| { | |
| "best_metric": 0.14487937092781067, | |
| "best_model_checkpoint": "limbxy_pose/checkpoint-1602", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 1780, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2808988764044944, | |
| "grad_norm": 1.6329148591013448e-13, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3282, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5617977528089888, | |
| "grad_norm": 609844.4375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3197, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8426966292134831, | |
| "grad_norm": 51261.34765625, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.2128, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.17451409995555878, | |
| "eval_rmse": 0.4177488386631012, | |
| "eval_runtime": 9.5905, | |
| "eval_samples_per_second": 104.27, | |
| "eval_steps_per_second": 1.668, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.1235955056179776, | |
| "grad_norm": 2084511.375, | |
| "learning_rate": 2e-05, | |
| "loss": 0.198, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.404494382022472, | |
| "grad_norm": 721293.5625, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1779, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6853932584269664, | |
| "grad_norm": 1280031.75, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1604, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9662921348314608, | |
| "grad_norm": 1277506.125, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1574, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.148615300655365, | |
| "eval_rmse": 0.3855065405368805, | |
| "eval_runtime": 9.3357, | |
| "eval_samples_per_second": 107.116, | |
| "eval_steps_per_second": 1.714, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.247191011235955, | |
| "grad_norm": 1975641.0, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1589, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5280898876404496, | |
| "grad_norm": 2331497.75, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.1757, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.808988764044944, | |
| "grad_norm": 4412882.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2045, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.15186643600463867, | |
| "eval_rmse": 0.3897004723548889, | |
| "eval_runtime": 9.3334, | |
| "eval_samples_per_second": 107.142, | |
| "eval_steps_per_second": 1.714, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.0898876404494384, | |
| "grad_norm": 1961631.5, | |
| "learning_rate": 4.918300653594771e-05, | |
| "loss": 0.1809, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.370786516853933, | |
| "grad_norm": 5652000.5, | |
| "learning_rate": 4.8366013071895424e-05, | |
| "loss": 0.1778, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.6516853932584272, | |
| "grad_norm": 1794376.5, | |
| "learning_rate": 4.7549019607843135e-05, | |
| "loss": 0.1861, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.932584269662921, | |
| "grad_norm": 1615815.75, | |
| "learning_rate": 4.673202614379085e-05, | |
| "loss": 0.1697, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.16323314607143402, | |
| "eval_rmse": 0.4040212333202362, | |
| "eval_runtime": 9.309, | |
| "eval_samples_per_second": 107.423, | |
| "eval_steps_per_second": 1.719, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.213483146067416, | |
| "grad_norm": 2846568.25, | |
| "learning_rate": 4.5915032679738564e-05, | |
| "loss": 0.1743, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.49438202247191, | |
| "grad_norm": 2384729.0, | |
| "learning_rate": 4.5098039215686275e-05, | |
| "loss": 0.1723, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.775280898876405, | |
| "grad_norm": 1569872.375, | |
| "learning_rate": 4.4281045751633986e-05, | |
| "loss": 0.1818, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.19486868381500244, | |
| "eval_rmse": 0.44143933057785034, | |
| "eval_runtime": 9.4254, | |
| "eval_samples_per_second": 106.096, | |
| "eval_steps_per_second": 1.698, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 5.056179775280899, | |
| "grad_norm": 343663.96875, | |
| "learning_rate": 4.3464052287581704e-05, | |
| "loss": 0.1845, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.337078651685394, | |
| "grad_norm": 1288543.625, | |
| "learning_rate": 4.2647058823529415e-05, | |
| "loss": 0.1941, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 5.617977528089888, | |
| "grad_norm": 4547344.0, | |
| "learning_rate": 4.1830065359477126e-05, | |
| "loss": 0.1685, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.898876404494382, | |
| "grad_norm": 185522.03125, | |
| "learning_rate": 4.101307189542484e-05, | |
| "loss": 0.1624, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.14749938249588013, | |
| "eval_rmse": 0.3840564787387848, | |
| "eval_runtime": 9.7128, | |
| "eval_samples_per_second": 102.957, | |
| "eval_steps_per_second": 1.647, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 6.179775280898877, | |
| "grad_norm": 2298808.25, | |
| "learning_rate": 4.0196078431372555e-05, | |
| "loss": 0.1578, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 6.460674157303371, | |
| "grad_norm": 1625852.375, | |
| "learning_rate": 3.9379084967320266e-05, | |
| "loss": 0.162, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 6.741573033707866, | |
| "grad_norm": 2281359.0, | |
| "learning_rate": 3.8562091503267977e-05, | |
| "loss": 0.1645, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.14835500717163086, | |
| "eval_rmse": 0.38516879081726074, | |
| "eval_runtime": 9.1497, | |
| "eval_samples_per_second": 109.293, | |
| "eval_steps_per_second": 1.749, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 7.022471910112359, | |
| "grad_norm": 673713.0625, | |
| "learning_rate": 3.774509803921569e-05, | |
| "loss": 0.1554, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 7.303370786516854, | |
| "grad_norm": 260786.171875, | |
| "learning_rate": 3.6928104575163405e-05, | |
| "loss": 0.1656, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 7.584269662921348, | |
| "grad_norm": 2860977.5, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 0.1642, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 7.865168539325842, | |
| "grad_norm": 1183203.125, | |
| "learning_rate": 3.529411764705883e-05, | |
| "loss": 0.1655, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.14708983898162842, | |
| "eval_rmse": 0.3835229277610779, | |
| "eval_runtime": 9.5185, | |
| "eval_samples_per_second": 105.059, | |
| "eval_steps_per_second": 1.681, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 8.146067415730338, | |
| "grad_norm": 2235303.5, | |
| "learning_rate": 3.447712418300654e-05, | |
| "loss": 0.1543, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 8.426966292134832, | |
| "grad_norm": 1311017.875, | |
| "learning_rate": 3.366013071895425e-05, | |
| "loss": 0.1507, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 8.707865168539326, | |
| "grad_norm": 1330708.25, | |
| "learning_rate": 3.284313725490196e-05, | |
| "loss": 0.1625, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 8.98876404494382, | |
| "grad_norm": 553780.625, | |
| "learning_rate": 3.202614379084967e-05, | |
| "loss": 0.1594, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.15354213118553162, | |
| "eval_rmse": 0.3918445110321045, | |
| "eval_runtime": 9.1616, | |
| "eval_samples_per_second": 109.151, | |
| "eval_steps_per_second": 1.746, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 9.269662921348315, | |
| "grad_norm": 208979.359375, | |
| "learning_rate": 3.120915032679739e-05, | |
| "loss": 0.1518, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 9.55056179775281, | |
| "grad_norm": 78840.34375, | |
| "learning_rate": 3.0392156862745097e-05, | |
| "loss": 0.1552, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 9.831460674157304, | |
| "grad_norm": 1873128.625, | |
| "learning_rate": 2.957516339869281e-05, | |
| "loss": 0.1513, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.1448940634727478, | |
| "eval_rmse": 0.38064953684806824, | |
| "eval_runtime": 9.364, | |
| "eval_samples_per_second": 106.792, | |
| "eval_steps_per_second": 1.709, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 10.112359550561798, | |
| "grad_norm": 1098680.625, | |
| "learning_rate": 2.8758169934640522e-05, | |
| "loss": 0.1551, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.393258426966293, | |
| "grad_norm": 1874154.875, | |
| "learning_rate": 2.7941176470588236e-05, | |
| "loss": 0.1516, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 10.674157303370787, | |
| "grad_norm": 53160.01953125, | |
| "learning_rate": 2.7124183006535947e-05, | |
| "loss": 0.1466, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 10.955056179775282, | |
| "grad_norm": 929561.25, | |
| "learning_rate": 2.630718954248366e-05, | |
| "loss": 0.1488, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.14547079801559448, | |
| "eval_rmse": 0.38140633702278137, | |
| "eval_runtime": 9.5104, | |
| "eval_samples_per_second": 105.148, | |
| "eval_steps_per_second": 1.682, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 11.235955056179776, | |
| "grad_norm": 1361347.875, | |
| "learning_rate": 2.5490196078431373e-05, | |
| "loss": 0.1457, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 11.51685393258427, | |
| "grad_norm": 72723.1953125, | |
| "learning_rate": 2.4673202614379087e-05, | |
| "loss": 0.1508, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 11.797752808988765, | |
| "grad_norm": 93677.0625, | |
| "learning_rate": 2.38562091503268e-05, | |
| "loss": 0.1507, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.1535731852054596, | |
| "eval_rmse": 0.3918841481208801, | |
| "eval_runtime": 9.5233, | |
| "eval_samples_per_second": 105.006, | |
| "eval_steps_per_second": 1.68, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 12.07865168539326, | |
| "grad_norm": 1383022.125, | |
| "learning_rate": 2.303921568627451e-05, | |
| "loss": 0.155, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 12.359550561797754, | |
| "grad_norm": 1435498.5, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.15, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 12.640449438202246, | |
| "grad_norm": 256395.265625, | |
| "learning_rate": 2.1405228758169934e-05, | |
| "loss": 0.1465, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 12.921348314606742, | |
| "grad_norm": 258689.03125, | |
| "learning_rate": 2.058823529411765e-05, | |
| "loss": 0.1522, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.14494504034519196, | |
| "eval_rmse": 0.380716472864151, | |
| "eval_runtime": 9.1341, | |
| "eval_samples_per_second": 109.48, | |
| "eval_steps_per_second": 1.752, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 13.202247191011235, | |
| "grad_norm": 1009406.375, | |
| "learning_rate": 1.977124183006536e-05, | |
| "loss": 0.1491, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 13.48314606741573, | |
| "grad_norm": 720892.125, | |
| "learning_rate": 1.895424836601307e-05, | |
| "loss": 0.1502, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 13.764044943820224, | |
| "grad_norm": 48925.546875, | |
| "learning_rate": 1.8137254901960785e-05, | |
| "loss": 0.1458, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.14527221024036407, | |
| "eval_rmse": 0.3811459243297577, | |
| "eval_runtime": 9.7085, | |
| "eval_samples_per_second": 103.003, | |
| "eval_steps_per_second": 1.648, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 14.044943820224718, | |
| "grad_norm": 421335.125, | |
| "learning_rate": 1.7320261437908496e-05, | |
| "loss": 0.1485, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 14.325842696629213, | |
| "grad_norm": 375955.40625, | |
| "learning_rate": 1.650326797385621e-05, | |
| "loss": 0.1457, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 14.606741573033707, | |
| "grad_norm": 624158.0, | |
| "learning_rate": 1.568627450980392e-05, | |
| "loss": 0.1498, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 14.887640449438202, | |
| "grad_norm": 115186.8984375, | |
| "learning_rate": 1.4869281045751634e-05, | |
| "loss": 0.1506, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.1455306112766266, | |
| "eval_rmse": 0.3814847767353058, | |
| "eval_runtime": 9.2528, | |
| "eval_samples_per_second": 108.075, | |
| "eval_steps_per_second": 1.729, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 15.168539325842696, | |
| "grad_norm": 1805258.0, | |
| "learning_rate": 1.4052287581699347e-05, | |
| "loss": 0.1467, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 15.44943820224719, | |
| "grad_norm": 520502.5625, | |
| "learning_rate": 1.323529411764706e-05, | |
| "loss": 0.1529, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 15.730337078651685, | |
| "grad_norm": 2020412.75, | |
| "learning_rate": 1.2418300653594772e-05, | |
| "loss": 0.1505, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.1451566517353058, | |
| "eval_rmse": 0.3809943199157715, | |
| "eval_runtime": 9.2736, | |
| "eval_samples_per_second": 107.833, | |
| "eval_steps_per_second": 1.725, | |
| "step": 1424 | |
| }, | |
| { | |
| "epoch": 16.01123595505618, | |
| "grad_norm": 252456.1875, | |
| "learning_rate": 1.1601307189542485e-05, | |
| "loss": 0.1481, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 16.292134831460675, | |
| "grad_norm": 267339.03125, | |
| "learning_rate": 1.0784313725490197e-05, | |
| "loss": 0.1464, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 16.573033707865168, | |
| "grad_norm": 482848.15625, | |
| "learning_rate": 9.96732026143791e-06, | |
| "loss": 0.1482, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 16.853932584269664, | |
| "grad_norm": 297641.71875, | |
| "learning_rate": 9.150326797385621e-06, | |
| "loss": 0.1463, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 0.1449102759361267, | |
| "eval_rmse": 0.38067084550857544, | |
| "eval_runtime": 9.457, | |
| "eval_samples_per_second": 105.742, | |
| "eval_steps_per_second": 1.692, | |
| "step": 1513 | |
| }, | |
| { | |
| "epoch": 17.134831460674157, | |
| "grad_norm": 353910.8125, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.1481, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 17.415730337078653, | |
| "grad_norm": 847917.6875, | |
| "learning_rate": 7.5163398692810456e-06, | |
| "loss": 0.1494, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 17.696629213483146, | |
| "grad_norm": 197619.375, | |
| "learning_rate": 6.699346405228758e-06, | |
| "loss": 0.145, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 17.97752808988764, | |
| "grad_norm": 934886.75, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.1463, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 0.14487937092781067, | |
| "eval_rmse": 0.380630224943161, | |
| "eval_runtime": 9.2621, | |
| "eval_samples_per_second": 107.967, | |
| "eval_steps_per_second": 1.727, | |
| "step": 1602 | |
| }, | |
| { | |
| "epoch": 18.258426966292134, | |
| "grad_norm": 178175.765625, | |
| "learning_rate": 5.065359477124184e-06, | |
| "loss": 0.1456, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 18.53932584269663, | |
| "grad_norm": 282639.03125, | |
| "learning_rate": 4.2483660130718954e-06, | |
| "loss": 0.145, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 18.820224719101123, | |
| "grad_norm": 338323.84375, | |
| "learning_rate": 3.431372549019608e-06, | |
| "loss": 0.1494, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 0.1456519514322281, | |
| "eval_rmse": 0.38164374232292175, | |
| "eval_runtime": 9.5216, | |
| "eval_samples_per_second": 105.024, | |
| "eval_steps_per_second": 1.68, | |
| "step": 1691 | |
| }, | |
| { | |
| "epoch": 19.10112359550562, | |
| "grad_norm": 340422.8125, | |
| "learning_rate": 2.6143790849673204e-06, | |
| "loss": 0.147, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 19.382022471910112, | |
| "grad_norm": 366319.65625, | |
| "learning_rate": 1.7973856209150326e-06, | |
| "loss": 0.1471, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 19.662921348314608, | |
| "grad_norm": 642705.4375, | |
| "learning_rate": 9.80392156862745e-07, | |
| "loss": 0.1467, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 19.9438202247191, | |
| "grad_norm": 455884.28125, | |
| "learning_rate": 1.6339869281045752e-07, | |
| "loss": 0.1454, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.14513316750526428, | |
| "eval_rmse": 0.3809635043144226, | |
| "eval_runtime": 9.5919, | |
| "eval_samples_per_second": 104.254, | |
| "eval_steps_per_second": 1.668, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 1780, | |
| "total_flos": 0.0, | |
| "train_loss": 0.16350786438149012, | |
| "train_runtime": 3277.5098, | |
| "train_samples_per_second": 34.551, | |
| "train_steps_per_second": 0.543 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1780, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |