| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.65848934796643, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.3856842517852783, | |
| "learning_rate": 1.9868055555555558e-05, | |
| "loss": 1.1287, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.4272608458995819, | |
| "learning_rate": 1.9736111111111115e-05, | |
| "loss": 1.0586, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.4174883961677551, | |
| "learning_rate": 1.9597222222222224e-05, | |
| "loss": 1.0166, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.35379958152770996, | |
| "learning_rate": 1.9458333333333333e-05, | |
| "loss": 0.8919, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.3820742964744568, | |
| "learning_rate": 1.9319444444444446e-05, | |
| "loss": 0.7971, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.34241142868995667, | |
| "learning_rate": 1.918055555555556e-05, | |
| "loss": 0.7466, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.48230335116386414, | |
| "learning_rate": 1.9041666666666668e-05, | |
| "loss": 0.7468, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.45057639479637146, | |
| "learning_rate": 1.890277777777778e-05, | |
| "loss": 0.7373, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.43228989839553833, | |
| "learning_rate": 1.876388888888889e-05, | |
| "loss": 0.756, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.460918664932251, | |
| "learning_rate": 1.8625000000000002e-05, | |
| "loss": 0.7186, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.8221462368965149, | |
| "learning_rate": 1.8486111111111115e-05, | |
| "loss": 0.6772, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.547721266746521, | |
| "learning_rate": 1.8347222222222224e-05, | |
| "loss": 0.6809, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 0.6005927920341492, | |
| "learning_rate": 1.8208333333333333e-05, | |
| "loss": 0.6651, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.7613247632980347, | |
| "learning_rate": 1.8069444444444446e-05, | |
| "loss": 0.641, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.668274462223053, | |
| "learning_rate": 1.793055555555556e-05, | |
| "loss": 0.6296, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.6101451516151428, | |
| "learning_rate": 1.7791666666666668e-05, | |
| "loss": 0.6033, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.8970944285392761, | |
| "learning_rate": 1.765277777777778e-05, | |
| "loss": 0.5939, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 0.9417675137519836, | |
| "learning_rate": 1.751388888888889e-05, | |
| "loss": 0.5966, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 1.3364688158035278, | |
| "learning_rate": 1.7375000000000002e-05, | |
| "loss": 0.5771, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.7763314843177795, | |
| "learning_rate": 1.7236111111111115e-05, | |
| "loss": 0.5588, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 1.156174659729004, | |
| "learning_rate": 1.7097222222222224e-05, | |
| "loss": 0.5571, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 0.8457236886024475, | |
| "learning_rate": 1.6958333333333333e-05, | |
| "loss": 0.5431, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 1.1372079849243164, | |
| "learning_rate": 1.6819444444444446e-05, | |
| "loss": 0.4839, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 1.0457851886749268, | |
| "learning_rate": 1.668055555555556e-05, | |
| "loss": 0.5041, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 1.1884819269180298, | |
| "learning_rate": 1.6541666666666668e-05, | |
| "loss": 0.468, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "grad_norm": 1.5204098224639893, | |
| "learning_rate": 1.6402777777777777e-05, | |
| "loss": 0.4778, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "grad_norm": 1.3218625783920288, | |
| "learning_rate": 1.626388888888889e-05, | |
| "loss": 0.4682, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "grad_norm": 1.2012380361557007, | |
| "learning_rate": 1.6125000000000002e-05, | |
| "loss": 0.4326, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "grad_norm": 1.4303064346313477, | |
| "learning_rate": 1.5986111111111115e-05, | |
| "loss": 0.4358, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 1.4230263233184814, | |
| "learning_rate": 1.5847222222222224e-05, | |
| "loss": 0.4384, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 1.5604521036148071, | |
| "learning_rate": 1.5708333333333333e-05, | |
| "loss": 0.3908, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "grad_norm": 1.7824558019638062, | |
| "learning_rate": 1.5569444444444446e-05, | |
| "loss": 0.4017, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "grad_norm": 1.4250575304031372, | |
| "learning_rate": 1.543055555555556e-05, | |
| "loss": 0.3759, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 1.5560147762298584, | |
| "learning_rate": 1.5291666666666668e-05, | |
| "loss": 0.3869, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "grad_norm": 1.7027453184127808, | |
| "learning_rate": 1.5152777777777779e-05, | |
| "loss": 0.3591, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.44, | |
| "grad_norm": 1.5872056484222412, | |
| "learning_rate": 1.501388888888889e-05, | |
| "loss": 0.3573, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "grad_norm": 1.5370341539382935, | |
| "learning_rate": 1.4875000000000002e-05, | |
| "loss": 0.3488, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 7.85, | |
| "grad_norm": 1.853333830833435, | |
| "learning_rate": 1.4736111111111113e-05, | |
| "loss": 0.3416, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "grad_norm": 1.5671972036361694, | |
| "learning_rate": 1.4597222222222223e-05, | |
| "loss": 0.3282, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 8.26, | |
| "grad_norm": 1.6593025922775269, | |
| "learning_rate": 1.4458333333333334e-05, | |
| "loss": 0.3112, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "grad_norm": 1.7826825380325317, | |
| "learning_rate": 1.4319444444444446e-05, | |
| "loss": 0.286, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 8.68, | |
| "grad_norm": 1.5925076007843018, | |
| "learning_rate": 1.4180555555555557e-05, | |
| "loss": 0.3281, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 8.88, | |
| "grad_norm": 1.7169785499572754, | |
| "learning_rate": 1.4041666666666666e-05, | |
| "loss": 0.3092, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "grad_norm": 1.8225599527359009, | |
| "learning_rate": 1.3902777777777779e-05, | |
| "loss": 0.2728, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 2.113538980484009, | |
| "learning_rate": 1.376388888888889e-05, | |
| "loss": 0.2893, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 2.0112457275390625, | |
| "learning_rate": 1.3625e-05, | |
| "loss": 0.277, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 9.71, | |
| "grad_norm": 1.9079582691192627, | |
| "learning_rate": 1.3486111111111113e-05, | |
| "loss": 0.261, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 9.92, | |
| "grad_norm": 1.793885350227356, | |
| "learning_rate": 1.3347222222222223e-05, | |
| "loss": 0.267, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 10.12, | |
| "grad_norm": 1.770896315574646, | |
| "learning_rate": 1.3208333333333334e-05, | |
| "loss": 0.2719, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 10.33, | |
| "grad_norm": 1.849857211112976, | |
| "learning_rate": 1.3069444444444446e-05, | |
| "loss": 0.2399, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.54, | |
| "grad_norm": 1.8785293102264404, | |
| "learning_rate": 1.2930555555555557e-05, | |
| "loss": 0.2316, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 10.74, | |
| "grad_norm": 1.7189656496047974, | |
| "learning_rate": 1.2791666666666666e-05, | |
| "loss": 0.243, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 10.95, | |
| "grad_norm": 1.6181199550628662, | |
| "learning_rate": 1.2652777777777779e-05, | |
| "loss": 0.2347, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 11.16, | |
| "grad_norm": 2.0756866931915283, | |
| "learning_rate": 1.251388888888889e-05, | |
| "loss": 0.2407, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "grad_norm": 1.9992755651474, | |
| "learning_rate": 1.2375000000000001e-05, | |
| "loss": 0.2205, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 11.57, | |
| "grad_norm": 2.19657039642334, | |
| "learning_rate": 1.2236111111111114e-05, | |
| "loss": 0.1992, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 11.78, | |
| "grad_norm": 1.8114335536956787, | |
| "learning_rate": 1.2097222222222223e-05, | |
| "loss": 0.219, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 11.98, | |
| "grad_norm": 1.841556429862976, | |
| "learning_rate": 1.1958333333333334e-05, | |
| "loss": 0.2051, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 12.19, | |
| "grad_norm": 1.9361953735351562, | |
| "learning_rate": 1.1819444444444446e-05, | |
| "loss": 0.2025, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 1.841424584388733, | |
| "learning_rate": 1.1680555555555557e-05, | |
| "loss": 0.2109, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 2.569114923477173, | |
| "learning_rate": 1.1541666666666667e-05, | |
| "loss": 0.1933, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 12.81, | |
| "grad_norm": 1.8263764381408691, | |
| "learning_rate": 1.1402777777777777e-05, | |
| "loss": 0.1814, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 1.7250192165374756, | |
| "learning_rate": 1.126388888888889e-05, | |
| "loss": 0.1842, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "grad_norm": 1.8750269412994385, | |
| "learning_rate": 1.1125000000000001e-05, | |
| "loss": 0.1803, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 13.43, | |
| "grad_norm": 2.2998244762420654, | |
| "learning_rate": 1.0986111111111114e-05, | |
| "loss": 0.167, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 13.63, | |
| "grad_norm": 1.663015604019165, | |
| "learning_rate": 1.0847222222222223e-05, | |
| "loss": 0.1723, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 13.84, | |
| "grad_norm": 2.0710666179656982, | |
| "learning_rate": 1.0708333333333334e-05, | |
| "loss": 0.1895, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 14.05, | |
| "grad_norm": 1.8963505029678345, | |
| "learning_rate": 1.0569444444444445e-05, | |
| "loss": 0.1783, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 14.25, | |
| "grad_norm": 2.3760714530944824, | |
| "learning_rate": 1.0430555555555557e-05, | |
| "loss": 0.1674, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 14.46, | |
| "grad_norm": 1.8337997198104858, | |
| "learning_rate": 1.0291666666666667e-05, | |
| "loss": 0.1677, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 14.67, | |
| "grad_norm": 1.8879417181015015, | |
| "learning_rate": 1.0152777777777778e-05, | |
| "loss": 0.1602, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 14.87, | |
| "grad_norm": 1.9163988828659058, | |
| "learning_rate": 1.001388888888889e-05, | |
| "loss": 0.1536, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 15.08, | |
| "grad_norm": 1.772133469581604, | |
| "learning_rate": 9.875000000000001e-06, | |
| "loss": 0.1422, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 15.29, | |
| "grad_norm": 2.173783540725708, | |
| "learning_rate": 9.736111111111112e-06, | |
| "loss": 0.1634, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 15.49, | |
| "grad_norm": 2.1147613525390625, | |
| "learning_rate": 9.597222222222223e-06, | |
| "loss": 0.14, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 15.7, | |
| "grad_norm": 2.0401387214660645, | |
| "learning_rate": 9.458333333333334e-06, | |
| "loss": 0.1498, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 15.91, | |
| "grad_norm": 2.0865604877471924, | |
| "learning_rate": 9.319444444444445e-06, | |
| "loss": 0.1395, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 16.11, | |
| "grad_norm": 1.884482502937317, | |
| "learning_rate": 9.180555555555556e-06, | |
| "loss": 0.1461, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 16.32, | |
| "grad_norm": 2.2012784481048584, | |
| "learning_rate": 9.041666666666667e-06, | |
| "loss": 0.1297, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 16.53, | |
| "grad_norm": 2.0317113399505615, | |
| "learning_rate": 8.902777777777778e-06, | |
| "loss": 0.15, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 16.73, | |
| "grad_norm": 1.4792850017547607, | |
| "learning_rate": 8.76388888888889e-06, | |
| "loss": 0.1359, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 16.94, | |
| "grad_norm": 2.3025472164154053, | |
| "learning_rate": 8.625000000000001e-06, | |
| "loss": 0.1298, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 17.15, | |
| "grad_norm": 1.8987499475479126, | |
| "learning_rate": 8.486111111111112e-06, | |
| "loss": 0.1287, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 17.35, | |
| "grad_norm": 3.6023619174957275, | |
| "learning_rate": 8.347222222222223e-06, | |
| "loss": 0.1584, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 17.56, | |
| "grad_norm": 1.6749480962753296, | |
| "learning_rate": 8.208333333333334e-06, | |
| "loss": 0.1071, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 17.77, | |
| "grad_norm": 1.894116759300232, | |
| "learning_rate": 8.069444444444445e-06, | |
| "loss": 0.1233, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 17.97, | |
| "grad_norm": 1.9405465126037598, | |
| "learning_rate": 7.930555555555556e-06, | |
| "loss": 0.1143, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 18.18, | |
| "grad_norm": 1.8571505546569824, | |
| "learning_rate": 7.791666666666667e-06, | |
| "loss": 0.1197, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 18.39, | |
| "grad_norm": 1.8494367599487305, | |
| "learning_rate": 7.652777777777778e-06, | |
| "loss": 0.133, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 18.59, | |
| "grad_norm": 1.8017449378967285, | |
| "learning_rate": 7.51388888888889e-06, | |
| "loss": 0.1036, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "grad_norm": 1.9734400510787964, | |
| "learning_rate": 7.375000000000001e-06, | |
| "loss": 0.1135, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 1.88029944896698, | |
| "learning_rate": 7.2361111111111115e-06, | |
| "loss": 0.1159, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 19.21, | |
| "grad_norm": 1.6627161502838135, | |
| "learning_rate": 7.097222222222223e-06, | |
| "loss": 0.1138, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 19.42, | |
| "grad_norm": 1.9287075996398926, | |
| "learning_rate": 6.958333333333333e-06, | |
| "loss": 0.1154, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 19.63, | |
| "grad_norm": 1.843787670135498, | |
| "learning_rate": 6.819444444444445e-06, | |
| "loss": 0.1052, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 19.83, | |
| "grad_norm": 2.0632147789001465, | |
| "learning_rate": 6.680555555555556e-06, | |
| "loss": 0.1157, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 20.04, | |
| "grad_norm": 1.6356507539749146, | |
| "learning_rate": 6.541666666666667e-06, | |
| "loss": 0.0993, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "grad_norm": 2.1572136878967285, | |
| "learning_rate": 6.402777777777778e-06, | |
| "loss": 0.0973, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 20.45, | |
| "grad_norm": 1.834170937538147, | |
| "learning_rate": 6.26388888888889e-06, | |
| "loss": 0.0947, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 20.66, | |
| "grad_norm": 1.80930495262146, | |
| "learning_rate": 6.125000000000001e-06, | |
| "loss": 0.1176, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 2880, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "total_flos": 4.626131694005453e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |