| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 0.7325745224952698, | |
| "learning_rate": 9.99993683466483e-05, | |
| "loss": 1.3543, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 0.8166712522506714, | |
| "learning_rate": 9.999747340255259e-05, | |
| "loss": 1.2915, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 1.328125238418579, | |
| "learning_rate": 9.999431521559082e-05, | |
| "loss": 1.2584, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 1.0592037439346313, | |
| "learning_rate": 9.998989386555814e-05, | |
| "loss": 1.2065, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 1.292716145515442, | |
| "learning_rate": 9.9984209464165e-05, | |
| "loss": 1.1616, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 1.1350401639938354, | |
| "learning_rate": 9.997726215503422e-05, | |
| "loss": 1.1175, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.0224, | |
| "grad_norm": 1.0363103151321411, | |
| "learning_rate": 9.996905211369748e-05, | |
| "loss": 1.1306, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 1.3373076915740967, | |
| "learning_rate": 9.995957954759071e-05, | |
| "loss": 1.1809, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 1.1969462633132935, | |
| "learning_rate": 9.994884469604912e-05, | |
| "loss": 1.2335, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.1673309803009033, | |
| "learning_rate": 9.993684783030088e-05, | |
| "loss": 1.1687, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.0352, | |
| "grad_norm": 1.849138617515564, | |
| "learning_rate": 9.99235892534604e-05, | |
| "loss": 1.1873, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 1.3689218759536743, | |
| "learning_rate": 9.990906930052064e-05, | |
| "loss": 1.093, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.0416, | |
| "grad_norm": 1.2357085943222046, | |
| "learning_rate": 9.989328833834471e-05, | |
| "loss": 1.1549, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 1.5436211824417114, | |
| "learning_rate": 9.987624676565652e-05, | |
| "loss": 1.1943, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 1.1809171438217163, | |
| "learning_rate": 9.98579450130307e-05, | |
| "loss": 1.1305, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 1.1661288738250732, | |
| "learning_rate": 9.983838354288181e-05, | |
| "loss": 1.0564, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.0544, | |
| "grad_norm": 2.7672770023345947, | |
| "learning_rate": 9.981756284945256e-05, | |
| "loss": 1.1576, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 1.6773067712783813, | |
| "learning_rate": 9.979548345880141e-05, | |
| "loss": 1.0685, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.0608, | |
| "grad_norm": 1.1834300756454468, | |
| "learning_rate": 9.977214592878916e-05, | |
| "loss": 1.107, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 1.0704506635665894, | |
| "learning_rate": 9.974755084906502e-05, | |
| "loss": 1.1127, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 2.3311216831207275, | |
| "learning_rate": 9.972169884105153e-05, | |
| "loss": 1.1119, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 1.1934360265731812, | |
| "learning_rate": 9.969459055792903e-05, | |
| "loss": 1.1084, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.0736, | |
| "grad_norm": 1.8318133354187012, | |
| "learning_rate": 9.9666226684619e-05, | |
| "loss": 1.1249, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 1.3005311489105225, | |
| "learning_rate": 9.963660793776688e-05, | |
| "loss": 1.0904, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.3241008520126343, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 1.0616, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 2.055724620819092, | |
| "learning_rate": 9.957360884852817e-05, | |
| "loss": 1.1737, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 1.22575044631958, | |
| "learning_rate": 9.954023009788504e-05, | |
| "loss": 1.0874, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.0968865156173706, | |
| "learning_rate": 9.950559965714648e-05, | |
| "loss": 1.0928, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.0928, | |
| "grad_norm": 1.220700740814209, | |
| "learning_rate": 9.946971840128981e-05, | |
| "loss": 1.1083, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.3711702823638916, | |
| "learning_rate": 9.94325872368957e-05, | |
| "loss": 1.1401, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.0992, | |
| "grad_norm": 1.261365532875061, | |
| "learning_rate": 9.939420710212511e-05, | |
| "loss": 1.159, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 1.2131611108779907, | |
| "learning_rate": 9.935457896669568e-05, | |
| "loss": 1.1364, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 1.0336030721664429, | |
| "learning_rate": 9.931370383185718e-05, | |
| "loss": 1.0561, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 2.2293291091918945, | |
| "learning_rate": 9.927158273036625e-05, | |
| "loss": 1.121, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 0.9839213490486145, | |
| "learning_rate": 9.922821672646027e-05, | |
| "loss": 1.1557, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 1.3354933261871338, | |
| "learning_rate": 9.918360691583056e-05, | |
| "loss": 1.0198, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.1184, | |
| "grad_norm": 1.2504842281341553, | |
| "learning_rate": 9.913775442559452e-05, | |
| "loss": 1.0997, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 1.236770749092102, | |
| "learning_rate": 9.909066041426733e-05, | |
| "loss": 1.1579, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 1.1531703472137451, | |
| "learning_rate": 9.904232607173262e-05, | |
| "loss": 1.1022, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 1.2251778841018677, | |
| "learning_rate": 9.899275261921234e-05, | |
| "loss": 1.1239, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.1312, | |
| "grad_norm": 1.0417462587356567, | |
| "learning_rate": 9.894194130923602e-05, | |
| "loss": 1.1896, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 1.3808753490447998, | |
| "learning_rate": 9.888989342560899e-05, | |
| "loss": 1.096, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 0.1376, | |
| "grad_norm": 1.349967360496521, | |
| "learning_rate": 9.883661028338008e-05, | |
| "loss": 1.077, | |
| "step": 1376 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 2.028076648712158, | |
| "learning_rate": 9.87820932288083e-05, | |
| "loss": 1.0932, | |
| "step": 1408 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.088742971420288, | |
| "learning_rate": 9.872634363932887e-05, | |
| "loss": 1.1665, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 1.1702725887298584, | |
| "learning_rate": 9.866936292351836e-05, | |
| "loss": 1.058, | |
| "step": 1472 | |
| }, | |
| { | |
| "epoch": 0.1504, | |
| "grad_norm": 1.2243529558181763, | |
| "learning_rate": 9.861115252105921e-05, | |
| "loss": 1.131, | |
| "step": 1504 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 1.4041122198104858, | |
| "learning_rate": 9.855171390270324e-05, | |
| "loss": 1.0844, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 0.1568, | |
| "grad_norm": 1.4032260179519653, | |
| "learning_rate": 9.849104857023455e-05, | |
| "loss": 1.1046, | |
| "step": 1568 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.4822256565093994, | |
| "learning_rate": 9.842915805643155e-05, | |
| "loss": 1.0779, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 1.7823238372802734, | |
| "learning_rate": 9.83660439250283e-05, | |
| "loss": 1.0, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.3723595142364502, | |
| "learning_rate": 9.830170777067485e-05, | |
| "loss": 1.0838, | |
| "step": 1664 | |
| }, | |
| { | |
| "epoch": 0.1696, | |
| "grad_norm": 1.4893419742584229, | |
| "learning_rate": 9.823615121889716e-05, | |
| "loss": 1.0734, | |
| "step": 1696 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.2856264114379883, | |
| "learning_rate": 9.816937592605579e-05, | |
| "loss": 1.0497, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 1.2529082298278809, | |
| "learning_rate": 9.81013835793043e-05, | |
| "loss": 1.0624, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.107729196548462, | |
| "learning_rate": 9.80321758965464e-05, | |
| "loss": 1.0491, | |
| "step": 1792 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 0.9313052296638489, | |
| "learning_rate": 9.796175462639272e-05, | |
| "loss": 1.1561, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 1.3460460901260376, | |
| "learning_rate": 9.789012154811647e-05, | |
| "loss": 1.0803, | |
| "step": 1856 | |
| }, | |
| { | |
| "epoch": 0.1888, | |
| "grad_norm": 1.5794706344604492, | |
| "learning_rate": 9.781727847160865e-05, | |
| "loss": 1.0698, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.2449215650558472, | |
| "learning_rate": 9.774322723733216e-05, | |
| "loss": 1.103, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.1952, | |
| "grad_norm": 1.1799278259277344, | |
| "learning_rate": 9.766796971627543e-05, | |
| "loss": 1.0284, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 1.1231826543807983, | |
| "learning_rate": 9.759150780990507e-05, | |
| "loss": 1.0863, | |
| "step": 1984 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 0.9852601289749146, | |
| "learning_rate": 9.751384345011787e-05, | |
| "loss": 1.0038, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.560398817062378, | |
| "learning_rate": 9.743497859919196e-05, | |
| "loss": 1.0669, | |
| "step": 2048 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 1.0659574270248413, | |
| "learning_rate": 9.735491524973722e-05, | |
| "loss": 1.1653, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.3178914785385132, | |
| "learning_rate": 9.727365542464497e-05, | |
| "loss": 1.0349, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 0.2144, | |
| "grad_norm": 1.0162935256958008, | |
| "learning_rate": 9.719379593129512e-05, | |
| "loss": 1.1365, | |
| "step": 2144 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.0954954624176025, | |
| "learning_rate": 9.711018657323799e-05, | |
| "loss": 1.0986, | |
| "step": 2176 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.1444238424301147, | |
| "learning_rate": 9.702538692289092e-05, | |
| "loss": 1.1172, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 1.112743616104126, | |
| "learning_rate": 9.693939912281324e-05, | |
| "loss": 1.0753, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.2272, | |
| "grad_norm": 2.3814074993133545, | |
| "learning_rate": 9.685222534558421e-05, | |
| "loss": 1.0974, | |
| "step": 2272 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.231828212738037, | |
| "learning_rate": 9.676386779374819e-05, | |
| "loss": 1.065, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 0.2336, | |
| "grad_norm": 1.3037365674972534, | |
| "learning_rate": 9.667432869975897e-05, | |
| "loss": 1.0593, | |
| "step": 2336 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 0.9929208755493164, | |
| "learning_rate": 9.658361032592323e-05, | |
| "loss": 1.0158, | |
| "step": 2368 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9741840362548828, | |
| "learning_rate": 9.649171496434361e-05, | |
| "loss": 1.1219, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.1726874113082886, | |
| "learning_rate": 9.639864493686061e-05, | |
| "loss": 1.1151, | |
| "step": 2432 | |
| }, | |
| { | |
| "epoch": 0.2464, | |
| "grad_norm": 0.9950255751609802, | |
| "learning_rate": 9.630440259499406e-05, | |
| "loss": 1.0553, | |
| "step": 2464 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.0302757024765015, | |
| "learning_rate": 9.620899031988359e-05, | |
| "loss": 0.9945, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 0.2528, | |
| "grad_norm": 1.5070362091064453, | |
| "learning_rate": 9.611241052222852e-05, | |
| "loss": 1.0367, | |
| "step": 2528 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.3058743476867676, | |
| "learning_rate": 9.601466564222697e-05, | |
| "loss": 1.0692, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 1.237705111503601, | |
| "learning_rate": 9.591575814951419e-05, | |
| "loss": 1.0626, | |
| "step": 2592 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 2.5512192249298096, | |
| "learning_rate": 9.581569054310016e-05, | |
| "loss": 1.0316, | |
| "step": 2624 | |
| }, | |
| { | |
| "epoch": 0.2656, | |
| "grad_norm": 1.5230085849761963, | |
| "learning_rate": 9.571446535130641e-05, | |
| "loss": 1.1311, | |
| "step": 2656 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9144539833068848, | |
| "learning_rate": 9.561208513170223e-05, | |
| "loss": 1.0661, | |
| "step": 2688 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 1.3663253784179688, | |
| "learning_rate": 9.550855247103998e-05, | |
| "loss": 1.0214, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 1.175469994544983, | |
| "learning_rate": 9.540386998518972e-05, | |
| "loss": 1.0807, | |
| "step": 2752 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.9978043437004089, | |
| "learning_rate": 9.529804031907319e-05, | |
| "loss": 0.9998, | |
| "step": 2784 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.9085396528244019, | |
| "learning_rate": 9.519106614659692e-05, | |
| "loss": 1.0589, | |
| "step": 2816 | |
| }, | |
| { | |
| "epoch": 0.2848, | |
| "grad_norm": 1.1169906854629517, | |
| "learning_rate": 9.50829501705847e-05, | |
| "loss": 1.0892, | |
| "step": 2848 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.0185884237289429, | |
| "learning_rate": 9.497369512270926e-05, | |
| "loss": 1.1336, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.2912, | |
| "grad_norm": 1.0060242414474487, | |
| "learning_rate": 9.48633037634233e-05, | |
| "loss": 1.0556, | |
| "step": 2912 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.1675645112991333, | |
| "learning_rate": 9.475177888188969e-05, | |
| "loss": 1.1435, | |
| "step": 2944 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 1.0665452480316162, | |
| "learning_rate": 9.463912329591105e-05, | |
| "loss": 1.0272, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 0.9090532064437866, | |
| "learning_rate": 9.452533985185852e-05, | |
| "loss": 1.0426, | |
| "step": 3008 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 0.8997248411178589, | |
| "learning_rate": 9.441043142459985e-05, | |
| "loss": 1.0359, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.1093074083328247, | |
| "learning_rate": 9.429440091742676e-05, | |
| "loss": 0.9781, | |
| "step": 3072 | |
| }, | |
| { | |
| "epoch": 0.3104, | |
| "grad_norm": 1.2096401453018188, | |
| "learning_rate": 9.41772512619816e-05, | |
| "loss": 1.0604, | |
| "step": 3104 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 0.9333838820457458, | |
| "learning_rate": 9.405898541818329e-05, | |
| "loss": 1.0607, | |
| "step": 3136 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 1.3407750129699707, | |
| "learning_rate": 9.393960637415248e-05, | |
| "loss": 1.0114, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.159203290939331, | |
| "learning_rate": 9.38191171461361e-05, | |
| "loss": 1.0373, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.3232, | |
| "grad_norm": 1.2726435661315918, | |
| "learning_rate": 9.369752077843114e-05, | |
| "loss": 1.1084, | |
| "step": 3232 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 1.0973173379898071, | |
| "learning_rate": 9.357482034330775e-05, | |
| "loss": 0.9722, | |
| "step": 3264 | |
| }, | |
| { | |
| "epoch": 0.3296, | |
| "grad_norm": 1.1249974966049194, | |
| "learning_rate": 9.345101894093154e-05, | |
| "loss": 1.0646, | |
| "step": 3296 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.109535574913025, | |
| "learning_rate": 9.332611969928536e-05, | |
| "loss": 1.0296, | |
| "step": 3328 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.9725342392921448, | |
| "learning_rate": 9.32001257740902e-05, | |
| "loss": 1.0389, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 1.2341935634613037, | |
| "learning_rate": 9.307304034872545e-05, | |
| "loss": 1.064, | |
| "step": 3392 | |
| }, | |
| { | |
| "epoch": 0.3424, | |
| "grad_norm": 0.9506546854972839, | |
| "learning_rate": 9.294486663414851e-05, | |
| "loss": 1.089, | |
| "step": 3424 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 1.4520303010940552, | |
| "learning_rate": 9.281560786881363e-05, | |
| "loss": 1.0143, | |
| "step": 3456 | |
| }, | |
| { | |
| "epoch": 0.3488, | |
| "grad_norm": 1.069808006286621, | |
| "learning_rate": 9.268526731859013e-05, | |
| "loss": 1.0328, | |
| "step": 3488 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.9005358815193176, | |
| "learning_rate": 9.25538482766798e-05, | |
| "loss": 1.0842, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.9558689594268799, | |
| "learning_rate": 9.242135406353378e-05, | |
| "loss": 0.9927, | |
| "step": 3552 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 1.1002886295318604, | |
| "learning_rate": 9.228778802676863e-05, | |
| "loss": 1.1007, | |
| "step": 3584 | |
| }, | |
| { | |
| "epoch": 0.3616, | |
| "grad_norm": 1.1814830303192139, | |
| "learning_rate": 9.215315354108174e-05, | |
| "loss": 1.0102, | |
| "step": 3616 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 1.4370577335357666, | |
| "learning_rate": 9.201745400816606e-05, | |
| "loss": 1.0723, | |
| "step": 3648 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 1.0058218240737915, | |
| "learning_rate": 9.18806928566242e-05, | |
| "loss": 1.1156, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 1.2105575799942017, | |
| "learning_rate": 9.174287354188174e-05, | |
| "loss": 1.0626, | |
| "step": 3712 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.8971224427223206, | |
| "learning_rate": 9.160399954609997e-05, | |
| "loss": 1.1357, | |
| "step": 3744 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 1.018344521522522, | |
| "learning_rate": 9.146407437808788e-05, | |
| "loss": 1.1171, | |
| "step": 3776 | |
| }, | |
| { | |
| "epoch": 0.3808, | |
| "grad_norm": 1.098207712173462, | |
| "learning_rate": 9.132310157321354e-05, | |
| "loss": 1.0556, | |
| "step": 3808 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.0569736957550049, | |
| "learning_rate": 9.11810846933147e-05, | |
| "loss": 0.9559, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.3872, | |
| "grad_norm": 1.0195281505584717, | |
| "learning_rate": 9.103802732660894e-05, | |
| "loss": 1.0586, | |
| "step": 3872 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 1.4709314107894897, | |
| "learning_rate": 9.089393308760283e-05, | |
| "loss": 1.0509, | |
| "step": 3904 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.8363422751426697, | |
| "learning_rate": 9.074880561700074e-05, | |
| "loss": 1.0672, | |
| "step": 3936 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.2150477170944214, | |
| "learning_rate": 9.06026485816128e-05, | |
| "loss": 1.0645, | |
| "step": 3968 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.0260250568389893, | |
| "learning_rate": 9.045546567426227e-05, | |
| "loss": 1.0307, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 1.3611576557159424, | |
| "learning_rate": 9.03072606136922e-05, | |
| "loss": 1.1087, | |
| "step": 4032 | |
| }, | |
| { | |
| "epoch": 0.4064, | |
| "grad_norm": 1.0070726871490479, | |
| "learning_rate": 9.015803714447153e-05, | |
| "loss": 1.0799, | |
| "step": 4064 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 1.0184143781661987, | |
| "learning_rate": 9.000779903690044e-05, | |
| "loss": 1.0447, | |
| "step": 4096 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.8251619935035706, | |
| "learning_rate": 8.985655008691512e-05, | |
| "loss": 1.0781, | |
| "step": 4128 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 1.1904375553131104, | |
| "learning_rate": 8.970429411599177e-05, | |
| "loss": 1.0679, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.4192, | |
| "grad_norm": 1.1670352220535278, | |
| "learning_rate": 8.955103497105021e-05, | |
| "loss": 1.0098, | |
| "step": 4192 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.018236517906189, | |
| "learning_rate": 8.93967765243565e-05, | |
| "loss": 1.0357, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 0.4256, | |
| "grad_norm": 1.187759518623352, | |
| "learning_rate": 8.924152267342529e-05, | |
| "loss": 1.1212, | |
| "step": 4256 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 0.9191340208053589, | |
| "learning_rate": 8.908527734092114e-05, | |
| "loss": 0.9963, | |
| "step": 4288 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.250663161277771, | |
| "learning_rate": 8.893297291025703e-05, | |
| "loss": 1.1243, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.1205859184265137, | |
| "learning_rate": 8.877478715861173e-05, | |
| "loss": 0.9712, | |
| "step": 4352 | |
| }, | |
| { | |
| "epoch": 0.4384, | |
| "grad_norm": 1.024400234222412, | |
| "learning_rate": 8.86156217179956e-05, | |
| "loss": 1.0184, | |
| "step": 4384 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 1.0629040002822876, | |
| "learning_rate": 8.845548060990401e-05, | |
| "loss": 1.0391, | |
| "step": 4416 | |
| }, | |
| { | |
| "epoch": 0.4448, | |
| "grad_norm": 1.0474681854248047, | |
| "learning_rate": 8.829436788048366e-05, | |
| "loss": 1.1721, | |
| "step": 4448 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.2960838079452515, | |
| "learning_rate": 8.813228760043037e-05, | |
| "loss": 1.0247, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 1.1051262617111206, | |
| "learning_rate": 8.796924386488624e-05, | |
| "loss": 1.068, | |
| "step": 4512 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 0.9894328713417053, | |
| "learning_rate": 8.780524079333615e-05, | |
| "loss": 0.9805, | |
| "step": 4544 | |
| }, | |
| { | |
| "epoch": 0.4576, | |
| "grad_norm": 1.0095499753952026, | |
| "learning_rate": 8.764028252950365e-05, | |
| "loss": 0.9994, | |
| "step": 4576 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.0299321413040161, | |
| "learning_rate": 8.74743732412464e-05, | |
| "loss": 1.0258, | |
| "step": 4608 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 1.11245858669281, | |
| "learning_rate": 8.73075171204507e-05, | |
| "loss": 1.0388, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 1.2084026336669922, | |
| "learning_rate": 8.713971838292569e-05, | |
| "loss": 1.1596, | |
| "step": 4672 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 1.1535048484802246, | |
| "learning_rate": 8.697098126829675e-05, | |
| "loss": 1.0674, | |
| "step": 4704 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.0976839065551758, | |
| "learning_rate": 8.680131003989842e-05, | |
| "loss": 1.1089, | |
| "step": 4736 | |
| }, | |
| { | |
| "epoch": 0.4768, | |
| "grad_norm": 1.1108759641647339, | |
| "learning_rate": 8.663070898466674e-05, | |
| "loss": 1.0047, | |
| "step": 4768 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.9953986406326294, | |
| "learning_rate": 8.645918241303084e-05, | |
| "loss": 1.0991, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.4832, | |
| "grad_norm": 1.0783305168151855, | |
| "learning_rate": 8.628673465880404e-05, | |
| "loss": 1.0449, | |
| "step": 4832 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.0670068264007568, | |
| "learning_rate": 8.611337007907448e-05, | |
| "loss": 1.0002, | |
| "step": 4864 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 1.4406965970993042, | |
| "learning_rate": 8.59390930540948e-05, | |
| "loss": 1.0825, | |
| "step": 4896 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 2.000100612640381, | |
| "learning_rate": 8.576390798717174e-05, | |
| "loss": 1.0658, | |
| "step": 4928 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 1.1239198446273804, | |
| "learning_rate": 8.558781930455464e-05, | |
| "loss": 1.0066, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.965144157409668, | |
| "learning_rate": 8.54108314553238e-05, | |
| "loss": 1.0965, | |
| "step": 4992 | |
| }, | |
| { | |
| "epoch": 0.5024, | |
| "grad_norm": 1.0297799110412598, | |
| "learning_rate": 8.523294891127794e-05, | |
| "loss": 1.0257, | |
| "step": 5024 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 1.1478264331817627, | |
| "learning_rate": 8.505417616682126e-05, | |
| "loss": 1.0601, | |
| "step": 5056 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 1.0132007598876953, | |
| "learning_rate": 8.487451773884987e-05, | |
| "loss": 1.0643, | |
| "step": 5088 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.5010863542556763, | |
| "learning_rate": 8.469397816663773e-05, | |
| "loss": 1.0577, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.5152, | |
| "grad_norm": 1.0446892976760864, | |
| "learning_rate": 8.451256201172186e-05, | |
| "loss": 1.0305, | |
| "step": 5152 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 1.0374213457107544, | |
| "learning_rate": 8.433027385778716e-05, | |
| "loss": 1.0254, | |
| "step": 5184 | |
| }, | |
| { | |
| "epoch": 0.5216, | |
| "grad_norm": 0.958988606929779, | |
| "learning_rate": 8.414711831055056e-05, | |
| "loss": 1.0157, | |
| "step": 5216 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.049494981765747, | |
| "learning_rate": 8.396309999764467e-05, | |
| "loss": 1.0241, | |
| "step": 5248 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.9103986620903015, | |
| "learning_rate": 8.377822356850084e-05, | |
| "loss": 1.0658, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 1.6454554796218872, | |
| "learning_rate": 8.359249369423177e-05, | |
| "loss": 1.0543, | |
| "step": 5312 | |
| }, | |
| { | |
| "epoch": 0.5344, | |
| "grad_norm": 1.1632812023162842, | |
| "learning_rate": 8.34059150675133e-05, | |
| "loss": 1.0576, | |
| "step": 5344 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.066264033317566, | |
| "learning_rate": 8.321849240246608e-05, | |
| "loss": 1.0488, | |
| "step": 5376 | |
| }, | |
| { | |
| "epoch": 0.5408, | |
| "grad_norm": 0.9884083867073059, | |
| "learning_rate": 8.303023043453624e-05, | |
| "loss": 1.054, | |
| "step": 5408 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 1.1581878662109375, | |
| "learning_rate": 8.284113392037593e-05, | |
| "loss": 1.0847, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 1.0645771026611328, | |
| "learning_rate": 8.265120763772303e-05, | |
| "loss": 0.9862, | |
| "step": 5472 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.2600454092025757, | |
| "learning_rate": 8.246045638528047e-05, | |
| "loss": 1.0295, | |
| "step": 5504 | |
| }, | |
| { | |
| "epoch": 0.5536, | |
| "grad_norm": 1.2756901979446411, | |
| "learning_rate": 8.226888498259496e-05, | |
| "loss": 0.9753, | |
| "step": 5536 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 1.0469154119491577, | |
| "learning_rate": 8.207649826993522e-05, | |
| "loss": 1.0993, | |
| "step": 5568 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.1633126735687256, | |
| "learning_rate": 8.188330110816976e-05, | |
| "loss": 0.9892, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.7112101316452026, | |
| "learning_rate": 8.168929837864395e-05, | |
| "loss": 0.9913, | |
| "step": 5632 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 1.0041791200637817, | |
| "learning_rate": 8.149449498305674e-05, | |
| "loss": 1.0494, | |
| "step": 5664 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 1.1538423299789429, | |
| "learning_rate": 8.12988958433369e-05, | |
| "loss": 1.0383, | |
| "step": 5696 | |
| }, | |
| { | |
| "epoch": 0.5728, | |
| "grad_norm": 0.9828271865844727, | |
| "learning_rate": 8.110250590151848e-05, | |
| "loss": 1.1132, | |
| "step": 5728 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.243087649345398, | |
| "learning_rate": 8.090533011961609e-05, | |
| "loss": 1.008, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.5792, | |
| "grad_norm": 1.0514239072799683, | |
| "learning_rate": 8.070737347949947e-05, | |
| "loss": 1.0286, | |
| "step": 5792 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 1.0970929861068726, | |
| "learning_rate": 8.050864098276762e-05, | |
| "loss": 1.1212, | |
| "step": 5824 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 1.0040539503097534, | |
| "learning_rate": 8.030913765062245e-05, | |
| "loss": 1.0395, | |
| "step": 5856 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.8210061192512512, | |
| "learning_rate": 8.010886852374191e-05, | |
| "loss": 1.1159, | |
| "step": 5888 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 1.0136836767196655, | |
| "learning_rate": 7.990783866215259e-05, | |
| "loss": 1.0392, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 1.1107640266418457, | |
| "learning_rate": 7.970605314510194e-05, | |
| "loss": 1.0279, | |
| "step": 5952 | |
| }, | |
| { | |
| "epoch": 0.5984, | |
| "grad_norm": 0.9535327553749084, | |
| "learning_rate": 7.950351707092987e-05, | |
| "loss": 1.0608, | |
| "step": 5984 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.3050202131271362, | |
| "learning_rate": 7.930023555693999e-05, | |
| "loss": 1.0714, | |
| "step": 6016 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 1.0925366878509521, | |
| "learning_rate": 7.909621373927029e-05, | |
| "loss": 0.9707, | |
| "step": 6048 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.9475853443145752, | |
| "learning_rate": 7.88914567727634e-05, | |
| "loss": 1.0056, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.6112, | |
| "grad_norm": 1.2536673545837402, | |
| "learning_rate": 7.868596983083623e-05, | |
| "loss": 1.0983, | |
| "step": 6112 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.1593080759048462, | |
| "learning_rate": 7.847975810534943e-05, | |
| "loss": 1.0214, | |
| "step": 6144 | |
| }, | |
| { | |
| "epoch": 0.6176, | |
| "grad_norm": 1.4903924465179443, | |
| "learning_rate": 7.82728268064761e-05, | |
| "loss": 1.0825, | |
| "step": 6176 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 1.31364905834198, | |
| "learning_rate": 7.80651811625702e-05, | |
| "loss": 1.0184, | |
| "step": 6208 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 1.1020359992980957, | |
| "learning_rate": 7.785682642003437e-05, | |
| "loss": 0.9785, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.680654525756836, | |
| "learning_rate": 7.764776784318751e-05, | |
| "loss": 1.0493, | |
| "step": 6272 | |
| }, | |
| { | |
| "epoch": 0.6304, | |
| "grad_norm": 0.8548070192337036, | |
| "learning_rate": 7.743801071413161e-05, | |
| "loss": 1.0325, | |
| "step": 6304 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 1.3193022012710571, | |
| "learning_rate": 7.722756033261844e-05, | |
| "loss": 1.0861, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 0.6368, | |
| "grad_norm": 1.1262884140014648, | |
| "learning_rate": 7.701642201591555e-05, | |
| "loss": 0.9799, | |
| "step": 6368 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0190273523330688, | |
| "learning_rate": 7.680460109867194e-05, | |
| "loss": 0.9806, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.9623986482620239, | |
| "learning_rate": 7.659210293278334e-05, | |
| "loss": 1.0146, | |
| "step": 6432 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 0.8106020092964172, | |
| "learning_rate": 7.637893288725688e-05, | |
| "loss": 1.1549, | |
| "step": 6464 | |
| }, | |
| { | |
| "epoch": 0.6496, | |
| "grad_norm": 1.0692909955978394, | |
| "learning_rate": 7.616509634807549e-05, | |
| "loss": 1.0515, | |
| "step": 6496 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.7676146626472473, | |
| "learning_rate": 7.595059871806187e-05, | |
| "loss": 1.0496, | |
| "step": 6528 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 1.4028490781784058, | |
| "learning_rate": 7.574217882816324e-05, | |
| "loss": 1.1564, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 2.0384438037872314, | |
| "learning_rate": 7.552639552903132e-05, | |
| "loss": 0.9668, | |
| "step": 6592 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 1.113044261932373, | |
| "learning_rate": 7.53099672765677e-05, | |
| "loss": 1.0345, | |
| "step": 6624 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.3547977209091187, | |
| "learning_rate": 7.509289953907758e-05, | |
| "loss": 1.0719, | |
| "step": 6656 | |
| }, | |
| { | |
| "epoch": 0.6688, | |
| "grad_norm": 0.9287874102592468, | |
| "learning_rate": 7.487519780102354e-05, | |
| "loss": 1.0301, | |
| "step": 6688 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 1.3750686645507812, | |
| "learning_rate": 7.46568675628869e-05, | |
| "loss": 1.0542, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.6752, | |
| "grad_norm": 0.5963271260261536, | |
| "learning_rate": 7.443791434102868e-05, | |
| "loss": 0.9945, | |
| "step": 6752 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.117193341255188, | |
| "learning_rate": 7.421834366755039e-05, | |
| "loss": 1.0214, | |
| "step": 6784 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 1.096929907798767, | |
| "learning_rate": 7.399816109015407e-05, | |
| "loss": 1.0439, | |
| "step": 6816 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 1.0610090494155884, | |
| "learning_rate": 7.377737217200226e-05, | |
| "loss": 1.041, | |
| "step": 6848 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 0.9771848320960999, | |
| "learning_rate": 7.355598249157734e-05, | |
| "loss": 1.1224, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.0380698442459106, | |
| "learning_rate": 7.333399764254068e-05, | |
| "loss": 1.0475, | |
| "step": 6912 | |
| }, | |
| { | |
| "epoch": 0.6944, | |
| "grad_norm": 1.1135938167572021, | |
| "learning_rate": 7.311142323359121e-05, | |
| "loss": 0.9665, | |
| "step": 6944 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 0.9427506327629089, | |
| "learning_rate": 7.288826488832384e-05, | |
| "loss": 1.0845, | |
| "step": 6976 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 1.020609736442566, | |
| "learning_rate": 7.266452824508719e-05, | |
| "loss": 1.0806, | |
| "step": 7008 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.3327020406723022, | |
| "learning_rate": 7.244021895684131e-05, | |
| "loss": 1.0456, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.7072, | |
| "grad_norm": 0.9490824937820435, | |
| "learning_rate": 7.221534269101474e-05, | |
| "loss": 1.0546, | |
| "step": 7072 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 1.043341875076294, | |
| "learning_rate": 7.198990512936135e-05, | |
| "loss": 0.9643, | |
| "step": 7104 | |
| }, | |
| { | |
| "epoch": 0.7136, | |
| "grad_norm": 1.0628856420516968, | |
| "learning_rate": 7.17639119678168e-05, | |
| "loss": 1.0433, | |
| "step": 7136 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.8244098424911499, | |
| "learning_rate": 7.153736891635463e-05, | |
| "loss": 1.0359, | |
| "step": 7168 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.1554003953933716, | |
| "learning_rate": 7.131028169884194e-05, | |
| "loss": 1.0216, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 1.1582995653152466, | |
| "learning_rate": 7.108265605289481e-05, | |
| "loss": 0.9845, | |
| "step": 7232 | |
| }, | |
| { | |
| "epoch": 0.7264, | |
| "grad_norm": 1.1655360460281372, | |
| "learning_rate": 7.085449772973333e-05, | |
| "loss": 1.0771, | |
| "step": 7264 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.28196382522583, | |
| "learning_rate": 7.062581249403627e-05, | |
| "loss": 1.0186, | |
| "step": 7296 | |
| }, | |
| { | |
| "epoch": 0.7328, | |
| "grad_norm": 1.149167537689209, | |
| "learning_rate": 7.039660612379546e-05, | |
| "loss": 0.9905, | |
| "step": 7328 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 1.0396078824996948, | |
| "learning_rate": 7.016688441016979e-05, | |
| "loss": 1.0196, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.8532673716545105, | |
| "learning_rate": 6.993665315733889e-05, | |
| "loss": 1.0197, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.03330659866333, | |
| "learning_rate": 6.970591818235641e-05, | |
| "loss": 1.0163, | |
| "step": 7424 | |
| }, | |
| { | |
| "epoch": 0.7456, | |
| "grad_norm": 1.5266470909118652, | |
| "learning_rate": 6.947468531500321e-05, | |
| "loss": 1.0247, | |
| "step": 7456 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 1.127951979637146, | |
| "learning_rate": 6.924296039763987e-05, | |
| "loss": 0.9851, | |
| "step": 7488 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 1.0132697820663452, | |
| "learning_rate": 6.901074928505928e-05, | |
| "loss": 1.0015, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.034342646598816, | |
| "learning_rate": 6.877805784433852e-05, | |
| "loss": 0.978, | |
| "step": 7552 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.9696159958839417, | |
| "learning_rate": 6.854489195469069e-05, | |
| "loss": 1.129, | |
| "step": 7584 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 1.056174874305725, | |
| "learning_rate": 6.831125750731646e-05, | |
| "loss": 1.0418, | |
| "step": 7616 | |
| }, | |
| { | |
| "epoch": 0.7648, | |
| "grad_norm": 0.9070044755935669, | |
| "learning_rate": 6.80771604052551e-05, | |
| "loss": 1.0073, | |
| "step": 7648 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.1860136985778809, | |
| "learning_rate": 6.784260656323533e-05, | |
| "loss": 1.0599, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.7712, | |
| "grad_norm": 1.0756847858428955, | |
| "learning_rate": 6.760760190752604e-05, | |
| "loss": 1.0392, | |
| "step": 7712 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 1.247762680053711, | |
| "learning_rate": 6.737215237578631e-05, | |
| "loss": 1.0265, | |
| "step": 7744 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 1.0265753269195557, | |
| "learning_rate": 6.71362639169156e-05, | |
| "loss": 1.031, | |
| "step": 7776 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 1.0263795852661133, | |
| "learning_rate": 6.689994249090333e-05, | |
| "loss": 0.9527, | |
| "step": 7808 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 1.2893991470336914, | |
| "learning_rate": 6.666319406867833e-05, | |
| "loss": 1.1626, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.958138644695282, | |
| "learning_rate": 6.642602463195799e-05, | |
| "loss": 1.1133, | |
| "step": 7872 | |
| }, | |
| { | |
| "epoch": 0.7904, | |
| "grad_norm": 1.257802128791809, | |
| "learning_rate": 6.618844017309708e-05, | |
| "loss": 1.0102, | |
| "step": 7904 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 1.1419870853424072, | |
| "learning_rate": 6.59504466949364e-05, | |
| "loss": 1.0997, | |
| "step": 7936 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.9523638486862183, | |
| "learning_rate": 6.571205021065108e-05, | |
| "loss": 1.0273, | |
| "step": 7968 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.1219632625579834, | |
| "learning_rate": 6.547325674359865e-05, | |
| "loss": 1.123, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8032, | |
| "grad_norm": 1.3210878372192383, | |
| "learning_rate": 6.523407232716684e-05, | |
| "loss": 0.9976, | |
| "step": 8032 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.176743984222412, | |
| "learning_rate": 6.499450300462121e-05, | |
| "loss": 1.0448, | |
| "step": 8064 | |
| }, | |
| { | |
| "epoch": 0.8096, | |
| "grad_norm": 1.2411494255065918, | |
| "learning_rate": 6.475455482895238e-05, | |
| "loss": 1.0001, | |
| "step": 8096 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 1.0539944171905518, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 1.122, | |
| "step": 8128 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 2.260613203048706, | |
| "learning_rate": 6.427354617791519e-05, | |
| "loss": 1.005, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.2137510776519775, | |
| "learning_rate": 6.403249785577589e-05, | |
| "loss": 0.9567, | |
| "step": 8192 | |
| }, | |
| { | |
| "epoch": 0.8224, | |
| "grad_norm": 1.1636831760406494, | |
| "learning_rate": 6.379109498666445e-05, | |
| "loss": 1.0428, | |
| "step": 8224 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 1.1331391334533691, | |
| "learning_rate": 6.354934366989812e-05, | |
| "loss": 1.0609, | |
| "step": 8256 | |
| }, | |
| { | |
| "epoch": 0.8288, | |
| "grad_norm": 1.5374737977981567, | |
| "learning_rate": 6.330725001359809e-05, | |
| "loss": 1.0728, | |
| "step": 8288 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.287787675857544, | |
| "learning_rate": 6.306482013453515e-05, | |
| "loss": 1.0416, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 1.130149006843567, | |
| "learning_rate": 6.28220601579751e-05, | |
| "loss": 1.0513, | |
| "step": 8352 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 0.9294027090072632, | |
| "learning_rate": 6.257897621752405e-05, | |
| "loss": 1.0551, | |
| "step": 8384 | |
| }, | |
| { | |
| "epoch": 0.8416, | |
| "grad_norm": 0.8640485405921936, | |
| "learning_rate": 6.233557445497345e-05, | |
| "loss": 1.0518, | |
| "step": 8416 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 1.1084208488464355, | |
| "learning_rate": 6.209186102014486e-05, | |
| "loss": 1.0359, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.8203976154327393, | |
| "learning_rate": 6.18478420707346e-05, | |
| "loss": 0.9709, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 1.507534384727478, | |
| "learning_rate": 6.160352377215816e-05, | |
| "loss": 0.9479, | |
| "step": 8512 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.8405012488365173, | |
| "learning_rate": 6.135891229739444e-05, | |
| "loss": 1.025, | |
| "step": 8544 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.9983368515968323, | |
| "learning_rate": 6.111401382682972e-05, | |
| "loss": 1.1023, | |
| "step": 8576 | |
| }, | |
| { | |
| "epoch": 0.8608, | |
| "grad_norm": 1.079447865486145, | |
| "learning_rate": 6.086883454810162e-05, | |
| "loss": 0.9684, | |
| "step": 8608 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.963784396648407, | |
| "learning_rate": 6.06310551852323e-05, | |
| "loss": 1.0703, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.8672, | |
| "grad_norm": 1.44817316532135, | |
| "learning_rate": 6.0385341175240205e-05, | |
| "loss": 1.0276, | |
| "step": 8672 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 1.0072044134140015, | |
| "learning_rate": 6.0139364767825626e-05, | |
| "loss": 1.0744, | |
| "step": 8704 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 1.328588604927063, | |
| "learning_rate": 5.9893132177861454e-05, | |
| "loss": 1.0823, | |
| "step": 8736 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 1.323585867881775, | |
| "learning_rate": 5.964664962669333e-05, | |
| "loss": 1.0011, | |
| "step": 8768 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.4633543491363525, | |
| "learning_rate": 5.939992334198242e-05, | |
| "loss": 0.9919, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 1.0282506942749023, | |
| "learning_rate": 5.9152959557548117e-05, | |
| "loss": 1.0215, | |
| "step": 8832 | |
| }, | |
| { | |
| "epoch": 0.8864, | |
| "grad_norm": 0.8649700284004211, | |
| "learning_rate": 5.89057645132105e-05, | |
| "loss": 1.0628, | |
| "step": 8864 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 0.9102625846862793, | |
| "learning_rate": 5.865834445463273e-05, | |
| "loss": 0.9597, | |
| "step": 8896 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 1.0294193029403687, | |
| "learning_rate": 5.841070563316315e-05, | |
| "loss": 1.0335, | |
| "step": 8928 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.122887372970581, | |
| "learning_rate": 5.8162854305677425e-05, | |
| "loss": 1.0743, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 0.8992, | |
| "grad_norm": 1.419608235359192, | |
| "learning_rate": 5.791479673442044e-05, | |
| "loss": 1.0136, | |
| "step": 8992 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 1.0360965728759766, | |
| "learning_rate": 5.7666539186848036e-05, | |
| "loss": 1.0314, | |
| "step": 9024 | |
| }, | |
| { | |
| "epoch": 0.9056, | |
| "grad_norm": 1.2409007549285889, | |
| "learning_rate": 5.74180879354687e-05, | |
| "loss": 0.903, | |
| "step": 9056 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.0799171924591064, | |
| "learning_rate": 5.716944925768505e-05, | |
| "loss": 1.0727, | |
| "step": 9088 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 1.0068849325180054, | |
| "learning_rate": 5.6920629435635256e-05, | |
| "loss": 0.9064, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 0.9761477708816528, | |
| "learning_rate": 5.6671634756034295e-05, | |
| "loss": 0.9928, | |
| "step": 9152 | |
| }, | |
| { | |
| "epoch": 0.9184, | |
| "grad_norm": 1.3264461755752563, | |
| "learning_rate": 5.642247151001515e-05, | |
| "loss": 1.0678, | |
| "step": 9184 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 1.6155526638031006, | |
| "learning_rate": 5.617314599296977e-05, | |
| "loss": 1.0057, | |
| "step": 9216 | |
| }, | |
| { | |
| "epoch": 0.9248, | |
| "grad_norm": 0.985884428024292, | |
| "learning_rate": 5.592366450439012e-05, | |
| "loss": 1.0783, | |
| "step": 9248 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 1.1194316148757935, | |
| "learning_rate": 5.567403334770891e-05, | |
| "loss": 1.086, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.9581426978111267, | |
| "learning_rate": 5.542425883014043e-05, | |
| "loss": 1.0819, | |
| "step": 9312 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 1.0260018110275269, | |
| "learning_rate": 5.517434726252113e-05, | |
| "loss": 1.0206, | |
| "step": 9344 | |
| }, | |
| { | |
| "epoch": 0.9376, | |
| "grad_norm": 1.3467062711715698, | |
| "learning_rate": 5.4924304959150175e-05, | |
| "loss": 1.0682, | |
| "step": 9376 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 1.030444622039795, | |
| "learning_rate": 5.467413823762993e-05, | |
| "loss": 1.0894, | |
| "step": 9408 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 1.1066439151763916, | |
| "learning_rate": 5.4423853418706327e-05, | |
| "loss": 0.938, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.088860034942627, | |
| "learning_rate": 5.417345682610914e-05, | |
| "loss": 1.0293, | |
| "step": 9472 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 1.4524608850479126, | |
| "learning_rate": 5.392295478639225e-05, | |
| "loss": 1.0259, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 1.0502616167068481, | |
| "learning_rate": 5.367235362877378e-05, | |
| "loss": 0.9685, | |
| "step": 9536 | |
| }, | |
| { | |
| "epoch": 0.9568, | |
| "grad_norm": 1.1287665367126465, | |
| "learning_rate": 5.3421659684976197e-05, | |
| "loss": 1.0295, | |
| "step": 9568 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.4596409797668457, | |
| "learning_rate": 5.317087928906627e-05, | |
| "loss": 1.0235, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.9632, | |
| "grad_norm": 1.3627421855926514, | |
| "learning_rate": 5.29200187772951e-05, | |
| "loss": 1.126, | |
| "step": 9632 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 1.2144567966461182, | |
| "learning_rate": 5.266908448793803e-05, | |
| "loss": 0.9882, | |
| "step": 9664 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 1.453833818435669, | |
| "learning_rate": 5.2418082761134445e-05, | |
| "loss": 1.0644, | |
| "step": 9696 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 1.1099966764450073, | |
| "learning_rate": 5.216701993872762e-05, | |
| "loss": 0.974, | |
| "step": 9728 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 0.8567425012588501, | |
| "learning_rate": 5.1915902364104506e-05, | |
| "loss": 1.0689, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 1.1577990055084229, | |
| "learning_rate": 5.166473638203539e-05, | |
| "loss": 1.0094, | |
| "step": 9792 | |
| }, | |
| { | |
| "epoch": 0.9824, | |
| "grad_norm": 0.8881478905677795, | |
| "learning_rate": 5.141352833851367e-05, | |
| "loss": 1.0945, | |
| "step": 9824 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.8964444994926453, | |
| "learning_rate": 5.116228458059543e-05, | |
| "loss": 1.0251, | |
| "step": 9856 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 1.2837964296340942, | |
| "learning_rate": 5.0911011456239157e-05, | |
| "loss": 1.1041, | |
| "step": 9888 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 1.0828759670257568, | |
| "learning_rate": 5.065971531414528e-05, | |
| "loss": 1.0765, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 0.9952, | |
| "grad_norm": 1.0157177448272705, | |
| "learning_rate": 5.0408402503595845e-05, | |
| "loss": 1.0109, | |
| "step": 9952 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 1.128143310546875, | |
| "learning_rate": 5.0157079374293983e-05, | |
| "loss": 1.0521, | |
| "step": 9984 | |
| } | |
| ], | |
| "logging_steps": 32, | |
| "max_steps": 20000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.411696496839885e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |