| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.4601171698963498, | |
| "eval_steps": 70, | |
| "global_step": 695, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003605227579990987, | |
| "grad_norm": 25.625, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 2.2117, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003605227579990987, | |
| "eval_loss": 2.2794668674468994, | |
| "eval_runtime": 172.286, | |
| "eval_samples_per_second": 5.038, | |
| "eval_steps_per_second": 2.519, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007210455159981974, | |
| "grad_norm": 25.875, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 2.3877, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010815682739972961, | |
| "grad_norm": 23.625, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 2.2174, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014420910319963948, | |
| "grad_norm": 23.75, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 2.0817, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.018026137899954935, | |
| "grad_norm": 23.375, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 2.1586, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021631365479945923, | |
| "grad_norm": 21.25, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 2.1557, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.025236593059936908, | |
| "grad_norm": 24.625, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 2.2864, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.028841820639927896, | |
| "grad_norm": 23.5, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 2.2588, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03244704821991888, | |
| "grad_norm": 24.25, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 2.2526, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03605227579990987, | |
| "grad_norm": 20.0, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.1614, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03965750337990086, | |
| "grad_norm": 18.75, | |
| "learning_rate": 2.2e-06, | |
| "loss": 2.195, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.043262730959891846, | |
| "grad_norm": 16.25, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.9388, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04686795853988283, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 2.6e-06, | |
| "loss": 1.9927, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.050473186119873815, | |
| "grad_norm": 10.5, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.9328, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.054078413699864804, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 3e-06, | |
| "loss": 2.1901, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05768364127985579, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.9234, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06128886885984678, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 1.8984, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06489409643983776, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.9257, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06849932401982875, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 1.841, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07210455159981974, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.8246, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07570977917981073, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 1.9791, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07931500675980171, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.8834, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0829202343397927, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 1.9163, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08652546191978369, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.7272, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.09013068949977468, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.771, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09373591707976565, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 5.2e-06, | |
| "loss": 1.8819, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09734114465975664, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 1.8637, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.10094637223974763, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.6805, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10455159981973862, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.8e-06, | |
| "loss": 1.7992, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10815682739972961, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7768, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1117620549797206, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 1.769, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11536728255971158, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.732, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11897251013970257, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 1.6727, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.12257773771969356, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.8317, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.12618296529968454, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7e-06, | |
| "loss": 1.6101, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12978819287967552, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.7853, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1333934204596665, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.4e-06, | |
| "loss": 1.942, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1369986480396575, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 1.7218, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1406038756196485, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 1.7993, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.14420910319963948, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.699, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14781433077963047, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 8.2e-06, | |
| "loss": 1.8091, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15141955835962145, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.7783, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.15502478593961244, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 8.6e-06, | |
| "loss": 1.7347, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.15863001351960343, | |
| "grad_norm": 1.75, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.6694, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.16223524109959442, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 9e-06, | |
| "loss": 1.6252, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1658404686795854, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 1.652, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1694456962595764, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 9.4e-06, | |
| "loss": 1.7591, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.17305092383956738, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.5724, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.17665615141955837, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 1.6818, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.18026137899954936, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7152, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18386660657954032, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.02e-05, | |
| "loss": 1.7289, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1874718341595313, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.04e-05, | |
| "loss": 1.7513, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1910770617395223, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.0600000000000002e-05, | |
| "loss": 1.6292, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.19468228931951329, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.0800000000000002e-05, | |
| "loss": 1.6733, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.19828751689950427, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 1.609, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.20189274447949526, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 1.6479, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.20549797205948625, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.14e-05, | |
| "loss": 1.5294, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.20910319963947724, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.16e-05, | |
| "loss": 1.7388, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.21270842721946823, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.18e-05, | |
| "loss": 1.5196, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.21631365479945922, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.7729, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2199188823794502, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.22e-05, | |
| "loss": 1.663, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2235241099594412, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.2400000000000002e-05, | |
| "loss": 1.7202, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.22712933753943218, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.2600000000000001e-05, | |
| "loss": 1.7875, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.23073456511942317, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 1.6535, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.23433979269941416, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 1.7317, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.23794502027940514, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 1.615, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.24155024785939613, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.3400000000000002e-05, | |
| "loss": 1.5796, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.24515547543938712, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 1.7032, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2487607030193781, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.38e-05, | |
| "loss": 1.6784, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.25236593059936907, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.5509, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25236593059936907, | |
| "eval_loss": 1.6160491704940796, | |
| "eval_runtime": 172.7827, | |
| "eval_samples_per_second": 5.024, | |
| "eval_steps_per_second": 2.512, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.25597115817936006, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.4200000000000001e-05, | |
| "loss": 1.7056, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.25957638575935105, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 1.5903, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.26318161333934204, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.46e-05, | |
| "loss": 1.7223, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.266786840919333, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.48e-05, | |
| "loss": 1.5832, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.270392068499324, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 1.6549, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.273997296079315, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 1.6179, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.277602523659306, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.54e-05, | |
| "loss": 1.4685, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.281207751239297, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 1.5328, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.28481297881928797, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.58e-05, | |
| "loss": 1.5563, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.28841820639927895, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.5276, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.29202343397926994, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.62e-05, | |
| "loss": 1.6386, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.29562866155926093, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.64e-05, | |
| "loss": 1.5232, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2992338891392519, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.66e-05, | |
| "loss": 1.7037, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3028391167192429, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 1.5738, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3064443442992339, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.7e-05, | |
| "loss": 1.7061, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3100495718792249, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.72e-05, | |
| "loss": 1.6186, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.31365479945921587, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.7400000000000003e-05, | |
| "loss": 1.7213, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.31726002703920686, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.76e-05, | |
| "loss": 1.5853, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.32086525461919785, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.7800000000000002e-05, | |
| "loss": 1.7614, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.32447048219918884, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.7156, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3280757097791798, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.8200000000000002e-05, | |
| "loss": 1.5067, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3316809373591708, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 1.4872, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3352861649391618, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.86e-05, | |
| "loss": 1.5932, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3388913925191528, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.88e-05, | |
| "loss": 1.5279, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3424966200991438, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.6528, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.34610184767913477, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 1.7462, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.34970707525912575, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.94e-05, | |
| "loss": 1.7114, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.35331230283911674, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 1.4267, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.35691753041910773, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.98e-05, | |
| "loss": 1.4906, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3605227579990987, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4999, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.36412798557908965, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.9999907650547006e-05, | |
| "loss": 1.682, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.36773321315908064, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.999963060389371e-05, | |
| "loss": 1.5293, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.37133844073907163, | |
| "grad_norm": 2.625, | |
| "learning_rate": 1.9999168865157137e-05, | |
| "loss": 1.4611, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3749436683190626, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.999852244286554e-05, | |
| "loss": 1.554, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3785488958990536, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.9997691348958278e-05, | |
| "loss": 1.586, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3821541234790446, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 1.999667559878556e-05, | |
| "loss": 1.7053, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3857593510590356, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.9995475211108183e-05, | |
| "loss": 1.5266, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.38936457863902657, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9994090208097176e-05, | |
| "loss": 1.6351, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.39296980621901756, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.9992520615333393e-05, | |
| "loss": 1.7012, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.39657503379900855, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9990766461807037e-05, | |
| "loss": 1.5982, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.40018026137899954, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.9988827779917138e-05, | |
| "loss": 1.6744, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4037854889589905, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9986704605470932e-05, | |
| "loss": 1.4991, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.4073907165389815, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.9984396977683223e-05, | |
| "loss": 1.4345, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4109959441189725, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.998190493917564e-05, | |
| "loss": 1.5644, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4146011716989635, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.9979228535975866e-05, | |
| "loss": 1.6319, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4182063992789545, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.9976367817516773e-05, | |
| "loss": 1.6227, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.42181162685894547, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9973322836635517e-05, | |
| "loss": 1.6121, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.42541685443893645, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.9970093649572567e-05, | |
| "loss": 1.5419, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.42902208201892744, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9966680315970647e-05, | |
| "loss": 1.5033, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.43262730959891843, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.996308289887366e-05, | |
| "loss": 1.6686, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4362325371789094, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9959301464725507e-05, | |
| "loss": 1.4628, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4398377647589004, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.995533608336886e-05, | |
| "loss": 1.6448, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4434429923388914, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.995118682804388e-05, | |
| "loss": 1.6058, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4470482199188824, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.9946853775386857e-05, | |
| "loss": 1.609, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.45065344749887337, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.9942337005428805e-05, | |
| "loss": 1.5685, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.45425867507886436, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9937636601593965e-05, | |
| "loss": 1.5097, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.45786390265885535, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.9932752650698285e-05, | |
| "loss": 1.5108, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.46146913023884634, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.9927685242947804e-05, | |
| "loss": 1.4884, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4650743578188373, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.9922434471936987e-05, | |
| "loss": 1.6259, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4686795853988283, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9917000434647e-05, | |
| "loss": 1.588, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4722848129788193, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.991138323144392e-05, | |
| "loss": 1.4927, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4758900405588103, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.990558296607687e-05, | |
| "loss": 1.3742, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4794952681388013, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9899599745676123e-05, | |
| "loss": 1.4905, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.48310049571879227, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 1.9893433680751105e-05, | |
| "loss": 1.4683, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.48670572329878325, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9887084885188354e-05, | |
| "loss": 1.5295, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.49031095087877424, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.9880553476249437e-05, | |
| "loss": 1.4426, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.49391617845876523, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.9873839574568756e-05, | |
| "loss": 1.6697, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4975214060387562, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9866943304151346e-05, | |
| "loss": 1.5263, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5011266336187472, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.9859864792370565e-05, | |
| "loss": 1.6241, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5047318611987381, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.985260416996575e-05, | |
| "loss": 1.5458, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5047318611987381, | |
| "eval_loss": 1.511067509651184, | |
| "eval_runtime": 172.8086, | |
| "eval_samples_per_second": 5.023, | |
| "eval_steps_per_second": 2.511, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5083370887787292, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9845161571039805e-05, | |
| "loss": 1.5775, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5119423163587201, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.983753713305672e-05, | |
| "loss": 1.5197, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5155475439387112, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.982973099683902e-05, | |
| "loss": 1.5779, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5191527715187021, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.98217433065652e-05, | |
| "loss": 1.4857, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5227579990986931, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.9813574209767013e-05, | |
| "loss": 1.352, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5263632266786841, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9805223857326794e-05, | |
| "loss": 1.5467, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5299684542586751, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.9796692403474632e-05, | |
| "loss": 1.5134, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.533573681838666, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.9787980005785553e-05, | |
| "loss": 1.4909, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5371789094186571, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.977908682517658e-05, | |
| "loss": 1.5335, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.540784136998648, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.9770013025903797e-05, | |
| "loss": 1.5908, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5443893645786391, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9760758775559275e-05, | |
| "loss": 1.5053, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.54799459215863, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9751324245068008e-05, | |
| "loss": 1.5144, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.551599819738621, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.974170960868474e-05, | |
| "loss": 1.4234, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.555205047318612, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.973191504399076e-05, | |
| "loss": 1.3503, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.558810274898603, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.97219407318906e-05, | |
| "loss": 1.5818, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.562415502478594, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9711786856608714e-05, | |
| "loss": 1.6081, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.566020730058585, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.970145360568607e-05, | |
| "loss": 1.53, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5696259576385759, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.969094116997668e-05, | |
| "loss": 1.466, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.573231185218567, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.968024974364408e-05, | |
| "loss": 1.5987, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5768364127985579, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.9669379524157755e-05, | |
| "loss": 1.5858, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.580441640378549, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9658330712289456e-05, | |
| "loss": 1.4567, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5840468679585399, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.9647103512109535e-05, | |
| "loss": 1.5982, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5876520955385308, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.9635698130983153e-05, | |
| "loss": 1.5962, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5912573231185219, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.962411477956645e-05, | |
| "loss": 1.5821, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5948625506985128, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.9612353671802658e-05, | |
| "loss": 1.6279, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5984677782785038, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.960041502491815e-05, | |
| "loss": 1.5282, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6020730058584948, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9588299059418434e-05, | |
| "loss": 1.4798, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6056782334384858, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.957600599908406e-05, | |
| "loss": 1.564, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6092834610184767, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.9563536070966513e-05, | |
| "loss": 1.4839, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6128886885984678, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.9550889505383996e-05, | |
| "loss": 1.3866, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6164939161784587, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.9538066535917196e-05, | |
| "loss": 1.5959, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6200991437584498, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 1.952506739940496e-05, | |
| "loss": 1.4689, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6237043713384407, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9511892335939904e-05, | |
| "loss": 1.4304, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6273095989184317, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9498541588864022e-05, | |
| "loss": 1.5569, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6309148264984227, | |
| "grad_norm": 2.375, | |
| "learning_rate": 1.948501540476414e-05, | |
| "loss": 1.5979, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6345200540784137, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.9471314033467413e-05, | |
| "loss": 1.5403, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6381252816584047, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.945743772803666e-05, | |
| "loss": 1.5151, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6417305092383957, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.9443386744765726e-05, | |
| "loss": 1.4184, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6453357368183866, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.942916134317473e-05, | |
| "loss": 1.5652, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6489409643983777, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.9414761786005293e-05, | |
| "loss": 1.5839, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6525461919783686, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.9400188339215657e-05, | |
| "loss": 1.5668, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6561514195583596, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.9385441271975786e-05, | |
| "loss": 1.53, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6597566471383506, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9370520856662406e-05, | |
| "loss": 1.4765, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6633618747183416, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.9355427368853946e-05, | |
| "loss": 1.4927, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6669671022983326, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.9340161087325483e-05, | |
| "loss": 1.5432, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6705723298783236, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.932472229404356e-05, | |
| "loss": 1.4757, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6741775574583145, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9309111274161005e-05, | |
| "loss": 1.6556, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6777827850383056, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.9293328316011645e-05, | |
| "loss": 1.362, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6813880126182965, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.927737371110499e-05, | |
| "loss": 1.4231, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6849932401982876, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9261247754120846e-05, | |
| "loss": 1.5703, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6885984677782785, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.924495074290388e-05, | |
| "loss": 1.5593, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6922036953582695, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.92284829784581e-05, | |
| "loss": 1.4376, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6958089229382605, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.9211844764941318e-05, | |
| "loss": 1.4063, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6994141505182515, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.919503640965951e-05, | |
| "loss": 1.4886, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7030193780982424, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.917805822306117e-05, | |
| "loss": 1.5224, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7066246056782335, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.916091051873154e-05, | |
| "loss": 1.4299, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7102298332582244, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.9143593613386845e-05, | |
| "loss": 1.6403, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7138350608382155, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.9126107826868436e-05, | |
| "loss": 1.5683, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7174402884182064, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.9108453482136866e-05, | |
| "loss": 1.4434, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7210455159981974, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.9090630905265963e-05, | |
| "loss": 1.4824, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7246507435781884, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 1.9072640425436762e-05, | |
| "loss": 1.5429, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7282559711581793, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.905448237493147e-05, | |
| "loss": 1.502, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.7318611987381703, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.9036157089127278e-05, | |
| "loss": 1.5172, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7354664263181613, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.901766490649022e-05, | |
| "loss": 1.8393, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7390716538981523, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.8999006168568883e-05, | |
| "loss": 1.589, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7426768814781433, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.8980181219988117e-05, | |
| "loss": 1.5266, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.7462821090581343, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.8961190408442662e-05, | |
| "loss": 1.5054, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7498873366381252, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.8942034084690727e-05, | |
| "loss": 1.55, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7534925642181163, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.8922712602547516e-05, | |
| "loss": 1.5821, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.7570977917981072, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.89032263188787e-05, | |
| "loss": 1.3585, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7570977917981072, | |
| "eval_loss": 1.456952452659607, | |
| "eval_runtime": 172.8993, | |
| "eval_samples_per_second": 5.02, | |
| "eval_steps_per_second": 2.51, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7607030193780983, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.8883575593593793e-05, | |
| "loss": 1.4735, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7643082469580892, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.8863760789639548e-05, | |
| "loss": 1.4517, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7679134745380802, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.8843782272993225e-05, | |
| "loss": 1.497, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7715187021180712, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.8823640412655844e-05, | |
| "loss": 1.5062, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7751239296980622, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.880333558064536e-05, | |
| "loss": 1.535, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7787291572780531, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.878286815198979e-05, | |
| "loss": 1.5964, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7823343848580442, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 1.876223850472032e-05, | |
| "loss": 1.43, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7859396124380351, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.8741447019864263e-05, | |
| "loss": 1.4559, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7895448400180262, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.872049408143808e-05, | |
| "loss": 1.654, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7931500675980171, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.8699380076440242e-05, | |
| "loss": 1.3525, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7967552951780081, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.8678105394844114e-05, | |
| "loss": 1.6202, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8003605227579991, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.8656670429590745e-05, | |
| "loss": 1.6778, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.8039657503379901, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.8635075576581587e-05, | |
| "loss": 1.4164, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.807570977917981, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.861332123467122e-05, | |
| "loss": 1.4554, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8111762054979721, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.859140780565996e-05, | |
| "loss": 1.5984, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.814781433077963, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.856933569428644e-05, | |
| "loss": 1.4117, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8183866606579541, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.8547105308220142e-05, | |
| "loss": 1.5561, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.821991888237945, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.852471705805387e-05, | |
| "loss": 1.5021, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.825597115817936, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.8502171357296144e-05, | |
| "loss": 1.3699, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.829202343397927, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.84794686223636e-05, | |
| "loss": 1.4277, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.832807570977918, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.8456609272573268e-05, | |
| "loss": 1.6487, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.836412798557909, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.8433593730134835e-05, | |
| "loss": 1.6057, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8400180261379, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.841042242014285e-05, | |
| "loss": 1.6268, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8436232537178909, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.838709577056888e-05, | |
| "loss": 1.3636, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.847228481297882, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.8363614212253585e-05, | |
| "loss": 1.447, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8508337088778729, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.833997817889878e-05, | |
| "loss": 1.612, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.854438936457864, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.8316188107059418e-05, | |
| "loss": 1.6382, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.8580441640378549, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.8292244436135517e-05, | |
| "loss": 1.5208, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8616493916178459, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.8268147608364068e-05, | |
| "loss": 1.5397, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8652546191978369, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.8243898068810833e-05, | |
| "loss": 1.4479, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8688598467778279, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.8219496265362164e-05, | |
| "loss": 1.4628, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8724650743578188, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.81949426487167e-05, | |
| "loss": 1.3236, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8760703019378098, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.8170237672377046e-05, | |
| "loss": 1.477, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8796755295178008, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.814538179264142e-05, | |
| "loss": 1.5331, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8832807570977917, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.81203754685952e-05, | |
| "loss": 1.4563, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8868859846777828, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.8095219162102453e-05, | |
| "loss": 1.3979, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8904912122577737, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.8069913337797414e-05, | |
| "loss": 1.4939, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8940964398377648, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.804445846307588e-05, | |
| "loss": 1.5643, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8977016674177557, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.801885500808661e-05, | |
| "loss": 1.6322, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9013068949977467, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.7993103445722615e-05, | |
| "loss": 1.4375, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9049121225777377, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.7967204251612432e-05, | |
| "loss": 1.4036, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9085173501577287, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.7941157904111346e-05, | |
| "loss": 1.3537, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.9121225777377197, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.7914964884292543e-05, | |
| "loss": 1.4312, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.9157278053177107, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.7888625675938237e-05, | |
| "loss": 1.4993, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.9193330328977016, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.7862140765530718e-05, | |
| "loss": 1.3346, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9229382604776927, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.783551064224339e-05, | |
| "loss": 1.6882, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.9265434880576836, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.7808735797931715e-05, | |
| "loss": 1.3881, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.9301487156376747, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.7781816727124138e-05, | |
| "loss": 1.4029, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.9337539432176656, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.7754753927012955e-05, | |
| "loss": 1.3836, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.9373591707976566, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.7727547897445117e-05, | |
| "loss": 1.5992, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9409643983776476, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.770019914091302e-05, | |
| "loss": 1.388, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9445696259576386, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 1.76727081625452e-05, | |
| "loss": 1.2611, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9481748535376295, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.7645075470097024e-05, | |
| "loss": 1.4325, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9517800811176206, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.7617301573941296e-05, | |
| "loss": 1.3901, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9553853086976115, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.758938698705884e-05, | |
| "loss": 1.4556, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9589905362776026, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.7561332225029022e-05, | |
| "loss": 1.5814, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9625957638575935, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.7533137806020226e-05, | |
| "loss": 1.3302, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9662009914375845, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.7504804250780292e-05, | |
| "loss": 1.5922, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9698062190175755, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 1.747633208262688e-05, | |
| "loss": 1.5888, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9734114465975665, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.744772182743782e-05, | |
| "loss": 1.6008, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9770166741775574, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.74189740136414e-05, | |
| "loss": 1.4963, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9806219017575485, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.7390089172206594e-05, | |
| "loss": 1.4549, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.9842271293375394, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.736106783663326e-05, | |
| "loss": 1.5186, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9878323569175305, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7331910542942298e-05, | |
| "loss": 1.5276, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9914375844975214, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.7302617829665725e-05, | |
| "loss": 1.4613, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9950428120775124, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7273190237836757e-05, | |
| "loss": 1.5259, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9986480396575034, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.7243628310979793e-05, | |
| "loss": 1.2902, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.0022532672374944, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.7213932595100384e-05, | |
| "loss": 1.4297, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.0058584948174853, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.7184103638675157e-05, | |
| "loss": 1.3548, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.0094637223974763, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.715414199264168e-05, | |
| "loss": 1.3931, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0094637223974763, | |
| "eval_loss": 1.413288950920105, | |
| "eval_runtime": 172.9216, | |
| "eval_samples_per_second": 5.02, | |
| "eval_steps_per_second": 2.51, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0130689499774674, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.7124048210388268e-05, | |
| "loss": 1.4812, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.0166741775574584, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.709382284774379e-05, | |
| "loss": 1.5801, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.0202794051374493, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.706346646296739e-05, | |
| "loss": 1.2822, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.0238846327174402, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.7032979616738167e-05, | |
| "loss": 1.4649, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.003605227579991, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.7002362872144843e-05, | |
| "loss": 1.3172, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.0072104551599819, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.697161679467534e-05, | |
| "loss": 1.2756, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.010815682739973, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.6940741952206342e-05, | |
| "loss": 1.2556, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.014420910319964, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.6909738914992812e-05, | |
| "loss": 1.206, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.0180261378999549, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 1.6878608255657457e-05, | |
| "loss": 1.2713, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.0216313654799458, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 1.6847350549180148e-05, | |
| "loss": 1.2716, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.025236593059937, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 1.6815966372887305e-05, | |
| "loss": 1.2531, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.028841820639928, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.6784456306441234e-05, | |
| "loss": 1.4101, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.0324470482199188, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 1.675282093182941e-05, | |
| "loss": 1.2461, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.0360522757999098, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.672106083335374e-05, | |
| "loss": 1.3521, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.039657503379901, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.6689176597619773e-05, | |
| "loss": 1.2929, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.0432627309598919, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.6657168813525855e-05, | |
| "loss": 1.3157, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.0468679585398828, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 1.662503807225225e-05, | |
| "loss": 1.2546, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.0504731861198737, | |
| "grad_norm": 2.5, | |
| "learning_rate": 1.659278496725024e-05, | |
| "loss": 1.3864, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.0540784136998649, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.6560410094231144e-05, | |
| "loss": 1.2212, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.0576836412798558, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.6527914051155328e-05, | |
| "loss": 1.4141, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0612888688598467, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.6495297438221145e-05, | |
| "loss": 1.3772, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.0648940964398377, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.6462560857853876e-05, | |
| "loss": 1.2321, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.0684993240198288, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.6429704914694573e-05, | |
| "loss": 1.1723, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.0721045515998198, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.6396730215588913e-05, | |
| "loss": 1.5368, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.0757097791798107, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.6363637369575984e-05, | |
| "loss": 1.2853, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0793150067598016, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.633042698787703e-05, | |
| "loss": 1.2167, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.0829202343397928, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 1.6297099683884163e-05, | |
| "loss": 1.308, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.0865254619197837, | |
| "grad_norm": 2.625, | |
| "learning_rate": 1.626365607314905e-05, | |
| "loss": 1.2853, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.0901306894997747, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.6230096773371514e-05, | |
| "loss": 1.3477, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.0937359170797656, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.619642240438816e-05, | |
| "loss": 1.2775, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0973411446597567, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.616263358816089e-05, | |
| "loss": 1.2302, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.1009463722397477, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 1.612873094876545e-05, | |
| "loss": 1.3578, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.1045515998197386, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.6094715112379874e-05, | |
| "loss": 1.2641, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.1081568273997295, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.6060586707272943e-05, | |
| "loss": 1.1083, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.1117620549797207, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.6026346363792565e-05, | |
| "loss": 1.224, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.1153672825597116, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 1.599199471435414e-05, | |
| "loss": 1.3007, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.1189725101397026, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.5957532393428872e-05, | |
| "loss": 1.276, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.1225777377196935, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.5922960037532057e-05, | |
| "loss": 1.7491, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.1261829652996846, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.588827828521133e-05, | |
| "loss": 1.2594, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.1297881928796756, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.585348777703486e-05, | |
| "loss": 1.3198, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.1333934204596665, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.581858915557953e-05, | |
| "loss": 1.1544, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.1369986480396574, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.5783583065419054e-05, | |
| "loss": 1.2363, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.1406038756196484, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.5748470153112093e-05, | |
| "loss": 1.2597, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.1442091031996395, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.57132510671903e-05, | |
| "loss": 1.2573, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.1478143307796305, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.5677926458146327e-05, | |
| "loss": 1.5368, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.1514195583596214, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.5642496978421842e-05, | |
| "loss": 1.4603, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.1550247859396126, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.560696328239547e-05, | |
| "loss": 1.1522, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.1586300135196035, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.5571326026370676e-05, | |
| "loss": 1.0869, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.1622352410995944, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.5535585868563688e-05, | |
| "loss": 1.2467, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.1658404686795854, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.5499743469091303e-05, | |
| "loss": 1.3221, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1694456962595763, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.5463799489958727e-05, | |
| "loss": 1.2027, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.1730509238395674, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.542775459504732e-05, | |
| "loss": 1.1461, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.1766561514195584, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.5391609450102346e-05, | |
| "loss": 1.2016, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.1802613789995493, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.5355364722720674e-05, | |
| "loss": 1.1223, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.1838666065795402, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5319021082338458e-05, | |
| "loss": 1.2911, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1874718341595314, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.5282579200218762e-05, | |
| "loss": 1.3201, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.1910770617395223, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.5246039749439159e-05, | |
| "loss": 1.3201, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.1946822893195133, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5209403404879305e-05, | |
| "loss": 1.2462, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.1982875168995042, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.5172670843208477e-05, | |
| "loss": 1.3508, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.2018927444794953, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5135842742873077e-05, | |
| "loss": 1.2456, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.2054979720594863, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5098919784084083e-05, | |
| "loss": 1.1362, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.2091031996394772, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.5061902648804503e-05, | |
| "loss": 1.298, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.2127084272194681, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.502479202073678e-05, | |
| "loss": 1.1552, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.2163136547994593, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.4987588585310154e-05, | |
| "loss": 1.2798, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.2199188823794502, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 1.4950293029668004e-05, | |
| "loss": 1.2284, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.2235241099594412, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.4912906042655164e-05, | |
| "loss": 1.4853, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.227129337539432, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.4875428314805195e-05, | |
| "loss": 1.3625, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.2307345651194233, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 1.483786053832763e-05, | |
| "loss": 1.4404, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.2343397926994142, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.4800203407095194e-05, | |
| "loss": 1.293, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.2379450202794051, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.4762457616630972e-05, | |
| "loss": 1.1668, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.2379450202794051, | |
| "eval_loss": 1.3923176527023315, | |
| "eval_runtime": 173.0224, | |
| "eval_samples_per_second": 5.017, | |
| "eval_steps_per_second": 2.508, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.241550247859396, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.4724623864095595e-05, | |
| "loss": 1.3352, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.2451554754393872, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.4686702848274328e-05, | |
| "loss": 1.2931, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.2487607030193781, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.4648695269564182e-05, | |
| "loss": 1.4859, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.252365930599369, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.461060182996098e-05, | |
| "loss": 1.2265, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.25597115817936, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.4572423233046386e-05, | |
| "loss": 1.197, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.2595763857593512, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 1.4534160183974908e-05, | |
| "loss": 1.357, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.263181613339342, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.4495813389460875e-05, | |
| "loss": 1.2742, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.266786840919333, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.4457383557765385e-05, | |
| "loss": 1.1569, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.270392068499324, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.4418871398683227e-05, | |
| "loss": 1.1054, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.273997296079315, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1.4380277623529766e-05, | |
| "loss": 1.192, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.277602523659306, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.4341602945127806e-05, | |
| "loss": 1.4075, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.281207751239297, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.4302848077794427e-05, | |
| "loss": 1.2089, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.284812978819288, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.426401373732779e-05, | |
| "loss": 1.1935, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.288418206399279, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.422510064099391e-05, | |
| "loss": 1.1739, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.29202343397927, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.4186109507513425e-05, | |
| "loss": 1.2352, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.295628661559261, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.4147041057048303e-05, | |
| "loss": 1.0588, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.2992338891392519, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.4107896011188546e-05, | |
| "loss": 1.2812, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.3028391167192428, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.4068675092938872e-05, | |
| "loss": 1.1794, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.306444344299234, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.4029379026705352e-05, | |
| "loss": 1.1947, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.3100495718792249, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.3990008538282027e-05, | |
| "loss": 1.24, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.3136547994592158, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.3950564354837512e-05, | |
| "loss": 1.1139, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.317260027039207, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 1.391104720490156e-05, | |
| "loss": 1.1621, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.320865254619198, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.387145781835161e-05, | |
| "loss": 1.1806, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.3244704821991888, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.3831796926399295e-05, | |
| "loss": 1.3742, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.3280757097791798, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.3792065261576953e-05, | |
| "loss": 1.4415, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.3316809373591707, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.3752263557724088e-05, | |
| "loss": 1.2285, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.3352861649391619, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.3712392549973814e-05, | |
| "loss": 1.189, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.3388913925191528, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.3672452974739278e-05, | |
| "loss": 1.2903, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.3424966200991437, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.3632445569700078e-05, | |
| "loss": 1.1714, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.3461018476791349, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.3592371073788595e-05, | |
| "loss": 1.4159, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.3497070752591258, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.355223022717639e-05, | |
| "loss": 1.2715, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.3533123028391167, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.3512023771260507e-05, | |
| "loss": 1.239, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.3569175304191077, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.347175244864979e-05, | |
| "loss": 1.4135, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.3605227579990986, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.3431417003151162e-05, | |
| "loss": 1.2575, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.3641279855790898, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.3391018179755886e-05, | |
| "loss": 1.2257, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.3677332131590807, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.3350556724625809e-05, | |
| "loss": 1.3427, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.3713384407390716, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.3310033385079589e-05, | |
| "loss": 1.3146, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.3749436683190626, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 1.3269448909578866e-05, | |
| "loss": 1.1914, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.3785488958990535, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.3228804047714462e-05, | |
| "loss": 1.1498, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.3821541234790446, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.3188099550192537e-05, | |
| "loss": 1.3878, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3857593510590356, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.31473361688207e-05, | |
| "loss": 1.4012, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.3893645786390265, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.3106514656494147e-05, | |
| "loss": 1.2258, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.3929698062190177, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 1.3065635767181748e-05, | |
| "loss": 1.1058, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.3965750337990086, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.302470025591211e-05, | |
| "loss": 1.3359, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.4001802613789995, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 1.2983708878759655e-05, | |
| "loss": 1.1056, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.4037854889589905, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.2942662392830632e-05, | |
| "loss": 1.142, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.4073907165389814, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 1.290156155624914e-05, | |
| "loss": 1.2136, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.4109959441189726, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.286040712814314e-05, | |
| "loss": 1.2239, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.4146011716989635, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.2819199868630419e-05, | |
| "loss": 1.3565, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.4182063992789544, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.2777940538804545e-05, | |
| "loss": 1.2145, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.4218116268589456, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.2736629900720832e-05, | |
| "loss": 1.4545, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.4254168544389365, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.2695268717382242e-05, | |
| "loss": 1.2354, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.4290220820189274, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.2653857752725305e-05, | |
| "loss": 1.1686, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.4326273095989184, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 1.2612397771606015e-05, | |
| "loss": 1.1085, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.4362325371789093, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.2570889539785683e-05, | |
| "loss": 1.2148, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.4398377647589005, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.2529333823916807e-05, | |
| "loss": 1.2704, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.4434429923388914, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.2487731391528919e-05, | |
| "loss": 1.0604, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.4470482199188823, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.2446083011014389e-05, | |
| "loss": 1.4105, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.4506534474988735, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.2404389451614253e-05, | |
| "loss": 1.1087, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.4542586750788644, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.2362651483403985e-05, | |
| "loss": 1.1673, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.4578639026588553, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.2320869877279297e-05, | |
| "loss": 1.2084, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.4614691302388463, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.2279045404941883e-05, | |
| "loss": 1.2761, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.4650743578188372, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 1.2237178838885168e-05, | |
| "loss": 1.2834, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.4686795853988284, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.2195270952380052e-05, | |
| "loss": 1.455, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.4722848129788193, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.215332251946061e-05, | |
| "loss": 1.4062, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.4758900405588102, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.2111334314909811e-05, | |
| "loss": 1.1746, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.4794952681388014, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.2069307114245197e-05, | |
| "loss": 1.121, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.4831004957187923, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.2027241693704567e-05, | |
| "loss": 1.3559, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.4867057232987833, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.1985138830231638e-05, | |
| "loss": 1.1604, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.4903109508787742, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.1942999301461694e-05, | |
| "loss": 1.3521, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4903109508787742, | |
| "eval_loss": 1.374480128288269, | |
| "eval_runtime": 172.9241, | |
| "eval_samples_per_second": 5.02, | |
| "eval_steps_per_second": 2.51, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4939161784587651, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.1900823885707216e-05, | |
| "loss": 1.306, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.4975214060387563, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.1858613361943518e-05, | |
| "loss": 1.5028, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.5011266336187472, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.1816368509794365e-05, | |
| "loss": 1.2328, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.5047318611987381, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.177409010951755e-05, | |
| "loss": 1.2548, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.5083370887787293, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.1731778941990497e-05, | |
| "loss": 1.3175, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.51194231635872, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.1689435788695844e-05, | |
| "loss": 1.397, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.5155475439387112, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.1647061431707e-05, | |
| "loss": 1.2416, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.519152771518702, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.1604656653673707e-05, | |
| "loss": 1.345, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.522757999098693, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.156222223780757e-05, | |
| "loss": 1.1482, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.5263632266786842, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.1519758967867608e-05, | |
| "loss": 1.2459, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.5299684542586751, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 1.1477267628145777e-05, | |
| "loss": 1.4684, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.533573681838666, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.1434749003452467e-05, | |
| "loss": 1.2233, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.5371789094186572, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.1392203879102027e-05, | |
| "loss": 1.3911, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.540784136998648, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.1349633040898246e-05, | |
| "loss": 1.2574, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.544389364578639, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.1307037275119854e-05, | |
| "loss": 1.3283, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.54799459215863, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.1264417368505981e-05, | |
| "loss": 1.3249, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.551599819738621, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 1.1221774108241646e-05, | |
| "loss": 1.344, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.555205047318612, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.117910828194319e-05, | |
| "loss": 1.3735, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.558810274898603, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.1136420677643763e-05, | |
| "loss": 1.3016, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.562415502478594, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.1093712083778748e-05, | |
| "loss": 1.1804, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.566020730058585, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.1050983289171195e-05, | |
| "loss": 1.2459, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.5696259576385758, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.1008235083017272e-05, | |
| "loss": 1.1279, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.573231185218567, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 1.096546825487167e-05, | |
| "loss": 1.0826, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.576836412798558, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.092268359463302e-05, | |
| "loss": 1.4331, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.5804416403785488, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.0879881892529325e-05, | |
| "loss": 1.2951, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.58404686795854, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.0837063939103332e-05, | |
| "loss": 1.1372, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.5876520955385307, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 1.0794230525197959e-05, | |
| "loss": 1.1218, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.5912573231185219, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.0751382441941677e-05, | |
| "loss": 1.1602, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.5948625506985128, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.0708520480733895e-05, | |
| "loss": 1.2669, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.5984677782785037, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.0665645433230345e-05, | |
| "loss": 1.2416, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.6020730058584949, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.0622758091328469e-05, | |
| "loss": 1.4618, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.6056782334384858, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.0579859247152774e-05, | |
| "loss": 1.2054, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.6092834610184767, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.0536949693040224e-05, | |
| "loss": 1.2841, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.612888688598468, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.0494030221525582e-05, | |
| "loss": 1.1989, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.6164939161784586, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.0451101625326798e-05, | |
| "loss": 1.252, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.6200991437584498, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.0408164697330348e-05, | |
| "loss": 1.1669, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.6237043713384407, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.0365220230576592e-05, | |
| "loss": 1.4606, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.6273095989184316, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.0322269018245128e-05, | |
| "loss": 1.2497, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.6309148264984228, | |
| "grad_norm": 1.75, | |
| "learning_rate": 1.0279311853640157e-05, | |
| "loss": 1.2096, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.6345200540784137, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.0236349530175807e-05, | |
| "loss": 1.2014, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.6381252816584047, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.019338284136149e-05, | |
| "loss": 1.0763, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.6417305092383958, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 1.015041258078725e-05, | |
| "loss": 1.222, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.6453357368183865, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.0107439542109097e-05, | |
| "loss": 1.2733, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.6489409643983777, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.0064464519034358e-05, | |
| "loss": 1.4022, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.6525461919783686, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.0021488305307003e-05, | |
| "loss": 1.11, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.6561514195583595, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 9.978511694692999e-06, | |
| "loss": 1.1927, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.6597566471383507, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 9.935535480965647e-06, | |
| "loss": 1.2483, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.6633618747183416, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 9.892560457890907e-06, | |
| "loss": 1.5043, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.6669671022983326, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 9.849587419212751e-06, | |
| "loss": 1.2133, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.6705723298783237, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 9.806617158638515e-06, | |
| "loss": 1.1917, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.6741775574583144, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 9.763650469824198e-06, | |
| "loss": 1.2497, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.6777827850383056, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.720688146359843e-06, | |
| "loss": 1.0885, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.6813880126182965, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 9.677730981754875e-06, | |
| "loss": 1.0702, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.6849932401982874, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 9.634779769423412e-06, | |
| "loss": 1.0635, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.6885984677782786, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 9.591835302669657e-06, | |
| "loss": 1.2321, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.6922036953582695, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 9.548898374673205e-06, | |
| "loss": 1.1834, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.6958089229382605, | |
| "grad_norm": 1.625, | |
| "learning_rate": 9.505969778474418e-06, | |
| "loss": 1.0861, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.6994141505182516, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 9.463050306959782e-06, | |
| "loss": 1.2529, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.7030193780982423, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 9.42014075284723e-06, | |
| "loss": 1.2823, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.7066246056782335, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 9.377241908671533e-06, | |
| "loss": 1.2319, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.7102298332582244, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 9.334354566769658e-06, | |
| "loss": 1.1869, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.7138350608382154, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 9.291479519266108e-06, | |
| "loss": 1.2677, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.7174402884182065, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 9.248617558058328e-06, | |
| "loss": 1.2629, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.7210455159981974, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 9.205769474802045e-06, | |
| "loss": 1.3352, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.7246507435781884, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.162936060896672e-06, | |
| "loss": 1.2224, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.7282559711581793, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 9.12011810747068e-06, | |
| "loss": 1.1854, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.7318611987381702, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.07731640536698e-06, | |
| "loss": 1.4678, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.7354664263181614, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 9.034531745128334e-06, | |
| "loss": 1.2149, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.7390716538981523, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 8.991764916982731e-06, | |
| "loss": 1.2232, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.7426768814781433, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 8.949016710828808e-06, | |
| "loss": 1.225, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.7426768814781433, | |
| "eval_loss": 1.35747492313385, | |
| "eval_runtime": 173.1266, | |
| "eval_samples_per_second": 5.014, | |
| "eval_steps_per_second": 2.507, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.7462821090581344, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 8.906287916221259e-06, | |
| "loss": 1.2286, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.7498873366381251, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 8.863579322356242e-06, | |
| "loss": 1.3773, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.7534925642181163, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 8.820891718056815e-06, | |
| "loss": 1.3011, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.7570977917981072, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 8.77822589175836e-06, | |
| "loss": 1.0843, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.7607030193780981, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.73558263149402e-06, | |
| "loss": 1.1688, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.7643082469580893, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.692962724880148e-06, | |
| "loss": 1.1874, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.7679134745380802, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 8.650366959101757e-06, | |
| "loss": 1.2458, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.7715187021180712, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 8.607796120897978e-06, | |
| "loss": 1.2683, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.7751239296980623, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 8.565250996547538e-06, | |
| "loss": 1.3619, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.778729157278053, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.522732371854228e-06, | |
| "loss": 1.2405, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.7823343848580442, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 8.480241032132394e-06, | |
| "loss": 1.2086, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.7859396124380351, | |
| "grad_norm": 1.75, | |
| "learning_rate": 8.437777762192434e-06, | |
| "loss": 1.3771, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.789544840018026, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 8.395343346326295e-06, | |
| "loss": 1.2608, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.7931500675980172, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 8.352938568293e-06, | |
| "loss": 1.3108, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.7967552951780081, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 8.310564211304159e-06, | |
| "loss": 1.3414, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.800360522757999, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.268221058009506e-06, | |
| "loss": 1.3927, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.8039657503379902, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 8.225909890482456e-06, | |
| "loss": 1.2549, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.807570977917981, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 8.183631490205636e-06, | |
| "loss": 1.1768, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.811176205497972, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 8.141386638056482e-06, | |
| "loss": 1.2259, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.814781433077963, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 8.09917611429279e-06, | |
| "loss": 1.1345, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.818386660657954, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 8.057000698538311e-06, | |
| "loss": 1.4693, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.8219918882379451, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 8.014861169768362e-06, | |
| "loss": 1.0846, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.825597115817936, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.972758306295436e-06, | |
| "loss": 1.2107, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.829202343397927, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 7.930692885754806e-06, | |
| "loss": 1.2296, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.8328075709779181, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.888665685090194e-06, | |
| "loss": 1.318, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.8364127985579088, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.846677480539392e-06, | |
| "loss": 1.273, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.8400180261379, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.80472904761995e-06, | |
| "loss": 1.4066, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.843623253717891, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 7.762821161114834e-06, | |
| "loss": 1.2398, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.8472284812978819, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 7.720954595058118e-06, | |
| "loss": 1.3608, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.850833708877873, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 7.679130122720704e-06, | |
| "loss": 1.1412, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.854438936457864, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.637348516596016e-06, | |
| "loss": 1.0725, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.8580441640378549, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 7.59561054838575e-06, | |
| "loss": 1.2413, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.861649391617846, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.5539169889856135e-06, | |
| "loss": 1.2436, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.8652546191978367, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 7.512268608471083e-06, | |
| "loss": 1.0819, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.868859846777828, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 7.470666176083193e-06, | |
| "loss": 1.1186, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.8724650743578188, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 7.42911046021432e-06, | |
| "loss": 1.2108, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.8760703019378098, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.387602228393987e-06, | |
| "loss": 1.2103, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.879675529517801, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.346142247274695e-06, | |
| "loss": 1.1988, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.8832807570977916, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 7.304731282617762e-06, | |
| "loss": 1.3596, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.8868859846777828, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 7.263370099279173e-06, | |
| "loss": 1.1653, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.8904912122577737, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 7.2220594611954606e-06, | |
| "loss": 1.3131, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.8940964398377647, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 7.1808001313695855e-06, | |
| "loss": 1.2719, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.8977016674177558, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.1395928718568605e-06, | |
| "loss": 1.1715, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.9013068949977467, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 7.098438443750865e-06, | |
| "loss": 1.1839, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.9049121225777377, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 7.057337607169373e-06, | |
| "loss": 1.1699, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.9085173501577288, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.016291121240346e-06, | |
| "loss": 1.3609, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.9121225777377195, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 6.975299744087891e-06, | |
| "loss": 1.2652, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.9157278053177107, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 6.934364232818254e-06, | |
| "loss": 1.1606, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.9193330328977016, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 6.8934853435058566e-06, | |
| "loss": 1.3524, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.9229382604776926, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 6.852663831179303e-06, | |
| "loss": 1.342, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.9265434880576837, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 6.811900449807465e-06, | |
| "loss": 1.1004, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.9301487156376747, | |
| "grad_norm": 1.75, | |
| "learning_rate": 6.771195952285541e-06, | |
| "loss": 1.2614, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.9337539432176656, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 6.730551090421137e-06, | |
| "loss": 1.1193, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.9373591707976567, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 6.689966614920414e-06, | |
| "loss": 1.3159, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.9409643983776474, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 6.6494432753741935e-06, | |
| "loss": 1.2379, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.9445696259576386, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 6.608981820244116e-06, | |
| "loss": 1.4973, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.9481748535376295, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 6.568582996848844e-06, | |
| "loss": 1.2734, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.9517800811176205, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 6.528247551350213e-06, | |
| "loss": 1.2755, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.9553853086976116, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.487976228739493e-06, | |
| "loss": 1.5293, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.9589905362776026, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 6.4477697728236146e-06, | |
| "loss": 1.3727, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.9625957638575935, | |
| "grad_norm": 1.875, | |
| "learning_rate": 6.407628926211409e-06, | |
| "loss": 1.1571, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.9662009914375846, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 6.367554430299924e-06, | |
| "loss": 1.3541, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.9698062190175754, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6.327547025260723e-06, | |
| "loss": 1.377, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.9734114465975665, | |
| "grad_norm": 1.75, | |
| "learning_rate": 6.287607450026189e-06, | |
| "loss": 1.2748, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.9770166741775574, | |
| "grad_norm": 2.0, | |
| "learning_rate": 6.247736442275918e-06, | |
| "loss": 1.0921, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.9806219017575484, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 6.2079347384230505e-06, | |
| "loss": 1.0822, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.9842271293375395, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 6.168203073600706e-06, | |
| "loss": 1.166, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.9878323569175305, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.128542181648395e-06, | |
| "loss": 1.4016, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.9914375844975214, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 6.088952795098442e-06, | |
| "loss": 1.0332, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.9950428120775126, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.049435645162487e-06, | |
| "loss": 1.0768, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.9950428120775126, | |
| "eval_loss": 1.3475145101547241, | |
| "eval_runtime": 173.071, | |
| "eval_samples_per_second": 5.015, | |
| "eval_steps_per_second": 2.508, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.9986480396575033, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 6.009991461717977e-06, | |
| "loss": 1.1225, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.0022532672374944, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 5.9706209732946495e-06, | |
| "loss": 1.2381, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.0058584948174856, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 5.931324907061131e-06, | |
| "loss": 1.1545, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.0094637223974763, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 5.892103988811457e-06, | |
| "loss": 1.1951, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.0130689499774674, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 5.852958942951701e-06, | |
| "loss": 1.1741, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.0166741775574586, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.8138904924865766e-06, | |
| "loss": 1.5341, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.0202794051374493, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 5.774899359006092e-06, | |
| "loss": 1.2724, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.0022532672374944, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 5.735986262672211e-06, | |
| "loss": 1.1883, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.0058584948174856, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.697151922205575e-06, | |
| "loss": 1.0542, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.0094637223974763, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 5.658397054872197e-06, | |
| "loss": 1.0822, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.0130689499774674, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 5.619722376470238e-06, | |
| "loss": 1.1295, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.016674177557458, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 5.581128601316774e-06, | |
| "loss": 1.1266, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.0202794051374493, | |
| "grad_norm": 2.0, | |
| "learning_rate": 5.542616442234618e-06, | |
| "loss": 1.1135, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.0238846327174405, | |
| "grad_norm": 1.875, | |
| "learning_rate": 5.504186610539131e-06, | |
| "loss": 1.0394, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.027489860297431, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 5.465839816025093e-06, | |
| "loss": 1.2965, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.0310950878774223, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 5.427576766953615e-06, | |
| "loss": 1.216, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.034700315457413, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 5.3893981700390215e-06, | |
| "loss": 1.1652, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.038305543037404, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 5.35130473043582e-06, | |
| "loss": 1.3548, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.0419107706173953, | |
| "grad_norm": 2.0, | |
| "learning_rate": 5.313297151725679e-06, | |
| "loss": 1.1593, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.045515998197386, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 5.275376135904408e-06, | |
| "loss": 1.0605, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.049121225777377, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 5.2375423833690255e-06, | |
| "loss": 1.045, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.0527264533573684, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 5.1997965929048125e-06, | |
| "loss": 1.0288, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.056331680937359, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 5.1621394616723705e-06, | |
| "loss": 1.1655, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.0599369085173502, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.124571685194804e-06, | |
| "loss": 1.2406, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.063542136097341, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 5.087093957344841e-06, | |
| "loss": 1.0553, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.067147363677332, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 5.049706970332e-06, | |
| "loss": 1.145, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.0707525912573232, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 5.0124114146898505e-06, | |
| "loss": 1.1514, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.074357818837314, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 4.9752079792632244e-06, | |
| "loss": 1.2022, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.077963046417305, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 4.938097351195499e-06, | |
| "loss": 1.0757, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.0815682739972963, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 4.9010802159159224e-06, | |
| "loss": 0.9544, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.085173501577287, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 4.864157257126928e-06, | |
| "loss": 1.1412, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.088778729157278, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 4.8273291567915225e-06, | |
| "loss": 1.4808, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.092383956737269, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 4.790596595120699e-06, | |
| "loss": 1.0057, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.09598918431726, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 4.753960250560843e-06, | |
| "loss": 1.08, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.099594411897251, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 4.7174207997812436e-06, | |
| "loss": 1.2154, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.103199639477242, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 4.680978917661544e-06, | |
| "loss": 1.0927, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.106804867057233, | |
| "grad_norm": 1.875, | |
| "learning_rate": 4.6446352772793256e-06, | |
| "loss": 1.0703, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.110410094637224, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 4.608390549897661e-06, | |
| "loss": 0.9291, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.114015322217215, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 4.5722454049526825e-06, | |
| "loss": 1.3928, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.117620549797206, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 4.536200510041271e-06, | |
| "loss": 0.9991, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.1212257773771968, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 4.5002565309087e-06, | |
| "loss": 1.1373, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.124831004957188, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 4.4644141314363165e-06, | |
| "loss": 1.36, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.128436232537179, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 4.4286739736293285e-06, | |
| "loss": 0.9845, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.1320414601171698, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 4.393036717604536e-06, | |
| "loss": 1.0419, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.135646687697161, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.357503021578158e-06, | |
| "loss": 1.2163, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.1392519152771516, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.322073541853677e-06, | |
| "loss": 0.965, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 4.286748932809707e-06, | |
| "loss": 0.9849, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.146462370437134, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 4.2515298468879064e-06, | |
| "loss": 1.2081, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.1500675980171247, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 4.216416934580947e-06, | |
| "loss": 0.9697, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.153672825597116, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 4.181410844420473e-06, | |
| "loss": 1.0377, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.157278053177107, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 4.146512222965144e-06, | |
| "loss": 1.3504, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.1608832807570977, | |
| "grad_norm": 1.875, | |
| "learning_rate": 4.111721714788671e-06, | |
| "loss": 1.0909, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.164488508337089, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 4.077039962467946e-06, | |
| "loss": 1.1935, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.16809373591708, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 4.042467606571134e-06, | |
| "loss": 1.0628, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.1716989634970707, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 4.008005285645863e-06, | |
| "loss": 1.178, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.175304191077062, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.973653636207437e-06, | |
| "loss": 0.9975, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.1789094186570526, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.939413292727061e-06, | |
| "loss": 1.2431, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.1825146462370437, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 3.9052848876201285e-06, | |
| "loss": 1.1352, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.186119873817035, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 3.8712690512345555e-06, | |
| "loss": 1.2565, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.1897251013970256, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 3.837366411839114e-06, | |
| "loss": 1.1465, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.1933303289770167, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.8035775956118416e-06, | |
| "loss": 1.0953, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.1969355565570075, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.7699032266284863e-06, | |
| "loss": 1.1816, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.2005407841369986, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 3.736343926850954e-06, | |
| "loss": 1.142, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.2041460117169898, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 3.702900316115836e-06, | |
| "loss": 1.1753, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.2077512392969805, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.6695730121229734e-06, | |
| "loss": 1.126, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.2113564668769716, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 3.6363626304240185e-06, | |
| "loss": 1.1409, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.214961694456963, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 3.6032697844110896e-06, | |
| "loss": 1.1814, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.2185669220369535, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 3.5702950853054284e-06, | |
| "loss": 1.1278, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.2221721496169446, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 3.5374391421461273e-06, | |
| "loss": 0.9474, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.2257773771969354, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 3.5047025617788578e-06, | |
| "loss": 1.1199, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.2257773771969354, | |
| "eval_loss": 1.3601243495941162, | |
| "eval_runtime": 173.0606, | |
| "eval_samples_per_second": 5.016, | |
| "eval_steps_per_second": 2.508, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.2293826047769265, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 3.4720859488446744e-06, | |
| "loss": 0.9183, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.2329878323569177, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 3.4395899057688575e-06, | |
| "loss": 1.1332, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.2365930599369084, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 3.407215032749763e-06, | |
| "loss": 1.2575, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.2401982875168995, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 3.374961927747751e-06, | |
| "loss": 1.0141, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.2438035150968907, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 3.342831186474149e-06, | |
| "loss": 1.0652, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.2474087426768814, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 3.31082340238023e-06, | |
| "loss": 1.1363, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.2510139702568726, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 3.27893916664626e-06, | |
| "loss": 1.0828, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.2546191978368633, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 3.2471790681705928e-06, | |
| "loss": 0.9991, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.2582244254168544, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 3.215543693558769e-06, | |
| "loss": 1.1052, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.2618296529968456, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 3.1840336271126935e-06, | |
| "loss": 1.126, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.2654348805768363, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 3.152649450819852e-06, | |
| "loss": 1.0175, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.2690401081568274, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 3.1213917443425456e-06, | |
| "loss": 1.0922, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.2726453357368186, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 3.0902610850071922e-06, | |
| "loss": 1.065, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.2762505633168093, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 3.0592580477936606e-06, | |
| "loss": 1.1629, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.2798557908968005, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.0283832053246644e-06, | |
| "loss": 1.0396, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.283461018476791, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.99763712785516e-06, | |
| "loss": 1.0549, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.2870662460567823, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.967020383261834e-06, | |
| "loss": 0.9981, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.2906714736367735, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 2.9365335370326143e-06, | |
| "loss": 1.2228, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.294276701216764, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 2.9061771522562143e-06, | |
| "loss": 1.2494, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.2978819287967553, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 2.875951789611734e-06, | |
| "loss": 1.0157, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.301487156376746, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 2.8458580073583262e-06, | |
| "loss": 1.0425, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.305092383956737, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 2.8158963613248437e-06, | |
| "loss": 1.3302, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.3086976115367284, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.7860674048996174e-06, | |
| "loss": 1.3854, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.312302839116719, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 2.756371689020214e-06, | |
| "loss": 0.8928, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.3159080666967102, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 2.7268097621632473e-06, | |
| "loss": 1.0802, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.3195132942767014, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 2.697382170334275e-06, | |
| "loss": 1.1749, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.323118521856692, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 2.6680894570577042e-06, | |
| "loss": 1.1211, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.3267237494366833, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 2.638932163366742e-06, | |
| "loss": 0.8447, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.3303289770166744, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.6099108277934105e-06, | |
| "loss": 0.9218, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.333934204596665, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 2.581025986358602e-06, | |
| "loss": 1.0153, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.3375394321766563, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 2.5522781725621814e-06, | |
| "loss": 1.0466, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.341144659756647, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 2.523667917373125e-06, | |
| "loss": 1.2164, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.344749887336638, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.4951957492197097e-06, | |
| "loss": 0.9825, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.3483551149166293, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 2.4668621939797745e-06, | |
| "loss": 1.0349, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.35196034249662, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.438667774970981e-06, | |
| "loss": 1.0732, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.355565570076611, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 2.4106130129411608e-06, | |
| "loss": 1.053, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.359170797656602, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.3826984260587084e-06, | |
| "loss": 0.9565, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.362776025236593, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 2.354924529902978e-06, | |
| "loss": 1.1232, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.366381252816584, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 2.327291837454799e-06, | |
| "loss": 1.1934, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.369986480396575, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 2.2998008590869838e-06, | |
| "loss": 1.1419, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.373591707976566, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 2.2724521025548828e-06, | |
| "loss": 1.1529, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.377196935556557, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 2.245246072987045e-06, | |
| "loss": 1.1641, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.380802163136548, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 2.2181832728758635e-06, | |
| "loss": 1.1397, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.384407390716539, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 2.191264202068286e-06, | |
| "loss": 1.1364, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.3880126182965298, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 2.1644893577566118e-06, | |
| "loss": 1.0603, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.391617845876521, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 2.137859234469286e-06, | |
| "loss": 1.1638, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.395223073456512, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 2.1113743240617668e-06, | |
| "loss": 1.1457, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.398828301036503, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.08503511570746e-06, | |
| "loss": 1.0903, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.402433528616494, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 2.058842095888658e-06, | |
| "loss": 1.0844, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.4060387561964847, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 2.0327957483875693e-06, | |
| "loss": 1.1195, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.409643983776476, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 2.006896554277388e-06, | |
| "loss": 1.0906, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.413249211356467, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.981144991913392e-06, | |
| "loss": 1.0255, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.4168544389364577, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.9555415369241228e-06, | |
| "loss": 1.2173, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.420459666516449, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.930086662202589e-06, | |
| "loss": 1.0533, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.42406489409644, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 1.9047808378975485e-06, | |
| "loss": 0.9335, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.4276701216764307, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.8796245314048046e-06, | |
| "loss": 1.2868, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.431275349256422, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.8546182073585828e-06, | |
| "loss": 0.9996, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.434880576836413, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 1.829762327622958e-06, | |
| "loss": 1.0173, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.4384858044164037, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.805057351283307e-06, | |
| "loss": 1.0216, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.442091031996395, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.7805037346378384e-06, | |
| "loss": 0.9223, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.4456962595763856, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.756101931189169e-06, | |
| "loss": 1.0502, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.4493014871563767, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7318523916359376e-06, | |
| "loss": 1.2346, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.452906714736368, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.7077555638644838e-06, | |
| "loss": 1.1281, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.4565119423163586, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 1.6838118929405856e-06, | |
| "loss": 0.9079, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.4601171698963498, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 1.660021821101222e-06, | |
| "loss": 1.0483, | |
| "step": 695 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 831, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 139, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.101230715042005e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |