| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9954545454545456, | |
| "eval_steps": 500, | |
| "global_step": 292, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006818181818181818, | |
| "grad_norm": 2.020231246948242, | |
| "learning_rate": 5.0000000000000004e-08, | |
| "loss": 0.8736, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013636363636363636, | |
| "grad_norm": 1.9368854761123657, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 0.8543, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020454545454545454, | |
| "grad_norm": 1.908379077911377, | |
| "learning_rate": 1.5000000000000002e-07, | |
| "loss": 0.8339, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02727272727272727, | |
| "grad_norm": 1.9313658475875854, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 0.834, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03409090909090909, | |
| "grad_norm": 1.983262062072754, | |
| "learning_rate": 2.5000000000000004e-07, | |
| "loss": 0.8726, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04090909090909091, | |
| "grad_norm": 1.9994115829467773, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 0.8135, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04772727272727273, | |
| "grad_norm": 1.8889026641845703, | |
| "learning_rate": 3.5000000000000004e-07, | |
| "loss": 0.8439, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05454545454545454, | |
| "grad_norm": 1.9258317947387695, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.84, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06136363636363636, | |
| "grad_norm": 1.9654443264007568, | |
| "learning_rate": 4.5000000000000003e-07, | |
| "loss": 0.8609, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06818181818181818, | |
| "grad_norm": 1.9278813600540161, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.8515, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 1.8698863983154297, | |
| "learning_rate": 5.5e-07, | |
| "loss": 0.8632, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08181818181818182, | |
| "grad_norm": 2.7078285217285156, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.8643, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08863636363636364, | |
| "grad_norm": 1.959272027015686, | |
| "learning_rate": 6.5e-07, | |
| "loss": 0.861, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09545454545454546, | |
| "grad_norm": 1.8952852487564087, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 0.8599, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10227272727272728, | |
| "grad_norm": 1.9469462633132935, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.8467, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10909090909090909, | |
| "grad_norm": 2.0111300945281982, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.8078, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1159090909090909, | |
| "grad_norm": 1.739009141921997, | |
| "learning_rate": 8.500000000000001e-07, | |
| "loss": 0.8114, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12272727272727273, | |
| "grad_norm": 1.7578809261322021, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 0.8489, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12954545454545455, | |
| "grad_norm": 1.9787132740020752, | |
| "learning_rate": 9.500000000000001e-07, | |
| "loss": 0.8052, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 1.6380164623260498, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.7882, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1431818181818182, | |
| "grad_norm": 1.686877727508545, | |
| "learning_rate": 1.0500000000000001e-06, | |
| "loss": 0.8126, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 1.690590739250183, | |
| "learning_rate": 1.1e-06, | |
| "loss": 0.8248, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15681818181818183, | |
| "grad_norm": 1.612255334854126, | |
| "learning_rate": 1.1500000000000002e-06, | |
| "loss": 0.8024, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16363636363636364, | |
| "grad_norm": 1.6669739484786987, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.818, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.17045454545454544, | |
| "grad_norm": 1.7726502418518066, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.8158, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17727272727272728, | |
| "grad_norm": 1.5930036306381226, | |
| "learning_rate": 1.3e-06, | |
| "loss": 0.7886, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18409090909090908, | |
| "grad_norm": 1.5480856895446777, | |
| "learning_rate": 1.3500000000000002e-06, | |
| "loss": 0.7747, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19090909090909092, | |
| "grad_norm": 1.3537116050720215, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.7891, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19772727272727272, | |
| "grad_norm": 1.3412829637527466, | |
| "learning_rate": 1.45e-06, | |
| "loss": 0.7799, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20454545454545456, | |
| "grad_norm": 1.237492322921753, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.7768, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21136363636363636, | |
| "grad_norm": 1.3467835187911987, | |
| "learning_rate": 1.5500000000000002e-06, | |
| "loss": 0.7344, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21818181818181817, | |
| "grad_norm": 1.1665314435958862, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.7152, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 1.2461743354797363, | |
| "learning_rate": 1.6500000000000003e-06, | |
| "loss": 0.758, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2318181818181818, | |
| "grad_norm": 1.0343059301376343, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 0.7362, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.23863636363636365, | |
| "grad_norm": 1.112117052078247, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.7771, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24545454545454545, | |
| "grad_norm": 1.104337453842163, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.7479, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.25227272727272726, | |
| "grad_norm": 0.9324578642845154, | |
| "learning_rate": 1.85e-06, | |
| "loss": 0.6936, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2590909090909091, | |
| "grad_norm": 1.0733048915863037, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 0.7302, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26590909090909093, | |
| "grad_norm": 0.8614453673362732, | |
| "learning_rate": 1.9500000000000004e-06, | |
| "loss": 0.7466, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.7729659080505371, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.725, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.27954545454545454, | |
| "grad_norm": 0.9341004490852356, | |
| "learning_rate": 2.05e-06, | |
| "loss": 0.6949, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2863636363636364, | |
| "grad_norm": 0.6668925881385803, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 0.6587, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.29318181818181815, | |
| "grad_norm": 0.681847333908081, | |
| "learning_rate": 2.15e-06, | |
| "loss": 0.6984, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.6222082376480103, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.7063, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3068181818181818, | |
| "grad_norm": 0.790607213973999, | |
| "learning_rate": 2.25e-06, | |
| "loss": 0.6804, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.31363636363636366, | |
| "grad_norm": 0.6110427379608154, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 0.6694, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.32045454545454544, | |
| "grad_norm": 0.6481993794441223, | |
| "learning_rate": 2.35e-06, | |
| "loss": 0.7321, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32727272727272727, | |
| "grad_norm": 0.5973438620567322, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.6508, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3340909090909091, | |
| "grad_norm": 0.6252740621566772, | |
| "learning_rate": 2.4500000000000003e-06, | |
| "loss": 0.6911, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 0.6826481819152832, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.6855, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3477272727272727, | |
| "grad_norm": 0.6307975649833679, | |
| "learning_rate": 2.55e-06, | |
| "loss": 0.714, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.35454545454545455, | |
| "grad_norm": 0.5900976061820984, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.6619, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3613636363636364, | |
| "grad_norm": 0.6203920841217041, | |
| "learning_rate": 2.6500000000000005e-06, | |
| "loss": 0.6822, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.36818181818181817, | |
| "grad_norm": 0.6197589039802551, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 0.6584, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.5921110510826111, | |
| "learning_rate": 2.7500000000000004e-06, | |
| "loss": 0.6739, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.38181818181818183, | |
| "grad_norm": 0.6215619444847107, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.6631, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3886363636363636, | |
| "grad_norm": 0.5495648980140686, | |
| "learning_rate": 2.85e-06, | |
| "loss": 0.6531, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39545454545454545, | |
| "grad_norm": 0.531947910785675, | |
| "learning_rate": 2.9e-06, | |
| "loss": 0.6365, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4022727272727273, | |
| "grad_norm": 0.559112548828125, | |
| "learning_rate": 2.95e-06, | |
| "loss": 0.6391, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 0.5377560257911682, | |
| "learning_rate": 3e-06, | |
| "loss": 0.6775, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4159090909090909, | |
| "grad_norm": 0.5167352557182312, | |
| "learning_rate": 3.05e-06, | |
| "loss": 0.6518, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.42272727272727273, | |
| "grad_norm": 0.5483390092849731, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 0.5972, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.42954545454545456, | |
| "grad_norm": 0.5687050223350525, | |
| "learning_rate": 3.1500000000000003e-06, | |
| "loss": 0.6515, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43636363636363634, | |
| "grad_norm": 0.9046968817710876, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.6497, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4431818181818182, | |
| "grad_norm": 0.5231667757034302, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.6364, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.49699342250823975, | |
| "learning_rate": 3.3000000000000006e-06, | |
| "loss": 0.6144, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.45681818181818185, | |
| "grad_norm": 0.5390080213546753, | |
| "learning_rate": 3.3500000000000005e-06, | |
| "loss": 0.6051, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4636363636363636, | |
| "grad_norm": 0.5252938270568848, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.6353, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.47045454545454546, | |
| "grad_norm": 0.6291780471801758, | |
| "learning_rate": 3.45e-06, | |
| "loss": 0.6326, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4772727272727273, | |
| "grad_norm": 0.5545375943183899, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.647, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.48409090909090907, | |
| "grad_norm": 0.47775956988334656, | |
| "learning_rate": 3.5500000000000003e-06, | |
| "loss": 0.5993, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4909090909090909, | |
| "grad_norm": 0.5016375184059143, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.6056, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.49772727272727274, | |
| "grad_norm": 0.4780517816543579, | |
| "learning_rate": 3.65e-06, | |
| "loss": 0.6189, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5045454545454545, | |
| "grad_norm": 0.5763716697692871, | |
| "learning_rate": 3.7e-06, | |
| "loss": 0.6262, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5113636363636364, | |
| "grad_norm": 0.5321404337882996, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.6336, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5181818181818182, | |
| "grad_norm": 0.4641991853713989, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.6317, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.45118412375450134, | |
| "learning_rate": 3.85e-06, | |
| "loss": 0.5927, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5318181818181819, | |
| "grad_norm": 0.5500645041465759, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 0.62, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5386363636363637, | |
| "grad_norm": 0.5197616219520569, | |
| "learning_rate": 3.95e-06, | |
| "loss": 0.6063, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.4672992527484894, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.6156, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5522727272727272, | |
| "grad_norm": 0.5995073318481445, | |
| "learning_rate": 4.05e-06, | |
| "loss": 0.6219, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5590909090909091, | |
| "grad_norm": 0.5593804717063904, | |
| "learning_rate": 4.1e-06, | |
| "loss": 0.6282, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5659090909090909, | |
| "grad_norm": 0.7544977068901062, | |
| "learning_rate": 4.15e-06, | |
| "loss": 0.6246, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5727272727272728, | |
| "grad_norm": 0.45378655195236206, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.5881, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5795454545454546, | |
| "grad_norm": 0.4407907724380493, | |
| "learning_rate": 4.25e-06, | |
| "loss": 0.6026, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5863636363636363, | |
| "grad_norm": 0.4333001673221588, | |
| "learning_rate": 4.3e-06, | |
| "loss": 0.6098, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5931818181818181, | |
| "grad_norm": 0.45020997524261475, | |
| "learning_rate": 4.350000000000001e-06, | |
| "loss": 0.5858, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.424003541469574, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.6186, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6068181818181818, | |
| "grad_norm": 0.427081435918808, | |
| "learning_rate": 4.450000000000001e-06, | |
| "loss": 0.6164, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6136363636363636, | |
| "grad_norm": 0.4430546164512634, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.6049, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6204545454545455, | |
| "grad_norm": 0.5469494462013245, | |
| "learning_rate": 4.5500000000000005e-06, | |
| "loss": 0.608, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6272727272727273, | |
| "grad_norm": 0.4481559693813324, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.5532, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.634090909090909, | |
| "grad_norm": 0.6563135981559753, | |
| "learning_rate": 4.65e-06, | |
| "loss": 0.6105, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6409090909090909, | |
| "grad_norm": 0.4572807252407074, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.5976, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6477272727272727, | |
| "grad_norm": 0.4423324167728424, | |
| "learning_rate": 4.75e-06, | |
| "loss": 0.616, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6545454545454545, | |
| "grad_norm": 0.6230632066726685, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.5879, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6613636363636364, | |
| "grad_norm": 0.461467981338501, | |
| "learning_rate": 4.85e-06, | |
| "loss": 0.5992, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6681818181818182, | |
| "grad_norm": 0.44455233216285706, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 0.6138, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.47281715273857117, | |
| "learning_rate": 4.95e-06, | |
| "loss": 0.59, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.46114516258239746, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6216, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6886363636363636, | |
| "grad_norm": 0.49900761246681213, | |
| "learning_rate": 4.9999795126530275e-06, | |
| "loss": 0.6222, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6954545454545454, | |
| "grad_norm": 0.4764839708805084, | |
| "learning_rate": 4.999918050947891e-06, | |
| "loss": 0.5829, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7022727272727273, | |
| "grad_norm": 0.43502742052078247, | |
| "learning_rate": 4.999815615891943e-06, | |
| "loss": 0.5982, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7090909090909091, | |
| "grad_norm": 0.43318599462509155, | |
| "learning_rate": 4.9996722091640805e-06, | |
| "loss": 0.5827, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7159090909090909, | |
| "grad_norm": 0.47218361496925354, | |
| "learning_rate": 4.9994878331147225e-06, | |
| "loss": 0.5907, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7227272727272728, | |
| "grad_norm": 0.48492228984832764, | |
| "learning_rate": 4.99926249076577e-06, | |
| "loss": 0.5459, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7295454545454545, | |
| "grad_norm": 0.4019850194454193, | |
| "learning_rate": 4.998996185810557e-06, | |
| "loss": 0.5957, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7363636363636363, | |
| "grad_norm": 0.47361984848976135, | |
| "learning_rate": 4.998688922613788e-06, | |
| "loss": 0.579, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7431818181818182, | |
| "grad_norm": 0.4612269401550293, | |
| "learning_rate": 4.9983407062114695e-06, | |
| "loss": 0.5814, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.45558398962020874, | |
| "learning_rate": 4.9979515423108255e-06, | |
| "loss": 0.6036, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7568181818181818, | |
| "grad_norm": 0.41492199897766113, | |
| "learning_rate": 4.997521437290205e-06, | |
| "loss": 0.5891, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7636363636363637, | |
| "grad_norm": 0.4484409689903259, | |
| "learning_rate": 4.997050398198977e-06, | |
| "loss": 0.6031, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7704545454545455, | |
| "grad_norm": 0.42671382427215576, | |
| "learning_rate": 4.996538432757414e-06, | |
| "loss": 0.6106, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7772727272727272, | |
| "grad_norm": 0.4222075343132019, | |
| "learning_rate": 4.995985549356568e-06, | |
| "loss": 0.5774, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7840909090909091, | |
| "grad_norm": 0.421021431684494, | |
| "learning_rate": 4.995391757058129e-06, | |
| "loss": 0.5754, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7909090909090909, | |
| "grad_norm": 0.4215952754020691, | |
| "learning_rate": 4.99475706559428e-06, | |
| "loss": 0.5895, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7977272727272727, | |
| "grad_norm": 0.42652982473373413, | |
| "learning_rate": 4.994081485367537e-06, | |
| "loss": 0.5646, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8045454545454546, | |
| "grad_norm": 0.4145742654800415, | |
| "learning_rate": 4.993365027450576e-06, | |
| "loss": 0.5862, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8113636363636364, | |
| "grad_norm": 0.4754612445831299, | |
| "learning_rate": 4.992607703586058e-06, | |
| "loss": 0.5645, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.4694959223270416, | |
| "learning_rate": 4.991809526186424e-06, | |
| "loss": 0.5989, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.4573237895965576, | |
| "learning_rate": 4.990970508333707e-06, | |
| "loss": 0.5769, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8318181818181818, | |
| "grad_norm": 0.4592607915401459, | |
| "learning_rate": 4.990090663779305e-06, | |
| "loss": 0.5529, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8386363636363636, | |
| "grad_norm": 0.4325920045375824, | |
| "learning_rate": 4.9891700069437635e-06, | |
| "loss": 0.5846, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8454545454545455, | |
| "grad_norm": 0.44623467326164246, | |
| "learning_rate": 4.988208552916535e-06, | |
| "loss": 0.5997, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8522727272727273, | |
| "grad_norm": 0.46071872115135193, | |
| "learning_rate": 4.987206317455734e-06, | |
| "loss": 0.5687, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8590909090909091, | |
| "grad_norm": 0.43121206760406494, | |
| "learning_rate": 4.986163316987877e-06, | |
| "loss": 0.5277, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.865909090909091, | |
| "grad_norm": 0.4711630046367645, | |
| "learning_rate": 4.985079568607613e-06, | |
| "loss": 0.592, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 0.4450221657752991, | |
| "learning_rate": 4.983955090077445e-06, | |
| "loss": 0.5656, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8795454545454545, | |
| "grad_norm": 0.6546564102172852, | |
| "learning_rate": 4.982789899827439e-06, | |
| "loss": 0.572, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8863636363636364, | |
| "grad_norm": 0.49622273445129395, | |
| "learning_rate": 4.9815840169549216e-06, | |
| "loss": 0.5719, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8931818181818182, | |
| "grad_norm": 0.5478544235229492, | |
| "learning_rate": 4.980337461224164e-06, | |
| "loss": 0.5774, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.4340018033981323, | |
| "learning_rate": 4.979050253066064e-06, | |
| "loss": 0.5444, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9068181818181819, | |
| "grad_norm": 0.4243793189525604, | |
| "learning_rate": 4.977722413577802e-06, | |
| "loss": 0.5759, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9136363636363637, | |
| "grad_norm": 0.45201942324638367, | |
| "learning_rate": 4.976353964522509e-06, | |
| "loss": 0.5746, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9204545454545454, | |
| "grad_norm": 0.42892220616340637, | |
| "learning_rate": 4.974944928328894e-06, | |
| "loss": 0.544, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9272727272727272, | |
| "grad_norm": 0.43185824155807495, | |
| "learning_rate": 4.973495328090891e-06, | |
| "loss": 0.5803, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9340909090909091, | |
| "grad_norm": 0.5512542724609375, | |
| "learning_rate": 4.972005187567267e-06, | |
| "loss": 0.5694, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9409090909090909, | |
| "grad_norm": 0.4836059808731079, | |
| "learning_rate": 4.970474531181245e-06, | |
| "loss": 0.5581, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9477272727272728, | |
| "grad_norm": 0.4508282244205475, | |
| "learning_rate": 4.968903384020095e-06, | |
| "loss": 0.5553, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9545454545454546, | |
| "grad_norm": 0.4764733612537384, | |
| "learning_rate": 4.967291771834727e-06, | |
| "loss": 0.5798, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9613636363636363, | |
| "grad_norm": 0.463468074798584, | |
| "learning_rate": 4.965639721039267e-06, | |
| "loss": 0.559, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9681818181818181, | |
| "grad_norm": 0.4655650854110718, | |
| "learning_rate": 4.963947258710626e-06, | |
| "loss": 0.5576, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.4651234745979309, | |
| "learning_rate": 4.962214412588053e-06, | |
| "loss": 0.5792, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9818181818181818, | |
| "grad_norm": 0.43080538511276245, | |
| "learning_rate": 4.960441211072686e-06, | |
| "loss": 0.5314, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9886363636363636, | |
| "grad_norm": 0.4403238594532013, | |
| "learning_rate": 4.9586276832270785e-06, | |
| "loss": 0.5592, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9954545454545455, | |
| "grad_norm": 0.43968647718429565, | |
| "learning_rate": 4.9567738587747314e-06, | |
| "loss": 0.5573, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0068181818181818, | |
| "grad_norm": 0.8337184190750122, | |
| "learning_rate": 4.954879768099599e-06, | |
| "loss": 1.1118, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0136363636363637, | |
| "grad_norm": 0.4042762815952301, | |
| "learning_rate": 4.952945442245598e-06, | |
| "loss": 0.5731, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0204545454545455, | |
| "grad_norm": 0.5666297078132629, | |
| "learning_rate": 4.95097091291609e-06, | |
| "loss": 0.5415, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0272727272727273, | |
| "grad_norm": 0.6148372888565063, | |
| "learning_rate": 4.948956212473371e-06, | |
| "loss": 0.5486, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0340909090909092, | |
| "grad_norm": 0.5780720710754395, | |
| "learning_rate": 4.946901373938132e-06, | |
| "loss": 0.5373, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.040909090909091, | |
| "grad_norm": 0.4648835361003876, | |
| "learning_rate": 4.944806430988927e-06, | |
| "loss": 0.5419, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0477272727272728, | |
| "grad_norm": 0.41131219267845154, | |
| "learning_rate": 4.942671417961615e-06, | |
| "loss": 0.5514, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0545454545454545, | |
| "grad_norm": 0.4733349084854126, | |
| "learning_rate": 4.940496369848795e-06, | |
| "loss": 0.5563, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0613636363636363, | |
| "grad_norm": 0.41179028153419495, | |
| "learning_rate": 4.938281322299243e-06, | |
| "loss": 0.5413, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0681818181818181, | |
| "grad_norm": 0.4770941138267517, | |
| "learning_rate": 4.936026311617316e-06, | |
| "loss": 0.5561, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.075, | |
| "grad_norm": 0.7752397060394287, | |
| "learning_rate": 4.933731374762361e-06, | |
| "loss": 0.5439, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0818181818181818, | |
| "grad_norm": 0.4323958456516266, | |
| "learning_rate": 4.931396549348115e-06, | |
| "loss": 0.5401, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0886363636363636, | |
| "grad_norm": 0.4497937858104706, | |
| "learning_rate": 4.9290218736420795e-06, | |
| "loss": 0.523, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0954545454545455, | |
| "grad_norm": 0.4881570339202881, | |
| "learning_rate": 4.926607386564898e-06, | |
| "loss": 0.5441, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1022727272727273, | |
| "grad_norm": 0.43357932567596436, | |
| "learning_rate": 4.9241531276897196e-06, | |
| "loss": 0.5499, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.1090909090909091, | |
| "grad_norm": 1.1125361919403076, | |
| "learning_rate": 4.921659137241544e-06, | |
| "loss": 0.5265, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.115909090909091, | |
| "grad_norm": 0.43150508403778076, | |
| "learning_rate": 4.919125456096574e-06, | |
| "loss": 0.5193, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.1227272727272728, | |
| "grad_norm": 0.40771809220314026, | |
| "learning_rate": 4.916552125781529e-06, | |
| "loss": 0.538, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.1295454545454546, | |
| "grad_norm": 0.5830567479133606, | |
| "learning_rate": 4.913939188472979e-06, | |
| "loss": 0.5548, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 0.4776526391506195, | |
| "learning_rate": 4.911286686996648e-06, | |
| "loss": 0.5328, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.143181818181818, | |
| "grad_norm": 0.41657742857933044, | |
| "learning_rate": 4.908594664826708e-06, | |
| "loss": 0.5336, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.4651948809623718, | |
| "learning_rate": 4.905863166085076e-06, | |
| "loss": 0.5812, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.1568181818181817, | |
| "grad_norm": 0.4399661719799042, | |
| "learning_rate": 4.903092235540679e-06, | |
| "loss": 0.5143, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.1636363636363636, | |
| "grad_norm": 0.4548935294151306, | |
| "learning_rate": 4.900281918608732e-06, | |
| "loss": 0.5425, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1704545454545454, | |
| "grad_norm": 0.43515875935554504, | |
| "learning_rate": 4.897432261349984e-06, | |
| "loss": 0.534, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1772727272727272, | |
| "grad_norm": 0.42954760789871216, | |
| "learning_rate": 4.894543310469968e-06, | |
| "loss": 0.5436, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.184090909090909, | |
| "grad_norm": 0.4063640534877777, | |
| "learning_rate": 4.891615113318236e-06, | |
| "loss": 0.5481, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.190909090909091, | |
| "grad_norm": 0.49416476488113403, | |
| "learning_rate": 4.888647717887582e-06, | |
| "loss": 0.5298, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1977272727272728, | |
| "grad_norm": 1.4082964658737183, | |
| "learning_rate": 4.8856411728132526e-06, | |
| "loss": 0.5366, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2045454545454546, | |
| "grad_norm": 0.44356465339660645, | |
| "learning_rate": 4.8825955273721524e-06, | |
| "loss": 0.5563, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2113636363636364, | |
| "grad_norm": 0.40925371646881104, | |
| "learning_rate": 4.879510831482039e-06, | |
| "loss": 0.5444, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2181818181818183, | |
| "grad_norm": 0.44379451870918274, | |
| "learning_rate": 4.876387135700701e-06, | |
| "loss": 0.5311, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.225, | |
| "grad_norm": 0.41487446427345276, | |
| "learning_rate": 4.873224491225128e-06, | |
| "loss": 0.5637, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.231818181818182, | |
| "grad_norm": 0.4520554542541504, | |
| "learning_rate": 4.870022949890676e-06, | |
| "loss": 0.5677, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2386363636363638, | |
| "grad_norm": 0.4528166949748993, | |
| "learning_rate": 4.866782564170217e-06, | |
| "loss": 0.5095, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2454545454545454, | |
| "grad_norm": 0.4409842789173126, | |
| "learning_rate": 4.863503387173276e-06, | |
| "loss": 0.5423, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2522727272727272, | |
| "grad_norm": 0.4355996251106262, | |
| "learning_rate": 4.860185472645161e-06, | |
| "loss": 0.5422, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.259090909090909, | |
| "grad_norm": 0.4871126115322113, | |
| "learning_rate": 4.856828874966086e-06, | |
| "loss": 0.5293, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2659090909090909, | |
| "grad_norm": 0.442348450422287, | |
| "learning_rate": 4.853433649150276e-06, | |
| "loss": 0.5616, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.44014668464660645, | |
| "learning_rate": 4.849999850845066e-06, | |
| "loss": 0.5354, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2795454545454545, | |
| "grad_norm": 0.4496249854564667, | |
| "learning_rate": 4.8465275363299905e-06, | |
| "loss": 0.5298, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2863636363636364, | |
| "grad_norm": 0.46060171723365784, | |
| "learning_rate": 4.84301676251586e-06, | |
| "loss": 0.554, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2931818181818182, | |
| "grad_norm": 0.7829347252845764, | |
| "learning_rate": 4.839467586943825e-06, | |
| "loss": 0.5606, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.45566731691360474, | |
| "learning_rate": 4.835880067784441e-06, | |
| "loss": 0.5118, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3068181818181819, | |
| "grad_norm": 0.43211469054222107, | |
| "learning_rate": 4.832254263836708e-06, | |
| "loss": 0.535, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3136363636363637, | |
| "grad_norm": 0.4350590705871582, | |
| "learning_rate": 4.828590234527107e-06, | |
| "loss": 0.5595, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3204545454545453, | |
| "grad_norm": 0.4056399464607239, | |
| "learning_rate": 4.82488803990863e-06, | |
| "loss": 0.55, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3272727272727272, | |
| "grad_norm": 0.4458341598510742, | |
| "learning_rate": 4.821147740659795e-06, | |
| "loss": 0.501, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.334090909090909, | |
| "grad_norm": 0.446504145860672, | |
| "learning_rate": 4.817369398083648e-06, | |
| "loss": 0.5393, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3409090909090908, | |
| "grad_norm": 0.47838401794433594, | |
| "learning_rate": 4.813553074106761e-06, | |
| "loss": 0.5354, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3477272727272727, | |
| "grad_norm": 0.4675792157649994, | |
| "learning_rate": 4.809698831278217e-06, | |
| "loss": 0.5202, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3545454545454545, | |
| "grad_norm": 0.4351370632648468, | |
| "learning_rate": 4.805806732768585e-06, | |
| "loss": 0.5256, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3613636363636363, | |
| "grad_norm": 0.4466511607170105, | |
| "learning_rate": 4.801876842368882e-06, | |
| "loss": 0.5238, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3681818181818182, | |
| "grad_norm": 0.45672544836997986, | |
| "learning_rate": 4.797909224489531e-06, | |
| "loss": 0.5426, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.422638863325119, | |
| "learning_rate": 4.793903944159303e-06, | |
| "loss": 0.5385, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.3818181818181818, | |
| "grad_norm": 0.4254551827907562, | |
| "learning_rate": 4.789861067024253e-06, | |
| "loss": 0.503, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.3886363636363637, | |
| "grad_norm": 0.43860334157943726, | |
| "learning_rate": 4.785780659346642e-06, | |
| "loss": 0.519, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.3954545454545455, | |
| "grad_norm": 0.43294480443000793, | |
| "learning_rate": 4.781662788003851e-06, | |
| "loss": 0.5537, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.4022727272727273, | |
| "grad_norm": 0.45006492733955383, | |
| "learning_rate": 4.777507520487289e-06, | |
| "loss": 0.5318, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4090909090909092, | |
| "grad_norm": 0.4178854823112488, | |
| "learning_rate": 4.773314924901281e-06, | |
| "loss": 0.5287, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.415909090909091, | |
| "grad_norm": 0.45662209391593933, | |
| "learning_rate": 4.769085069961955e-06, | |
| "loss": 0.5609, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.4227272727272728, | |
| "grad_norm": 0.41578733921051025, | |
| "learning_rate": 4.764818024996117e-06, | |
| "loss": 0.5249, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.4295454545454547, | |
| "grad_norm": 0.48305758833885193, | |
| "learning_rate": 4.760513859940112e-06, | |
| "loss": 0.515, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.4363636363636363, | |
| "grad_norm": 0.4564506709575653, | |
| "learning_rate": 4.756172645338675e-06, | |
| "loss": 0.535, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4431818181818181, | |
| "grad_norm": 0.4499240815639496, | |
| "learning_rate": 4.751794452343785e-06, | |
| "loss": 0.5236, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.41851088404655457, | |
| "learning_rate": 4.747379352713489e-06, | |
| "loss": 0.5254, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.4568181818181818, | |
| "grad_norm": 0.4690937399864197, | |
| "learning_rate": 4.7429274188107275e-06, | |
| "loss": 0.5312, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.4636363636363636, | |
| "grad_norm": 0.5516958832740784, | |
| "learning_rate": 4.738438723602154e-06, | |
| "loss": 0.5315, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.4704545454545455, | |
| "grad_norm": 1.2040799856185913, | |
| "learning_rate": 4.733913340656933e-06, | |
| "loss": 0.544, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4772727272727273, | |
| "grad_norm": 0.445425808429718, | |
| "learning_rate": 4.729351344145536e-06, | |
| "loss": 0.5309, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4840909090909091, | |
| "grad_norm": 0.4186754822731018, | |
| "learning_rate": 4.7247528088385296e-06, | |
| "loss": 0.5138, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.490909090909091, | |
| "grad_norm": 0.43284520506858826, | |
| "learning_rate": 4.720117810105341e-06, | |
| "loss": 0.521, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.4977272727272728, | |
| "grad_norm": 0.4308505058288574, | |
| "learning_rate": 4.715446423913036e-06, | |
| "loss": 0.5436, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5045454545454544, | |
| "grad_norm": 0.4471941888332367, | |
| "learning_rate": 4.710738726825059e-06, | |
| "loss": 0.5528, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5113636363636362, | |
| "grad_norm": 0.4374965727329254, | |
| "learning_rate": 4.705994795999991e-06, | |
| "loss": 0.5238, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.518181818181818, | |
| "grad_norm": 0.6026428937911987, | |
| "learning_rate": 4.701214709190277e-06, | |
| "loss": 0.5268, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.525, | |
| "grad_norm": 0.4413348138332367, | |
| "learning_rate": 4.696398544740955e-06, | |
| "loss": 0.5534, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5318181818181817, | |
| "grad_norm": 0.45917898416519165, | |
| "learning_rate": 4.69154638158837e-06, | |
| "loss": 0.511, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5386363636363636, | |
| "grad_norm": 0.5339583158493042, | |
| "learning_rate": 4.686658299258881e-06, | |
| "loss": 0.5395, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 0.4617757201194763, | |
| "learning_rate": 4.681734377867562e-06, | |
| "loss": 0.5425, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5522727272727272, | |
| "grad_norm": 0.4513031840324402, | |
| "learning_rate": 4.67677469811688e-06, | |
| "loss": 0.5241, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.559090909090909, | |
| "grad_norm": 0.43548423051834106, | |
| "learning_rate": 4.671779341295378e-06, | |
| "loss": 0.5224, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.565909090909091, | |
| "grad_norm": 0.5516494512557983, | |
| "learning_rate": 4.666748389276344e-06, | |
| "loss": 0.5346, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5727272727272728, | |
| "grad_norm": 0.4441164433956146, | |
| "learning_rate": 4.661681924516466e-06, | |
| "loss": 0.5208, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5795454545454546, | |
| "grad_norm": 0.4057452380657196, | |
| "learning_rate": 4.6565800300544805e-06, | |
| "loss": 0.5173, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5863636363636364, | |
| "grad_norm": 0.4162004590034485, | |
| "learning_rate": 4.651442789509813e-06, | |
| "loss": 0.5307, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.5931818181818183, | |
| "grad_norm": 0.43314146995544434, | |
| "learning_rate": 4.646270287081208e-06, | |
| "loss": 0.5359, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5786082148551941, | |
| "learning_rate": 4.641062607545347e-06, | |
| "loss": 0.5307, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.606818181818182, | |
| "grad_norm": 0.43811941146850586, | |
| "learning_rate": 4.6358198362554585e-06, | |
| "loss": 0.5556, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6136363636363638, | |
| "grad_norm": 0.4374074935913086, | |
| "learning_rate": 4.630542059139923e-06, | |
| "loss": 0.5253, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6204545454545456, | |
| "grad_norm": 0.4387815594673157, | |
| "learning_rate": 4.625229362700863e-06, | |
| "loss": 0.534, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6272727272727274, | |
| "grad_norm": 0.4976162016391754, | |
| "learning_rate": 4.61988183401272e-06, | |
| "loss": 0.5255, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.634090909090909, | |
| "grad_norm": 0.4603639543056488, | |
| "learning_rate": 4.614499560720837e-06, | |
| "loss": 0.511, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6409090909090909, | |
| "grad_norm": 0.43136921525001526, | |
| "learning_rate": 4.609082631040012e-06, | |
| "loss": 0.5099, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.6477272727272727, | |
| "grad_norm": 0.4573090970516205, | |
| "learning_rate": 4.603631133753061e-06, | |
| "loss": 0.5515, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.6545454545454545, | |
| "grad_norm": 0.4292807877063751, | |
| "learning_rate": 4.598145158209356e-06, | |
| "loss": 0.5048, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.6613636363636364, | |
| "grad_norm": 0.6377942562103271, | |
| "learning_rate": 4.592624794323366e-06, | |
| "loss": 0.5333, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6681818181818182, | |
| "grad_norm": 0.49924615025520325, | |
| "learning_rate": 4.587070132573178e-06, | |
| "loss": 0.5127, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.675, | |
| "grad_norm": 0.4723946154117584, | |
| "learning_rate": 4.581481263999019e-06, | |
| "loss": 0.5127, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.6818181818181817, | |
| "grad_norm": 1.1846898794174194, | |
| "learning_rate": 4.575858280201761e-06, | |
| "loss": 0.5367, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.6886363636363635, | |
| "grad_norm": 0.4306166172027588, | |
| "learning_rate": 4.570201273341418e-06, | |
| "loss": 0.52, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.6954545454545453, | |
| "grad_norm": 0.45308929681777954, | |
| "learning_rate": 4.564510336135642e-06, | |
| "loss": 0.5589, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7022727272727272, | |
| "grad_norm": 0.4482448399066925, | |
| "learning_rate": 4.558785561858196e-06, | |
| "loss": 0.5161, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.709090909090909, | |
| "grad_norm": 0.44250935316085815, | |
| "learning_rate": 4.5530270443374305e-06, | |
| "loss": 0.5342, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7159090909090908, | |
| "grad_norm": 0.8177540898323059, | |
| "learning_rate": 4.547234877954741e-06, | |
| "loss": 0.5402, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.7227272727272727, | |
| "grad_norm": 0.5169854760169983, | |
| "learning_rate": 4.541409157643027e-06, | |
| "loss": 0.5301, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.7295454545454545, | |
| "grad_norm": 0.4544285237789154, | |
| "learning_rate": 4.535549978885132e-06, | |
| "loss": 0.5314, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.7363636363636363, | |
| "grad_norm": 0.42860639095306396, | |
| "learning_rate": 4.5296574377122765e-06, | |
| "loss": 0.5249, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7431818181818182, | |
| "grad_norm": 0.526492178440094, | |
| "learning_rate": 4.5237316307024895e-06, | |
| "loss": 0.5142, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.4208003878593445, | |
| "learning_rate": 4.517772654979024e-06, | |
| "loss": 0.5151, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7568181818181818, | |
| "grad_norm": 0.4974212944507599, | |
| "learning_rate": 4.51178060820876e-06, | |
| "loss": 0.5161, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7636363636363637, | |
| "grad_norm": 0.4148060083389282, | |
| "learning_rate": 4.505755588600613e-06, | |
| "loss": 0.5099, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7704545454545455, | |
| "grad_norm": 0.4106168746948242, | |
| "learning_rate": 4.499697694903915e-06, | |
| "loss": 0.5376, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.7772727272727273, | |
| "grad_norm": 0.49017220735549927, | |
| "learning_rate": 4.493607026406802e-06, | |
| "loss": 0.525, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7840909090909092, | |
| "grad_norm": 0.4141218662261963, | |
| "learning_rate": 4.487483682934587e-06, | |
| "loss": 0.5303, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.790909090909091, | |
| "grad_norm": 0.44826585054397583, | |
| "learning_rate": 4.481327764848118e-06, | |
| "loss": 0.4886, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.7977272727272728, | |
| "grad_norm": 0.42013809084892273, | |
| "learning_rate": 4.47513937304214e-06, | |
| "loss": 0.5204, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8045454545454547, | |
| "grad_norm": 0.44327741861343384, | |
| "learning_rate": 4.4689186089436365e-06, | |
| "loss": 0.5097, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8113636363636365, | |
| "grad_norm": 0.39498746395111084, | |
| "learning_rate": 4.462665574510169e-06, | |
| "loss": 0.5102, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.48296087980270386, | |
| "learning_rate": 4.456380372228208e-06, | |
| "loss": 0.5299, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.825, | |
| "grad_norm": 0.4398634433746338, | |
| "learning_rate": 4.450063105111447e-06, | |
| "loss": 0.5419, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8318181818181818, | |
| "grad_norm": 0.43333035707473755, | |
| "learning_rate": 4.443713876699124e-06, | |
| "loss": 0.5047, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8386363636363636, | |
| "grad_norm": 0.5823659300804138, | |
| "learning_rate": 4.4373327910543125e-06, | |
| "loss": 0.5082, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.8454545454545455, | |
| "grad_norm": 0.44385817646980286, | |
| "learning_rate": 4.430919952762226e-06, | |
| "loss": 0.5171, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8522727272727273, | |
| "grad_norm": 0.4444347023963928, | |
| "learning_rate": 4.424475466928499e-06, | |
| "loss": 0.5351, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8590909090909091, | |
| "grad_norm": 0.549674391746521, | |
| "learning_rate": 4.417999439177465e-06, | |
| "loss": 0.498, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.865909090909091, | |
| "grad_norm": 0.4340274930000305, | |
| "learning_rate": 4.4114919756504275e-06, | |
| "loss": 0.5429, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.8727272727272726, | |
| "grad_norm": 0.44019854068756104, | |
| "learning_rate": 4.404953183003916e-06, | |
| "loss": 0.5133, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.8795454545454544, | |
| "grad_norm": 0.4473486542701721, | |
| "learning_rate": 4.398383168407941e-06, | |
| "loss": 0.5413, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8863636363636362, | |
| "grad_norm": 0.44450780749320984, | |
| "learning_rate": 4.391782039544239e-06, | |
| "loss": 0.4998, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.893181818181818, | |
| "grad_norm": 0.4310123920440674, | |
| "learning_rate": 4.385149904604502e-06, | |
| "loss": 0.5134, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.48953917622566223, | |
| "learning_rate": 4.378486872288611e-06, | |
| "loss": 0.5102, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9068181818181817, | |
| "grad_norm": 0.49219778180122375, | |
| "learning_rate": 4.371793051802849e-06, | |
| "loss": 0.5213, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9136363636363636, | |
| "grad_norm": 0.818645715713501, | |
| "learning_rate": 4.365068552858116e-06, | |
| "loss": 0.5251, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9204545454545454, | |
| "grad_norm": 0.4407028257846832, | |
| "learning_rate": 4.358313485668124e-06, | |
| "loss": 0.5191, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9272727272727272, | |
| "grad_norm": 0.5301506519317627, | |
| "learning_rate": 4.3515279609476e-06, | |
| "loss": 0.5003, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.934090909090909, | |
| "grad_norm": 0.4673819839954376, | |
| "learning_rate": 4.3447120899104615e-06, | |
| "loss": 0.521, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.940909090909091, | |
| "grad_norm": 0.4106168746948242, | |
| "learning_rate": 4.337865984268002e-06, | |
| "loss": 0.5199, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9477272727272728, | |
| "grad_norm": 0.43583595752716064, | |
| "learning_rate": 4.3309897562270525e-06, | |
| "loss": 0.5367, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9545454545454546, | |
| "grad_norm": 0.4984910488128662, | |
| "learning_rate": 4.324083518488151e-06, | |
| "loss": 0.5202, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9613636363636364, | |
| "grad_norm": 0.4516090452671051, | |
| "learning_rate": 4.317147384243688e-06, | |
| "loss": 0.5432, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.9681818181818183, | |
| "grad_norm": 0.43753015995025635, | |
| "learning_rate": 4.3101814671760546e-06, | |
| "loss": 0.5309, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.975, | |
| "grad_norm": 0.4472252130508423, | |
| "learning_rate": 4.303185881455778e-06, | |
| "loss": 0.5344, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.981818181818182, | |
| "grad_norm": 1.178377389907837, | |
| "learning_rate": 4.296160741739652e-06, | |
| "loss": 0.5283, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.9886363636363638, | |
| "grad_norm": 0.5085738301277161, | |
| "learning_rate": 4.289106163168858e-06, | |
| "loss": 0.5133, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.9954545454545456, | |
| "grad_norm": 0.52077716588974, | |
| "learning_rate": 4.282022261367074e-06, | |
| "loss": 0.5155, | |
| "step": 292 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 876, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 146, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.880633908528572e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |