| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 8415, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0023767082590612004, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 2.3767082590612003e-06, | |
| "loss": 3.6948, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.004753416518122401, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 4.753416518122401e-06, | |
| "loss": 3.8478, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0071301247771836, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 7.130124777183601e-06, | |
| "loss": 3.6042, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.009506833036244802, | |
| "grad_norm": 2.625, | |
| "learning_rate": 9.506833036244801e-06, | |
| "loss": 3.5338, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.011883541295306001, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.1883541295306002e-05, | |
| "loss": 3.2504, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0142602495543672, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.4260249554367203e-05, | |
| "loss": 2.956, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.016636957813428402, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.66369578134284e-05, | |
| "loss": 2.9073, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.019013666072489603, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.9013666072489602e-05, | |
| "loss": 2.6436, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0213903743315508, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 2.1390374331550803e-05, | |
| "loss": 2.6018, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.023767082590612002, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 2.3767082590612004e-05, | |
| "loss": 2.4289, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.026143790849673203, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 2.6143790849673204e-05, | |
| "loss": 2.5239, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.0285204991087344, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 2.8520499108734405e-05, | |
| "loss": 2.4617, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.030897207367795602, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 3.08972073677956e-05, | |
| "loss": 2.396, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.033273915626856804, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 3.32739156268568e-05, | |
| "loss": 2.2682, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.035650623885918005, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 3.5650623885918004e-05, | |
| "loss": 2.3627, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.038027332144979206, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 3.8027332144979205e-05, | |
| "loss": 2.3072, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 4.0404040404040405e-05, | |
| "loss": 2.2124, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.0427807486631016, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 4.2780748663101606e-05, | |
| "loss": 2.2619, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.0451574569221628, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 4.515745692216281e-05, | |
| "loss": 2.1928, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.047534165181224004, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 4.753416518122401e-05, | |
| "loss": 2.0428, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.049910873440285206, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 4.991087344028521e-05, | |
| "loss": 1.9929, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.05228758169934641, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 5.228758169934641e-05, | |
| "loss": 2.146, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.05466428995840761, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 5.466428995840761e-05, | |
| "loss": 2.0289, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.0570409982174688, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 5.704099821746881e-05, | |
| "loss": 1.9554, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.059417706476530004, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 5.941770647653001e-05, | |
| "loss": 1.9871, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.061794414735591205, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 6.17944147355912e-05, | |
| "loss": 2.0099, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.06417112299465241, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.41711229946524e-05, | |
| "loss": 1.9199, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.06654783125371361, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 6.65478312537136e-05, | |
| "loss": 2.0074, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.0689245395127748, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 6.892453951277481e-05, | |
| "loss": 1.8052, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.07130124777183601, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.130124777183601e-05, | |
| "loss": 2.0366, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.0736779560308972, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 7.367795603089721e-05, | |
| "loss": 1.8426, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.07605466428995841, | |
| "grad_norm": 3.75, | |
| "learning_rate": 7.605466428995841e-05, | |
| "loss": 1.6836, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.0784313725490196, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 7.843137254901961e-05, | |
| "loss": 2.0483, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 8.080808080808081e-05, | |
| "loss": 1.902, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.08318478906714201, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.318478906714201e-05, | |
| "loss": 1.9887, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.0855614973262032, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 8.556149732620321e-05, | |
| "loss": 1.6542, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.08793820558526441, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 8.793820558526441e-05, | |
| "loss": 1.6718, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.0903149138443256, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 9.031491384432561e-05, | |
| "loss": 1.706, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.09269162210338681, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 9.269162210338681e-05, | |
| "loss": 1.629, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.09506833036244801, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 9.506833036244802e-05, | |
| "loss": 1.6748, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0974450386215092, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 9.744503862150922e-05, | |
| "loss": 1.8269, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.09982174688057041, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.982174688057042e-05, | |
| "loss": 1.8727, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.1021984551396316, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 0.00010219845513963162, | |
| "loss": 1.7455, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.10457516339869281, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 0.00010457516339869282, | |
| "loss": 1.7395, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.10695187165775401, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 0.00010695187165775402, | |
| "loss": 1.7935, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.10932857991681522, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 0.00010932857991681522, | |
| "loss": 1.8009, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.11170528817587641, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 0.00011170528817587642, | |
| "loss": 1.5582, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.1140819964349376, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00011408199643493762, | |
| "loss": 1.6402, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.11645870469399881, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 0.00011645870469399882, | |
| "loss": 1.665, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.11883541295306001, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 0.00011883541295306002, | |
| "loss": 1.8407, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 0.00012121212121212122, | |
| "loss": 1.7665, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.12358882947118241, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 0.0001235888294711824, | |
| "loss": 1.8593, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.1259655377302436, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 0.00012596553773024362, | |
| "loss": 1.7738, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.12834224598930483, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.0001283422459893048, | |
| "loss": 1.8227, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.13071895424836602, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 0.00013071895424836603, | |
| "loss": 1.7159, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.13309566250742721, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 0.0001330956625074272, | |
| "loss": 1.6953, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.1354723707664884, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 0.00013547237076648843, | |
| "loss": 1.5567, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.1378490790255496, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.00013784907902554961, | |
| "loss": 1.7637, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.14022578728461083, | |
| "grad_norm": 2.375, | |
| "learning_rate": 0.00014022578728461083, | |
| "loss": 1.5884, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.14260249554367202, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.00014260249554367202, | |
| "loss": 1.5542, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.14497920380273321, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 0.00014497920380273323, | |
| "loss": 1.5124, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.1473559120617944, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 0.00014735591206179442, | |
| "loss": 1.5822, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.1497326203208556, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 0.00014973262032085563, | |
| "loss": 1.6216, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.15210932857991682, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00015210932857991682, | |
| "loss": 1.7366, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.15448603683897802, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 0.00015448603683897803, | |
| "loss": 1.3673, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.1568627450980392, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 0.00015686274509803922, | |
| "loss": 1.4071, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.1592394533571004, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 0.00015923945335710043, | |
| "loss": 1.6515, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 0.00016161616161616162, | |
| "loss": 1.7604, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.16399286987522282, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 0.0001639928698752228, | |
| "loss": 1.5925, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.16636957813428402, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 0.00016636957813428402, | |
| "loss": 1.5874, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.1687462863933452, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 0.0001687462863933452, | |
| "loss": 1.6116, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.1711229946524064, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 0.00017112299465240642, | |
| "loss": 1.544, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.17349970291146763, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 0.0001734997029114676, | |
| "loss": 1.5015, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.17587641117052882, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 0.00017587641117052883, | |
| "loss": 1.5858, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.17825311942959002, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 0.00017825311942959, | |
| "loss": 1.5469, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1806298276886512, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 0.00018062982768865123, | |
| "loss": 1.5092, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.1830065359477124, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 0.00018300653594771241, | |
| "loss": 1.3362, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.18538324420677363, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 0.00018538324420677363, | |
| "loss": 1.5088, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.18775995246583482, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 0.00018775995246583482, | |
| "loss": 1.5518, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.19013666072489602, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 0.00019013666072489603, | |
| "loss": 1.4344, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.1925133689839572, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 0.00019251336898395722, | |
| "loss": 1.5411, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.1948900772430184, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.00019489007724301843, | |
| "loss": 1.5832, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.19726678550207963, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 0.00019726678550207962, | |
| "loss": 1.5468, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.19964349376114082, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.00019964349376114083, | |
| "loss": 1.3053, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 0.0001999968531423333, | |
| "loss": 1.5303, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.2043969102792632, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.0001999850935511802, | |
| "loss": 1.6331, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.20677361853832443, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 0.0001999646242513938, | |
| "loss": 1.5621, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.20915032679738563, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.00019993544702605638, | |
| "loss": 1.4932, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.21152703505644682, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 0.00019989756441679795, | |
| "loss": 1.3936, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.21390374331550802, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 0.00019985097972357547, | |
| "loss": 1.3227, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.2162804515745692, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 0.0001997956970043848, | |
| "loss": 1.4253, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.21865715983363043, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 0.00019973172107490758, | |
| "loss": 1.3946, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.22103386809269163, | |
| "grad_norm": 1.625, | |
| "learning_rate": 0.00019965905750809158, | |
| "loss": 1.2779, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.22341057635175282, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 0.00019957771263366526, | |
| "loss": 1.5051, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.22578728461081402, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 0.00019948769353758653, | |
| "loss": 1.3954, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.2281639928698752, | |
| "grad_norm": 2.25, | |
| "learning_rate": 0.00019938900806142518, | |
| "loss": 1.2049, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.23054070112893643, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 0.00019928166480168011, | |
| "loss": 1.5335, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.23291740938799763, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 0.00019916567310903035, | |
| "loss": 1.3868, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.0001990410430875205, | |
| "loss": 1.3677, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.23767082590612001, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.0001989077855936807, | |
| "loss": 1.2757, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.24004753416518124, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 0.0001987659122355806, | |
| "loss": 1.4893, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 0.00019861543537181867, | |
| "loss": 1.2123, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.24480095068330363, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.00019845636811044521, | |
| "loss": 1.4714, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.24717765894236482, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 0.0001982887243078207, | |
| "loss": 1.463, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.24955436720142601, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 0.00019811251856740873, | |
| "loss": 1.5086, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2519310754604872, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.000197927766238504, | |
| "loss": 1.3604, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.25430778371954843, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 0.00019773448341489495, | |
| "loss": 1.2226, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.25668449197860965, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 0.00019753268693346225, | |
| "loss": 1.2003, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.2590612002376708, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 0.0001973223943727117, | |
| "loss": 1.192, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.26143790849673204, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 0.00019710362405124334, | |
| "loss": 1.2616, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2638146167557932, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.0001968763950261554, | |
| "loss": 1.2493, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.26619132501485443, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 0.0001966407270913846, | |
| "loss": 1.278, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.26856803327391565, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.00019639664077598142, | |
| "loss": 1.3797, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.2709447415329768, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.0001961441573423223, | |
| "loss": 1.1243, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.27332144979203804, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.0001958832987842571, | |
| "loss": 1.324, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.2756981580510992, | |
| "grad_norm": 2.375, | |
| "learning_rate": 0.00019561408782519345, | |
| "loss": 1.3572, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.27807486631016043, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 0.0001953365479161172, | |
| "loss": 1.3404, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.28045157456922165, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 0.00019505070323354965, | |
| "loss": 1.3237, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 0.0001947565786774415, | |
| "loss": 1.2894, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.28520499108734404, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 0.0001944541998690038, | |
| "loss": 1.3445, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2875816993464052, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 0.0001941435931484761, | |
| "loss": 1.3859, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.28995840760546643, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 0.00019382478557283204, | |
| "loss": 1.2185, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.29233511586452765, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.00019349780491342223, | |
| "loss": 1.3545, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.2947118241235888, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00019316267965355528, | |
| "loss": 1.2694, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.29708853238265004, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 0.00019281943898601645, | |
| "loss": 1.0126, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2994652406417112, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 0.00019246811281052487, | |
| "loss": 1.2593, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.3018419489007724, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.00019210873173112865, | |
| "loss": 1.3609, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.30421865715983365, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.0001917413270535393, | |
| "loss": 0.9368, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.3065953654188948, | |
| "grad_norm": 2.25, | |
| "learning_rate": 0.0001913659307824045, | |
| "loss": 1.1893, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.30897207367795604, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 0.0001909825756185202, | |
| "loss": 1.1345, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.3113487819370172, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.0001905912949559821, | |
| "loss": 1.2925, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.3137254901960784, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.00019019212287927663, | |
| "loss": 1.0842, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.31610219845513965, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 0.00018978509416031186, | |
| "loss": 1.2722, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.3184789067142008, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.00018937024425538855, | |
| "loss": 1.3566, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.32085561497326204, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 0.0001889476093021115, | |
| "loss": 0.9899, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 0.00018851722611624164, | |
| "loss": 1.0713, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.3256090314913844, | |
| "grad_norm": 2.25, | |
| "learning_rate": 0.00018807913218848906, | |
| "loss": 1.1957, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.32798573975044565, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 0.000187633365681247, | |
| "loss": 1.3258, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.3303624480095068, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 0.00018717996542526777, | |
| "loss": 1.1146, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.33273915626856804, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 0.00018671897091627993, | |
| "loss": 1.0215, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.33511586452762926, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.00018625042231154817, | |
| "loss": 1.1107, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.3374925727866904, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 0.00018577436042637477, | |
| "loss": 1.1655, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.33986928104575165, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 0.00018529082673054457, | |
| "loss": 1.1197, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.3422459893048128, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 0.0001847998633447123, | |
| "loss": 1.1792, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.34462269756387404, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 0.0001843015130367335, | |
| "loss": 1.1933, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.34699940582293526, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 0.00018379581921793914, | |
| "loss": 1.1603, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.3493761140819964, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00018328282593935377, | |
| "loss": 1.1653, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.35175282234105765, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 0.00018276257788785855, | |
| "loss": 1.0807, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.3541295306001188, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 0.00018223512038229833, | |
| "loss": 1.0923, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.35650623885918004, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 0.00018170049936953406, | |
| "loss": 1.3022, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.35888294711824126, | |
| "grad_norm": 1.625, | |
| "learning_rate": 0.00018115876142044032, | |
| "loss": 1.2415, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.3612596553773024, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 0.0001806099537258485, | |
| "loss": 1.0849, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 0.00018005412409243606, | |
| "loss": 1.1712, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.3660130718954248, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.000179491320938562, | |
| "loss": 1.2795, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.36838978015448604, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.00017892159329004916, | |
| "loss": 1.1304, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.37076648841354726, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 0.00017834499077591374, | |
| "loss": 1.277, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.3731431966726084, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 0.00017776156362404186, | |
| "loss": 1.2222, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.37551990493166965, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.0001771713626568143, | |
| "loss": 1.2639, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.3778966131907308, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 0.0001765744392866795, | |
| "loss": 1.1679, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.38027332144979203, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 0.00017597084551167476, | |
| "loss": 1.0657, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.38265002970885326, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00017536063391089697, | |
| "loss": 1.1886, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.3850267379679144, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.00017474385763992212, | |
| "loss": 1.1526, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.38740344622697565, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 0.00017412057042617525, | |
| "loss": 1.0648, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.3897801544860368, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 0.00017349082656424995, | |
| "loss": 1.2938, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.39215686274509803, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 0.00017285468091117904, | |
| "loss": 1.1246, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.39453357100415926, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 0.00017221218888165572, | |
| "loss": 1.0179, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.3969102792632204, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 0.0001715634064432065, | |
| "loss": 0.914, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.39928698752228164, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 0.0001709083901113159, | |
| "loss": 1.1192, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.40166369578134287, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 0.00017024719694450337, | |
| "loss": 1.1462, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 0.00016957988453935276, | |
| "loss": 0.9452, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.40641711229946526, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 0.00016890651102549538, | |
| "loss": 1.1837, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.4087938205585264, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 0.00016822713506054604, | |
| "loss": 0.8889, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.41117052881758764, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.0001675418158249935, | |
| "loss": 1.0873, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.41354723707664887, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 0.0001668506130170453, | |
| "loss": 0.995, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.41592394533571003, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 0.0001661535868474273, | |
| "loss": 1.1595, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.41830065359477125, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 0.00016545079803413892, | |
| "loss": 1.1134, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.4206773618538324, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 0.00016474230779716384, | |
| "loss": 0.7802, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.42305407011289364, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 0.00016402817785313712, | |
| "loss": 0.9085, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.42543077837195487, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 0.00016330847040996915, | |
| "loss": 1.0042, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.42780748663101603, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 0.00016258324816142668, | |
| "loss": 1.0318, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.43018419489007725, | |
| "grad_norm": 3.375, | |
| "learning_rate": 0.00016185257428167143, | |
| "loss": 1.0833, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.4325609031491384, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.00016111651241975734, | |
| "loss": 1.0208, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.43493761140819964, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00016037512669408565, | |
| "loss": 1.0074, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.43731431966726086, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.0001596284816868198, | |
| "loss": 1.2403, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.43969102792632203, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 0.00015887664243825967, | |
| "loss": 1.0694, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.44206773618538325, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.0001581196744411759, | |
| "loss": 1.0033, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 0.0001573576436351046, | |
| "loss": 1.0574, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.44682115270350564, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 0.00015659061640060378, | |
| "loss": 1.2204, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.44919786096256686, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 0.0001558186595534705, | |
| "loss": 1.157, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.45157456922162803, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 0.0001550418403389208, | |
| "loss": 0.9542, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.45395127748068925, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 0.00015426022642573193, | |
| "loss": 0.9281, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.4563279857397504, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 0.00015347388590034757, | |
| "loss": 0.9331, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.45870469399881164, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 0.00015268288726094705, | |
| "loss": 0.7978, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.46108140225787286, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 0.00015188729941147824, | |
| "loss": 0.9807, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.46345811051693403, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 0.0001510871916556555, | |
| "loss": 1.0974, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.46583481877599525, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 0.00015028263369092253, | |
| "loss": 0.9807, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.4682115270350565, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 0.00014947369560238104, | |
| "loss": 1.0588, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 0.00014866044785668563, | |
| "loss": 1.0442, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.47296494355317886, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 0.00014784296129590548, | |
| "loss": 0.8837, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.47534165181224003, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 0.00014702130713135317, | |
| "loss": 1.129, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.47771836007130125, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 0.00014619555693738166, | |
| "loss": 1.0496, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.4800950683303625, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 0.0001453657826451493, | |
| "loss": 1.0136, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.48247177658942364, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 0.00014453205653635376, | |
| "loss": 0.9794, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 0.00014369445123693596, | |
| "loss": 0.7651, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.48722519310754603, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 0.0001428530397107533, | |
| "loss": 0.9667, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.48960190136660725, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 0.0001420078952532238, | |
| "loss": 0.8897, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.4919786096256685, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 0.0001411590914849415, | |
| "loss": 1.0208, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.49435531788472964, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00014030670234526323, | |
| "loss": 1.0138, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.49673202614379086, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00013945080208586775, | |
| "loss": 0.8435, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.49910873440285203, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 0.0001385914652642877, | |
| "loss": 0.7405, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5014854426619133, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 0.00013772876673741498, | |
| "loss": 0.8999, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.5038621509209744, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 0.00013686278165497977, | |
| "loss": 0.9063, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.5062388591800356, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 0.00013599358545300438, | |
| "loss": 0.8485, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.5086155674390969, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 0.00013512125384723204, | |
| "loss": 0.7843, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.5109922756981581, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 0.00013424586282653116, | |
| "loss": 0.7548, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.5133689839572193, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 0.00013336748864627593, | |
| "loss": 0.9609, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.5157456922162804, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 0.00013248620782170396, | |
| "loss": 0.9358, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.5181224004753416, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 0.00013160209712125074, | |
| "loss": 1.0639, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.5204991087344029, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.0001307152335598624, | |
| "loss": 1.0147, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.5228758169934641, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.00012982569439228713, | |
| "loss": 0.7749, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 0.0001289335571063453, | |
| "loss": 0.642, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.5276292335115864, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 0.00012803889941617944, | |
| "loss": 0.863, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.5300059417706476, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 0.0001271417992554849, | |
| "loss": 0.8817, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.5323826500297089, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 0.00012624233477072057, | |
| "loss": 0.9788, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.5347593582887701, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.00012534058431430198, | |
| "loss": 0.8273, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5371360665478313, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.0001244366264377757, | |
| "loss": 0.848, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.5395127748068924, | |
| "grad_norm": 2.625, | |
| "learning_rate": 0.00012353053988497684, | |
| "loss": 0.8556, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.5418894830659536, | |
| "grad_norm": 2.375, | |
| "learning_rate": 0.00012262240358516967, | |
| "loss": 0.8137, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.5442661913250149, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00012171229664617208, | |
| "loss": 0.7362, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.5466428995840761, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 0.00012080029834746447, | |
| "loss": 0.7504, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.5490196078431373, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 0.00011988648813328367, | |
| "loss": 1.0486, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.5513963161021984, | |
| "grad_norm": 1.875, | |
| "learning_rate": 0.00011897094560570265, | |
| "loss": 0.7487, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.5537730243612596, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 0.00011805375051769636, | |
| "loss": 0.9325, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.5561497326203209, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 0.00011713498276619432, | |
| "loss": 0.906, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.5585264408793821, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 0.0001162147223851209, | |
| "loss": 0.7214, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.5609031491384433, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.0001152930495384236, | |
| "loss": 1.01, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.5632798573975044, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 0.00011437004451308983, | |
| "loss": 1.0897, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 0.00011344578771215319, | |
| "loss": 0.8222, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.5680332739156269, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 0.00011252035964768961, | |
| "loss": 0.8411, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.5704099821746881, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 0.00011159384093380377, | |
| "loss": 0.747, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5727866904337493, | |
| "grad_norm": 2.375, | |
| "learning_rate": 0.00011066631227960693, | |
| "loss": 0.7352, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.5751633986928104, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.00010973785448218639, | |
| "loss": 0.901, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.5775401069518716, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 0.00010880854841956712, | |
| "loss": 0.6489, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.5799168152109329, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 0.00010787847504366649, | |
| "loss": 0.7034, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.5822935234699941, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 0.00010694771537324269, | |
| "loss": 0.7427, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.5846702317290553, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 0.00010601635048683698, | |
| "loss": 0.7655, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.5870469399881164, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 0.00010508446151571109, | |
| "loss": 0.7467, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.5894236482471776, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 0.0001041521296367798, | |
| "loss": 0.7942, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.5918003565062389, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 0.00010321943606553961, | |
| "loss": 1.033, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.5941770647653001, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 0.00010228646204899401, | |
| "loss": 0.7049, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5965537730243613, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 0.0001013532888585762, | |
| "loss": 0.7016, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.5989304812834224, | |
| "grad_norm": 2.875, | |
| "learning_rate": 0.00010041999778306936, | |
| "loss": 0.9511, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.6013071895424836, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 9.948667012152566e-05, | |
| "loss": 0.6895, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.6036838978015449, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 9.855338717618432e-05, | |
| "loss": 0.8696, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 2.625, | |
| "learning_rate": 9.762023024538926e-05, | |
| "loss": 0.8621, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.6084373143196673, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 9.668728061650733e-05, | |
| "loss": 0.9548, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.6108140225787284, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 9.575461955884726e-05, | |
| "loss": 0.8048, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.6131907308377896, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.482232831658034e-05, | |
| "loss": 0.8866, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.6155674390968509, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 9.389048810166317e-05, | |
| "loss": 0.7404, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.6179441473559121, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 9.29591800867634e-05, | |
| "loss": 0.6817, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.6203208556149733, | |
| "grad_norm": 5.0, | |
| "learning_rate": 9.202848539818865e-05, | |
| "loss": 0.7504, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.6226975638740344, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 9.109848510881961e-05, | |
| "loss": 0.7767, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.6250742721330956, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 9.016926023104789e-05, | |
| "loss": 0.7846, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.6274509803921569, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 8.924089170971887e-05, | |
| "loss": 0.8327, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.6298276886512181, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 8.831346041508069e-05, | |
| "loss": 0.8035, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.6322043969102793, | |
| "grad_norm": 2.25, | |
| "learning_rate": 8.738704713573959e-05, | |
| "loss": 0.9569, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.6345811051693404, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.646173257162245e-05, | |
| "loss": 0.6629, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.6369578134284016, | |
| "grad_norm": 2.875, | |
| "learning_rate": 8.553759732694696e-05, | |
| "loss": 0.6597, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.6393345216874629, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 8.461472190320021e-05, | |
| "loss": 0.6857, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.6417112299465241, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 8.369318669212625e-05, | |
| "loss": 0.7234, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.6440879382055853, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 8.277307196872303e-05, | |
| "loss": 0.7483, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 8.185445788424974e-05, | |
| "loss": 0.7538, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.6488413547237076, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 8.093742445924491e-05, | |
| "loss": 0.7882, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.6512180629827689, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 8.002205157655554e-05, | |
| "loss": 0.6984, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.6535947712418301, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.910841897437875e-05, | |
| "loss": 0.7044, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6559714795008913, | |
| "grad_norm": 3.390625, | |
| "learning_rate": 7.819660623931575e-05, | |
| "loss": 0.6569, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.6583481877599525, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 7.728669279943897e-05, | |
| "loss": 0.8069, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.6607248960190136, | |
| "grad_norm": 3.125, | |
| "learning_rate": 7.637875791737299e-05, | |
| "loss": 0.759, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.6631016042780749, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 7.547288068339025e-05, | |
| "loss": 0.6806, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.6654783125371361, | |
| "grad_norm": 2.875, | |
| "learning_rate": 7.456914000852123e-05, | |
| "loss": 0.7777, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.6678550207961973, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 7.366761461768052e-05, | |
| "loss": 0.7375, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.6702317290552585, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.276838304280935e-05, | |
| "loss": 0.6862, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.6726084373143196, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 7.187152361603432e-05, | |
| "loss": 0.6195, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.6749851455733809, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 7.097711446284405e-05, | |
| "loss": 0.5486, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.6773618538324421, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.008523349528377e-05, | |
| "loss": 0.7452, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6797385620915033, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 6.919595840516815e-05, | |
| "loss": 0.7417, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.6821152703505645, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 6.830936665731371e-05, | |
| "loss": 0.6798, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.6844919786096256, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 6.742553548279095e-05, | |
| "loss": 0.7222, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 6.654454187219649e-05, | |
| "loss": 0.6629, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.6892453951277481, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 6.56664625689466e-05, | |
| "loss": 0.7808, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.6916221033868093, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 6.479137406259206e-05, | |
| "loss": 0.7898, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.6939988116458705, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 6.39193525821551e-05, | |
| "loss": 0.8595, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.6963755199049316, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 6.305047408948908e-05, | |
| "loss": 0.8169, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.6987522281639929, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 6.21848142726615e-05, | |
| "loss": 0.7892, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.7011289364230541, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 6.13224485393608e-05, | |
| "loss": 0.8179, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.7035056446821153, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 6.046345201032748e-05, | |
| "loss": 0.6881, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 5.960789951281052e-05, | |
| "loss": 0.9263, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.7082590612002376, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.8755865574049016e-05, | |
| "loss": 0.7661, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.7106357694592988, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 5.7907424414780135e-05, | |
| "loss": 0.7932, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.7130124777183601, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 5.706264994277386e-05, | |
| "loss": 0.8045, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7153891859774213, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 5.6221615746394644e-05, | |
| "loss": 0.6996, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.7177658942364825, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 5.538439508819139e-05, | |
| "loss": 0.7662, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.7201426024955436, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 5.4551060898515404e-05, | |
| "loss": 0.6049, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.7225193107546048, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.372168576916732e-05, | |
| "loss": 0.6502, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.7248960190136661, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 5.289634194707387e-05, | |
| "loss": 0.6785, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 5.207510132799436e-05, | |
| "loss": 0.7902, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.7296494355317885, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 5.125803545025758e-05, | |
| "loss": 0.6948, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.7320261437908496, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 5.0445215488530525e-05, | |
| "loss": 0.7113, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.7344028520499108, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 4.963671224761808e-05, | |
| "loss": 0.6249, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.7367795603089721, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 4.883259615629515e-05, | |
| "loss": 0.7295, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.7391562685680333, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 4.8032937261171896e-05, | |
| "loss": 0.8041, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.7415329768270945, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 4.7237805220591744e-05, | |
| "loss": 0.6296, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.7439096850861556, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 4.644726929856342e-05, | |
| "loss": 0.8887, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.7462863933452168, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 4.5661398358727524e-05, | |
| "loss": 0.6501, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.7486631016042781, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 4.4880260858357746e-05, | |
| "loss": 0.5981, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.7510398098633393, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 4.4103924842397395e-05, | |
| "loss": 0.5993, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.7534165181224005, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 4.3332457937532246e-05, | |
| "loss": 0.8187, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.7557932263814616, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 4.256592734629947e-05, | |
| "loss": 0.826, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.7581699346405228, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 4.18043998412335e-05, | |
| "loss": 0.5091, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.7605466428995841, | |
| "grad_norm": 0.78125, | |
| "learning_rate": 4.104794175904966e-05, | |
| "loss": 0.609, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.7629233511586453, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 4.02966189948655e-05, | |
| "loss": 0.7031, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.7653000594177065, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 3.955049699646054e-05, | |
| "loss": 0.5454, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 3.880964075857535e-05, | |
| "loss": 0.629, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.7700534759358288, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 3.80741148172497e-05, | |
| "loss": 0.6376, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.7724301841948901, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 3.734398324420073e-05, | |
| "loss": 0.6744, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7748068924539513, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 3.661930964124193e-05, | |
| "loss": 0.6133, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.7771836007130125, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.5900157134742574e-05, | |
| "loss": 0.5723, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.7795603089720736, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 3.5186588370128746e-05, | |
| "loss": 0.6248, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.7819370172311348, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 3.447866550642649e-05, | |
| "loss": 0.6323, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 3.377645021084701e-05, | |
| "loss": 0.7705, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7866904337492573, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 3.3080003653414724e-05, | |
| "loss": 0.5787, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.7890671420083185, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 3.238938650163899e-05, | |
| "loss": 0.7117, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.7914438502673797, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 3.17046589152292e-05, | |
| "loss": 0.5979, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.7938205585264408, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 3.10258805408542e-05, | |
| "loss": 0.6202, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.7961972667855021, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 3.0353110506946647e-05, | |
| "loss": 0.6462, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.7985739750445633, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 2.968640741855223e-05, | |
| "loss": 0.576, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.8009506833036245, | |
| "grad_norm": 1.875, | |
| "learning_rate": 2.9025829352224477e-05, | |
| "loss": 0.7645, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.8033273915626857, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 2.8371433850965922e-05, | |
| "loss": 0.6348, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.8057040998217468, | |
| "grad_norm": 2.875, | |
| "learning_rate": 2.7723277919215397e-05, | |
| "loss": 0.7165, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 2.70814180178823e-05, | |
| "loss": 0.6772, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.8104575163398693, | |
| "grad_norm": 3.25, | |
| "learning_rate": 2.644591005942846e-05, | |
| "loss": 0.6995, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.8128342245989305, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 2.5816809402997522e-05, | |
| "loss": 0.6519, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.8152109328579917, | |
| "grad_norm": 4.375, | |
| "learning_rate": 2.5194170849592492e-05, | |
| "loss": 0.612, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.8175876411170528, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 2.4578048637302208e-05, | |
| "loss": 0.8347, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.8199643493761141, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 2.396849643657657e-05, | |
| "loss": 0.7846, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.8223410576351753, | |
| "grad_norm": 3.0, | |
| "learning_rate": 2.3365567345551233e-05, | |
| "loss": 0.6799, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 0.8247177658942365, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.276931388542235e-05, | |
| "loss": 0.704, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 0.8270944741532977, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 2.2179787995871403e-05, | |
| "loss": 0.5832, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.8294711824123588, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 2.1597041030540643e-05, | |
| "loss": 0.6464, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 0.8318478906714201, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 2.1021123752559836e-05, | |
| "loss": 0.6596, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8342245989304813, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 2.0452086330124164e-05, | |
| "loss": 0.6124, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 0.8366013071895425, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.988997833212406e-05, | |
| "loss": 0.8721, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.8389780154486037, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.933484872382737e-05, | |
| "loss": 0.6729, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 0.8413547237076648, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 1.8786745862613885e-05, | |
| "loss": 0.7747, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 0.8437314319667261, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 1.8245717493762925e-05, | |
| "loss": 0.6807, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.8461081402257873, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 1.7711810746294312e-05, | |
| "loss": 0.851, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.7185072128862933e-05, | |
| "loss": 0.6449, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 0.8508615567439097, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.6665547525707316e-05, | |
| "loss": 0.6683, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 0.8532382650029708, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.6153282192652698e-05, | |
| "loss": 0.5818, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 0.8556149732620321, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.5648320753168844e-05, | |
| "loss": 0.6614, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.8579916815210933, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.5150707194482695e-05, | |
| "loss": 0.7086, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 0.8603683897801545, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 1.4660484863746938e-05, | |
| "loss": 0.778, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 0.8627450980392157, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.4177696464263723e-05, | |
| "loss": 0.7629, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 0.8651218062982768, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 1.3702384051765005e-05, | |
| "loss": 0.8923, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 0.8674985145573381, | |
| "grad_norm": 4.125, | |
| "learning_rate": 1.3234589030748956e-05, | |
| "loss": 0.6005, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.8698752228163993, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 1.2774352150873203e-05, | |
| "loss": 0.741, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 0.8722519310754605, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 1.2321713503405208e-05, | |
| "loss": 0.5276, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 0.8746286393345217, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 1.187671251772987e-05, | |
| "loss": 0.5997, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.8770053475935828, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.143938795791476e-05, | |
| "loss": 0.6653, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 0.8793820558526441, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 1.1009777919333507e-05, | |
| "loss": 0.5658, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.8817587641117053, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.0587919825347236e-05, | |
| "loss": 0.6695, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 0.8841354723707665, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.0173850424044596e-05, | |
| "loss": 0.68, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.8865121806298277, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 9.76760578504068e-06, | |
| "loss": 0.8137, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 9.369221296335006e-06, | |
| "loss": 0.7129, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 0.8912655971479501, | |
| "grad_norm": 4.625, | |
| "learning_rate": 8.978731661228768e-06, | |
| "loss": 0.6628, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8936423054070113, | |
| "grad_norm": 4.125, | |
| "learning_rate": 8.596170895301959e-06, | |
| "loss": 0.6621, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.8960190136660725, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 8.221572323450222e-06, | |
| "loss": 0.6171, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 0.8983957219251337, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 7.854968576981824e-06, | |
| "loss": 0.5361, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 0.9007724301841948, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 7.49639159077532e-06, | |
| "loss": 0.6402, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 0.9031491384432561, | |
| "grad_norm": 3.25, | |
| "learning_rate": 7.145872600497561e-06, | |
| "loss": 0.5515, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.9055258467023173, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.8034421398827765e-06, | |
| "loss": 0.6165, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 0.9079025549613785, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 6.469130038072835e-06, | |
| "loss": 0.8606, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 0.9102792632204397, | |
| "grad_norm": 2.875, | |
| "learning_rate": 6.142965417018798e-06, | |
| "loss": 0.6364, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 0.9126559714795008, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 5.824976688944051e-06, | |
| "loss": 0.578, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.9150326797385621, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.515191553869381e-06, | |
| "loss": 0.5698, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.9174093879976233, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 5.213636997200044e-06, | |
| "loss": 0.6748, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 0.9197860962566845, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 4.920339287374942e-06, | |
| "loss": 0.7089, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 0.9221628045157457, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 4.635323973578543e-06, | |
| "loss": 0.7648, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 0.9245395127748068, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 4.3586158835151495e-06, | |
| "loss": 0.6778, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 0.9269162210338681, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 4.090239121246231e-06, | |
| "loss": 0.7377, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 3.5, | |
| "learning_rate": 3.830217065090702e-06, | |
| "loss": 0.6488, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 0.9316696375519905, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 3.5785723655884287e-06, | |
| "loss": 0.5778, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.9340463458110517, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 3.335326943527117e-06, | |
| "loss": 0.6642, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 0.936423054070113, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 3.100501988032878e-06, | |
| "loss": 0.5973, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 0.9387997623291741, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 2.874117954724309e-06, | |
| "loss": 0.618, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 2.656194563930714e-06, | |
| "loss": 0.7053, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.9435531788472965, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 2.446750798974229e-06, | |
| "loss": 0.6074, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 0.9459298871063577, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 2.2458049045161244e-06, | |
| "loss": 0.5246, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 0.948306595365419, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 2.0533743849676436e-06, | |
| "loss": 0.648, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 0.9506833036244801, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 1.869476002965065e-06, | |
| "loss": 0.6625, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9530600118835413, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.69412577790955e-06, | |
| "loss": 0.7165, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 0.9554367201426025, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.5273389845717245e-06, | |
| "loss": 0.6024, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 0.9578134284016637, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 1.3691301517610554e-06, | |
| "loss": 0.6466, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 0.960190136660725, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.2195130610602623e-06, | |
| "loss": 0.6671, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 0.9625668449197861, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.0785007456247886e-06, | |
| "loss": 0.628, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.9649435531788473, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 9.461054890474996e-07, | |
| "loss": 0.776, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 0.9673202614379085, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 8.223388242886265e-07, | |
| "loss": 0.6368, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 2.625, | |
| "learning_rate": 7.072115326711704e-07, | |
| "loss": 0.5938, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.972073677956031, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 6.00733642941742e-07, | |
| "loss": 0.6331, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 0.9744503862150921, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 5.029144303968724e-07, | |
| "loss": 0.566, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.9768270944741533, | |
| "grad_norm": 3.5, | |
| "learning_rate": 4.1376241607518074e-07, | |
| "loss": 0.6529, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 0.9792038027332145, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 3.332853660149904e-07, | |
| "loss": 0.7165, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 0.9815805109922757, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 2.6149029057785936e-07, | |
| "loss": 0.6302, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 0.983957219251337, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 1.9838344383793505e-07, | |
| "loss": 0.6525, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 0.9863339275103981, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 1.4397032303715697e-07, | |
| "loss": 0.6024, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.9887106357694593, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.825566810633958e-08, | |
| "loss": 0.4994, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.9910873440285205, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 6.124346125233604e-08, | |
| "loss": 0.7707, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 0.9934640522875817, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 3.2936926611149e-08, | |
| "loss": 0.6724, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 0.995840760546643, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.3338529967010793e-08, | |
| "loss": 0.6268, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 0.9982174688057041, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 2.449978537655273e-09, | |
| "loss": 0.8173, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 8415, | |
| "total_flos": 9.688250026819584e+16, | |
| "train_loss": 1.1022938394574842, | |
| "train_runtime": 7319.3616, | |
| "train_samples_per_second": 1.15, | |
| "train_steps_per_second": 1.15 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 8415, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.688250026819584e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |