| { | |
| "best_metric": 0.7276496291160583, | |
| "best_model_checkpoint": "/mnt/bn/qingyi-bn-lq/llama/saved-alpaca-cot13b/checkpoint-2400", | |
| "epoch": 2.4628014366341713, | |
| "global_step": 2400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 2.1559, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 1.736, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 1.2009, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 0.9775, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.0003, | |
| "loss": 0.8852, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 0.0002978738483345145, | |
| "loss": 0.8508, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.000295747696669029, | |
| "loss": 0.8276, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 0.00029362154500354353, | |
| "loss": 0.8102, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 0.0002914953933380581, | |
| "loss": 0.7959, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 0.0002893692416725726, | |
| "loss": 0.7816, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_loss": 0.7917931079864502, | |
| "eval_runtime": 25.1681, | |
| "eval_samples_per_second": 79.466, | |
| "eval_steps_per_second": 1.271, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.00028724309000708714, | |
| "loss": 0.7793, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.00028511693834160166, | |
| "loss": 0.7778, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 0.0002829907866761162, | |
| "loss": 0.7828, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.00028086463501063075, | |
| "loss": 0.7716, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.00027873848334514527, | |
| "loss": 0.7686, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 0.0002766123316796598, | |
| "loss": 0.7618, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 0.0002744861800141743, | |
| "loss": 0.7568, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 0.0002723600283486889, | |
| "loss": 0.7679, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 0.0002702338766832034, | |
| "loss": 0.7713, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.0002681077250177179, | |
| "loss": 0.7516, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_loss": 0.7647445201873779, | |
| "eval_runtime": 25.1482, | |
| "eval_samples_per_second": 79.529, | |
| "eval_steps_per_second": 1.272, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.00026598157335223243, | |
| "loss": 0.7608, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 0.00026385542168674695, | |
| "loss": 0.758, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 0.0002617292700212615, | |
| "loss": 0.7547, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 0.00025960311835577604, | |
| "loss": 0.7593, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.00025747696669029056, | |
| "loss": 0.7537, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 0.0002553508150248051, | |
| "loss": 0.7586, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 0.0002532246633593196, | |
| "loss": 0.7726, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0002510985116938341, | |
| "loss": 0.756, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 0.0002489723600283487, | |
| "loss": 0.7437, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 0.0002468462083628632, | |
| "loss": 0.7498, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "eval_loss": 0.7551296353340149, | |
| "eval_runtime": 25.165, | |
| "eval_samples_per_second": 79.475, | |
| "eval_steps_per_second": 1.272, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.0002447200566973777, | |
| "loss": 0.749, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 0.00024259390503189224, | |
| "loss": 0.7431, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 0.0002404677533664068, | |
| "loss": 0.7533, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.0002383416017009213, | |
| "loss": 0.7486, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 0.00023621545003543583, | |
| "loss": 0.737, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 0.0002340892983699504, | |
| "loss": 0.7463, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 0.00023196314670446492, | |
| "loss": 0.7419, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.00022983699503897943, | |
| "loss": 0.7414, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.00022771084337349395, | |
| "loss": 0.7502, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.00022558469170800847, | |
| "loss": 0.7397, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 0.7478033900260925, | |
| "eval_runtime": 25.1278, | |
| "eval_samples_per_second": 79.593, | |
| "eval_steps_per_second": 1.273, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 0.00022345854004252302, | |
| "loss": 0.7456, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.00022133238837703754, | |
| "loss": 0.7317, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.00021920623671155208, | |
| "loss": 0.7377, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 0.0002170800850460666, | |
| "loss": 0.7439, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 0.00021495393338058114, | |
| "loss": 0.7394, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.00021282778171509566, | |
| "loss": 0.7397, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 0.00021070163004961018, | |
| "loss": 0.7369, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 0.0002085754783841247, | |
| "loss": 0.7365, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.00020644932671863922, | |
| "loss": 0.7443, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 0.0002043231750531538, | |
| "loss": 0.7364, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "eval_loss": 0.7435225248336792, | |
| "eval_runtime": 25.1054, | |
| "eval_samples_per_second": 79.664, | |
| "eval_steps_per_second": 1.275, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 0.0002021970233876683, | |
| "loss": 0.7369, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.00020007087172218283, | |
| "loss": 0.7318, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 0.00019794472005669735, | |
| "loss": 0.734, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.0001958185683912119, | |
| "loss": 0.7299, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.0001936924167257264, | |
| "loss": 0.7352, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 0.00019156626506024093, | |
| "loss": 0.727, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 0.0001894401133947555, | |
| "loss": 0.7262, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 0.00018731396172927002, | |
| "loss": 0.7332, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 0.00018518781006378454, | |
| "loss": 0.7315, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 0.00018306165839829906, | |
| "loss": 0.7391, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_loss": 0.739651620388031, | |
| "eval_runtime": 25.1539, | |
| "eval_samples_per_second": 79.511, | |
| "eval_steps_per_second": 1.272, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 0.00018093550673281358, | |
| "loss": 0.726, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 0.00017880935506732812, | |
| "loss": 0.7358, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 0.00017668320340184267, | |
| "loss": 0.7195, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 0.00017455705173635719, | |
| "loss": 0.7304, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 0.0001724309000708717, | |
| "loss": 0.7281, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 0.00017030474840538625, | |
| "loss": 0.7216, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 0.00016817859673990077, | |
| "loss": 0.7247, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 0.00016605244507441529, | |
| "loss": 0.7251, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 0.0001639262934089298, | |
| "loss": 0.7266, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 0.00016180014174344438, | |
| "loss": 0.7339, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "eval_loss": 0.737014651298523, | |
| "eval_runtime": 25.1254, | |
| "eval_samples_per_second": 79.601, | |
| "eval_steps_per_second": 1.274, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 0.0001596739900779589, | |
| "loss": 0.7336, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 0.00015754783841247341, | |
| "loss": 0.7316, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 0.00015542168674698793, | |
| "loss": 0.7317, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 0.00015329553508150245, | |
| "loss": 0.7289, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 0.000151169383416017, | |
| "loss": 0.7321, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 0.00014904323175053151, | |
| "loss": 0.7213, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 0.00014691708008504606, | |
| "loss": 0.7313, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.00014479092841956058, | |
| "loss": 0.7232, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 0.00014266477675407512, | |
| "loss": 0.7298, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 0.00014053862508858964, | |
| "loss": 0.7203, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_loss": 0.7359894514083862, | |
| "eval_runtime": 25.1784, | |
| "eval_samples_per_second": 79.433, | |
| "eval_steps_per_second": 1.271, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 0.00013841247342310416, | |
| "loss": 0.7187, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 0.0001362863217576187, | |
| "loss": 0.708, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.00013416017009213323, | |
| "loss": 0.7252, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 0.00013203401842664774, | |
| "loss": 0.717, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 0.0001299078667611623, | |
| "loss": 0.7273, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 0.0001277817150956768, | |
| "loss": 0.7244, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 0.00012565556343019135, | |
| "loss": 0.7261, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 0.00012352941176470587, | |
| "loss": 0.7191, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 0.0001214032600992204, | |
| "loss": 0.724, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 0.00011927710843373494, | |
| "loss": 0.714, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_loss": 0.7326391935348511, | |
| "eval_runtime": 25.1862, | |
| "eval_samples_per_second": 79.409, | |
| "eval_steps_per_second": 1.271, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 0.00011715095676824945, | |
| "loss": 0.7196, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 0.000115024805102764, | |
| "loss": 0.7188, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 0.00011289865343727852, | |
| "loss": 0.7223, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 0.00011077250177179304, | |
| "loss": 0.7311, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 0.00010864635010630757, | |
| "loss": 0.7217, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 0.00010652019844082211, | |
| "loss": 0.7205, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 0.00010439404677533663, | |
| "loss": 0.7175, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 0.00010226789510985115, | |
| "loss": 0.7235, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 0.0001001417434443657, | |
| "loss": 0.7158, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 9.801559177888021e-05, | |
| "loss": 0.7084, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_loss": 0.730781614780426, | |
| "eval_runtime": 25.1365, | |
| "eval_samples_per_second": 79.565, | |
| "eval_steps_per_second": 1.273, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 9.588944011339475e-05, | |
| "loss": 0.7159, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 9.376328844790927e-05, | |
| "loss": 0.7207, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 9.163713678242381e-05, | |
| "loss": 0.7107, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 8.951098511693833e-05, | |
| "loss": 0.7151, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 8.738483345145286e-05, | |
| "loss": 0.7157, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 8.525868178596739e-05, | |
| "loss": 0.7163, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 8.313253012048193e-05, | |
| "loss": 0.7152, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 8.100637845499644e-05, | |
| "loss": 0.7159, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 7.888022678951099e-05, | |
| "loss": 0.709, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 7.675407512402551e-05, | |
| "loss": 0.7247, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_loss": 0.7292709350585938, | |
| "eval_runtime": 25.1853, | |
| "eval_samples_per_second": 79.412, | |
| "eval_steps_per_second": 1.271, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 7.462792345854004e-05, | |
| "loss": 0.7214, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 7.250177179305457e-05, | |
| "loss": 0.7067, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 7.037562012756909e-05, | |
| "loss": 0.7264, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 6.824946846208362e-05, | |
| "loss": 0.7144, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 6.612331679659815e-05, | |
| "loss": 0.7184, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 6.399716513111269e-05, | |
| "loss": 0.7181, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 6.187101346562722e-05, | |
| "loss": 0.7165, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 5.9744861800141736e-05, | |
| "loss": 0.7229, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 5.761871013465627e-05, | |
| "loss": 0.7176, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 5.5492558469170794e-05, | |
| "loss": 0.7049, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "eval_loss": 0.7276496291160583, | |
| "eval_runtime": 25.206, | |
| "eval_samples_per_second": 79.346, | |
| "eval_steps_per_second": 1.27, | |
| "step": 2400 | |
| } | |
| ], | |
| "max_steps": 2922, | |
| "num_train_epochs": 3, | |
| "total_flos": 1.2133612842645455e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |