 1720 + }, + { + "epoch": 0.42, + "learning_rate": 8.701035933959211e-06, + "loss": 0.1085, + "step": 1730 + }, + { + "epoch": 0.42, + "learning_rate": 8.692942699902881e-06, + "loss": 0.0752, + "step": 1740 + }, + { + "epoch": 0.42, + "learning_rate": 8.684849465846553e-06, + "loss": 0.0635, + "step": 1750 + }, + { + "epoch": 0.42, + "learning_rate": 8.676756231790225e-06, + "loss": 0.075, + "step": 1760 + }, + { + "epoch": 0.43, + "learning_rate": 8.668662997733895e-06, + "loss": 0.1272, + "step": 1770 + }, + { + "epoch": 0.43, + "learning_rate": 8.660569763677566e-06, + "loss": 0.1021, + "step": 1780 + }, + { + "epoch": 0.43, + "learning_rate": 8.652476529621238e-06, + "loss": 0.1668, + "step": 1790 + }, + { + "epoch": 0.43, + "learning_rate": 8.644383295564908e-06, + "loss": 0.0798, + "step": 1800 + }, + { + "epoch": 0.44, + "learning_rate": 8.63629006150858e-06, + "loss": 0.0927, + "step": 1810 + }, + { + "epoch": 0.44, + "learning_rate": 8.628196827452251e-06, + "loss": 0.0739, + "step": 1820 + }, + { + "epoch": 0.44, + "learning_rate": 8.620103593395921e-06, + "loss": 0.0829, + "step": 1830 + }, + { + "epoch": 0.44, + "learning_rate": 8.612010359339593e-06, + "loss": 0.1396, + "step": 1840 + }, + { + "epoch": 0.44, + "learning_rate": 8.603917125283265e-06, + "loss": 0.0811, + "step": 1850 + }, + { + "epoch": 0.45, + "learning_rate": 8.595823891226935e-06, + "loss": 0.0915, + "step": 1860 + }, + { + "epoch": 0.45, + "learning_rate": 8.587730657170607e-06, + "loss": 0.1205, + "step": 1870 + }, + { + "epoch": 0.45, + "learning_rate": 8.579637423114278e-06, + "loss": 0.0549, + "step": 1880 + }, + { + "epoch": 0.45, + "learning_rate": 8.571544189057948e-06, + "loss": 0.1083, + "step": 1890 + }, + { + "epoch": 0.46, + "learning_rate": 8.56345095500162e-06, + "loss": 0.0424, + "step": 1900 + }, + { + "epoch": 0.46, + "learning_rate": 8.555357720945292e-06, + "loss": 0.0689, + "step": 1910 + }, + { + "epoch": 0.46, + "learning_rate": 8.547264486888962e-06, + "loss": 0.0414, + "step": 1920 + }, + { + "epoch": 0.46, + "learning_rate": 8.539171252832633e-06, + "loss": 0.0659, + "step": 1930 + }, + { + "epoch": 0.47, + "learning_rate": 8.531078018776305e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.47, + "learning_rate": 8.522984784719975e-06, + "loss": 0.1117, + "step": 1950 + }, + { + "epoch": 0.47, + "learning_rate": 8.514891550663645e-06, + "loss": 0.0832, + "step": 1960 + }, + { + "epoch": 0.47, + "learning_rate": 8.506798316607317e-06, + "loss": 0.1311, + "step": 1970 + }, + { + "epoch": 0.48, + "learning_rate": 8.498705082550988e-06, + "loss": 0.0835, + "step": 1980 + }, + { + "epoch": 0.48, + "learning_rate": 8.490611848494658e-06, + "loss": 0.0672, + "step": 1990 + }, + { + "epoch": 0.48, + "learning_rate": 8.48251861443833e-06, + "loss": 0.0709, + "step": 2000 + }, + { + "epoch": 0.48, + "eval_loss": 0.08033791929483414, + "eval_runtime": 419.5502, + "eval_samples_per_second": 14.871, + "eval_steps_per_second": 2.479, + "step": 2000 + }, + { + "epoch": 0.48, + "learning_rate": 8.474425380382002e-06, + "loss": 0.0657, + "step": 2010 + }, + { + "epoch": 0.49, + "learning_rate": 8.466332146325672e-06, + "loss": 0.0599, + "step": 2020 + }, + { + "epoch": 0.49, + "learning_rate": 8.458238912269343e-06, + "loss": 0.0615, + "step": 2030 + }, + { + "epoch": 0.49, + "learning_rate": 8.450145678213015e-06, + "loss": 0.0525, + "step": 2040 + }, + { + "epoch": 0.49, + "learning_rate": 8.442052444156685e-06, + "loss": 0.0472, + "step": 2050 + }, + { + "epoch": 0.5, + "learning_rate": 8.433959210100357e-06, + "loss": 0.0493, + "step": 2060 + }, + { + "epoch": 0.5, + "learning_rate": 8.425865976044028e-06, + "loss": 0.0658, + "step": 2070 + }, + { + "epoch": 0.5, + "learning_rate": 8.417772741987698e-06, + "loss": 0.0658, + "step": 2080 + }, + { + "epoch": 0.5, + "learning_rate": 8.40967950793137e-06, + "loss": 0.0507, + "step": 2090 + }, + { + "epoch": 0.5, + "learning_rate": 8.401586273875042e-06, + "loss": 0.1675, + "step": 2100 + }, + { + "epoch": 0.51, + "learning_rate": 8.393493039818712e-06, + "loss": 0.0991, + "step": 2110 + }, + { + "epoch": 0.51, + "learning_rate": 8.385399805762383e-06, + "loss": 0.1147, + "step": 2120 + }, + { + "epoch": 0.51, + "learning_rate": 8.377306571706055e-06, + "loss": 0.1269, + "step": 2130 + }, + { + "epoch": 0.51, + "learning_rate": 8.369213337649725e-06, + "loss": 0.0442, + "step": 2140 + }, + { + "epoch": 0.52, + "learning_rate": 8.361120103593397e-06, + "loss": 0.0722, + "step": 2150 + }, + { + "epoch": 0.52, + "learning_rate": 8.353026869537068e-06, + "loss": 0.1512, + "step": 2160 + }, + { + "epoch": 0.52, + "learning_rate": 8.344933635480738e-06, + "loss": 0.0903, + "step": 2170 + }, + { + "epoch": 0.52, + "learning_rate": 8.33684040142441e-06, + "loss": 0.0442, + "step": 2180 + }, + { + "epoch": 0.53, + "learning_rate": 8.328747167368082e-06, + "loss": 0.095, + "step": 2190 + }, + { + "epoch": 0.53, + "learning_rate": 8.320653933311752e-06, + "loss": 0.1374, + "step": 2200 + }, + { + "epoch": 0.53, + "learning_rate": 8.312560699255423e-06, + "loss": 0.0676, + "step": 2210 + }, + { + "epoch": 0.53, + "learning_rate": 8.304467465199095e-06, + "loss": 0.0835, + "step": 2220 + }, + { + "epoch": 0.54, + "learning_rate": 8.296374231142765e-06, + "loss": 0.1064, + "step": 2230 + }, + { + "epoch": 0.54, + "learning_rate": 8.288280997086437e-06, + "loss": 0.1068, + "step": 2240 + }, + { + "epoch": 0.54, + "learning_rate": 8.280187763030108e-06, + "loss": 0.0637, + "step": 2250 + }, + { + "epoch": 0.54, + "learning_rate": 8.272094528973778e-06, + "loss": 0.1137, + "step": 2260 + }, + { + "epoch": 0.55, + "learning_rate": 8.26400129491745e-06, + "loss": 0.1051, + "step": 2270 + }, + { + "epoch": 0.55, + "learning_rate": 8.255908060861122e-06, + "loss": 0.0872, + "step": 2280 + }, + { + "epoch": 0.55, + "learning_rate": 8.247814826804792e-06, + "loss": 0.0676, + "step": 2290 + }, + { + "epoch": 0.55, + "learning_rate": 8.239721592748463e-06, + "loss": 0.1471, + "step": 2300 + }, + { + "epoch": 0.56, + "learning_rate": 8.231628358692133e-06, + "loss": 0.13, + "step": 2310 + }, + { + "epoch": 0.56, + "learning_rate": 8.223535124635805e-06, + "loss": 0.0815, + "step": 2320 + }, + { + "epoch": 0.56, + "learning_rate": 8.215441890579477e-06, + "loss": 0.0811, + "step": 2330 + }, + { + "epoch": 0.56, + "learning_rate": 8.207348656523147e-06, + "loss": 0.1489, + "step": 2340 + }, + { + "epoch": 0.56, + "learning_rate": 8.199255422466818e-06, + "loss": 0.1038, + "step": 2350 + }, + { + "epoch": 0.57, + "learning_rate": 8.19116218841049e-06, + "loss": 0.0437, + "step": 2360 + }, + { + "epoch": 0.57, + "learning_rate": 8.18306895435416e-06, + "loss": 0.0659, + "step": 2370 + }, + { + "epoch": 0.57, + "learning_rate": 8.174975720297832e-06, + "loss": 0.1125, + "step": 2380 + }, + { + "epoch": 0.57, + "learning_rate": 8.166882486241503e-06, + "loss": 0.0875, + "step": 2390 + }, + { + "epoch": 0.58, + "learning_rate": 8.158789252185173e-06, + "loss": 0.1194, + "step": 2400 + }, + { + "epoch": 0.58, + "learning_rate": 8.150696018128845e-06, + "loss": 0.085, + "step": 2410 + }, + { + "epoch": 0.58, + "learning_rate": 8.142602784072517e-06, + "loss": 0.0998, + "step": 2420 + }, + { + "epoch": 0.58, + "learning_rate": 8.134509550016187e-06, + "loss": 0.0805, + "step": 2430 + }, + { + "epoch": 0.59, + "learning_rate": 8.126416315959858e-06, + "loss": 0.0551, + "step": 2440 + }, + { + "epoch": 0.59, + "learning_rate": 8.11832308190353e-06, + "loss": 0.0644, + "step": 2450 + }, + { + "epoch": 0.59, + "learning_rate": 8.1102298478472e-06, + "loss": 0.0822, + "step": 2460 + }, + { + "epoch": 0.59, + "learning_rate": 8.102136613790872e-06, + "loss": 0.1049, + "step": 2470 + }, + { + "epoch": 0.6, + "learning_rate": 8.094043379734543e-06, + "loss": 0.0556, + "step": 2480 + }, + { + "epoch": 0.6, + "learning_rate": 8.085950145678213e-06, + "loss": 0.0539, + "step": 2490 + }, + { + "epoch": 0.6, + "learning_rate": 8.077856911621885e-06, + "loss": 0.0947, + "step": 2500 + }, + { + "epoch": 0.6, + "eval_loss": 0.07605639100074768, + "eval_runtime": 420.2955, + "eval_samples_per_second": 14.844, + "eval_steps_per_second": 2.474, + "step": 2500 + }, + { + "epoch": 0.6, + "learning_rate": 8.069763677565557e-06, + "loss": 0.0563, + "step": 2510 + }, + { + "epoch": 0.61, + "learning_rate": 8.061670443509227e-06, + "loss": 0.0718, + "step": 2520 + }, + { + "epoch": 0.61, + "learning_rate": 8.053577209452898e-06, + "loss": 0.0855, + "step": 2530 + }, + { + "epoch": 0.61, + "learning_rate": 8.04548397539657e-06, + "loss": 0.1026, + "step": 2540 + }, + { + "epoch": 0.61, + "learning_rate": 8.03739074134024e-06, + "loss": 0.0649, + "step": 2550 + }, + { + "epoch": 0.62, + "learning_rate": 8.02929750728391e-06, + "loss": 0.0671, + "step": 2560 + }, + { + "epoch": 0.62, + "learning_rate": 8.021204273227582e-06, + "loss": 0.1167, + "step": 2570 + }, + { + "epoch": 0.62, + "learning_rate": 8.013111039171253e-06, + "loss": 0.0706, + "step": 2580 + }, + { + "epoch": 0.62, + "learning_rate": 8.005017805114923e-06, + "loss": 0.0706, + "step": 2590 + }, + { + "epoch": 0.62, + "learning_rate": 7.996924571058595e-06, + "loss": 0.0706, + "step": 2600 + }, + { + "epoch": 0.63, + "learning_rate": 7.988831337002267e-06, + "loss": 0.0581, + "step": 2610 + }, + { + "epoch": 0.63, + "learning_rate": 7.980738102945937e-06, + "loss": 0.0372, + "step": 2620 + }, + { + "epoch": 0.63, + "learning_rate": 7.972644868889608e-06, + "loss": 0.0666, + "step": 2630 + }, + { + "epoch": 0.63, + "learning_rate": 7.96455163483328e-06, + "loss": 0.0601, + "step": 2640 + }, + { + "epoch": 0.64, + "learning_rate": 7.95645840077695e-06, + "loss": 0.0773, + "step": 2650 + }, + { + "epoch": 0.64, + "learning_rate": 7.948365166720622e-06, + "loss": 0.1009, + "step": 2660 + }, + { + "epoch": 0.64, + "learning_rate": 7.940271932664293e-06, + "loss": 0.1277, + "step": 2670 + }, + { + "epoch": 0.64, + "learning_rate": 7.932178698607963e-06, + "loss": 0.0673, + "step": 2680 + }, + { + "epoch": 0.65, + "learning_rate": 7.924085464551635e-06, + "loss": 0.0264, + "step": 2690 + }, + { + "epoch": 0.65, + "learning_rate": 7.915992230495307e-06, + "loss": 0.0646, + "step": 2700 + }, + { + "epoch": 0.65, + "learning_rate": 7.907898996438977e-06, + "loss": 0.0649, + "step": 2710 + }, + { + "epoch": 0.65, + "learning_rate": 7.899805762382648e-06, + "loss": 0.0272, + "step": 2720 + }, + { + "epoch": 0.66, + "learning_rate": 7.89171252832632e-06, + "loss": 0.143, + "step": 2730 + }, + { + "epoch": 0.66, + "learning_rate": 7.88361929426999e-06, + "loss": 0.1017, + "step": 2740 + }, + { + "epoch": 0.66, + "learning_rate": 7.875526060213662e-06, + "loss": 0.0122, + "step": 2750 + }, + { + "epoch": 0.66, + "learning_rate": 7.867432826157334e-06, + "loss": 0.0752, + "step": 2760 + }, + { + "epoch": 0.67, + "learning_rate": 7.859339592101004e-06, + "loss": 0.0944, + "step": 2770 + }, + { + "epoch": 0.67, + "learning_rate": 7.851246358044675e-06, + "loss": 0.0836, + "step": 2780 + }, + { + "epoch": 0.67, + "learning_rate": 7.843153123988347e-06, + "loss": 0.0772, + "step": 2790 + }, + { + "epoch": 0.67, + "learning_rate": 7.835059889932017e-06, + "loss": 0.0723, + "step": 2800 + }, + { + "epoch": 0.68, + "learning_rate": 7.826966655875689e-06, + "loss": 0.0766, + "step": 2810 + }, + { + "epoch": 0.68, + "learning_rate": 7.81887342181936e-06, + "loss": 0.1145, + "step": 2820 + }, + { + "epoch": 0.68, + "learning_rate": 7.81078018776303e-06, + "loss": 0.1332, + "step": 2830 + }, + { + "epoch": 0.68, + "learning_rate": 7.802686953706702e-06, + "loss": 0.0541, + "step": 2840 + }, + { + "epoch": 0.69, + "learning_rate": 7.794593719650374e-06, + "loss": 0.1186, + "step": 2850 + }, + { + "epoch": 0.69, + "learning_rate": 7.786500485594044e-06, + "loss": 0.0482, + "step": 2860 + }, + { + "epoch": 0.69, + "learning_rate": 7.778407251537715e-06, + "loss": 0.1562, + "step": 2870 + }, + { + "epoch": 0.69, + "learning_rate": 7.770314017481387e-06, + "loss": 0.066, + "step": 2880 + }, + { + "epoch": 0.69, + "learning_rate": 7.762220783425057e-06, + "loss": 0.048, + "step": 2890 + }, + { + "epoch": 0.7, + "learning_rate": 7.754127549368729e-06, + "loss": 0.0834, + "step": 2900 + }, + { + "epoch": 0.7, + "learning_rate": 7.7460343153124e-06, + "loss": 0.1008, + "step": 2910 + }, + { + "epoch": 0.7, + "learning_rate": 7.73794108125607e-06, + "loss": 0.0632, + "step": 2920 + }, + { + "epoch": 0.7, + "learning_rate": 7.729847847199742e-06, + "loss": 0.0674, + "step": 2930 + }, + { + "epoch": 0.71, + "learning_rate": 7.721754613143414e-06, + "loss": 0.105, + "step": 2940 + }, + { + "epoch": 0.71, + "learning_rate": 7.713661379087084e-06, + "loss": 0.0866, + "step": 2950 + }, + { + "epoch": 0.71, + "learning_rate": 7.705568145030755e-06, + "loss": 0.1476, + "step": 2960 + }, + { + "epoch": 0.71, + "learning_rate": 7.697474910974427e-06, + "loss": 0.0523, + "step": 2970 + }, + { + "epoch": 0.72, + "learning_rate": 7.689381676918097e-06, + "loss": 0.0827, + "step": 2980 + }, + { + "epoch": 0.72, + "learning_rate": 7.681288442861769e-06, + "loss": 0.0721, + "step": 2990 + }, + { + "epoch": 0.72, + "learning_rate": 7.67319520880544e-06, + "loss": 0.083, + "step": 3000 + }, + { + "epoch": 0.72, + "eval_loss": 0.07007448375225067, + "eval_runtime": 421.7523, + "eval_samples_per_second": 14.793, + "eval_steps_per_second": 2.466, + "step": 3000 + }, + { + "epoch": 0.72, + "learning_rate": 7.66510197474911e-06, + "loss": 0.0455, + "step": 3010 + }, + { + "epoch": 0.73, + "learning_rate": 7.657008740692782e-06, + "loss": 0.0492, + "step": 3020 + }, + { + "epoch": 0.73, + "learning_rate": 7.648915506636454e-06, + "loss": 0.0628, + "step": 3030 + }, + { + "epoch": 0.73, + "learning_rate": 7.640822272580124e-06, + "loss": 0.1527, + "step": 3040 + }, + { + "epoch": 0.73, + "learning_rate": 7.632729038523795e-06, + "loss": 0.0649, + "step": 3050 + }, + { + "epoch": 0.74, + "learning_rate": 7.624635804467466e-06, + "loss": 0.0879, + "step": 3060 + }, + { + "epoch": 0.74, + "learning_rate": 7.616542570411137e-06, + "loss": 0.0881, + "step": 3070 + }, + { + "epoch": 0.74, + "learning_rate": 7.608449336354809e-06, + "loss": 0.0698, + "step": 3080 + }, + { + "epoch": 0.74, + "learning_rate": 7.6003561022984795e-06, + "loss": 0.0234, + "step": 3090 + }, + { + "epoch": 0.75, + "learning_rate": 7.59226286824215e-06, + "loss": 0.0553, + "step": 3100 + }, + { + "epoch": 0.75, + "learning_rate": 7.584169634185822e-06, + "loss": 0.0314, + "step": 3110 + }, + { + "epoch": 0.75, + "learning_rate": 7.576076400129493e-06, + "loss": 0.0445, + "step": 3120 + }, + { + "epoch": 0.75, + "learning_rate": 7.567983166073164e-06, + "loss": 0.0733, + "step": 3130 + }, + { + "epoch": 0.75, + "learning_rate": 7.559889932016835e-06, + "loss": 0.1025, + "step": 3140 + }, + { + "epoch": 0.76, + "learning_rate": 7.551796697960506e-06, + "loss": 0.0537, + "step": 3150 + }, + { + "epoch": 0.76, + "learning_rate": 7.543703463904176e-06, + "loss": 0.1012, + "step": 3160 + }, + { + "epoch": 0.76, + "learning_rate": 7.535610229847847e-06, + "loss": 0.1665, + "step": 3170 + }, + { + "epoch": 0.76, + "learning_rate": 7.527516995791519e-06, + "loss": 0.0747, + "step": 3180 + }, + { + "epoch": 0.77, + "learning_rate": 7.5194237617351895e-06, + "loss": 0.0589, + "step": 3190 + }, + { + "epoch": 0.77, + "learning_rate": 7.51133052767886e-06, + "loss": 0.1095, + "step": 3200 + }, + { + "epoch": 0.77, + "learning_rate": 7.503237293622532e-06, + "loss": 0.1046, + "step": 3210 + }, + { + "epoch": 0.77, + "learning_rate": 7.495144059566203e-06, + "loss": 0.0555, + "step": 3220 + }, + { + "epoch": 0.78, + "learning_rate": 7.487050825509874e-06, + "loss": 0.0881, + "step": 3230 + }, + { + "epoch": 0.78, + "learning_rate": 7.478957591453545e-06, + "loss": 0.1073, + "step": 3240 + }, + { + "epoch": 0.78, + "learning_rate": 7.470864357397216e-06, + "loss": 0.1232, + "step": 3250 + }, + { + "epoch": 0.78, + "learning_rate": 7.462771123340887e-06, + "loss": 0.0583, + "step": 3260 + }, + { + "epoch": 0.79, + "learning_rate": 7.454677889284559e-06, + "loss": 0.0309, + "step": 3270 + }, + { + "epoch": 0.79, + "learning_rate": 7.4465846552282296e-06, + "loss": 0.0778, + "step": 3280 + }, + { + "epoch": 0.79, + "learning_rate": 7.4384914211719e-06, + "loss": 0.0337, + "step": 3290 + }, + { + "epoch": 0.79, + "learning_rate": 7.430398187115572e-06, + "loss": 0.0699, + "step": 3300 + }, + { + "epoch": 0.8, + "learning_rate": 7.422304953059243e-06, + "loss": 0.0904, + "step": 3310 + }, + { + "epoch": 0.8, + "learning_rate": 7.414211719002914e-06, + "loss": 0.0711, + "step": 3320 + }, + { + "epoch": 0.8, + "learning_rate": 7.406118484946585e-06, + "loss": 0.0739, + "step": 3330 + }, + { + "epoch": 0.8, + "learning_rate": 7.398025250890256e-06, + "loss": 0.0648, + "step": 3340 + }, + { + "epoch": 0.81, + "learning_rate": 7.389932016833927e-06, + "loss": 0.0557, + "step": 3350 + }, + { + "epoch": 0.81, + "learning_rate": 7.381838782777599e-06, + "loss": 0.0771, + "step": 3360 + }, + { + "epoch": 0.81, + "learning_rate": 7.37374554872127e-06, + "loss": 0.0522, + "step": 3370 + }, + { + "epoch": 0.81, + "learning_rate": 7.3656523146649404e-06, + "loss": 0.0399, + "step": 3380 + }, + { + "epoch": 0.81, + "learning_rate": 7.357559080608612e-06, + "loss": 0.0246, + "step": 3390 + }, + { + "epoch": 0.82, + "learning_rate": 7.349465846552283e-06, + "loss": 0.0744, + "step": 3400 + }, + { + "epoch": 0.82, + "learning_rate": 7.341372612495954e-06, + "loss": 0.03, + "step": 3410 + }, + { + "epoch": 0.82, + "learning_rate": 7.333279378439625e-06, + "loss": 0.0631, + "step": 3420 + }, + { + "epoch": 0.82, + "learning_rate": 7.325186144383296e-06, + "loss": 0.0545, + "step": 3430 + }, + { + "epoch": 0.83, + "learning_rate": 7.317092910326967e-06, + "loss": 0.091, + "step": 3440 + }, + { + "epoch": 0.83, + "learning_rate": 7.308999676270638e-06, + "loss": 0.0478, + "step": 3450 + }, + { + "epoch": 0.83, + "learning_rate": 7.30090644221431e-06, + "loss": 0.0798, + "step": 3460 + }, + { + "epoch": 0.83, + "learning_rate": 7.2928132081579805e-06, + "loss": 0.0909, + "step": 3470 + }, + { + "epoch": 0.84, + "learning_rate": 7.284719974101651e-06, + "loss": 0.0662, + "step": 3480 + }, + { + "epoch": 0.84, + "learning_rate": 7.276626740045323e-06, + "loss": 0.081, + "step": 3490 + }, + { + "epoch": 0.84, + "learning_rate": 7.268533505988994e-06, + "loss": 0.0582, + "step": 3500 + }, + { + "epoch": 0.84, + "eval_loss": 0.07141905277967453, + "eval_runtime": 418.3794, + "eval_samples_per_second": 14.912, + "eval_steps_per_second": 2.486, + "step": 3500 + }, + { + "epoch": 0.84, + "learning_rate": 7.260440271932665e-06, + "loss": 0.0321, + "step": 3510 + }, + { + "epoch": 0.85, + "learning_rate": 7.252347037876336e-06, + "loss": 0.0503, + "step": 3520 + }, + { + "epoch": 0.85, + "learning_rate": 7.244253803820007e-06, + "loss": 0.1218, + "step": 3530 + }, + { + "epoch": 0.85, + "learning_rate": 7.236160569763678e-06, + "loss": 0.1038, + "step": 3540 + }, + { + "epoch": 0.85, + "learning_rate": 7.22806733570735e-06, + "loss": 0.0805, + "step": 3550 + }, + { + "epoch": 0.86, + "learning_rate": 7.2199741016510205e-06, + "loss": 0.0562, + "step": 3560 + }, + { + "epoch": 0.86, + "learning_rate": 7.211880867594691e-06, + "loss": 0.1049, + "step": 3570 + }, + { + "epoch": 0.86, + "learning_rate": 7.203787633538363e-06, + "loss": 0.0581, + "step": 3580 + }, + { + "epoch": 0.86, + "learning_rate": 7.195694399482034e-06, + "loss": 0.0768, + "step": 3590 + }, + { + "epoch": 0.87, + "learning_rate": 7.187601165425705e-06, + "loss": 0.0927, + "step": 3600 + }, + { + "epoch": 0.87, + "learning_rate": 7.179507931369376e-06, + "loss": 0.0995, + "step": 3610 + }, + { + "epoch": 0.87, + "learning_rate": 7.171414697313047e-06, + "loss": 0.0728, + "step": 3620 + }, + { + "epoch": 0.87, + "learning_rate": 7.163321463256718e-06, + "loss": 0.0643, + "step": 3630 + }, + { + "epoch": 0.88, + "learning_rate": 7.15522822920039e-06, + "loss": 0.0445, + "step": 3640 + }, + { + "epoch": 0.88, + "learning_rate": 7.1471349951440605e-06, + "loss": 0.0285, + "step": 3650 + }, + { + "epoch": 0.88, + "learning_rate": 7.139041761087731e-06, + "loss": 0.1066, + "step": 3660 + }, + { + "epoch": 0.88, + "learning_rate": 7.130948527031403e-06, + "loss": 0.0584, + "step": 3670 + }, + { + "epoch": 0.88, + "learning_rate": 7.122855292975074e-06, + "loss": 0.0826, + "step": 3680 + }, + { + "epoch": 0.89, + "learning_rate": 7.114762058918745e-06, + "loss": 0.0652, + "step": 3690 + }, + { + "epoch": 0.89, + "learning_rate": 7.106668824862416e-06, + "loss": 0.0705, + "step": 3700 + }, + { + "epoch": 0.89, + "learning_rate": 7.098575590806087e-06, + "loss": 0.1125, + "step": 3710 + }, + { + "epoch": 0.89, + "learning_rate": 7.090482356749758e-06, + "loss": 0.0496, + "step": 3720 + }, + { + "epoch": 0.9, + "learning_rate": 7.08238912269343e-06, + "loss": 0.1235, + "step": 3730 + }, + { + "epoch": 0.9, + "learning_rate": 7.074295888637101e-06, + "loss": 0.05, + "step": 3740 + }, + { + "epoch": 0.9, + "learning_rate": 7.066202654580771e-06, + "loss": 0.0946, + "step": 3750 + }, + { + "epoch": 0.9, + "learning_rate": 7.058109420524441e-06, + "loss": 0.0747, + "step": 3760 + }, + { + "epoch": 0.91, + "learning_rate": 7.050016186468113e-06, + "loss": 0.0465, + "step": 3770 + }, + { + "epoch": 0.91, + "learning_rate": 7.041922952411784e-06, + "loss": 0.0584, + "step": 3780 + }, + { + "epoch": 0.91, + "learning_rate": 7.033829718355455e-06, + "loss": 0.0869, + "step": 3790 + }, + { + "epoch": 0.91, + "learning_rate": 7.0257364842991264e-06, + "loss": 0.1663, + "step": 3800 + }, + { + "epoch": 0.92, + "learning_rate": 7.017643250242797e-06, + "loss": 0.0674, + "step": 3810 + }, + { + "epoch": 0.92, + "learning_rate": 7.009550016186468e-06, + "loss": 0.0591, + "step": 3820 + }, + { + "epoch": 0.92, + "learning_rate": 7.00145678213014e-06, + "loss": 0.066, + "step": 3830 + }, + { + "epoch": 0.92, + "learning_rate": 6.993363548073811e-06, + "loss": 0.0543, + "step": 3840 + }, + { + "epoch": 0.93, + "learning_rate": 6.9852703140174814e-06, + "loss": 0.0911, + "step": 3850 + }, + { + "epoch": 0.93, + "learning_rate": 6.977177079961152e-06, + "loss": 0.0704, + "step": 3860 + }, + { + "epoch": 0.93, + "learning_rate": 6.969083845904824e-06, + "loss": 0.0216, + "step": 3870 + }, + { + "epoch": 0.93, + "learning_rate": 6.960990611848495e-06, + "loss": 0.0337, + "step": 3880 + }, + { + "epoch": 0.94, + "learning_rate": 6.952897377792166e-06, + "loss": 0.0832, + "step": 3890 + }, + { + "epoch": 0.94, + "learning_rate": 6.944804143735837e-06, + "loss": 0.079, + "step": 3900 + }, + { + "epoch": 0.94, + "learning_rate": 6.936710909679508e-06, + "loss": 0.103, + "step": 3910 + }, + { + "epoch": 0.94, + "learning_rate": 6.928617675623179e-06, + "loss": 0.1251, + "step": 3920 + }, + { + "epoch": 0.94, + "learning_rate": 6.920524441566851e-06, + "loss": 0.0507, + "step": 3930 + }, + { + "epoch": 0.95, + "learning_rate": 6.9124312075105215e-06, + "loss": 0.0729, + "step": 3940 + }, + { + "epoch": 0.95, + "learning_rate": 6.904337973454192e-06, + "loss": 0.0672, + "step": 3950 + }, + { + "epoch": 0.95, + "learning_rate": 6.896244739397864e-06, + "loss": 0.0507, + "step": 3960 + }, + { + "epoch": 0.95, + "learning_rate": 6.888151505341535e-06, + "loss": 0.0317, + "step": 3970 + }, + { + "epoch": 0.96, + "learning_rate": 6.880058271285206e-06, + "loss": 0.1081, + "step": 3980 + }, + { + "epoch": 0.96, + "learning_rate": 6.871965037228877e-06, + "loss": 0.0791, + "step": 3990 + }, + { + "epoch": 0.96, + "learning_rate": 6.863871803172548e-06, + "loss": 0.0538, + "step": 4000 + }, + { + "epoch": 0.96, + "eval_loss": 0.062354035675525665, + "eval_runtime": 420.5379, + "eval_samples_per_second": 14.836, + "eval_steps_per_second": 2.473, + "step": 4000 + }, + { + "epoch": 0.96, + "learning_rate": 6.855778569116219e-06, + "loss": 0.0375, + "step": 4010 + }, + { + "epoch": 0.97, + "learning_rate": 6.847685335059891e-06, + "loss": 0.0496, + "step": 4020 + }, + { + "epoch": 0.97, + "learning_rate": 6.8395921010035615e-06, + "loss": 0.0513, + "step": 4030 + }, + { + "epoch": 0.97, + "learning_rate": 6.831498866947232e-06, + "loss": 0.0532, + "step": 4040 + }, + { + "epoch": 0.97, + "learning_rate": 6.823405632890904e-06, + "loss": 0.1018, + "step": 4050 + }, + { + "epoch": 0.98, + "learning_rate": 6.815312398834575e-06, + "loss": 0.1197, + "step": 4060 + }, + { + "epoch": 0.98, + "learning_rate": 6.807219164778246e-06, + "loss": 0.0292, + "step": 4070 + }, + { + "epoch": 0.98, + "learning_rate": 6.799125930721917e-06, + "loss": 0.0293, + "step": 4080 + }, + { + "epoch": 0.98, + "learning_rate": 6.791032696665588e-06, + "loss": 0.0745, + "step": 4090 + }, + { + "epoch": 0.99, + "learning_rate": 6.782939462609259e-06, + "loss": 0.0772, + "step": 4100 + }, + { + "epoch": 0.99, + "learning_rate": 6.774846228552931e-06, + "loss": 0.0682, + "step": 4110 + }, + { + "epoch": 0.99, + "learning_rate": 6.7667529944966016e-06, + "loss": 0.0241, + "step": 4120 + }, + { + "epoch": 0.99, + "learning_rate": 6.758659760440272e-06, + "loss": 0.0742, + "step": 4130 + }, + { + "epoch": 1.0, + "learning_rate": 6.750566526383944e-06, + "loss": 0.0476, + "step": 4140 + }, + { + "epoch": 1.0, + "learning_rate": 6.742473292327615e-06, + "loss": 0.0545, + "step": 4150 + }, + { + "epoch": 1.0, + "learning_rate": 6.734380058271286e-06, + "loss": 0.1159, + "step": 4160 + }, + { + "epoch": 1.0, + "learning_rate": 6.726286824214957e-06, + "loss": 0.0373, + "step": 4170 + }, + { + "epoch": 1.0, + "learning_rate": 6.718193590158628e-06, + "loss": 0.0347, + "step": 4180 + }, + { + "epoch": 1.01, + "learning_rate": 6.710100356102299e-06, + "loss": 0.077, + "step": 4190 + }, + { + "epoch": 1.01, + "learning_rate": 6.702007122045971e-06, + "loss": 0.0615, + "step": 4200 + }, + { + "epoch": 1.01, + "learning_rate": 6.693913887989642e-06, + "loss": 0.0324, + "step": 4210 + }, + { + "epoch": 1.01, + "learning_rate": 6.6858206539333124e-06, + "loss": 0.0654, + "step": 4220 + }, + { + "epoch": 1.02, + "learning_rate": 6.677727419876984e-06, + "loss": 0.0241, + "step": 4230 + }, + { + "epoch": 1.02, + "learning_rate": 6.669634185820655e-06, + "loss": 0.0492, + "step": 4240 + }, + { + "epoch": 1.02, + "learning_rate": 6.661540951764326e-06, + "loss": 0.071, + "step": 4250 + }, + { + "epoch": 1.02, + "learning_rate": 6.653447717707997e-06, + "loss": 0.0264, + "step": 4260 + }, + { + "epoch": 1.03, + "learning_rate": 6.645354483651668e-06, + "loss": 0.0806, + "step": 4270 + }, + { + "epoch": 1.03, + "learning_rate": 6.637261249595339e-06, + "loss": 0.0686, + "step": 4280 + }, + { + "epoch": 1.03, + "learning_rate": 6.62916801553901e-06, + "loss": 0.0456, + "step": 4290 + }, + { + "epoch": 1.03, + "learning_rate": 6.621074781482682e-06, + "loss": 0.0755, + "step": 4300 + }, + { + "epoch": 1.04, + "learning_rate": 6.6129815474263525e-06, + "loss": 0.0603, + "step": 4310 + }, + { + "epoch": 1.04, + "learning_rate": 6.604888313370023e-06, + "loss": 0.0633, + "step": 4320 + }, + { + "epoch": 1.04, + "learning_rate": 6.596795079313695e-06, + "loss": 0.0405, + "step": 4330 + }, + { + "epoch": 1.04, + "learning_rate": 6.588701845257366e-06, + "loss": 0.0792, + "step": 4340 + }, + { + "epoch": 1.05, + "learning_rate": 6.580608611201037e-06, + "loss": 0.0321, + "step": 4350 + }, + { + "epoch": 1.05, + "learning_rate": 6.572515377144708e-06, + "loss": 0.0656, + "step": 4360 + }, + { + "epoch": 1.05, + "learning_rate": 6.564422143088378e-06, + "loss": 0.1121, + "step": 4370 + }, + { + "epoch": 1.05, + "learning_rate": 6.556328909032049e-06, + "loss": 0.0571, + "step": 4380 + }, + { + "epoch": 1.06, + "learning_rate": 6.54823567497572e-06, + "loss": 0.0718, + "step": 4390 + }, + { + "epoch": 1.06, + "learning_rate": 6.540142440919392e-06, + "loss": 0.0872, + "step": 4400 + }, + { + "epoch": 1.06, + "learning_rate": 6.5320492068630625e-06, + "loss": 0.0857, + "step": 4410 + }, + { + "epoch": 1.06, + "learning_rate": 6.523955972806733e-06, + "loss": 0.0732, + "step": 4420 + }, + { + "epoch": 1.06, + "learning_rate": 6.515862738750405e-06, + "loss": 0.0503, + "step": 4430 + }, + { + "epoch": 1.07, + "learning_rate": 6.507769504694076e-06, + "loss": 0.0663, + "step": 4440 + }, + { + "epoch": 1.07, + "learning_rate": 6.499676270637747e-06, + "loss": 0.0379, + "step": 4450 + }, + { + "epoch": 1.07, + "learning_rate": 6.491583036581418e-06, + "loss": 0.0526, + "step": 4460 + }, + { + "epoch": 1.07, + "learning_rate": 6.483489802525089e-06, + "loss": 0.0289, + "step": 4470 + }, + { + "epoch": 1.08, + "learning_rate": 6.47539656846876e-06, + "loss": 0.0576, + "step": 4480 + }, + { + "epoch": 1.08, + "learning_rate": 6.467303334412432e-06, + "loss": 0.08, + "step": 4490 + }, + { + "epoch": 1.08, + "learning_rate": 6.4592101003561025e-06, + "loss": 0.0725, + "step": 4500 + }, + { + "epoch": 1.08, + "eval_loss": 0.06298486888408661, + "eval_runtime": 423.5912, + "eval_samples_per_second": 14.729, + "eval_steps_per_second": 2.455, + "step": 4500 + }, + { + "epoch": 1.08, + "learning_rate": 6.451116866299773e-06, + "loss": 0.0493, + "step": 4510 + }, + { + "epoch": 1.09, + "learning_rate": 6.443023632243445e-06, + "loss": 0.1188, + "step": 4520 + }, + { + "epoch": 1.09, + "learning_rate": 6.434930398187116e-06, + "loss": 0.0328, + "step": 4530 + }, + { + "epoch": 1.09, + "learning_rate": 6.426837164130787e-06, + "loss": 0.0304, + "step": 4540 + }, + { + "epoch": 1.09, + "learning_rate": 6.418743930074458e-06, + "loss": 0.0231, + "step": 4550 + }, + { + "epoch": 1.1, + "learning_rate": 6.410650696018129e-06, + "loss": 0.1232, + "step": 4560 + }, + { + "epoch": 1.1, + "learning_rate": 6.4025574619618e-06, + "loss": 0.0372, + "step": 4570 + }, + { + "epoch": 1.1, + "learning_rate": 6.394464227905472e-06, + "loss": 0.0218, + "step": 4580 + }, + { + "epoch": 1.1, + "learning_rate": 6.3863709938491426e-06, + "loss": 0.0668, + "step": 4590 + }, + { + "epoch": 1.11, + "learning_rate": 6.378277759792813e-06, + "loss": 0.0772, + "step": 4600 + }, + { + "epoch": 1.11, + "learning_rate": 6.370184525736485e-06, + "loss": 0.0436, + "step": 4610 + }, + { + "epoch": 1.11, + "learning_rate": 6.362091291680156e-06, + "loss": 0.1143, + "step": 4620 + }, + { + "epoch": 1.11, + "learning_rate": 6.353998057623827e-06, + "loss": 0.097, + "step": 4630 + }, + { + "epoch": 1.12, + "learning_rate": 6.3459048235674984e-06, + "loss": 0.0443, + "step": 4640 + }, + { + "epoch": 1.12, + "learning_rate": 6.337811589511169e-06, + "loss": 0.0512, + "step": 4650 + }, + { + "epoch": 1.12, + "learning_rate": 6.32971835545484e-06, + "loss": 0.0669, + "step": 4660 + }, + { + "epoch": 1.12, + "learning_rate": 6.321625121398512e-06, + "loss": 0.0289, + "step": 4670 + }, + { + "epoch": 1.12, + "learning_rate": 6.313531887342183e-06, + "loss": 0.0962, + "step": 4680 + }, + { + "epoch": 1.13, + "learning_rate": 6.3054386532858534e-06, + "loss": 0.0693, + "step": 4690 + }, + { + "epoch": 1.13, + "learning_rate": 6.297345419229524e-06, + "loss": 0.0607, + "step": 4700 + }, + { + "epoch": 1.13, + "learning_rate": 6.289252185173196e-06, + "loss": 0.0253, + "step": 4710 + }, + { + "epoch": 1.13, + "learning_rate": 6.281158951116867e-06, + "loss": 0.0795, + "step": 4720 + }, + { + "epoch": 1.14, + "learning_rate": 6.273065717060538e-06, + "loss": 0.0582, + "step": 4730 + }, + { + "epoch": 1.14, + "learning_rate": 6.264972483004209e-06, + "loss": 0.0502, + "step": 4740 + }, + { + "epoch": 1.14, + "learning_rate": 6.25687924894788e-06, + "loss": 0.0792, + "step": 4750 + }, + { + "epoch": 1.14, + "learning_rate": 6.248786014891551e-06, + "loss": 0.0944, + "step": 4760 + }, + { + "epoch": 1.15, + "learning_rate": 6.240692780835223e-06, + "loss": 0.048, + "step": 4770 + }, + { + "epoch": 1.15, + "learning_rate": 6.2325995467788935e-06, + "loss": 0.0969, + "step": 4780 + }, + { + "epoch": 1.15, + "learning_rate": 6.224506312722564e-06, + "loss": 0.0488, + "step": 4790 + }, + { + "epoch": 1.15, + "learning_rate": 6.216413078666236e-06, + "loss": 0.0324, + "step": 4800 + }, + { + "epoch": 1.16, + "learning_rate": 6.208319844609907e-06, + "loss": 0.0607, + "step": 4810 + }, + { + "epoch": 1.16, + "learning_rate": 6.200226610553578e-06, + "loss": 0.066, + "step": 4820 + }, + { + "epoch": 1.16, + "learning_rate": 6.192133376497249e-06, + "loss": 0.1019, + "step": 4830 + }, + { + "epoch": 1.16, + "learning_rate": 6.18404014244092e-06, + "loss": 0.0544, + "step": 4840 + }, + { + "epoch": 1.17, + "learning_rate": 6.175946908384591e-06, + "loss": 0.0604, + "step": 4850 + }, + { + "epoch": 1.17, + "learning_rate": 6.167853674328263e-06, + "loss": 0.06, + "step": 4860 + }, + { + "epoch": 1.17, + "learning_rate": 6.1597604402719335e-06, + "loss": 0.0683, + "step": 4870 + }, + { + "epoch": 1.17, + "learning_rate": 6.151667206215604e-06, + "loss": 0.0427, + "step": 4880 + }, + { + "epoch": 1.18, + "learning_rate": 6.143573972159276e-06, + "loss": 0.0667, + "step": 4890 + }, + { + "epoch": 1.18, + "learning_rate": 6.135480738102947e-06, + "loss": 0.0418, + "step": 4900 + }, + { + "epoch": 1.18, + "learning_rate": 6.127387504046618e-06, + "loss": 0.0582, + "step": 4910 + }, + { + "epoch": 1.18, + "learning_rate": 6.119294269990289e-06, + "loss": 0.0528, + "step": 4920 + }, + { + "epoch": 1.19, + "learning_rate": 6.11120103593396e-06, + "loss": 0.0269, + "step": 4930 + }, + { + "epoch": 1.19, + "learning_rate": 6.103107801877631e-06, + "loss": 0.0734, + "step": 4940 + }, + { + "epoch": 1.19, + "learning_rate": 6.095014567821303e-06, + "loss": 0.0628, + "step": 4950 + }, + { + "epoch": 1.19, + "learning_rate": 6.0869213337649736e-06, + "loss": 0.0433, + "step": 4960 + }, + { + "epoch": 1.19, + "learning_rate": 6.078828099708644e-06, + "loss": 0.0796, + "step": 4970 + }, + { + "epoch": 1.2, + "learning_rate": 6.070734865652314e-06, + "loss": 0.0393, + "step": 4980 + }, + { + "epoch": 1.2, + "learning_rate": 6.062641631595986e-06, + "loss": 0.0338, + "step": 4990 + }, + { + "epoch": 1.2, + "learning_rate": 6.054548397539657e-06, + "loss": 0.0551, + "step": 5000 + }, + { + "epoch": 1.2, + "eval_loss": 0.06391571462154388, + "eval_runtime": 422.2798, + "eval_samples_per_second": 14.775, + "eval_steps_per_second": 2.463, + "step": 5000 + }, + { + "epoch": 1.2, + "learning_rate": 6.046455163483328e-06, + "loss": 0.0382, + "step": 5010 + }, + { + "epoch": 1.21, + "learning_rate": 6.038361929426999e-06, + "loss": 0.0267, + "step": 5020 + }, + { + "epoch": 1.21, + "learning_rate": 6.03026869537067e-06, + "loss": 0.073, + "step": 5030 + }, + { + "epoch": 1.21, + "learning_rate": 6.022175461314341e-06, + "loss": 0.0461, + "step": 5040 + }, + { + "epoch": 1.21, + "learning_rate": 6.014082227258013e-06, + "loss": 0.0565, + "step": 5050 + }, + { + "epoch": 1.22, + "learning_rate": 6.005988993201684e-06, + "loss": 0.1028, + "step": 5060 + }, + { + "epoch": 1.22, + "learning_rate": 5.997895759145354e-06, + "loss": 0.0487, + "step": 5070 + }, + { + "epoch": 1.22, + "learning_rate": 5.989802525089026e-06, + "loss": 0.0595, + "step": 5080 + }, + { + "epoch": 1.22, + "learning_rate": 5.981709291032697e-06, + "loss": 0.073, + "step": 5090 + }, + { + "epoch": 1.23, + "learning_rate": 5.973616056976368e-06, + "loss": 0.0907, + "step": 5100 + }, + { + "epoch": 1.23, + "learning_rate": 5.9655228229200394e-06, + "loss": 0.0602, + "step": 5110 + }, + { + "epoch": 1.23, + "learning_rate": 5.95742958886371e-06, + "loss": 0.0659, + "step": 5120 + }, + { + "epoch": 1.23, + "learning_rate": 5.949336354807381e-06, + "loss": 0.0558, + "step": 5130 + }, + { + "epoch": 1.24, + "learning_rate": 5.941243120751052e-06, + "loss": 0.0578, + "step": 5140 + }, + { + "epoch": 1.24, + "learning_rate": 5.933149886694724e-06, + "loss": 0.0522, + "step": 5150 + }, + { + "epoch": 1.24, + "learning_rate": 5.9250566526383945e-06, + "loss": 0.0249, + "step": 5160 + }, + { + "epoch": 1.24, + "learning_rate": 5.916963418582065e-06, + "loss": 0.0463, + "step": 5170 + }, + { + "epoch": 1.25, + "learning_rate": 5.908870184525737e-06, + "loss": 0.0394, + "step": 5180 + }, + { + "epoch": 1.25, + "learning_rate": 5.900776950469408e-06, + "loss": 0.0894, + "step": 5190 + }, + { + "epoch": 1.25, + "learning_rate": 5.892683716413079e-06, + "loss": 0.0551, + "step": 5200 + }, + { + "epoch": 1.25, + "learning_rate": 5.88459048235675e-06, + "loss": 0.0413, + "step": 5210 + }, + { + "epoch": 1.25, + "learning_rate": 5.876497248300421e-06, + "loss": 0.0754, + "step": 5220 + }, + { + "epoch": 1.26, + "learning_rate": 5.868404014244092e-06, + "loss": 0.0368, + "step": 5230 + }, + { + "epoch": 1.26, + "learning_rate": 5.860310780187764e-06, + "loss": 0.0759, + "step": 5240 + }, + { + "epoch": 1.26, + "learning_rate": 5.8522175461314345e-06, + "loss": 0.0272, + "step": 5250 + }, + { + "epoch": 1.26, + "learning_rate": 5.844124312075105e-06, + "loss": 0.0471, + "step": 5260 + }, + { + "epoch": 1.27, + "learning_rate": 5.836031078018777e-06, + "loss": 0.0552, + "step": 5270 + }, + { + "epoch": 1.27, + "learning_rate": 5.827937843962448e-06, + "loss": 0.0327, + "step": 5280 + }, + { + "epoch": 1.27, + "learning_rate": 5.819844609906119e-06, + "loss": 0.0246, + "step": 5290 + }, + { + "epoch": 1.27, + "learning_rate": 5.81175137584979e-06, + "loss": 0.0409, + "step": 5300 + }, + { + "epoch": 1.28, + "learning_rate": 5.803658141793461e-06, + "loss": 0.1141, + "step": 5310 + }, + { + "epoch": 1.28, + "learning_rate": 5.795564907737132e-06, + "loss": 0.015, + "step": 5320 + }, + { + "epoch": 1.28, + "learning_rate": 5.787471673680804e-06, + "loss": 0.0513, + "step": 5330 + }, + { + "epoch": 1.28, + "learning_rate": 5.7793784396244745e-06, + "loss": 0.0619, + "step": 5340 + }, + { + "epoch": 1.29, + "learning_rate": 5.771285205568145e-06, + "loss": 0.0296, + "step": 5350 + }, + { + "epoch": 1.29, + "learning_rate": 5.763191971511817e-06, + "loss": 0.0313, + "step": 5360 + }, + { + "epoch": 1.29, + "learning_rate": 5.755098737455488e-06, + "loss": 0.0445, + "step": 5370 + }, + { + "epoch": 1.29, + "learning_rate": 5.747005503399159e-06, + "loss": 0.0267, + "step": 5380 + }, + { + "epoch": 1.3, + "learning_rate": 5.73891226934283e-06, + "loss": 0.0459, + "step": 5390 + }, + { + "epoch": 1.3, + "learning_rate": 5.730819035286501e-06, + "loss": 0.0078, + "step": 5400 + }, + { + "epoch": 1.3, + "learning_rate": 5.722725801230172e-06, + "loss": 0.1484, + "step": 5410 + }, + { + "epoch": 1.3, + "learning_rate": 5.714632567173844e-06, + "loss": 0.0942, + "step": 5420 + }, + { + "epoch": 1.31, + "learning_rate": 5.7065393331175146e-06, + "loss": 0.0504, + "step": 5430 + }, + { + "epoch": 1.31, + "learning_rate": 5.698446099061185e-06, + "loss": 0.037, + "step": 5440 + }, + { + "epoch": 1.31, + "learning_rate": 5.690352865004857e-06, + "loss": 0.0666, + "step": 5450 + }, + { + "epoch": 1.31, + "learning_rate": 5.682259630948528e-06, + "loss": 0.0476, + "step": 5460 + }, + { + "epoch": 1.31, + "learning_rate": 5.674166396892199e-06, + "loss": 0.0295, + "step": 5470 + }, + { + "epoch": 1.32, + "learning_rate": 5.6660731628358704e-06, + "loss": 0.0387, + "step": 5480 + }, + { + "epoch": 1.32, + "learning_rate": 5.657979928779541e-06, + "loss": 0.1298, + "step": 5490 + }, + { + "epoch": 1.32, + "learning_rate": 5.649886694723212e-06, + "loss": 0.065, + "step": 5500 + }, + { + "epoch": 1.32, + "eval_loss": 0.060380518436431885, + "eval_runtime": 421.7182, + "eval_samples_per_second": 14.794, + "eval_steps_per_second": 2.466, + "step": 5500 + }, + { + "epoch": 1.32, + "learning_rate": 5.641793460666884e-06, + "loss": 0.0346, + "step": 5510 + }, + { + "epoch": 1.33, + "learning_rate": 5.633700226610555e-06, + "loss": 0.0539, + "step": 5520 + }, + { + "epoch": 1.33, + "learning_rate": 5.6256069925542254e-06, + "loss": 0.0384, + "step": 5530 + }, + { + "epoch": 1.33, + "learning_rate": 5.617513758497897e-06, + "loss": 0.0466, + "step": 5540 + }, + { + "epoch": 1.33, + "learning_rate": 5.609420524441568e-06, + "loss": 0.0526, + "step": 5550 + }, + { + "epoch": 1.34, + "learning_rate": 5.601327290385239e-06, + "loss": 0.0724, + "step": 5560 + }, + { + "epoch": 1.34, + "learning_rate": 5.59323405632891e-06, + "loss": 0.0912, + "step": 5570 + }, + { + "epoch": 1.34, + "learning_rate": 5.58514082227258e-06, + "loss": 0.0925, + "step": 5580 + }, + { + "epoch": 1.34, + "learning_rate": 5.577047588216251e-06, + "loss": 0.055, + "step": 5590 + }, + { + "epoch": 1.35, + "learning_rate": 5.568954354159922e-06, + "loss": 0.0689, + "step": 5600 + }, + { + "epoch": 1.35, + "learning_rate": 5.560861120103593e-06, + "loss": 0.0406, + "step": 5610 + }, + { + "epoch": 1.35, + "learning_rate": 5.552767886047265e-06, + "loss": 0.0298, + "step": 5620 + }, + { + "epoch": 1.35, + "learning_rate": 5.5446746519909355e-06, + "loss": 0.0713, + "step": 5630 + }, + { + "epoch": 1.36, + "learning_rate": 5.536581417934606e-06, + "loss": 0.0415, + "step": 5640 + }, + { + "epoch": 1.36, + "learning_rate": 5.528488183878278e-06, + "loss": 0.0824, + "step": 5650 + }, + { + "epoch": 1.36, + "learning_rate": 5.520394949821949e-06, + "loss": 0.0334, + "step": 5660 + }, + { + "epoch": 1.36, + "learning_rate": 5.51230171576562e-06, + "loss": 0.0371, + "step": 5670 + }, + { + "epoch": 1.37, + "learning_rate": 5.504208481709291e-06, + "loss": 0.1487, + "step": 5680 + }, + { + "epoch": 1.37, + "learning_rate": 5.496115247652962e-06, + "loss": 0.0388, + "step": 5690 + }, + { + "epoch": 1.37, + "learning_rate": 5.488022013596633e-06, + "loss": 0.099, + "step": 5700 + }, + { + "epoch": 1.37, + "learning_rate": 5.479928779540305e-06, + "loss": 0.0654, + "step": 5710 + }, + { + "epoch": 1.38, + "learning_rate": 5.4718355454839755e-06, + "loss": 0.1034, + "step": 5720 + }, + { + "epoch": 1.38, + "learning_rate": 5.463742311427646e-06, + "loss": 0.0367, + "step": 5730 + }, + { + "epoch": 1.38, + "learning_rate": 5.455649077371318e-06, + "loss": 0.0567, + "step": 5740 + }, + { + "epoch": 1.38, + "learning_rate": 5.447555843314989e-06, + "loss": 0.068, + "step": 5750 + }, + { + "epoch": 1.38, + "learning_rate": 5.43946260925866e-06, + "loss": 0.0346, + "step": 5760 + }, + { + "epoch": 1.39, + "learning_rate": 5.431369375202331e-06, + "loss": 0.0562, + "step": 5770 + }, + { + "epoch": 1.39, + "learning_rate": 5.423276141146002e-06, + "loss": 0.0961, + "step": 5780 + }, + { + "epoch": 1.39, + "learning_rate": 5.415182907089673e-06, + "loss": 0.0564, + "step": 5790 + }, + { + "epoch": 1.39, + "learning_rate": 5.407089673033345e-06, + "loss": 0.0845, + "step": 5800 + }, + { + "epoch": 1.4, + "learning_rate": 5.3989964389770155e-06, + "loss": 0.0264, + "step": 5810 + }, + { + "epoch": 1.4, + "learning_rate": 5.390903204920686e-06, + "loss": 0.0952, + "step": 5820 + }, + { + "epoch": 1.4, + "learning_rate": 5.382809970864358e-06, + "loss": 0.0338, + "step": 5830 + }, + { + "epoch": 1.4, + "learning_rate": 5.374716736808029e-06, + "loss": 0.0424, + "step": 5840 + }, + { + "epoch": 1.41, + "learning_rate": 5.3666235027517e-06, + "loss": 0.0343, + "step": 5850 + }, + { + "epoch": 1.41, + "learning_rate": 5.358530268695371e-06, + "loss": 0.0154, + "step": 5860 + }, + { + "epoch": 1.41, + "learning_rate": 5.350437034639042e-06, + "loss": 0.0572, + "step": 5870 + }, + { + "epoch": 1.41, + "learning_rate": 5.342343800582713e-06, + "loss": 0.0626, + "step": 5880 + }, + { + "epoch": 1.42, + "learning_rate": 5.334250566526385e-06, + "loss": 0.0388, + "step": 5890 + }, + { + "epoch": 1.42, + "learning_rate": 5.326157332470056e-06, + "loss": 0.0502, + "step": 5900 + }, + { + "epoch": 1.42, + "learning_rate": 5.318064098413726e-06, + "loss": 0.0355, + "step": 5910 + }, + { + "epoch": 1.42, + "learning_rate": 5.309970864357398e-06, + "loss": 0.0266, + "step": 5920 + }, + { + "epoch": 1.43, + "learning_rate": 5.301877630301069e-06, + "loss": 0.0291, + "step": 5930 + }, + { + "epoch": 1.43, + "learning_rate": 5.29378439624474e-06, + "loss": 0.0969, + "step": 5940 + }, + { + "epoch": 1.43, + "learning_rate": 5.2856911621884114e-06, + "loss": 0.0716, + "step": 5950 + }, + { + "epoch": 1.43, + "learning_rate": 5.277597928132082e-06, + "loss": 0.0652, + "step": 5960 + }, + { + "epoch": 1.44, + "learning_rate": 5.269504694075753e-06, + "loss": 0.049, + "step": 5970 + }, + { + "epoch": 1.44, + "learning_rate": 5.261411460019425e-06, + "loss": 0.0579, + "step": 5980 + }, + { + "epoch": 1.44, + "learning_rate": 5.253318225963096e-06, + "loss": 0.0687, + "step": 5990 + }, + { + "epoch": 1.44, + "learning_rate": 5.2452249919067665e-06, + "loss": 0.1287, + "step": 6000 + }, + { + "epoch": 1.44, + "eval_loss": 0.05871313065290451, + "eval_runtime": 422.5328, + "eval_samples_per_second": 14.766, + "eval_steps_per_second": 2.461, + "step": 6000 + }, + { + "epoch": 1.44, + "learning_rate": 5.237131757850437e-06, + "loss": 0.0854, + "step": 6010 + }, + { + "epoch": 1.45, + "learning_rate": 5.229038523794109e-06, + "loss": 0.067, + "step": 6020 + }, + { + "epoch": 1.45, + "learning_rate": 5.22094528973778e-06, + "loss": 0.0497, + "step": 6030 + }, + { + "epoch": 1.45, + "learning_rate": 5.212852055681451e-06, + "loss": 0.0371, + "step": 6040 + }, + { + "epoch": 1.45, + "learning_rate": 5.204758821625122e-06, + "loss": 0.0543, + "step": 6050 + }, + { + "epoch": 1.46, + "learning_rate": 5.196665587568793e-06, + "loss": 0.0866, + "step": 6060 + }, + { + "epoch": 1.46, + "learning_rate": 5.188572353512464e-06, + "loss": 0.082, + "step": 6070 + }, + { + "epoch": 1.46, + "learning_rate": 5.180479119456136e-06, + "loss": 0.0578, + "step": 6080 + }, + { + "epoch": 1.46, + "learning_rate": 5.1723858853998065e-06, + "loss": 0.0379, + "step": 6090 + }, + { + "epoch": 1.47, + "learning_rate": 5.164292651343477e-06, + "loss": 0.0401, + "step": 6100 + }, + { + "epoch": 1.47, + "learning_rate": 5.156199417287149e-06, + "loss": 0.0433, + "step": 6110 + }, + { + "epoch": 1.47, + "learning_rate": 5.14810618323082e-06, + "loss": 0.0796, + "step": 6120 + }, + { + "epoch": 1.47, + "learning_rate": 5.140012949174491e-06, + "loss": 0.0806, + "step": 6130 + }, + { + "epoch": 1.48, + "learning_rate": 5.131919715118162e-06, + "loss": 0.0504, + "step": 6140 + }, + { + "epoch": 1.48, + "learning_rate": 5.123826481061833e-06, + "loss": 0.1104, + "step": 6150 + }, + { + "epoch": 1.48, + "learning_rate": 5.115733247005504e-06, + "loss": 0.091, + "step": 6160 + }, + { + "epoch": 1.48, + "learning_rate": 5.107640012949176e-06, + "loss": 0.0286, + "step": 6170 + }, + { + "epoch": 1.49, + "learning_rate": 5.0995467788928465e-06, + "loss": 0.037, + "step": 6180 + }, + { + "epoch": 1.49, + "learning_rate": 5.0914535448365165e-06, + "loss": 0.0839, + "step": 6190 + }, + { + "epoch": 1.49, + "learning_rate": 5.083360310780187e-06, + "loss": 0.0796, + "step": 6200 + }, + { + "epoch": 1.49, + "learning_rate": 5.075267076723859e-06, + "loss": 0.0865, + "step": 6210 + }, + { + "epoch": 1.5, + "learning_rate": 5.06717384266753e-06, + "loss": 0.0612, + "step": 6220 + }, + { + "epoch": 1.5, + "learning_rate": 5.059080608611201e-06, + "loss": 0.0351, + "step": 6230 + }, + { + "epoch": 1.5, + "learning_rate": 5.050987374554872e-06, + "loss": 0.0713, + "step": 6240 + }, + { + "epoch": 1.5, + "learning_rate": 5.042894140498543e-06, + "loss": 0.0394, + "step": 6250 + }, + { + "epoch": 1.5, + "learning_rate": 5.034800906442214e-06, + "loss": 0.0667, + "step": 6260 + }, + { + "epoch": 1.51, + "learning_rate": 5.026707672385886e-06, + "loss": 0.069, + "step": 6270 + }, + { + "epoch": 1.51, + "learning_rate": 5.0186144383295566e-06, + "loss": 0.0735, + "step": 6280 + }, + { + "epoch": 1.51, + "learning_rate": 5.010521204273227e-06, + "loss": 0.0531, + "step": 6290 + }, + { + "epoch": 1.51, + "learning_rate": 5.002427970216899e-06, + "loss": 0.0821, + "step": 6300 + }, + { + "epoch": 1.52, + "learning_rate": 4.994334736160571e-06, + "loss": 0.0888, + "step": 6310 + }, + { + "epoch": 1.52, + "learning_rate": 4.986241502104242e-06, + "loss": 0.0351, + "step": 6320 + }, + { + "epoch": 1.52, + "learning_rate": 4.978148268047912e-06, + "loss": 0.0539, + "step": 6330 + }, + { + "epoch": 1.52, + "learning_rate": 4.970055033991583e-06, + "loss": 0.0789, + "step": 6340 + }, + { + "epoch": 1.53, + "learning_rate": 4.961961799935254e-06, + "loss": 0.0773, + "step": 6350 + }, + { + "epoch": 1.53, + "learning_rate": 4.953868565878926e-06, + "loss": 0.0708, + "step": 6360 + }, + { + "epoch": 1.53, + "learning_rate": 4.945775331822597e-06, + "loss": 0.0292, + "step": 6370 + }, + { + "epoch": 1.53, + "learning_rate": 4.9376820977662674e-06, + "loss": 0.0275, + "step": 6380 + }, + { + "epoch": 1.54, + "learning_rate": 4.929588863709939e-06, + "loss": 0.069, + "step": 6390 + }, + { + "epoch": 1.54, + "learning_rate": 4.92149562965361e-06, + "loss": 0.0662, + "step": 6400 + }, + { + "epoch": 1.54, + "learning_rate": 4.913402395597281e-06, + "loss": 0.0576, + "step": 6410 + }, + { + "epoch": 1.54, + "learning_rate": 4.9053091615409525e-06, + "loss": 0.0424, + "step": 6420 + }, + { + "epoch": 1.55, + "learning_rate": 4.897215927484623e-06, + "loss": 0.0681, + "step": 6430 + }, + { + "epoch": 1.55, + "learning_rate": 4.889122693428294e-06, + "loss": 0.0696, + "step": 6440 + }, + { + "epoch": 1.55, + "learning_rate": 4.881029459371965e-06, + "loss": 0.0489, + "step": 6450 + }, + { + "epoch": 1.55, + "learning_rate": 4.872936225315637e-06, + "loss": 0.0202, + "step": 6460 + }, + { + "epoch": 1.56, + "learning_rate": 4.8648429912593075e-06, + "loss": 0.0902, + "step": 6470 + }, + { + "epoch": 1.56, + "learning_rate": 4.856749757202978e-06, + "loss": 0.0927, + "step": 6480 + }, + { + "epoch": 1.56, + "learning_rate": 4.84865652314665e-06, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.56, + "learning_rate": 4.840563289090321e-06, + "loss": 0.07, + "step": 6500 + }, + { + "epoch": 1.56, + "eval_loss": 0.06623455882072449, + "eval_runtime": 425.1763, + "eval_samples_per_second": 14.674, + "eval_steps_per_second": 2.446, + "step": 6500 + }, + { + "epoch": 1.56, + "learning_rate": 4.832470055033992e-06, + "loss": 0.0689, + "step": 6510 + }, + { + "epoch": 1.57, + "learning_rate": 4.824376820977663e-06, + "loss": 0.0537, + "step": 6520 + }, + { + "epoch": 1.57, + "learning_rate": 4.816283586921334e-06, + "loss": 0.0282, + "step": 6530 + }, + { + "epoch": 1.57, + "learning_rate": 4.808190352865005e-06, + "loss": 0.0445, + "step": 6540 + }, + { + "epoch": 1.57, + "learning_rate": 4.800097118808677e-06, + "loss": 0.0624, + "step": 6550 + }, + { + "epoch": 1.58, + "learning_rate": 4.7920038847523475e-06, + "loss": 0.0681, + "step": 6560 + }, + { + "epoch": 1.58, + "learning_rate": 4.783910650696018e-06, + "loss": 0.0253, + "step": 6570 + }, + { + "epoch": 1.58, + "learning_rate": 4.77581741663969e-06, + "loss": 0.0468, + "step": 6580 + }, + { + "epoch": 1.58, + "learning_rate": 4.767724182583361e-06, + "loss": 0.0389, + "step": 6590 + }, + { + "epoch": 1.59, + "learning_rate": 4.759630948527032e-06, + "loss": 0.0646, + "step": 6600 + }, + { + "epoch": 1.59, + "learning_rate": 4.751537714470703e-06, + "loss": 0.0458, + "step": 6610 + }, + { + "epoch": 1.59, + "learning_rate": 4.743444480414374e-06, + "loss": 0.0541, + "step": 6620 + }, + { + "epoch": 1.59, + "learning_rate": 4.735351246358045e-06, + "loss": 0.086, + "step": 6630 + }, + { + "epoch": 1.6, + "learning_rate": 4.727258012301716e-06, + "loss": 0.0413, + "step": 6640 + }, + { + "epoch": 1.6, + "learning_rate": 4.719164778245387e-06, + "loss": 0.1058, + "step": 6650 + }, + { + "epoch": 1.6, + "learning_rate": 4.711071544189058e-06, + "loss": 0.057, + "step": 6660 + }, + { + "epoch": 1.6, + "learning_rate": 4.702978310132729e-06, + "loss": 0.0331, + "step": 6670 + }, + { + "epoch": 1.61, + "learning_rate": 4.6948850760764e-06, + "loss": 0.063, + "step": 6680 + }, + { + "epoch": 1.61, + "learning_rate": 4.686791842020072e-06, + "loss": 0.1448, + "step": 6690 + }, + { + "epoch": 1.61, + "learning_rate": 4.6786986079637426e-06, + "loss": 0.0554, + "step": 6700 + }, + { + "epoch": 1.61, + "learning_rate": 4.670605373907413e-06, + "loss": 0.0202, + "step": 6710 + }, + { + "epoch": 1.62, + "learning_rate": 4.662512139851085e-06, + "loss": 0.0419, + "step": 6720 + }, + { + "epoch": 1.62, + "learning_rate": 4.654418905794756e-06, + "loss": 0.0634, + "step": 6730 + }, + { + "epoch": 1.62, + "learning_rate": 4.646325671738427e-06, + "loss": 0.0252, + "step": 6740 + }, + { + "epoch": 1.62, + "learning_rate": 4.638232437682098e-06, + "loss": 0.0494, + "step": 6750 + }, + { + "epoch": 1.62, + "learning_rate": 4.630139203625769e-06, + "loss": 0.0464, + "step": 6760 + }, + { + "epoch": 1.63, + "learning_rate": 4.62204596956944e-06, + "loss": 0.071, + "step": 6770 + }, + { + "epoch": 1.63, + "learning_rate": 4.613952735513112e-06, + "loss": 0.0503, + "step": 6780 + }, + { + "epoch": 1.63, + "learning_rate": 4.605859501456783e-06, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.63, + "learning_rate": 4.5977662674004534e-06, + "loss": 0.0297, + "step": 6800 + }, + { + "epoch": 1.64, + "learning_rate": 4.589673033344125e-06, + "loss": 0.0667, + "step": 6810 + }, + { + "epoch": 1.64, + "learning_rate": 4.581579799287796e-06, + "loss": 0.0284, + "step": 6820 + }, + { + "epoch": 1.64, + "learning_rate": 4.573486565231467e-06, + "loss": 0.0392, + "step": 6830 + }, + { + "epoch": 1.64, + "learning_rate": 4.5653933311751385e-06, + "loss": 0.0686, + "step": 6840 + }, + { + "epoch": 1.65, + "learning_rate": 4.557300097118809e-06, + "loss": 0.0599, + "step": 6850 + }, + { + "epoch": 1.65, + "learning_rate": 4.54920686306248e-06, + "loss": 0.0776, + "step": 6860 + }, + { + "epoch": 1.65, + "learning_rate": 4.541113629006151e-06, + "loss": 0.0654, + "step": 6870 + }, + { + "epoch": 1.65, + "learning_rate": 4.533020394949823e-06, + "loss": 0.0383, + "step": 6880 + }, + { + "epoch": 1.66, + "learning_rate": 4.5249271608934935e-06, + "loss": 0.089, + "step": 6890 + }, + { + "epoch": 1.66, + "learning_rate": 4.516833926837164e-06, + "loss": 0.0522, + "step": 6900 + }, + { + "epoch": 1.66, + "learning_rate": 4.508740692780836e-06, + "loss": 0.0414, + "step": 6910 + }, + { + "epoch": 1.66, + "learning_rate": 4.500647458724507e-06, + "loss": 0.0608, + "step": 6920 + }, + { + "epoch": 1.67, + "learning_rate": 4.492554224668178e-06, + "loss": 0.0383, + "step": 6930 + }, + { + "epoch": 1.67, + "learning_rate": 4.4844609906118485e-06, + "loss": 0.0698, + "step": 6940 + }, + { + "epoch": 1.67, + "learning_rate": 4.476367756555519e-06, + "loss": 0.0545, + "step": 6950 + }, + { + "epoch": 1.67, + "learning_rate": 4.468274522499191e-06, + "loss": 0.0969, + "step": 6960 + }, + { + "epoch": 1.68, + "learning_rate": 4.460181288442862e-06, + "loss": 0.1183, + "step": 6970 + }, + { + "epoch": 1.68, + "learning_rate": 4.452088054386533e-06, + "loss": 0.0602, + "step": 6980 + }, + { + "epoch": 1.68, + "learning_rate": 4.443994820330204e-06, + "loss": 0.076, + "step": 6990 + }, + { + "epoch": 1.68, + "learning_rate": 4.435901586273875e-06, + "loss": 0.0696, + "step": 7000 + }, + { + "epoch": 1.68, + "eval_loss": 0.06365075707435608, + "eval_runtime": 421.4563, + "eval_samples_per_second": 14.803, + "eval_steps_per_second": 2.468, + "step": 7000 + }, + { + "epoch": 1.69, + "learning_rate": 4.427808352217546e-06, + "loss": 0.0342, + "step": 7010 + }, + { + "epoch": 1.69, + "learning_rate": 4.419715118161218e-06, + "loss": 0.0384, + "step": 7020 + }, + { + "epoch": 1.69, + "learning_rate": 4.4116218841048885e-06, + "loss": 0.0539, + "step": 7030 + }, + { + "epoch": 1.69, + "learning_rate": 4.403528650048559e-06, + "loss": 0.0311, + "step": 7040 + }, + { + "epoch": 1.69, + "learning_rate": 4.395435415992231e-06, + "loss": 0.0499, + "step": 7050 + }, + { + "epoch": 1.7, + "learning_rate": 4.387342181935902e-06, + "loss": 0.0806, + "step": 7060 + }, + { + "epoch": 1.7, + "learning_rate": 4.379248947879573e-06, + "loss": 0.0477, + "step": 7070 + }, + { + "epoch": 1.7, + "learning_rate": 4.371155713823244e-06, + "loss": 0.0743, + "step": 7080 + }, + { + "epoch": 1.7, + "learning_rate": 4.363062479766915e-06, + "loss": 0.0536, + "step": 7090 + }, + { + "epoch": 1.71, + "learning_rate": 4.354969245710586e-06, + "loss": 0.0416, + "step": 7100 + }, + { + "epoch": 1.71, + "learning_rate": 4.346876011654258e-06, + "loss": 0.0444, + "step": 7110 + }, + { + "epoch": 1.71, + "learning_rate": 4.3387827775979286e-06, + "loss": 0.0477, + "step": 7120 + }, + { + "epoch": 1.71, + "learning_rate": 4.330689543541599e-06, + "loss": 0.0378, + "step": 7130 + }, + { + "epoch": 1.72, + "learning_rate": 4.322596309485271e-06, + "loss": 0.0766, + "step": 7140 + }, + { + "epoch": 1.72, + "learning_rate": 4.314503075428942e-06, + "loss": 0.0799, + "step": 7150 + }, + { + "epoch": 1.72, + "learning_rate": 4.306409841372613e-06, + "loss": 0.0544, + "step": 7160 + }, + { + "epoch": 1.72, + "learning_rate": 4.2983166073162844e-06, + "loss": 0.038, + "step": 7170 + }, + { + "epoch": 1.73, + "learning_rate": 4.290223373259955e-06, + "loss": 0.0563, + "step": 7180 + }, + { + "epoch": 1.73, + "learning_rate": 4.282130139203626e-06, + "loss": 0.0489, + "step": 7190 + }, + { + "epoch": 1.73, + "learning_rate": 4.274036905147298e-06, + "loss": 0.1399, + "step": 7200 + }, + { + "epoch": 1.73, + "learning_rate": 4.265943671090969e-06, + "loss": 0.0635, + "step": 7210 + }, + { + "epoch": 1.74, + "learning_rate": 4.2578504370346394e-06, + "loss": 0.096, + "step": 7220 + }, + { + "epoch": 1.74, + "learning_rate": 4.249757202978311e-06, + "loss": 0.0733, + "step": 7230 + }, + { + "epoch": 1.74, + "learning_rate": 4.241663968921982e-06, + "loss": 0.0199, + "step": 7240 + }, + { + "epoch": 1.74, + "learning_rate": 4.233570734865653e-06, + "loss": 0.0771, + "step": 7250 + }, + { + "epoch": 1.75, + "learning_rate": 4.225477500809324e-06, + "loss": 0.0161, + "step": 7260 + }, + { + "epoch": 1.75, + "learning_rate": 4.2173842667529944e-06, + "loss": 0.0533, + "step": 7270 + }, + { + "epoch": 1.75, + "learning_rate": 4.209291032696666e-06, + "loss": 0.0657, + "step": 7280 + }, + { + "epoch": 1.75, + "learning_rate": 4.201197798640337e-06, + "loss": 0.0864, + "step": 7290 + }, + { + "epoch": 1.75, + "learning_rate": 4.193104564584008e-06, + "loss": 0.0785, + "step": 7300 + }, + { + "epoch": 1.76, + "learning_rate": 4.185011330527679e-06, + "loss": 0.0581, + "step": 7310 + }, + { + "epoch": 1.76, + "learning_rate": 4.17691809647135e-06, + "loss": 0.0191, + "step": 7320 + }, + { + "epoch": 1.76, + "learning_rate": 4.168824862415021e-06, + "loss": 0.0368, + "step": 7330 + }, + { + "epoch": 1.76, + "learning_rate": 4.160731628358692e-06, + "loss": 0.038, + "step": 7340 + }, + { + "epoch": 1.77, + "learning_rate": 4.152638394302364e-06, + "loss": 0.059, + "step": 7350 + }, + { + "epoch": 1.77, + "learning_rate": 4.1445451602460345e-06, + "loss": 0.0383, + "step": 7360 + }, + { + "epoch": 1.77, + "learning_rate": 4.136451926189705e-06, + "loss": 0.0275, + "step": 7370 + }, + { + "epoch": 1.77, + "learning_rate": 4.128358692133377e-06, + "loss": 0.0698, + "step": 7380 + }, + { + "epoch": 1.78, + "learning_rate": 4.120265458077048e-06, + "loss": 0.0537, + "step": 7390 + }, + { + "epoch": 1.78, + "learning_rate": 4.112172224020719e-06, + "loss": 0.0629, + "step": 7400 + }, + { + "epoch": 1.78, + "learning_rate": 4.10407898996439e-06, + "loss": 0.0461, + "step": 7410 + }, + { + "epoch": 1.78, + "learning_rate": 4.095985755908061e-06, + "loss": 0.073, + "step": 7420 + }, + { + "epoch": 1.79, + "learning_rate": 4.087892521851732e-06, + "loss": 0.0601, + "step": 7430 + }, + { + "epoch": 1.79, + "learning_rate": 4.079799287795404e-06, + "loss": 0.0704, + "step": 7440 + }, + { + "epoch": 1.79, + "learning_rate": 4.0717060537390745e-06, + "loss": 0.0685, + "step": 7450 + }, + { + "epoch": 1.79, + "learning_rate": 4.063612819682745e-06, + "loss": 0.0378, + "step": 7460 + }, + { + "epoch": 1.8, + "learning_rate": 4.055519585626417e-06, + "loss": 0.0568, + "step": 7470 + }, + { + "epoch": 1.8, + "learning_rate": 4.047426351570088e-06, + "loss": 0.0474, + "step": 7480 + }, + { + "epoch": 1.8, + "learning_rate": 4.039333117513759e-06, + "loss": 0.0264, + "step": 7490 + }, + { + "epoch": 1.8, + "learning_rate": 4.03123988345743e-06, + "loss": 0.0395, + "step": 7500 + }, + { + "epoch": 1.8, + "eval_loss": 0.05503068119287491, + "eval_runtime": 420.8663, + "eval_samples_per_second": 14.824, + "eval_steps_per_second": 2.471, + "step": 7500 + }, + { + "epoch": 1.81, + "learning_rate": 4.023146649401101e-06, + "loss": 0.0627, + "step": 7510 + }, + { + "epoch": 1.81, + "learning_rate": 4.015053415344772e-06, + "loss": 0.1148, + "step": 7520 + }, + { + "epoch": 1.81, + "learning_rate": 4.006960181288444e-06, + "loss": 0.0942, + "step": 7530 + }, + { + "epoch": 1.81, + "learning_rate": 3.9988669472321146e-06, + "loss": 0.0564, + "step": 7540 + }, + { + "epoch": 1.81, + "learning_rate": 3.990773713175785e-06, + "loss": 0.0268, + "step": 7550 + }, + { + "epoch": 1.82, + "learning_rate": 3.982680479119456e-06, + "loss": 0.0532, + "step": 7560 + }, + { + "epoch": 1.82, + "learning_rate": 3.974587245063127e-06, + "loss": 0.0496, + "step": 7570 + }, + { + "epoch": 1.82, + "learning_rate": 3.966494011006799e-06, + "loss": 0.0749, + "step": 7580 + }, + { + "epoch": 1.82, + "learning_rate": 3.95840077695047e-06, + "loss": 0.0598, + "step": 7590 + }, + { + "epoch": 1.83, + "learning_rate": 3.95030754289414e-06, + "loss": 0.0352, + "step": 7600 + }, + { + "epoch": 1.83, + "learning_rate": 3.942214308837812e-06, + "loss": 0.0369, + "step": 7610 + }, + { + "epoch": 1.83, + "learning_rate": 3.934121074781483e-06, + "loss": 0.0853, + "step": 7620 + }, + { + "epoch": 1.83, + "learning_rate": 3.926027840725154e-06, + "loss": 0.0183, + "step": 7630 + }, + { + "epoch": 1.84, + "learning_rate": 3.9179346066688254e-06, + "loss": 0.0534, + "step": 7640 + }, + { + "epoch": 1.84, + "learning_rate": 3.909841372612496e-06, + "loss": 0.0357, + "step": 7650 + }, + { + "epoch": 1.84, + "learning_rate": 3.901748138556167e-06, + "loss": 0.0346, + "step": 7660 + }, + { + "epoch": 1.84, + "learning_rate": 3.893654904499839e-06, + "loss": 0.0632, + "step": 7670 + }, + { + "epoch": 1.85, + "learning_rate": 3.88556167044351e-06, + "loss": 0.0799, + "step": 7680 + }, + { + "epoch": 1.85, + "learning_rate": 3.8774684363871805e-06, + "loss": 0.0709, + "step": 7690 + }, + { + "epoch": 1.85, + "learning_rate": 3.869375202330852e-06, + "loss": 0.0475, + "step": 7700 + }, + { + "epoch": 1.85, + "learning_rate": 3.861281968274523e-06, + "loss": 0.0478, + "step": 7710 + }, + { + "epoch": 1.86, + "learning_rate": 3.853188734218194e-06, + "loss": 0.0281, + "step": 7720 + }, + { + "epoch": 1.86, + "learning_rate": 3.8450955001618655e-06, + "loss": 0.0493, + "step": 7730 + }, + { + "epoch": 1.86, + "learning_rate": 3.837002266105536e-06, + "loss": 0.0141, + "step": 7740 + }, + { + "epoch": 1.86, + "learning_rate": 3.828909032049207e-06, + "loss": 0.0184, + "step": 7750 + }, + { + "epoch": 1.87, + "learning_rate": 3.820815797992878e-06, + "loss": 0.0905, + "step": 7760 + }, + { + "epoch": 1.87, + "learning_rate": 3.8127225639365497e-06, + "loss": 0.0583, + "step": 7770 + }, + { + "epoch": 1.87, + "learning_rate": 3.8046293298802205e-06, + "loss": 0.0186, + "step": 7780 + }, + { + "epoch": 1.87, + "learning_rate": 3.7965360958238917e-06, + "loss": 0.0486, + "step": 7790 + }, + { + "epoch": 1.88, + "learning_rate": 3.788442861767563e-06, + "loss": 0.0567, + "step": 7800 + }, + { + "epoch": 1.88, + "learning_rate": 3.780349627711234e-06, + "loss": 0.0904, + "step": 7810 + }, + { + "epoch": 1.88, + "learning_rate": 3.772256393654905e-06, + "loss": 0.0301, + "step": 7820 + }, + { + "epoch": 1.88, + "learning_rate": 3.7641631595985763e-06, + "loss": 0.064, + "step": 7830 + }, + { + "epoch": 1.88, + "learning_rate": 3.756069925542247e-06, + "loss": 0.0224, + "step": 7840 + }, + { + "epoch": 1.89, + "learning_rate": 3.747976691485918e-06, + "loss": 0.0837, + "step": 7850 + }, + { + "epoch": 1.89, + "learning_rate": 3.739883457429589e-06, + "loss": 0.058, + "step": 7860 + }, + { + "epoch": 1.89, + "learning_rate": 3.73179022337326e-06, + "loss": 0.0503, + "step": 7870 + }, + { + "epoch": 1.89, + "learning_rate": 3.723696989316931e-06, + "loss": 0.0421, + "step": 7880 + }, + { + "epoch": 1.9, + "learning_rate": 3.715603755260602e-06, + "loss": 0.0227, + "step": 7890 + }, + { + "epoch": 1.9, + "learning_rate": 3.7075105212042735e-06, + "loss": 0.0897, + "step": 7900 + }, + { + "epoch": 1.9, + "learning_rate": 3.6994172871479443e-06, + "loss": 0.0251, + "step": 7910 + }, + { + "epoch": 1.9, + "learning_rate": 3.6913240530916155e-06, + "loss": 0.0472, + "step": 7920 + }, + { + "epoch": 1.91, + "learning_rate": 3.683230819035287e-06, + "loss": 0.057, + "step": 7930 + }, + { + "epoch": 1.91, + "learning_rate": 3.6751375849789576e-06, + "loss": 0.0341, + "step": 7940 + }, + { + "epoch": 1.91, + "learning_rate": 3.667044350922629e-06, + "loss": 0.032, + "step": 7950 + }, + { + "epoch": 1.91, + "learning_rate": 3.6589511168663e-06, + "loss": 0.0716, + "step": 7960 + }, + { + "epoch": 1.92, + "learning_rate": 3.650857882809971e-06, + "loss": 0.0776, + "step": 7970 + }, + { + "epoch": 1.92, + "learning_rate": 3.6427646487536422e-06, + "loss": 0.0638, + "step": 7980 + }, + { + "epoch": 1.92, + "learning_rate": 3.6346714146973135e-06, + "loss": 0.0467, + "step": 7990 + }, + { + "epoch": 1.92, + "learning_rate": 3.6265781806409843e-06, + "loss": 0.0847, + "step": 8000 + }, + { + "epoch": 1.92, + "eval_loss": 0.05776594951748848, + "eval_runtime": 419.9667, + "eval_samples_per_second": 14.856, + "eval_steps_per_second": 2.476, + "step": 8000 + }, + { + "epoch": 1.93, + "learning_rate": 3.6184849465846556e-06, + "loss": 0.0776, + "step": 8010 + }, + { + "epoch": 1.93, + "learning_rate": 3.610391712528327e-06, + "loss": 0.0423, + "step": 8020 + }, + { + "epoch": 1.93, + "learning_rate": 3.6022984784719977e-06, + "loss": 0.0674, + "step": 8030 + }, + { + "epoch": 1.93, + "learning_rate": 3.594205244415669e-06, + "loss": 0.1314, + "step": 8040 + }, + { + "epoch": 1.94, + "learning_rate": 3.58611201035934e-06, + "loss": 0.0338, + "step": 8050 + }, + { + "epoch": 1.94, + "learning_rate": 3.578018776303011e-06, + "loss": 0.0743, + "step": 8060 + }, + { + "epoch": 1.94, + "learning_rate": 3.5699255422466823e-06, + "loss": 0.0898, + "step": 8070 + }, + { + "epoch": 1.94, + "learning_rate": 3.561832308190353e-06, + "loss": 0.0677, + "step": 8080 + }, + { + "epoch": 1.94, + "learning_rate": 3.5537390741340244e-06, + "loss": 0.0241, + "step": 8090 + }, + { + "epoch": 1.95, + "learning_rate": 3.5456458400776956e-06, + "loss": 0.048, + "step": 8100 + }, + { + "epoch": 1.95, + "learning_rate": 3.5375526060213665e-06, + "loss": 0.0368, + "step": 8110 + }, + { + "epoch": 1.95, + "learning_rate": 3.5294593719650377e-06, + "loss": 0.042, + "step": 8120 + }, + { + "epoch": 1.95, + "learning_rate": 3.521366137908709e-06, + "loss": 0.034, + "step": 8130 + }, + { + "epoch": 1.96, + "learning_rate": 3.51327290385238e-06, + "loss": 0.0329, + "step": 8140 + }, + { + "epoch": 1.96, + "learning_rate": 3.505179669796051e-06, + "loss": 0.0327, + "step": 8150 + }, + { + "epoch": 1.96, + "learning_rate": 3.4970864357397215e-06, + "loss": 0.0725, + "step": 8160 + }, + { + "epoch": 1.96, + "learning_rate": 3.4889932016833927e-06, + "loss": 0.0155, + "step": 8170 + }, + { + "epoch": 1.97, + "learning_rate": 3.480899967627064e-06, + "loss": 0.0364, + "step": 8180 + }, + { + "epoch": 1.97, + "learning_rate": 3.472806733570735e-06, + "loss": 0.0293, + "step": 8190 + }, + { + "epoch": 1.97, + "learning_rate": 3.464713499514406e-06, + "loss": 0.0474, + "step": 8200 + }, + { + "epoch": 1.97, + "learning_rate": 3.4566202654580773e-06, + "loss": 0.0385, + "step": 8210 + }, + { + "epoch": 1.98, + "learning_rate": 3.448527031401748e-06, + "loss": 0.049, + "step": 8220 + }, + { + "epoch": 1.98, + "learning_rate": 3.4404337973454194e-06, + "loss": 0.0733, + "step": 8230 + }, + { + "epoch": 1.98, + "learning_rate": 3.4323405632890907e-06, + "loss": 0.0627, + "step": 8240 + }, + { + "epoch": 1.98, + "learning_rate": 3.4242473292327615e-06, + "loss": 0.0517, + "step": 8250 + }, + { + "epoch": 1.99, + "learning_rate": 3.4161540951764328e-06, + "loss": 0.0784, + "step": 8260 + }, + { + "epoch": 1.99, + "learning_rate": 3.408060861120104e-06, + "loss": 0.03, + "step": 8270 + }, + { + "epoch": 1.99, + "learning_rate": 3.399967627063775e-06, + "loss": 0.0475, + "step": 8280 + }, + { + "epoch": 1.99, + "learning_rate": 3.391874393007446e-06, + "loss": 0.0269, + "step": 8290 + }, + { + "epoch": 2.0, + "learning_rate": 3.383781158951117e-06, + "loss": 0.0631, + "step": 8300 + }, + { + "epoch": 2.0, + "learning_rate": 3.375687924894788e-06, + "loss": 0.0298, + "step": 8310 + }, + { + "epoch": 2.0, + "learning_rate": 3.3675946908384595e-06, + "loss": 0.0735, + "step": 8320 + }, + { + "epoch": 2.0, + "learning_rate": 3.3595014567821303e-06, + "loss": 0.0316, + "step": 8330 + }, + { + "epoch": 2.0, + "learning_rate": 3.3514082227258015e-06, + "loss": 0.0923, + "step": 8340 + }, + { + "epoch": 2.01, + "learning_rate": 3.343314988669473e-06, + "loss": 0.0292, + "step": 8350 + }, + { + "epoch": 2.01, + "learning_rate": 3.3352217546131436e-06, + "loss": 0.0544, + "step": 8360 + }, + { + "epoch": 2.01, + "learning_rate": 3.327128520556815e-06, + "loss": 0.033, + "step": 8370 + }, + { + "epoch": 2.01, + "learning_rate": 3.319035286500486e-06, + "loss": 0.0477, + "step": 8380 + }, + { + "epoch": 2.02, + "learning_rate": 3.310942052444157e-06, + "loss": 0.0361, + "step": 8390 + }, + { + "epoch": 2.02, + "learning_rate": 3.3028488183878282e-06, + "loss": 0.0755, + "step": 8400 + }, + { + "epoch": 2.02, + "learning_rate": 3.2947555843314995e-06, + "loss": 0.0704, + "step": 8410 + }, + { + "epoch": 2.02, + "learning_rate": 3.2866623502751703e-06, + "loss": 0.0622, + "step": 8420 + }, + { + "epoch": 2.03, + "learning_rate": 3.2785691162188416e-06, + "loss": 0.0602, + "step": 8430 + }, + { + "epoch": 2.03, + "learning_rate": 3.270475882162513e-06, + "loss": 0.0669, + "step": 8440 + }, + { + "epoch": 2.03, + "learning_rate": 3.2623826481061837e-06, + "loss": 0.1246, + "step": 8450 + }, + { + "epoch": 2.03, + "learning_rate": 3.2542894140498545e-06, + "loss": 0.0573, + "step": 8460 + }, + { + "epoch": 2.04, + "learning_rate": 3.2461961799935253e-06, + "loss": 0.0502, + "step": 8470 + }, + { + "epoch": 2.04, + "learning_rate": 3.2381029459371966e-06, + "loss": 0.0629, + "step": 8480 + }, + { + "epoch": 2.04, + "learning_rate": 3.230009711880868e-06, + "loss": 0.0653, + "step": 8490 + }, + { + "epoch": 2.04, + "learning_rate": 3.2219164778245387e-06, + "loss": 0.0619, + "step": 8500 + }, + { + "epoch": 2.04, + "eval_loss": 0.05417681485414505, + "eval_runtime": 419.908, + "eval_samples_per_second": 14.858, + "eval_steps_per_second": 2.477, + "step": 8500 + }, + { + "epoch": 2.05, + "learning_rate": 3.21382324376821e-06, + "loss": 0.0128, + "step": 8510 + }, + { + "epoch": 2.05, + "learning_rate": 3.2057300097118808e-06, + "loss": 0.0547, + "step": 8520 + }, + { + "epoch": 2.05, + "learning_rate": 3.197636775655552e-06, + "loss": 0.037, + "step": 8530 + }, + { + "epoch": 2.05, + "learning_rate": 3.1895435415992233e-06, + "loss": 0.0507, + "step": 8540 + }, + { + "epoch": 2.06, + "learning_rate": 3.181450307542894e-06, + "loss": 0.0464, + "step": 8550 + }, + { + "epoch": 2.06, + "learning_rate": 3.1733570734865654e-06, + "loss": 0.0226, + "step": 8560 + }, + { + "epoch": 2.06, + "learning_rate": 3.1652638394302366e-06, + "loss": 0.0486, + "step": 8570 + }, + { + "epoch": 2.06, + "learning_rate": 3.1571706053739075e-06, + "loss": 0.0426, + "step": 8580 + }, + { + "epoch": 2.06, + "learning_rate": 3.1490773713175787e-06, + "loss": 0.0512, + "step": 8590 + }, + { + "epoch": 2.07, + "learning_rate": 3.14098413726125e-06, + "loss": 0.0682, + "step": 8600 + }, + { + "epoch": 2.07, + "learning_rate": 3.132890903204921e-06, + "loss": 0.0854, + "step": 8610 + }, + { + "epoch": 2.07, + "learning_rate": 3.124797669148592e-06, + "loss": 0.0305, + "step": 8620 + }, + { + "epoch": 2.07, + "learning_rate": 3.1167044350922633e-06, + "loss": 0.0462, + "step": 8630 + }, + { + "epoch": 2.08, + "learning_rate": 3.108611201035934e-06, + "loss": 0.056, + "step": 8640 + }, + { + "epoch": 2.08, + "learning_rate": 3.1005179669796054e-06, + "loss": 0.0322, + "step": 8650 + }, + { + "epoch": 2.08, + "learning_rate": 3.0924247329232767e-06, + "loss": 0.034, + "step": 8660 + }, + { + "epoch": 2.08, + "learning_rate": 3.0843314988669475e-06, + "loss": 0.0381, + "step": 8670 + }, + { + "epoch": 2.09, + "learning_rate": 3.0762382648106188e-06, + "loss": 0.0242, + "step": 8680 + }, + { + "epoch": 2.09, + "learning_rate": 3.06814503075429e-06, + "loss": 0.0278, + "step": 8690 + }, + { + "epoch": 2.09, + "learning_rate": 3.060051796697961e-06, + "loss": 0.0292, + "step": 8700 + }, + { + "epoch": 2.09, + "learning_rate": 3.051958562641632e-06, + "loss": 0.0189, + "step": 8710 + }, + { + "epoch": 2.1, + "learning_rate": 3.043865328585303e-06, + "loss": 0.0176, + "step": 8720 + }, + { + "epoch": 2.1, + "learning_rate": 3.035772094528974e-06, + "loss": 0.0662, + "step": 8730 + }, + { + "epoch": 2.1, + "learning_rate": 3.0276788604726455e-06, + "loss": 0.0554, + "step": 8740 + }, + { + "epoch": 2.1, + "learning_rate": 3.0195856264163163e-06, + "loss": 0.0749, + "step": 8750 + }, + { + "epoch": 2.11, + "learning_rate": 3.011492392359987e-06, + "loss": 0.0604, + "step": 8760 + }, + { + "epoch": 2.11, + "learning_rate": 3.003399158303658e-06, + "loss": 0.0231, + "step": 8770 + }, + { + "epoch": 2.11, + "learning_rate": 2.995305924247329e-06, + "loss": 0.0374, + "step": 8780 + }, + { + "epoch": 2.11, + "learning_rate": 2.9872126901910005e-06, + "loss": 0.0246, + "step": 8790 + }, + { + "epoch": 2.12, + "learning_rate": 2.9791194561346713e-06, + "loss": 0.0416, + "step": 8800 + }, + { + "epoch": 2.12, + "learning_rate": 2.9710262220783426e-06, + "loss": 0.0318, + "step": 8810 + }, + { + "epoch": 2.12, + "learning_rate": 2.962932988022014e-06, + "loss": 0.0543, + "step": 8820 + }, + { + "epoch": 2.12, + "learning_rate": 2.9548397539656846e-06, + "loss": 0.0128, + "step": 8830 + }, + { + "epoch": 2.12, + "learning_rate": 2.946746519909356e-06, + "loss": 0.0213, + "step": 8840 + }, + { + "epoch": 2.13, + "learning_rate": 2.938653285853027e-06, + "loss": 0.0349, + "step": 8850 + }, + { + "epoch": 2.13, + "learning_rate": 2.930560051796698e-06, + "loss": 0.015, + "step": 8860 + }, + { + "epoch": 2.13, + "learning_rate": 2.9224668177403692e-06, + "loss": 0.0576, + "step": 8870 + }, + { + "epoch": 2.13, + "learning_rate": 2.9143735836840405e-06, + "loss": 0.0612, + "step": 8880 + }, + { + "epoch": 2.14, + "learning_rate": 2.9062803496277113e-06, + "loss": 0.0291, + "step": 8890 + }, + { + "epoch": 2.14, + "learning_rate": 2.8981871155713826e-06, + "loss": 0.0548, + "step": 8900 + }, + { + "epoch": 2.14, + "learning_rate": 2.890093881515054e-06, + "loss": 0.0102, + "step": 8910 + }, + { + "epoch": 2.14, + "learning_rate": 2.8820006474587247e-06, + "loss": 0.0655, + "step": 8920 + }, + { + "epoch": 2.15, + "learning_rate": 2.873907413402396e-06, + "loss": 0.0401, + "step": 8930 + }, + { + "epoch": 2.15, + "learning_rate": 2.8658141793460668e-06, + "loss": 0.0465, + "step": 8940 + }, + { + "epoch": 2.15, + "learning_rate": 2.857720945289738e-06, + "loss": 0.0634, + "step": 8950 + }, + { + "epoch": 2.15, + "learning_rate": 2.8496277112334093e-06, + "loss": 0.0129, + "step": 8960 + }, + { + "epoch": 2.16, + "learning_rate": 2.84153447717708e-06, + "loss": 0.119, + "step": 8970 + }, + { + "epoch": 2.16, + "learning_rate": 2.8334412431207514e-06, + "loss": 0.0495, + "step": 8980 + }, + { + "epoch": 2.16, + "learning_rate": 2.8253480090644226e-06, + "loss": 0.023, + "step": 8990 + }, + { + "epoch": 2.16, + "learning_rate": 2.8172547750080935e-06, + "loss": 0.0359, + "step": 9000 + }, + { + "epoch": 2.16, + "eval_loss": 0.055740032345056534, + "eval_runtime": 424.4643, + "eval_samples_per_second": 14.699, + "eval_steps_per_second": 2.45, + "step": 9000 + }, + { + "epoch": 2.17, + "learning_rate": 2.8091615409517647e-06, + "loss": 0.03, + "step": 9010 + }, + { + "epoch": 2.17, + "learning_rate": 2.801068306895436e-06, + "loss": 0.0264, + "step": 9020 + }, + { + "epoch": 2.17, + "learning_rate": 2.792975072839107e-06, + "loss": 0.0503, + "step": 9030 + }, + { + "epoch": 2.17, + "learning_rate": 2.784881838782778e-06, + "loss": 0.0524, + "step": 9040 + }, + { + "epoch": 2.18, + "learning_rate": 2.7767886047264493e-06, + "loss": 0.0188, + "step": 9050 + }, + { + "epoch": 2.18, + "learning_rate": 2.7686953706701197e-06, + "loss": 0.0353, + "step": 9060 + }, + { + "epoch": 2.18, + "learning_rate": 2.760602136613791e-06, + "loss": 0.0275, + "step": 9070 + }, + { + "epoch": 2.18, + "learning_rate": 2.752508902557462e-06, + "loss": 0.0241, + "step": 9080 + }, + { + "epoch": 2.19, + "learning_rate": 2.744415668501133e-06, + "loss": 0.0504, + "step": 9090 + }, + { + "epoch": 2.19, + "learning_rate": 2.7363224344448043e-06, + "loss": 0.0317, + "step": 9100 + }, + { + "epoch": 2.19, + "learning_rate": 2.728229200388475e-06, + "loss": 0.0395, + "step": 9110 + }, + { + "epoch": 2.19, + "learning_rate": 2.7201359663321464e-06, + "loss": 0.0673, + "step": 9120 + }, + { + "epoch": 2.19, + "learning_rate": 2.7120427322758177e-06, + "loss": 0.0977, + "step": 9130 + }, + { + "epoch": 2.2, + "learning_rate": 2.7039494982194885e-06, + "loss": 0.0232, + "step": 9140 + }, + { + "epoch": 2.2, + "learning_rate": 2.6958562641631598e-06, + "loss": 0.0813, + "step": 9150 + }, + { + "epoch": 2.2, + "learning_rate": 2.6877630301068306e-06, + "loss": 0.057, + "step": 9160 + }, + { + "epoch": 2.2, + "learning_rate": 2.679669796050502e-06, + "loss": 0.0599, + "step": 9170 + }, + { + "epoch": 2.21, + "learning_rate": 2.671576561994173e-06, + "loss": 0.0184, + "step": 9180 + }, + { + "epoch": 2.21, + "learning_rate": 2.663483327937844e-06, + "loss": 0.0408, + "step": 9190 + }, + { + "epoch": 2.21, + "learning_rate": 2.655390093881515e-06, + "loss": 0.0318, + "step": 9200 + }, + { + "epoch": 2.21, + "learning_rate": 2.6472968598251865e-06, + "loss": 0.0325, + "step": 9210 + }, + { + "epoch": 2.22, + "learning_rate": 2.6392036257688573e-06, + "loss": 0.0316, + "step": 9220 + }, + { + "epoch": 2.22, + "learning_rate": 2.6311103917125286e-06, + "loss": 0.0534, + "step": 9230 + }, + { + "epoch": 2.22, + "learning_rate": 2.6230171576562e-06, + "loss": 0.0278, + "step": 9240 + }, + { + "epoch": 2.22, + "learning_rate": 2.6149239235998706e-06, + "loss": 0.0131, + "step": 9250 + }, + { + "epoch": 2.23, + "learning_rate": 2.606830689543542e-06, + "loss": 0.0922, + "step": 9260 + }, + { + "epoch": 2.23, + "learning_rate": 2.598737455487213e-06, + "loss": 0.0545, + "step": 9270 + }, + { + "epoch": 2.23, + "learning_rate": 2.590644221430884e-06, + "loss": 0.036, + "step": 9280 + }, + { + "epoch": 2.23, + "learning_rate": 2.5825509873745553e-06, + "loss": 0.0464, + "step": 9290 + }, + { + "epoch": 2.24, + "learning_rate": 2.5744577533182265e-06, + "loss": 0.0464, + "step": 9300 + }, + { + "epoch": 2.24, + "learning_rate": 2.5663645192618973e-06, + "loss": 0.0742, + "step": 9310 + }, + { + "epoch": 2.24, + "learning_rate": 2.5582712852055686e-06, + "loss": 0.0433, + "step": 9320 + }, + { + "epoch": 2.24, + "learning_rate": 2.55017805114924e-06, + "loss": 0.0383, + "step": 9330 + }, + { + "epoch": 2.25, + "learning_rate": 2.5420848170929107e-06, + "loss": 0.0927, + "step": 9340 + }, + { + "epoch": 2.25, + "learning_rate": 2.533991583036582e-06, + "loss": 0.0426, + "step": 9350 + }, + { + "epoch": 2.25, + "learning_rate": 2.525898348980253e-06, + "loss": 0.0205, + "step": 9360 + }, + { + "epoch": 2.25, + "learning_rate": 2.5178051149239236e-06, + "loss": 0.074, + "step": 9370 + }, + { + "epoch": 2.25, + "learning_rate": 2.5097118808675944e-06, + "loss": 0.0595, + "step": 9380 + }, + { + "epoch": 2.26, + "learning_rate": 2.5016186468112657e-06, + "loss": 0.0529, + "step": 9390 + }, + { + "epoch": 2.26, + "learning_rate": 2.493525412754937e-06, + "loss": 0.0288, + "step": 9400 + }, + { + "epoch": 2.26, + "learning_rate": 2.4854321786986082e-06, + "loss": 0.0928, + "step": 9410 + }, + { + "epoch": 2.26, + "learning_rate": 2.4773389446422795e-06, + "loss": 0.0521, + "step": 9420 + }, + { + "epoch": 2.27, + "learning_rate": 2.4692457105859503e-06, + "loss": 0.0226, + "step": 9430 + }, + { + "epoch": 2.27, + "learning_rate": 2.461152476529621e-06, + "loss": 0.0657, + "step": 9440 + }, + { + "epoch": 2.27, + "learning_rate": 2.4530592424732924e-06, + "loss": 0.0642, + "step": 9450 + }, + { + "epoch": 2.27, + "learning_rate": 2.4449660084169636e-06, + "loss": 0.0337, + "step": 9460 + }, + { + "epoch": 2.28, + "learning_rate": 2.4368727743606345e-06, + "loss": 0.0866, + "step": 9470 + }, + { + "epoch": 2.28, + "learning_rate": 2.4287795403043057e-06, + "loss": 0.0831, + "step": 9480 + }, + { + "epoch": 2.28, + "learning_rate": 2.420686306247977e-06, + "loss": 0.0633, + "step": 9490 + }, + { + "epoch": 2.28, + "learning_rate": 2.412593072191648e-06, + "loss": 0.0436, + "step": 9500 + }, + { + "epoch": 2.28, + "eval_loss": 0.051913052797317505, + "eval_runtime": 425.0656, + "eval_samples_per_second": 14.678, + "eval_steps_per_second": 2.447, + "step": 9500 + }, + { + "epoch": 2.29, + "learning_rate": 2.404499838135319e-06, + "loss": 0.0245, + "step": 9510 + }, + { + "epoch": 2.29, + "learning_rate": 2.3964066040789903e-06, + "loss": 0.0476, + "step": 9520 + }, + { + "epoch": 2.29, + "learning_rate": 2.388313370022661e-06, + "loss": 0.0921, + "step": 9530 + }, + { + "epoch": 2.29, + "learning_rate": 2.3802201359663324e-06, + "loss": 0.0519, + "step": 9540 + }, + { + "epoch": 2.3, + "learning_rate": 2.3721269019100037e-06, + "loss": 0.0207, + "step": 9550 + }, + { + "epoch": 2.3, + "learning_rate": 2.3640336678536745e-06, + "loss": 0.0676, + "step": 9560 + }, + { + "epoch": 2.3, + "learning_rate": 2.3559404337973458e-06, + "loss": 0.0178, + "step": 9570 + }, + { + "epoch": 2.3, + "learning_rate": 2.347847199741017e-06, + "loss": 0.033, + "step": 9580 + }, + { + "epoch": 2.31, + "learning_rate": 2.339753965684688e-06, + "loss": 0.0179, + "step": 9590 + }, + { + "epoch": 2.31, + "learning_rate": 2.3316607316283587e-06, + "loss": 0.0349, + "step": 9600 + }, + { + "epoch": 2.31, + "learning_rate": 2.32356749757203e-06, + "loss": 0.0972, + "step": 9610 + }, + { + "epoch": 2.31, + "learning_rate": 2.315474263515701e-06, + "loss": 0.012, + "step": 9620 + }, + { + "epoch": 2.31, + "learning_rate": 2.307381029459372e-06, + "loss": 0.0256, + "step": 9630 + }, + { + "epoch": 2.32, + "learning_rate": 2.2992877954030433e-06, + "loss": 0.0304, + "step": 9640 + }, + { + "epoch": 2.32, + "learning_rate": 2.291194561346714e-06, + "loss": 0.0662, + "step": 9650 + }, + { + "epoch": 2.32, + "learning_rate": 2.2831013272903854e-06, + "loss": 0.0758, + "step": 9660 + }, + { + "epoch": 2.32, + "learning_rate": 2.2750080932340566e-06, + "loss": 0.0623, + "step": 9670 + }, + { + "epoch": 2.33, + "learning_rate": 2.2669148591777275e-06, + "loss": 0.0128, + "step": 9680 + }, + { + "epoch": 2.33, + "learning_rate": 2.2588216251213987e-06, + "loss": 0.0472, + "step": 9690 + }, + { + "epoch": 2.33, + "learning_rate": 2.25072839106507e-06, + "loss": 0.0314, + "step": 9700 + }, + { + "epoch": 2.33, + "learning_rate": 2.242635157008741e-06, + "loss": 0.0679, + "step": 9710 + }, + { + "epoch": 2.34, + "learning_rate": 2.234541922952412e-06, + "loss": 0.0336, + "step": 9720 + }, + { + "epoch": 2.34, + "learning_rate": 2.2264486888960833e-06, + "loss": 0.0449, + "step": 9730 + }, + { + "epoch": 2.34, + "learning_rate": 2.218355454839754e-06, + "loss": 0.0273, + "step": 9740 + }, + { + "epoch": 2.34, + "learning_rate": 2.210262220783425e-06, + "loss": 0.0463, + "step": 9750 + }, + { + "epoch": 2.35, + "learning_rate": 2.2021689867270963e-06, + "loss": 0.0686, + "step": 9760 + }, + { + "epoch": 2.35, + "learning_rate": 2.1940757526707675e-06, + "loss": 0.0652, + "step": 9770 + }, + { + "epoch": 2.35, + "learning_rate": 2.1859825186144384e-06, + "loss": 0.0325, + "step": 9780 + }, + { + "epoch": 2.35, + "learning_rate": 2.1778892845581096e-06, + "loss": 0.0434, + "step": 9790 + }, + { + "epoch": 2.36, + "learning_rate": 2.169796050501781e-06, + "loss": 0.0151, + "step": 9800 + }, + { + "epoch": 2.36, + "learning_rate": 2.1617028164454517e-06, + "loss": 0.0431, + "step": 9810 + }, + { + "epoch": 2.36, + "learning_rate": 2.153609582389123e-06, + "loss": 0.1106, + "step": 9820 + }, + { + "epoch": 2.36, + "learning_rate": 2.145516348332794e-06, + "loss": 0.0638, + "step": 9830 + }, + { + "epoch": 2.37, + "learning_rate": 2.137423114276465e-06, + "loss": 0.0426, + "step": 9840 + }, + { + "epoch": 2.37, + "learning_rate": 2.1293298802201363e-06, + "loss": 0.0161, + "step": 9850 + }, + { + "epoch": 2.37, + "learning_rate": 2.121236646163807e-06, + "loss": 0.0403, + "step": 9860 + }, + { + "epoch": 2.37, + "learning_rate": 2.1131434121074784e-06, + "loss": 0.0475, + "step": 9870 + }, + { + "epoch": 2.38, + "learning_rate": 2.1050501780511496e-06, + "loss": 0.05, + "step": 9880 + }, + { + "epoch": 2.38, + "learning_rate": 2.0969569439948205e-06, + "loss": 0.0556, + "step": 9890 + }, + { + "epoch": 2.38, + "learning_rate": 2.0888637099384913e-06, + "loss": 0.0607, + "step": 9900 + }, + { + "epoch": 2.38, + "learning_rate": 2.0807704758821626e-06, + "loss": 0.0408, + "step": 9910 + }, + { + "epoch": 2.38, + "learning_rate": 2.072677241825834e-06, + "loss": 0.0765, + "step": 9920 + }, + { + "epoch": 2.39, + "learning_rate": 2.0645840077695047e-06, + "loss": 0.0561, + "step": 9930 + }, + { + "epoch": 2.39, + "learning_rate": 2.056490773713176e-06, + "loss": 0.0246, + "step": 9940 + }, + { + "epoch": 2.39, + "learning_rate": 2.048397539656847e-06, + "loss": 0.0363, + "step": 9950 + }, + { + "epoch": 2.39, + "learning_rate": 2.040304305600518e-06, + "loss": 0.0954, + "step": 9960 + }, + { + "epoch": 2.4, + "learning_rate": 2.0322110715441893e-06, + "loss": 0.021, + "step": 9970 + }, + { + "epoch": 2.4, + "learning_rate": 2.0241178374878605e-06, + "loss": 0.0391, + "step": 9980 + }, + { + "epoch": 2.4, + "learning_rate": 2.0160246034315314e-06, + "loss": 0.019, + "step": 9990 + }, + { + "epoch": 2.4, + "learning_rate": 2.0079313693752026e-06, + "loss": 0.0438, + "step": 10000 + }, + { + "epoch": 2.4, + "eval_loss": 0.05346548184752464, + "eval_runtime": 424.4978, + "eval_samples_per_second": 14.697, + "eval_steps_per_second": 2.45, + "step": 10000 + }, + { + "epoch": 2.41, + "learning_rate": 1.999838135318874e-06, + "loss": 0.0486, + "step": 10010 + }, + { + "epoch": 2.41, + "learning_rate": 1.9917449012625447e-06, + "loss": 0.0236, + "step": 10020 + }, + { + "epoch": 2.41, + "learning_rate": 1.983651667206216e-06, + "loss": 0.0801, + "step": 10030 + }, + { + "epoch": 2.41, + "learning_rate": 1.975558433149887e-06, + "loss": 0.0376, + "step": 10040 + }, + { + "epoch": 2.42, + "learning_rate": 1.9674651990935576e-06, + "loss": 0.0567, + "step": 10050 + }, + { + "epoch": 2.42, + "learning_rate": 1.959371965037229e-06, + "loss": 0.0389, + "step": 10060 + }, + { + "epoch": 2.42, + "learning_rate": 1.9512787309809e-06, + "loss": 0.0227, + "step": 10070 + }, + { + "epoch": 2.42, + "learning_rate": 1.943185496924571e-06, + "loss": 0.0069, + "step": 10080 + }, + { + "epoch": 2.43, + "learning_rate": 1.9350922628682422e-06, + "loss": 0.1, + "step": 10090 + }, + { + "epoch": 2.43, + "learning_rate": 1.9269990288119135e-06, + "loss": 0.0304, + "step": 10100 + }, + { + "epoch": 2.43, + "learning_rate": 1.9189057947555843e-06, + "loss": 0.0552, + "step": 10110 + }, + { + "epoch": 2.43, + "learning_rate": 1.9108125606992556e-06, + "loss": 0.0346, + "step": 10120 + }, + { + "epoch": 2.44, + "learning_rate": 1.9027193266429266e-06, + "loss": 0.0506, + "step": 10130 + }, + { + "epoch": 2.44, + "learning_rate": 1.8946260925865979e-06, + "loss": 0.025, + "step": 10140 + }, + { + "epoch": 2.44, + "learning_rate": 1.886532858530269e-06, + "loss": 0.0482, + "step": 10150 + }, + { + "epoch": 2.44, + "learning_rate": 1.87843962447394e-06, + "loss": 0.0622, + "step": 10160 + }, + { + "epoch": 2.44, + "learning_rate": 1.8703463904176112e-06, + "loss": 0.031, + "step": 10170 + }, + { + "epoch": 2.45, + "learning_rate": 1.8622531563612823e-06, + "loss": 0.0519, + "step": 10180 + }, + { + "epoch": 2.45, + "learning_rate": 1.8541599223049533e-06, + "loss": 0.0365, + "step": 10190 + }, + { + "epoch": 2.45, + "learning_rate": 1.8460666882486241e-06, + "loss": 0.0353, + "step": 10200 + }, + { + "epoch": 2.45, + "learning_rate": 1.8379734541922952e-06, + "loss": 0.038, + "step": 10210 + }, + { + "epoch": 2.46, + "learning_rate": 1.8298802201359664e-06, + "loss": 0.0489, + "step": 10220 + }, + { + "epoch": 2.46, + "learning_rate": 1.8217869860796375e-06, + "loss": 0.0352, + "step": 10230 + }, + { + "epoch": 2.46, + "learning_rate": 1.8136937520233085e-06, + "loss": 0.0484, + "step": 10240 + }, + { + "epoch": 2.46, + "learning_rate": 1.8056005179669798e-06, + "loss": 0.045, + "step": 10250 + }, + { + "epoch": 2.47, + "learning_rate": 1.7975072839106508e-06, + "loss": 0.0399, + "step": 10260 + }, + { + "epoch": 2.47, + "learning_rate": 1.7894140498543219e-06, + "loss": 0.0185, + "step": 10270 + }, + { + "epoch": 2.47, + "learning_rate": 1.7813208157979931e-06, + "loss": 0.0591, + "step": 10280 + }, + { + "epoch": 2.47, + "learning_rate": 1.7732275817416642e-06, + "loss": 0.0341, + "step": 10290 + }, + { + "epoch": 2.48, + "learning_rate": 1.7651343476853352e-06, + "loss": 0.0752, + "step": 10300 + }, + { + "epoch": 2.48, + "learning_rate": 1.7570411136290063e-06, + "loss": 0.0192, + "step": 10310 + }, + { + "epoch": 2.48, + "learning_rate": 1.7489478795726775e-06, + "loss": 0.0385, + "step": 10320 + }, + { + "epoch": 2.48, + "learning_rate": 1.7408546455163486e-06, + "loss": 0.0273, + "step": 10330 + }, + { + "epoch": 2.49, + "learning_rate": 1.7327614114600196e-06, + "loss": 0.0268, + "step": 10340 + }, + { + "epoch": 2.49, + "learning_rate": 1.7246681774036905e-06, + "loss": 0.0514, + "step": 10350 + }, + { + "epoch": 2.49, + "learning_rate": 1.7165749433473617e-06, + "loss": 0.0516, + "step": 10360 + }, + { + "epoch": 2.49, + "learning_rate": 1.7084817092910328e-06, + "loss": 0.0264, + "step": 10370 + }, + { + "epoch": 2.5, + "learning_rate": 1.7003884752347038e-06, + "loss": 0.0497, + "step": 10380 + }, + { + "epoch": 2.5, + "learning_rate": 1.692295241178375e-06, + "loss": 0.0577, + "step": 10390 + }, + { + "epoch": 2.5, + "learning_rate": 1.684202007122046e-06, + "loss": 0.0508, + "step": 10400 + }, + { + "epoch": 2.5, + "learning_rate": 1.6761087730657171e-06, + "loss": 0.0394, + "step": 10410 + }, + { + "epoch": 2.5, + "learning_rate": 1.6680155390093882e-06, + "loss": 0.0134, + "step": 10420 + }, + { + "epoch": 2.51, + "learning_rate": 1.6599223049530594e-06, + "loss": 0.1044, + "step": 10430 + }, + { + "epoch": 2.51, + "learning_rate": 1.6518290708967305e-06, + "loss": 0.0621, + "step": 10440 + }, + { + "epoch": 2.51, + "learning_rate": 1.6437358368404015e-06, + "loss": 0.008, + "step": 10450 + }, + { + "epoch": 2.51, + "learning_rate": 1.6356426027840728e-06, + "loss": 0.0397, + "step": 10460 + }, + { + "epoch": 2.52, + "learning_rate": 1.6275493687277438e-06, + "loss": 0.0527, + "step": 10470 + }, + { + "epoch": 2.52, + "learning_rate": 1.6194561346714149e-06, + "loss": 0.0272, + "step": 10480 + }, + { + "epoch": 2.52, + "learning_rate": 1.6113629006150861e-06, + "loss": 0.0338, + "step": 10490 + }, + { + "epoch": 2.52, + "learning_rate": 1.603269666558757e-06, + "loss": 0.0708, + "step": 10500 + }, + { + "epoch": 2.52, + "eval_loss": 0.053738877177238464, + "eval_runtime": 419.1802, + "eval_samples_per_second": 14.884, + "eval_steps_per_second": 2.481, + "step": 10500 + }, + { + "epoch": 2.53, + "learning_rate": 1.595176432502428e-06, + "loss": 0.0526, + "step": 10510 + }, + { + "epoch": 2.53, + "learning_rate": 1.587083198446099e-06, + "loss": 0.035, + "step": 10520 + }, + { + "epoch": 2.53, + "learning_rate": 1.5789899643897701e-06, + "loss": 0.0287, + "step": 10530 + }, + { + "epoch": 2.53, + "learning_rate": 1.5708967303334414e-06, + "loss": 0.0316, + "step": 10540 + }, + { + "epoch": 2.54, + "learning_rate": 1.5628034962771124e-06, + "loss": 0.0577, + "step": 10550 + }, + { + "epoch": 2.54, + "learning_rate": 1.5547102622207835e-06, + "loss": 0.0449, + "step": 10560 + }, + { + "epoch": 2.54, + "learning_rate": 1.5466170281644547e-06, + "loss": 0.0123, + "step": 10570 + }, + { + "epoch": 2.54, + "learning_rate": 1.5385237941081258e-06, + "loss": 0.0483, + "step": 10580 + }, + { + "epoch": 2.55, + "learning_rate": 1.5304305600517968e-06, + "loss": 0.0655, + "step": 10590 + }, + { + "epoch": 2.55, + "learning_rate": 1.522337325995468e-06, + "loss": 0.0382, + "step": 10600 + }, + { + "epoch": 2.55, + "learning_rate": 1.514244091939139e-06, + "loss": 0.0532, + "step": 10610 + }, + { + "epoch": 2.55, + "learning_rate": 1.5061508578828101e-06, + "loss": 0.0381, + "step": 10620 + }, + { + "epoch": 2.56, + "learning_rate": 1.4980576238264812e-06, + "loss": 0.0642, + "step": 10630 + }, + { + "epoch": 2.56, + "learning_rate": 1.4899643897701524e-06, + "loss": 0.0348, + "step": 10640 + }, + { + "epoch": 2.56, + "learning_rate": 1.4818711557138233e-06, + "loss": 0.0408, + "step": 10650 + }, + { + "epoch": 2.56, + "learning_rate": 1.4737779216574943e-06, + "loss": 0.0572, + "step": 10660 + }, + { + "epoch": 2.56, + "learning_rate": 1.4656846876011654e-06, + "loss": 0.0274, + "step": 10670 + }, + { + "epoch": 2.57, + "learning_rate": 1.4575914535448366e-06, + "loss": 0.0522, + "step": 10680 + }, + { + "epoch": 2.57, + "learning_rate": 1.4494982194885077e-06, + "loss": 0.039, + "step": 10690 + }, + { + "epoch": 2.57, + "learning_rate": 1.4414049854321787e-06, + "loss": 0.0186, + "step": 10700 + }, + { + "epoch": 2.57, + "learning_rate": 1.43331175137585e-06, + "loss": 0.0489, + "step": 10710 + }, + { + "epoch": 2.58, + "learning_rate": 1.425218517319521e-06, + "loss": 0.0254, + "step": 10720 + }, + { + "epoch": 2.58, + "learning_rate": 1.417125283263192e-06, + "loss": 0.0267, + "step": 10730 + }, + { + "epoch": 2.58, + "learning_rate": 1.4090320492068631e-06, + "loss": 0.0366, + "step": 10740 + }, + { + "epoch": 2.58, + "learning_rate": 1.4009388151505344e-06, + "loss": 0.0588, + "step": 10750 + }, + { + "epoch": 2.59, + "learning_rate": 1.3928455810942054e-06, + "loss": 0.0401, + "step": 10760 + }, + { + "epoch": 2.59, + "learning_rate": 1.3847523470378765e-06, + "loss": 0.0379, + "step": 10770 + }, + { + "epoch": 2.59, + "learning_rate": 1.3766591129815477e-06, + "loss": 0.0597, + "step": 10780 + }, + { + "epoch": 2.59, + "learning_rate": 1.3685658789252188e-06, + "loss": 0.0661, + "step": 10790 + }, + { + "epoch": 2.6, + "learning_rate": 1.3604726448688898e-06, + "loss": 0.056, + "step": 10800 + }, + { + "epoch": 2.6, + "learning_rate": 1.3523794108125606e-06, + "loss": 0.0509, + "step": 10810 + }, + { + "epoch": 2.6, + "learning_rate": 1.3442861767562319e-06, + "loss": 0.0478, + "step": 10820 + }, + { + "epoch": 2.6, + "learning_rate": 1.336192942699903e-06, + "loss": 0.1052, + "step": 10830 + }, + { + "epoch": 2.61, + "learning_rate": 1.328099708643574e-06, + "loss": 0.0269, + "step": 10840 + }, + { + "epoch": 2.61, + "learning_rate": 1.320006474587245e-06, + "loss": 0.0569, + "step": 10850 + }, + { + "epoch": 2.61, + "learning_rate": 1.3119132405309163e-06, + "loss": 0.0703, + "step": 10860 + }, + { + "epoch": 2.61, + "learning_rate": 1.3038200064745873e-06, + "loss": 0.046, + "step": 10870 + }, + { + "epoch": 2.62, + "learning_rate": 1.2957267724182584e-06, + "loss": 0.0231, + "step": 10880 + }, + { + "epoch": 2.62, + "learning_rate": 1.2876335383619296e-06, + "loss": 0.0164, + "step": 10890 + }, + { + "epoch": 2.62, + "learning_rate": 1.2795403043056007e-06, + "loss": 0.018, + "step": 10900 + }, + { + "epoch": 2.62, + "learning_rate": 1.2714470702492717e-06, + "loss": 0.0732, + "step": 10910 + }, + { + "epoch": 2.62, + "learning_rate": 1.263353836192943e-06, + "loss": 0.0269, + "step": 10920 + }, + { + "epoch": 2.63, + "learning_rate": 1.255260602136614e-06, + "loss": 0.0345, + "step": 10930 + }, + { + "epoch": 2.63, + "learning_rate": 1.247167368080285e-06, + "loss": 0.059, + "step": 10940 + }, + { + "epoch": 2.63, + "learning_rate": 1.2390741340239561e-06, + "loss": 0.0209, + "step": 10950 + }, + { + "epoch": 2.63, + "learning_rate": 1.2309808999676272e-06, + "loss": 0.034, + "step": 10960 + }, + { + "epoch": 2.64, + "learning_rate": 1.2228876659112982e-06, + "loss": 0.048, + "step": 10970 + }, + { + "epoch": 2.64, + "learning_rate": 1.2147944318549695e-06, + "loss": 0.0502, + "step": 10980 + }, + { + "epoch": 2.64, + "learning_rate": 1.2067011977986403e-06, + "loss": 0.0557, + "step": 10990 + }, + { + "epoch": 2.64, + "learning_rate": 1.1986079637423115e-06, + "loss": 0.0466, + "step": 11000 + }, + { + "epoch": 2.64, + "eval_loss": 0.05311308801174164, + "eval_runtime": 427.0174, + "eval_samples_per_second": 14.611, + "eval_steps_per_second": 2.435, + "step": 11000 + }, + { + "epoch": 2.65, + "learning_rate": 1.1905147296859826e-06, + "loss": 0.0215, + "step": 11010 + }, + { + "epoch": 2.65, + "learning_rate": 1.1824214956296536e-06, + "loss": 0.0223, + "step": 11020 + }, + { + "epoch": 2.65, + "learning_rate": 1.1743282615733249e-06, + "loss": 0.0395, + "step": 11030 + }, + { + "epoch": 2.65, + "learning_rate": 1.166235027516996e-06, + "loss": 0.0381, + "step": 11040 + }, + { + "epoch": 2.66, + "learning_rate": 1.158141793460667e-06, + "loss": 0.0333, + "step": 11050 + }, + { + "epoch": 2.66, + "learning_rate": 1.150048559404338e-06, + "loss": 0.0715, + "step": 11060 + }, + { + "epoch": 2.66, + "learning_rate": 1.141955325348009e-06, + "loss": 0.0349, + "step": 11070 + }, + { + "epoch": 2.66, + "learning_rate": 1.1338620912916801e-06, + "loss": 0.0403, + "step": 11080 + }, + { + "epoch": 2.67, + "learning_rate": 1.1257688572353514e-06, + "loss": 0.01, + "step": 11090 + }, + { + "epoch": 2.67, + "learning_rate": 1.1176756231790224e-06, + "loss": 0.0349, + "step": 11100 + }, + { + "epoch": 2.67, + "learning_rate": 1.1095823891226935e-06, + "loss": 0.0348, + "step": 11110 + }, + { + "epoch": 2.67, + "learning_rate": 1.1014891550663647e-06, + "loss": 0.0407, + "step": 11120 + }, + { + "epoch": 2.68, + "learning_rate": 1.0933959210100358e-06, + "loss": 0.0489, + "step": 11130 + }, + { + "epoch": 2.68, + "learning_rate": 1.0853026869537068e-06, + "loss": 0.0312, + "step": 11140 + }, + { + "epoch": 2.68, + "learning_rate": 1.0772094528973779e-06, + "loss": 0.0251, + "step": 11150 + }, + { + "epoch": 2.68, + "learning_rate": 1.069116218841049e-06, + "loss": 0.0104, + "step": 11160 + }, + { + "epoch": 2.69, + "learning_rate": 1.06102298478472e-06, + "loss": 0.0234, + "step": 11170 + }, + { + "epoch": 2.69, + "learning_rate": 1.0529297507283912e-06, + "loss": 0.0771, + "step": 11180 + }, + { + "epoch": 2.69, + "learning_rate": 1.0448365166720622e-06, + "loss": 0.0459, + "step": 11190 + }, + { + "epoch": 2.69, + "learning_rate": 1.0367432826157333e-06, + "loss": 0.0531, + "step": 11200 + }, + { + "epoch": 2.69, + "learning_rate": 1.0286500485594045e-06, + "loss": 0.0425, + "step": 11210 + }, + { + "epoch": 2.7, + "learning_rate": 1.0205568145030754e-06, + "loss": 0.065, + "step": 11220 + }, + { + "epoch": 2.7, + "learning_rate": 1.0124635804467466e-06, + "loss": 0.0433, + "step": 11230 + }, + { + "epoch": 2.7, + "learning_rate": 1.0043703463904177e-06, + "loss": 0.0411, + "step": 11240 + }, + { + "epoch": 2.7, + "learning_rate": 9.962771123340887e-07, + "loss": 0.0491, + "step": 11250 + }, + { + "epoch": 2.71, + "learning_rate": 9.8818387827776e-07, + "loss": 0.0458, + "step": 11260 + }, + { + "epoch": 2.71, + "learning_rate": 9.80090644221431e-07, + "loss": 0.0471, + "step": 11270 + }, + { + "epoch": 2.71, + "learning_rate": 9.71997410165102e-07, + "loss": 0.0827, + "step": 11280 + }, + { + "epoch": 2.71, + "learning_rate": 9.639041761087731e-07, + "loss": 0.0444, + "step": 11290 + }, + { + "epoch": 2.72, + "learning_rate": 9.558109420524442e-07, + "loss": 0.0544, + "step": 11300 + }, + { + "epoch": 2.72, + "learning_rate": 9.477177079961153e-07, + "loss": 0.0131, + "step": 11310 + }, + { + "epoch": 2.72, + "learning_rate": 9.396244739397864e-07, + "loss": 0.0207, + "step": 11320 + }, + { + "epoch": 2.72, + "learning_rate": 9.315312398834575e-07, + "loss": 0.0294, + "step": 11330 + }, + { + "epoch": 2.73, + "learning_rate": 9.234380058271287e-07, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 2.73, + "learning_rate": 9.153447717707997e-07, + "loss": 0.0498, + "step": 11350 + }, + { + "epoch": 2.73, + "learning_rate": 9.072515377144709e-07, + "loss": 0.0559, + "step": 11360 + }, + { + "epoch": 2.73, + "learning_rate": 8.991583036581418e-07, + "loss": 0.0404, + "step": 11370 + }, + { + "epoch": 2.74, + "learning_rate": 8.910650696018129e-07, + "loss": 0.0696, + "step": 11380 + }, + { + "epoch": 2.74, + "learning_rate": 8.82971835545484e-07, + "loss": 0.0604, + "step": 11390 + }, + { + "epoch": 2.74, + "learning_rate": 8.748786014891551e-07, + "loss": 0.0296, + "step": 11400 + }, + { + "epoch": 2.74, + "learning_rate": 8.667853674328262e-07, + "loss": 0.0334, + "step": 11410 + }, + { + "epoch": 2.75, + "learning_rate": 8.586921333764973e-07, + "loss": 0.0485, + "step": 11420 + }, + { + "epoch": 2.75, + "learning_rate": 8.505988993201685e-07, + "loss": 0.0668, + "step": 11430 + }, + { + "epoch": 2.75, + "learning_rate": 8.425056652638395e-07, + "loss": 0.0432, + "step": 11440 + }, + { + "epoch": 2.75, + "learning_rate": 8.344124312075106e-07, + "loss": 0.0224, + "step": 11450 + }, + { + "epoch": 2.75, + "learning_rate": 8.263191971511816e-07, + "loss": 0.0168, + "step": 11460 + }, + { + "epoch": 2.76, + "learning_rate": 8.182259630948528e-07, + "loss": 0.0436, + "step": 11470 + }, + { + "epoch": 2.76, + "learning_rate": 8.101327290385238e-07, + "loss": 0.0671, + "step": 11480 + }, + { + "epoch": 2.76, + "learning_rate": 8.02039494982195e-07, + "loss": 0.0364, + "step": 11490 + }, + { + "epoch": 2.76, + "learning_rate": 7.939462609258661e-07, + "loss": 0.0382, + "step": 11500 + }, + { + "epoch": 2.76, + "eval_loss": 0.05557642877101898, + "eval_runtime": 424.4281, + "eval_samples_per_second": 14.7, + "eval_steps_per_second": 2.45, + "step": 11500 + }, + { + "epoch": 2.77, + "learning_rate": 7.858530268695372e-07, + "loss": 0.0535, + "step": 11510 + }, + { + "epoch": 2.77, + "learning_rate": 7.777597928132082e-07, + "loss": 0.0424, + "step": 11520 + }, + { + "epoch": 2.77, + "learning_rate": 7.696665587568793e-07, + "loss": 0.0247, + "step": 11530 + }, + { + "epoch": 2.77, + "learning_rate": 7.615733247005504e-07, + "loss": 0.0392, + "step": 11540 + }, + { + "epoch": 2.78, + "learning_rate": 7.534800906442214e-07, + "loss": 0.0214, + "step": 11550 + }, + { + "epoch": 2.78, + "learning_rate": 7.453868565878926e-07, + "loss": 0.054, + "step": 11560 + }, + { + "epoch": 2.78, + "learning_rate": 7.372936225315637e-07, + "loss": 0.0176, + "step": 11570 + }, + { + "epoch": 2.78, + "learning_rate": 7.292003884752348e-07, + "loss": 0.0525, + "step": 11580 + }, + { + "epoch": 2.79, + "learning_rate": 7.211071544189059e-07, + "loss": 0.0422, + "step": 11590 + }, + { + "epoch": 2.79, + "learning_rate": 7.130139203625769e-07, + "loss": 0.0418, + "step": 11600 + }, + { + "epoch": 2.79, + "learning_rate": 7.04920686306248e-07, + "loss": 0.0488, + "step": 11610 + }, + { + "epoch": 2.79, + "learning_rate": 6.968274522499191e-07, + "loss": 0.0446, + "step": 11620 + }, + { + "epoch": 2.8, + "learning_rate": 6.887342181935902e-07, + "loss": 0.0435, + "step": 11630 + }, + { + "epoch": 2.8, + "learning_rate": 6.806409841372613e-07, + "loss": 0.0678, + "step": 11640 + }, + { + "epoch": 2.8, + "learning_rate": 6.725477500809324e-07, + "loss": 0.0361, + "step": 11650 + }, + { + "epoch": 2.8, + "learning_rate": 6.644545160246036e-07, + "loss": 0.0161, + "step": 11660 + }, + { + "epoch": 2.81, + "learning_rate": 6.563612819682746e-07, + "loss": 0.0114, + "step": 11670 + }, + { + "epoch": 2.81, + "learning_rate": 6.482680479119457e-07, + "loss": 0.028, + "step": 11680 + }, + { + "epoch": 2.81, + "learning_rate": 6.401748138556167e-07, + "loss": 0.0659, + "step": 11690 + }, + { + "epoch": 2.81, + "learning_rate": 6.320815797992879e-07, + "loss": 0.0111, + "step": 11700 + }, + { + "epoch": 2.81, + "learning_rate": 6.239883457429589e-07, + "loss": 0.055, + "step": 11710 + }, + { + "epoch": 2.82, + "learning_rate": 6.158951116866301e-07, + "loss": 0.1518, + "step": 11720 + }, + { + "epoch": 2.82, + "learning_rate": 6.078018776303011e-07, + "loss": 0.0504, + "step": 11730 + }, + { + "epoch": 2.82, + "learning_rate": 5.997086435739721e-07, + "loss": 0.0357, + "step": 11740 + }, + { + "epoch": 2.82, + "learning_rate": 5.916154095176433e-07, + "loss": 0.0824, + "step": 11750 + }, + { + "epoch": 2.83, + "learning_rate": 5.835221754613144e-07, + "loss": 0.0532, + "step": 11760 + }, + { + "epoch": 2.83, + "learning_rate": 5.754289414049855e-07, + "loss": 0.0604, + "step": 11770 + }, + { + "epoch": 2.83, + "learning_rate": 5.673357073486565e-07, + "loss": 0.0205, + "step": 11780 + }, + { + "epoch": 2.83, + "learning_rate": 5.592424732923277e-07, + "loss": 0.0383, + "step": 11790 + }, + { + "epoch": 2.84, + "learning_rate": 5.511492392359987e-07, + "loss": 0.0626, + "step": 11800 + }, + { + "epoch": 2.84, + "learning_rate": 5.430560051796698e-07, + "loss": 0.0554, + "step": 11810 + }, + { + "epoch": 2.84, + "learning_rate": 5.349627711233409e-07, + "loss": 0.0269, + "step": 11820 + }, + { + "epoch": 2.84, + "learning_rate": 5.268695370670121e-07, + "loss": 0.0196, + "step": 11830 + }, + { + "epoch": 2.85, + "learning_rate": 5.187763030106831e-07, + "loss": 0.021, + "step": 11840 + }, + { + "epoch": 2.85, + "learning_rate": 5.106830689543542e-07, + "loss": 0.0399, + "step": 11850 + }, + { + "epoch": 2.85, + "learning_rate": 5.025898348980253e-07, + "loss": 0.0266, + "step": 11860 + }, + { + "epoch": 2.85, + "learning_rate": 4.944966008416964e-07, + "loss": 0.0532, + "step": 11870 + }, + { + "epoch": 2.86, + "learning_rate": 4.864033667853674e-07, + "loss": 0.0496, + "step": 11880 + }, + { + "epoch": 2.86, + "learning_rate": 4.783101327290386e-07, + "loss": 0.0515, + "step": 11890 + }, + { + "epoch": 2.86, + "learning_rate": 4.7021689867270966e-07, + "loss": 0.0316, + "step": 11900 + }, + { + "epoch": 2.86, + "learning_rate": 4.6212366461638076e-07, + "loss": 0.0348, + "step": 11910 + }, + { + "epoch": 2.87, + "learning_rate": 4.540304305600518e-07, + "loss": 0.043, + "step": 11920 + }, + { + "epoch": 2.87, + "learning_rate": 4.459371965037229e-07, + "loss": 0.0377, + "step": 11930 + }, + { + "epoch": 2.87, + "learning_rate": 4.37843962447394e-07, + "loss": 0.0494, + "step": 11940 + }, + { + "epoch": 2.87, + "learning_rate": 4.2975072839106515e-07, + "loss": 0.0746, + "step": 11950 + }, + { + "epoch": 2.88, + "learning_rate": 4.216574943347362e-07, + "loss": 0.046, + "step": 11960 + }, + { + "epoch": 2.88, + "learning_rate": 4.135642602784073e-07, + "loss": 0.0195, + "step": 11970 + }, + { + "epoch": 2.88, + "learning_rate": 4.054710262220784e-07, + "loss": 0.0578, + "step": 11980 + }, + { + "epoch": 2.88, + "learning_rate": 3.973777921657495e-07, + "loss": 0.0308, + "step": 11990 + }, + { + "epoch": 2.88, + "learning_rate": 3.8928455810942053e-07, + "loss": 0.0896, + "step": 12000 + }, + { + "epoch": 2.88, + "eval_loss": 0.0539301373064518, + "eval_runtime": 422.0934, + "eval_samples_per_second": 14.781, + "eval_steps_per_second": 2.464, + "step": 12000 + } + ], + "logging_steps": 10, + "max_steps": 12480, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "total_flos": 130144689930240.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}