{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "global_step": 1438, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.999403411726636e-05, "loss": 1.8493, "step": 10 }, { "epoch": 0.03, "learning_rate": 4.997613931640598e-05, "loss": 1.613, "step": 20 }, { "epoch": 0.04, "learning_rate": 4.9946324138081534e-05, "loss": 1.5193, "step": 30 }, { "epoch": 0.06, "learning_rate": 4.990460281220164e-05, "loss": 1.4805, "step": 40 }, { "epoch": 0.07, "learning_rate": 4.98509952511293e-05, "loss": 1.443, "step": 50 }, { "epoch": 0.08, "learning_rate": 4.978552704017835e-05, "loss": 1.4186, "step": 60 }, { "epoch": 0.1, "learning_rate": 4.9708229425402354e-05, "loss": 1.394, "step": 70 }, { "epoch": 0.11, "learning_rate": 4.961913929868172e-05, "loss": 1.38, "step": 80 }, { "epoch": 0.13, "learning_rate": 4.9518299180116366e-05, "loss": 1.3715, "step": 90 }, { "epoch": 0.14, "learning_rate": 4.940575719773205e-05, "loss": 1.3548, "step": 100 }, { "epoch": 0.15, "learning_rate": 4.928156706451035e-05, "loss": 1.3536, "step": 110 }, { "epoch": 0.17, "learning_rate": 4.9145788052752975e-05, "loss": 1.3535, "step": 120 }, { "epoch": 0.18, "learning_rate": 4.899848496579287e-05, "loss": 1.3414, "step": 130 }, { "epoch": 0.19, "learning_rate": 4.883972810706548e-05, "loss": 1.3275, "step": 140 }, { "epoch": 0.21, "learning_rate": 4.866959324655501e-05, "loss": 1.3342, "step": 150 }, { "epoch": 0.22, "learning_rate": 4.8488161584631574e-05, "loss": 1.3234, "step": 160 }, { "epoch": 0.24, "learning_rate": 4.829551971329672e-05, "loss": 1.3235, "step": 170 }, { "epoch": 0.25, "learning_rate": 4.809175957485558e-05, "loss": 1.3245, "step": 180 }, { "epoch": 0.26, "learning_rate": 4.7876978418035455e-05, "loss": 1.3132, "step": 190 }, { "epoch": 0.28, "learning_rate": 4.7651278751571984e-05, "loss": 1.3193, "step": 200 }, { "epoch": 0.29, "learning_rate": 4.7414768295284594e-05, "loss": 1.3158, "step": 210 }, { "epoch": 0.31, "learning_rate": 4.716755992866509e-05, "loss": 1.3089, "step": 220 }, { "epoch": 0.32, "learning_rate": 4.6909771637003556e-05, "loss": 1.3232, "step": 230 }, { "epoch": 0.33, "learning_rate": 4.664152645507744e-05, "loss": 1.307, "step": 240 }, { "epoch": 0.35, "learning_rate": 4.636295240843068e-05, "loss": 1.3057, "step": 250 }, { "epoch": 0.36, "learning_rate": 4.607418245227089e-05, "loss": 1.2985, "step": 260 }, { "epoch": 0.38, "learning_rate": 4.577535440801368e-05, "loss": 1.3111, "step": 270 }, { "epoch": 0.39, "learning_rate": 4.546661089750462e-05, "loss": 1.2948, "step": 280 }, { "epoch": 0.4, "learning_rate": 4.5148099274950005e-05, "loss": 1.3037, "step": 290 }, { "epoch": 0.42, "learning_rate": 4.481997155658898e-05, "loss": 1.3042, "step": 300 }, { "epoch": 0.43, "learning_rate": 4.4482384348140695e-05, "loss": 1.2899, "step": 310 }, { "epoch": 0.45, "learning_rate": 4.413549877006099e-05, "loss": 1.3006, "step": 320 }, { "epoch": 0.46, "learning_rate": 4.3779480380644337e-05, "loss": 1.2908, "step": 330 }, { "epoch": 0.47, "learning_rate": 4.34144990970077e-05, "loss": 1.2898, "step": 340 }, { "epoch": 0.49, "learning_rate": 4.304072911399415e-05, "loss": 1.2822, "step": 350 }, { "epoch": 0.5, "learning_rate": 4.2658348821034724e-05, "loss": 1.2927, "step": 360 }, { "epoch": 0.51, "learning_rate": 4.226754071700841e-05, "loss": 1.2847, "step": 370 }, { "epoch": 0.53, "learning_rate": 4.1868491323140816e-05, "loss": 1.2873, "step": 380 }, { "epoch": 0.54, "learning_rate": 4.146139109398303e-05, "loss": 1.288, "step": 390 }, { "epoch": 0.56, "learning_rate": 4.1046434326513305e-05, "loss": 1.2849, "step": 400 }, { "epoch": 0.57, "learning_rate": 4.062381906740477e-05, "loss": 1.2819, "step": 410 }, { "epoch": 0.58, "learning_rate": 4.019374701850362e-05, "loss": 1.2694, "step": 420 }, { "epoch": 0.6, "learning_rate": 3.975642344056271e-05, "loss": 1.2843, "step": 430 }, { "epoch": 0.61, "learning_rate": 3.931205705527664e-05, "loss": 1.2646, "step": 440 }, { "epoch": 0.63, "learning_rate": 3.886085994566506e-05, "loss": 1.277, "step": 450 }, { "epoch": 0.64, "learning_rate": 3.840304745485162e-05, "loss": 1.2769, "step": 460 }, { "epoch": 0.65, "learning_rate": 3.7938838083287056e-05, "loss": 1.2797, "step": 470 }, { "epoch": 0.67, "learning_rate": 3.746845338446533e-05, "loss": 1.2776, "step": 480 }, { "epoch": 0.68, "learning_rate": 3.6992117859182674e-05, "loss": 1.2847, "step": 490 }, { "epoch": 0.7, "learning_rate": 3.651005884838995e-05, "loss": 1.2762, "step": 500 }, { "epoch": 0.71, "learning_rate": 3.6022506424689485e-05, "loss": 1.2834, "step": 510 }, { "epoch": 0.72, "learning_rate": 3.552969328252817e-05, "loss": 1.2754, "step": 520 }, { "epoch": 0.74, "learning_rate": 3.5031854627139284e-05, "loss": 1.2807, "step": 530 }, { "epoch": 0.75, "learning_rate": 3.452922806228587e-05, "loss": 1.2783, "step": 540 }, { "epoch": 0.76, "learning_rate": 3.402205347685953e-05, "loss": 1.2748, "step": 550 }, { "epoch": 0.78, "learning_rate": 3.351057293038841e-05, "loss": 1.2737, "step": 560 }, { "epoch": 0.79, "learning_rate": 3.2995030537509396e-05, "loss": 1.2729, "step": 570 }, { "epoch": 0.81, "learning_rate": 3.247567235145929e-05, "loss": 1.2712, "step": 580 }, { "epoch": 0.82, "learning_rate": 3.195274624664086e-05, "loss": 1.2716, "step": 590 }, { "epoch": 0.83, "learning_rate": 3.14265018003197e-05, "loss": 1.2823, "step": 600 }, { "epoch": 0.85, "learning_rate": 3.089719017350828e-05, "loss": 1.2627, "step": 610 }, { "epoch": 0.86, "learning_rate": 3.03650639910942e-05, "loss": 1.2658, "step": 620 }, { "epoch": 0.88, "learning_rate": 2.983037722126978e-05, "loss": 1.2803, "step": 630 }, { "epoch": 0.89, "learning_rate": 2.929338505432045e-05, "loss": 1.2731, "step": 640 }, { "epoch": 0.9, "learning_rate": 2.8754343780829968e-05, "loss": 1.2652, "step": 650 }, { "epoch": 0.92, "learning_rate": 2.821351066936043e-05, "loss": 1.2717, "step": 660 }, { "epoch": 0.93, "learning_rate": 2.767114384366555e-05, "loss": 1.259, "step": 670 }, { "epoch": 0.95, "learning_rate": 2.7127502159495782e-05, "loss": 1.2753, "step": 680 }, { "epoch": 0.96, "learning_rate": 2.6582845081054096e-05, "loss": 1.2744, "step": 690 }, { "epoch": 0.97, "learning_rate": 2.6037432557161285e-05, "loss": 1.2634, "step": 700 }, { "epoch": 0.99, "learning_rate": 2.5491524897190068e-05, "loss": 1.2668, "step": 710 }, { "epoch": 1.0, "learning_rate": 2.4945382646827058e-05, "loss": 1.2642, "step": 720 }, { "epoch": 1.02, "learning_rate": 2.4399266463721993e-05, "loss": 1.2636, "step": 730 }, { "epoch": 1.03, "learning_rate": 2.3853436993083453e-05, "loss": 1.269, "step": 740 }, { "epoch": 1.04, "learning_rate": 2.3308154743280595e-05, "loss": 1.2489, "step": 750 }, { "epoch": 1.06, "learning_rate": 2.2763679961510137e-05, "loss": 1.2586, "step": 760 }, { "epoch": 1.07, "learning_rate": 2.222027250958805e-05, "loss": 1.2633, "step": 770 }, { "epoch": 1.08, "learning_rate": 2.1678191739925107e-05, "loss": 1.2576, "step": 780 }, { "epoch": 1.1, "learning_rate": 2.1137696371745608e-05, "loss": 1.259, "step": 790 }, { "epoch": 1.11, "learning_rate": 2.059904436760835e-05, "loss": 1.2697, "step": 800 }, { "epoch": 1.13, "learning_rate": 2.006249281028859e-05, "loss": 1.2505, "step": 810 }, { "epoch": 1.14, "learning_rate": 1.9528297780080067e-05, "loss": 1.2651, "step": 820 }, { "epoch": 1.15, "learning_rate": 1.899671423257534e-05, "loss": 1.2569, "step": 830 }, { "epoch": 1.17, "learning_rate": 1.8467995876983002e-05, "loss": 1.2617, "step": 840 }, { "epoch": 1.18, "learning_rate": 1.7942395055039756e-05, "loss": 1.2501, "step": 850 }, { "epoch": 1.2, "learning_rate": 1.742016262057507e-05, "loss": 1.2575, "step": 860 }, { "epoch": 1.21, "learning_rate": 1.690154781978604e-05, "loss": 1.2583, "step": 870 }, { "epoch": 1.22, "learning_rate": 1.638679817227951e-05, "loss": 1.2671, "step": 880 }, { "epoch": 1.24, "learning_rate": 1.5876159352938202e-05, "loss": 1.2547, "step": 890 }, { "epoch": 1.25, "learning_rate": 1.5369875074667365e-05, "loss": 1.2585, "step": 900 }, { "epoch": 1.27, "learning_rate": 1.4868186972077708e-05, "loss": 1.2496, "step": 910 }, { "epoch": 1.28, "learning_rate": 1.4371334486160354e-05, "loss": 1.2524, "step": 920 }, { "epoch": 1.29, "learning_rate": 1.3879554750008653e-05, "loss": 1.2605, "step": 930 }, { "epoch": 1.31, "learning_rate": 1.3393082475641544e-05, "loss": 1.247, "step": 940 }, { "epoch": 1.32, "learning_rate": 1.2912149841982388e-05, "loss": 1.2629, "step": 950 }, { "epoch": 1.34, "learning_rate": 1.2436986384046786e-05, "loss": 1.2493, "step": 960 }, { "epoch": 1.35, "learning_rate": 1.1967818883392306e-05, "loss": 1.2556, "step": 970 }, { "epoch": 1.36, "learning_rate": 1.150487125988224e-05, "loss": 1.2531, "step": 980 }, { "epoch": 1.38, "learning_rate": 1.1048364464815297e-05, "loss": 1.2549, "step": 990 }, { "epoch": 1.39, "learning_rate": 1.0598516375471978e-05, "loss": 1.2554, "step": 1000 }, { "epoch": 1.4, "learning_rate": 1.0155541691128212e-05, "loss": 1.247, "step": 1010 }, { "epoch": 1.42, "learning_rate": 9.719651830585668e-06, "loss": 1.2541, "step": 1020 }, { "epoch": 1.43, "learning_rate": 9.291054831267749e-06, "loss": 1.2567, "step": 1030 }, { "epoch": 1.45, "learning_rate": 8.869955249929509e-06, "loss": 1.2455, "step": 1040 }, { "epoch": 1.46, "learning_rate": 8.456554065028651e-06, "loss": 1.2525, "step": 1050 }, { "epoch": 1.47, "learning_rate": 8.051048580804457e-06, "loss": 1.2583, "step": 1060 }, { "epoch": 1.49, "learning_rate": 7.65363233311028e-06, "loss": 1.255, "step": 1070 }, { "epoch": 1.5, "learning_rate": 7.264494997044522e-06, "loss": 1.2548, "step": 1080 }, { "epoch": 1.52, "learning_rate": 6.88382229642433e-06, "loss": 1.2549, "step": 1090 }, { "epoch": 1.53, "learning_rate": 6.51179591514505e-06, "loss": 1.2577, "step": 1100 }, { "epoch": 1.54, "learning_rate": 6.1485934104678285e-06, "loss": 1.2471, "step": 1110 }, { "epoch": 1.56, "learning_rate": 5.794388128276798e-06, "loss": 1.2502, "step": 1120 }, { "epoch": 1.57, "learning_rate": 5.449349120346137e-06, "loss": 1.2581, "step": 1130 }, { "epoch": 1.59, "learning_rate": 5.113641063656618e-06, "loss": 1.252, "step": 1140 }, { "epoch": 1.6, "learning_rate": 4.787424181800174e-06, "loss": 1.2497, "step": 1150 }, { "epoch": 1.61, "learning_rate": 4.470854168509819e-06, "loss": 1.2539, "step": 1160 }, { "epoch": 1.63, "learning_rate": 4.164082113351672e-06, "loss": 1.2643, "step": 1170 }, { "epoch": 1.64, "learning_rate": 3.867254429614281e-06, "loss": 1.2448, "step": 1180 }, { "epoch": 1.66, "learning_rate": 3.580512784429921e-06, "loss": 1.2487, "step": 1190 }, { "epoch": 1.67, "learning_rate": 3.3039940311609868e-06, "loss": 1.2544, "step": 1200 }, { "epoch": 1.68, "learning_rate": 3.0378301440839253e-06, "loss": 1.2491, "step": 1210 }, { "epoch": 1.7, "learning_rate": 2.7821481554018065e-06, "loss": 1.2545, "step": 1220 }, { "epoch": 1.71, "learning_rate": 2.5370700946155456e-06, "loss": 1.2477, "step": 1230 }, { "epoch": 1.72, "learning_rate": 2.3027129302828494e-06, "loss": 1.2564, "step": 1240 }, { "epoch": 1.74, "learning_rate": 2.0791885141925344e-06, "loss": 1.2481, "step": 1250 }, { "epoch": 1.75, "learning_rate": 1.8666035279809535e-06, "loss": 1.2476, "step": 1260 }, { "epoch": 1.77, "learning_rate": 1.6650594322160097e-06, "loss": 1.2538, "step": 1270 }, { "epoch": 1.78, "learning_rate": 1.4746524179729742e-06, "loss": 1.249, "step": 1280 }, { "epoch": 1.79, "learning_rate": 1.295473360925345e-06, "loss": 1.2518, "step": 1290 }, { "epoch": 1.81, "learning_rate": 1.1276077779725335e-06, "loss": 1.2569, "step": 1300 }, { "epoch": 1.82, "learning_rate": 9.711357864251647e-07, "loss": 1.252, "step": 1310 }, { "epoch": 1.84, "learning_rate": 8.261320657674626e-07, "loss": 1.2554, "step": 1320 }, { "epoch": 1.85, "learning_rate": 6.926658220148929e-07, "loss": 1.2504, "step": 1330 }, { "epoch": 1.86, "learning_rate": 5.708007546841893e-07, "loss": 1.2466, "step": 1340 }, { "epoch": 1.88, "learning_rate": 4.6059502639142773e-07, "loss": 1.2581, "step": 1350 }, { "epoch": 1.89, "learning_rate": 3.6210123509273775e-07, "loss": 1.2511, "step": 1360 }, { "epoch": 1.91, "learning_rate": 2.753663889808328e-07, "loss": 1.2505, "step": 1370 }, { "epoch": 1.92, "learning_rate": 2.0043188404937075e-07, "loss": 1.2503, "step": 1380 }, { "epoch": 1.93, "learning_rate": 1.3733348433589143e-07, "loss": 1.2462, "step": 1390 }, { "epoch": 1.95, "learning_rate": 8.610130485265655e-08, "loss": 1.2578, "step": 1400 }, { "epoch": 1.96, "learning_rate": 4.67597972136663e-08, "loss": 1.2563, "step": 1410 }, { "epoch": 1.97, "learning_rate": 1.9327737964616733e-08, "loss": 1.2589, "step": 1420 }, { "epoch": 1.99, "learning_rate": 3.8182196213903955e-09, "loss": 1.2553, "step": 1430 }, { "epoch": 2.0, "step": 1438, "total_flos": 3.15739188563542e+18, "train_loss": 1.285984360292989, "train_runtime": 22790.2039, "train_samples_per_second": 6.057, "train_steps_per_second": 0.063 } ], "max_steps": 1438, "num_train_epochs": 2, "total_flos": 3.15739188563542e+18, "trial_name": null, "trial_params": null }