| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 1124, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0017793594306049821, | |
| "grad_norm": 2.2989236207973964, | |
| "learning_rate": 9.99998046979289e-06, | |
| "loss": 0.1564, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0035587188612099642, | |
| "grad_norm": 1.7813665559851424, | |
| "learning_rate": 9.999921879324127e-06, | |
| "loss": 0.1348, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005338078291814947, | |
| "grad_norm": 2.5223532997899487, | |
| "learning_rate": 9.999824229051425e-06, | |
| "loss": 0.1772, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0071174377224199285, | |
| "grad_norm": 2.5322245124330283, | |
| "learning_rate": 9.999687519737639e-06, | |
| "loss": 0.1603, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.008896797153024912, | |
| "grad_norm": 2.071273564888649, | |
| "learning_rate": 9.99951175245075e-06, | |
| "loss": 0.1158, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010676156583629894, | |
| "grad_norm": 2.4567606621659355, | |
| "learning_rate": 9.999296928563868e-06, | |
| "loss": 0.1791, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012455516014234875, | |
| "grad_norm": 2.337937390816148, | |
| "learning_rate": 9.999043049755216e-06, | |
| "loss": 0.1575, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014234875444839857, | |
| "grad_norm": 1.9047286151803622, | |
| "learning_rate": 9.998750118008117e-06, | |
| "loss": 0.1057, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01601423487544484, | |
| "grad_norm": 2.1933590076226603, | |
| "learning_rate": 9.998418135610974e-06, | |
| "loss": 0.1169, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.017793594306049824, | |
| "grad_norm": 3.1577608582010637, | |
| "learning_rate": 9.998047105157265e-06, | |
| "loss": 0.2148, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.019572953736654804, | |
| "grad_norm": 1.9338044455617387, | |
| "learning_rate": 9.997637029545509e-06, | |
| "loss": 0.1098, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.021352313167259787, | |
| "grad_norm": 2.6727098140974093, | |
| "learning_rate": 9.997187911979252e-06, | |
| "loss": 0.1353, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023131672597864767, | |
| "grad_norm": 2.5630616989369788, | |
| "learning_rate": 9.996699755967035e-06, | |
| "loss": 0.1324, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.02491103202846975, | |
| "grad_norm": 2.243001220039926, | |
| "learning_rate": 9.996172565322375e-06, | |
| "loss": 0.1125, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.026690391459074734, | |
| "grad_norm": 2.1788414018689366, | |
| "learning_rate": 9.995606344163728e-06, | |
| "loss": 0.1182, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.028469750889679714, | |
| "grad_norm": 1.7357467795887234, | |
| "learning_rate": 9.995001096914462e-06, | |
| "loss": 0.1138, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.030249110320284697, | |
| "grad_norm": 2.6896113619572843, | |
| "learning_rate": 9.994356828302818e-06, | |
| "loss": 0.1672, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03202846975088968, | |
| "grad_norm": 1.7911064017725098, | |
| "learning_rate": 9.993673543361874e-06, | |
| "loss": 0.1002, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.033807829181494664, | |
| "grad_norm": 2.490568209484001, | |
| "learning_rate": 9.992951247429512e-06, | |
| "loss": 0.1829, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03558718861209965, | |
| "grad_norm": 2.059339740578818, | |
| "learning_rate": 9.992189946148366e-06, | |
| "loss": 0.1215, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037366548042704624, | |
| "grad_norm": 1.9477880146273077, | |
| "learning_rate": 9.991389645465786e-06, | |
| "loss": 0.1003, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03914590747330961, | |
| "grad_norm": 2.0587659872794632, | |
| "learning_rate": 9.990550351633784e-06, | |
| "loss": 0.1379, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04092526690391459, | |
| "grad_norm": 1.783918075574652, | |
| "learning_rate": 9.989672071208993e-06, | |
| "loss": 0.1253, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.042704626334519574, | |
| "grad_norm": 1.9720942968915145, | |
| "learning_rate": 9.988754811052616e-06, | |
| "loss": 0.1459, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04448398576512456, | |
| "grad_norm": 2.2104818855262334, | |
| "learning_rate": 9.987798578330365e-06, | |
| "loss": 0.1397, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.046263345195729534, | |
| "grad_norm": 1.7282596203846439, | |
| "learning_rate": 9.986803380512406e-06, | |
| "loss": 0.1201, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04804270462633452, | |
| "grad_norm": 1.7087644587760773, | |
| "learning_rate": 9.98576922537331e-06, | |
| "loss": 0.1441, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0498220640569395, | |
| "grad_norm": 1.731749171911367, | |
| "learning_rate": 9.984696120991979e-06, | |
| "loss": 0.1225, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.051601423487544484, | |
| "grad_norm": 1.7526326860644732, | |
| "learning_rate": 9.983584075751598e-06, | |
| "loss": 0.1014, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05338078291814947, | |
| "grad_norm": 2.0343594011497363, | |
| "learning_rate": 9.982433098339553e-06, | |
| "loss": 0.1372, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05516014234875445, | |
| "grad_norm": 3.081538786183411, | |
| "learning_rate": 9.981243197747375e-06, | |
| "loss": 0.2113, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05693950177935943, | |
| "grad_norm": 3.4354658184110782, | |
| "learning_rate": 9.980014383270668e-06, | |
| "loss": 0.1944, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05871886120996441, | |
| "grad_norm": 2.157748190936568, | |
| "learning_rate": 9.978746664509032e-06, | |
| "loss": 0.1351, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.060498220640569395, | |
| "grad_norm": 1.8718868631035068, | |
| "learning_rate": 9.97744005136599e-06, | |
| "loss": 0.1161, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06227758007117438, | |
| "grad_norm": 1.4655541105594427, | |
| "learning_rate": 9.976094554048912e-06, | |
| "loss": 0.0979, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06405693950177936, | |
| "grad_norm": 2.117845274244393, | |
| "learning_rate": 9.974710183068935e-06, | |
| "loss": 0.1164, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06583629893238434, | |
| "grad_norm": 1.926427126810416, | |
| "learning_rate": 9.97328694924088e-06, | |
| "loss": 0.1228, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06761565836298933, | |
| "grad_norm": 2.1198504798700197, | |
| "learning_rate": 9.971824863683168e-06, | |
| "loss": 0.1712, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0693950177935943, | |
| "grad_norm": 1.991340344103373, | |
| "learning_rate": 9.970323937817732e-06, | |
| "loss": 0.0991, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0711743772241993, | |
| "grad_norm": 2.079135240864282, | |
| "learning_rate": 9.968784183369929e-06, | |
| "loss": 0.1136, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07295373665480427, | |
| "grad_norm": 2.3267271381989945, | |
| "learning_rate": 9.96720561236845e-06, | |
| "loss": 0.1208, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07473309608540925, | |
| "grad_norm": 2.4754565587419393, | |
| "learning_rate": 9.965588237145219e-06, | |
| "loss": 0.1488, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07651245551601424, | |
| "grad_norm": 1.9285896200644408, | |
| "learning_rate": 9.963932070335307e-06, | |
| "loss": 0.1159, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07829181494661921, | |
| "grad_norm": 2.335239387323971, | |
| "learning_rate": 9.962237124876828e-06, | |
| "loss": 0.146, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0800711743772242, | |
| "grad_norm": 2.3542771758551684, | |
| "learning_rate": 9.960503414010833e-06, | |
| "loss": 0.1304, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08185053380782918, | |
| "grad_norm": 1.996782869676472, | |
| "learning_rate": 9.958730951281218e-06, | |
| "loss": 0.1187, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08362989323843416, | |
| "grad_norm": 2.1747953190780733, | |
| "learning_rate": 9.956919750534607e-06, | |
| "loss": 0.1265, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08540925266903915, | |
| "grad_norm": 2.5381294454533347, | |
| "learning_rate": 9.955069825920249e-06, | |
| "loss": 0.1874, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08718861209964412, | |
| "grad_norm": 2.0812346967531603, | |
| "learning_rate": 9.953181191889913e-06, | |
| "loss": 0.1606, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.08896797153024912, | |
| "grad_norm": 1.9152691137659266, | |
| "learning_rate": 9.95125386319776e-06, | |
| "loss": 0.119, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09074733096085409, | |
| "grad_norm": 2.0986176409037864, | |
| "learning_rate": 9.949287854900243e-06, | |
| "loss": 0.1264, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.09252669039145907, | |
| "grad_norm": 2.142463501049564, | |
| "learning_rate": 9.947283182355982e-06, | |
| "loss": 0.133, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09430604982206406, | |
| "grad_norm": 2.0682496142570437, | |
| "learning_rate": 9.945239861225644e-06, | |
| "loss": 0.1456, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09608540925266904, | |
| "grad_norm": 2.3856853105760556, | |
| "learning_rate": 9.943157907471825e-06, | |
| "loss": 0.173, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.09786476868327403, | |
| "grad_norm": 1.8057000804766883, | |
| "learning_rate": 9.941037337358918e-06, | |
| "loss": 0.1371, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.099644128113879, | |
| "grad_norm": 2.1632884509987345, | |
| "learning_rate": 9.938878167452991e-06, | |
| "loss": 0.1524, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10142348754448399, | |
| "grad_norm": 2.5979880161803512, | |
| "learning_rate": 9.936680414621663e-06, | |
| "loss": 0.1851, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10320284697508897, | |
| "grad_norm": 1.7308399143466189, | |
| "learning_rate": 9.934444096033958e-06, | |
| "loss": 0.107, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10498220640569395, | |
| "grad_norm": 1.8929950621519485, | |
| "learning_rate": 9.932169229160183e-06, | |
| "loss": 0.1484, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10676156583629894, | |
| "grad_norm": 1.671527163129104, | |
| "learning_rate": 9.929855831771787e-06, | |
| "loss": 0.1081, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10854092526690391, | |
| "grad_norm": 2.169224986497327, | |
| "learning_rate": 9.927503921941218e-06, | |
| "loss": 0.1509, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1103202846975089, | |
| "grad_norm": 2.076172080188326, | |
| "learning_rate": 9.925113518041796e-06, | |
| "loss": 0.158, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11209964412811388, | |
| "grad_norm": 2.0892770140637413, | |
| "learning_rate": 9.922684638747551e-06, | |
| "loss": 0.1491, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11387900355871886, | |
| "grad_norm": 1.702615330991061, | |
| "learning_rate": 9.920217303033091e-06, | |
| "loss": 0.1154, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11565836298932385, | |
| "grad_norm": 1.9106931615502134, | |
| "learning_rate": 9.917711530173444e-06, | |
| "loss": 0.1364, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11743772241992882, | |
| "grad_norm": 2.6217097893544694, | |
| "learning_rate": 9.91516733974392e-06, | |
| "loss": 0.1512, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.11921708185053381, | |
| "grad_norm": 1.746017063983458, | |
| "learning_rate": 9.912584751619943e-06, | |
| "loss": 0.1113, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12099644128113879, | |
| "grad_norm": 1.6574813874021803, | |
| "learning_rate": 9.909963785976902e-06, | |
| "loss": 0.1159, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12277580071174377, | |
| "grad_norm": 2.2338454281739675, | |
| "learning_rate": 9.907304463290004e-06, | |
| "loss": 0.1779, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12455516014234876, | |
| "grad_norm": 1.7894754538103521, | |
| "learning_rate": 9.904606804334094e-06, | |
| "loss": 0.106, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12633451957295375, | |
| "grad_norm": 1.8272634966400607, | |
| "learning_rate": 9.901870830183506e-06, | |
| "loss": 0.1108, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.12811387900355872, | |
| "grad_norm": 1.8567383873220387, | |
| "learning_rate": 9.899096562211902e-06, | |
| "loss": 0.1253, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.1298932384341637, | |
| "grad_norm": 1.9771339695311383, | |
| "learning_rate": 9.896284022092088e-06, | |
| "loss": 0.1478, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13167259786476868, | |
| "grad_norm": 2.466326186140443, | |
| "learning_rate": 9.893433231795864e-06, | |
| "loss": 0.1629, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13345195729537365, | |
| "grad_norm": 2.672768567870085, | |
| "learning_rate": 9.890544213593838e-06, | |
| "loss": 0.1711, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13523131672597866, | |
| "grad_norm": 2.0364796130770575, | |
| "learning_rate": 9.887616990055262e-06, | |
| "loss": 0.1601, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.13701067615658363, | |
| "grad_norm": 2.283866986835037, | |
| "learning_rate": 9.884651584047845e-06, | |
| "loss": 0.1559, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1387900355871886, | |
| "grad_norm": 1.7519874907325979, | |
| "learning_rate": 9.881648018737587e-06, | |
| "loss": 0.1301, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14056939501779359, | |
| "grad_norm": 1.6619822725053455, | |
| "learning_rate": 9.878606317588588e-06, | |
| "loss": 0.1128, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1423487544483986, | |
| "grad_norm": 2.4986248745658335, | |
| "learning_rate": 9.875526504362868e-06, | |
| "loss": 0.1747, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14412811387900357, | |
| "grad_norm": 2.1998045948724023, | |
| "learning_rate": 9.872408603120187e-06, | |
| "loss": 0.149, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14590747330960854, | |
| "grad_norm": 2.1215629230985598, | |
| "learning_rate": 9.869252638217846e-06, | |
| "loss": 0.1349, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.14768683274021352, | |
| "grad_norm": 2.1920865823592663, | |
| "learning_rate": 9.866058634310503e-06, | |
| "loss": 0.1585, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1494661921708185, | |
| "grad_norm": 2.029607848823659, | |
| "learning_rate": 9.862826616349981e-06, | |
| "loss": 0.1434, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1512455516014235, | |
| "grad_norm": 2.3335641293765876, | |
| "learning_rate": 9.859556609585075e-06, | |
| "loss": 0.1552, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15302491103202848, | |
| "grad_norm": 1.8563757616233483, | |
| "learning_rate": 9.856248639561346e-06, | |
| "loss": 0.1264, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15480427046263345, | |
| "grad_norm": 2.037073463186406, | |
| "learning_rate": 9.85290273212093e-06, | |
| "loss": 0.1284, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.15658362989323843, | |
| "grad_norm": 2.2859924397379645, | |
| "learning_rate": 9.849518913402334e-06, | |
| "loss": 0.1696, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1583629893238434, | |
| "grad_norm": 2.0922642238357145, | |
| "learning_rate": 9.84609720984023e-06, | |
| "loss": 0.1334, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1601423487544484, | |
| "grad_norm": 1.9480268347042389, | |
| "learning_rate": 9.84263764816525e-06, | |
| "loss": 0.1338, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1619217081850534, | |
| "grad_norm": 1.678422407540989, | |
| "learning_rate": 9.839140255403776e-06, | |
| "loss": 0.1067, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16370106761565836, | |
| "grad_norm": 2.2563006092322784, | |
| "learning_rate": 9.83560505887773e-06, | |
| "loss": 0.1559, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.16548042704626334, | |
| "grad_norm": 1.6246367566498707, | |
| "learning_rate": 9.83203208620436e-06, | |
| "loss": 0.1215, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.16725978647686832, | |
| "grad_norm": 2.237516887966557, | |
| "learning_rate": 9.828421365296023e-06, | |
| "loss": 0.1552, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.16903914590747332, | |
| "grad_norm": 2.5083976834062, | |
| "learning_rate": 9.824772924359974e-06, | |
| "loss": 0.1672, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1708185053380783, | |
| "grad_norm": 2.0571417323812478, | |
| "learning_rate": 9.821086791898133e-06, | |
| "loss": 0.1543, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17259786476868327, | |
| "grad_norm": 1.6458847629077484, | |
| "learning_rate": 9.817362996706872e-06, | |
| "loss": 0.1223, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.17437722419928825, | |
| "grad_norm": 1.765576916674382, | |
| "learning_rate": 9.81360156787679e-06, | |
| "loss": 0.113, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.17615658362989323, | |
| "grad_norm": 1.781819931683478, | |
| "learning_rate": 9.809802534792477e-06, | |
| "loss": 0.1243, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.17793594306049823, | |
| "grad_norm": 2.048821774210223, | |
| "learning_rate": 9.805965927132294e-06, | |
| "loss": 0.1222, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1797153024911032, | |
| "grad_norm": 1.8527010656895775, | |
| "learning_rate": 9.802091774868143e-06, | |
| "loss": 0.1268, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.18149466192170818, | |
| "grad_norm": 1.7331501942412082, | |
| "learning_rate": 9.798180108265218e-06, | |
| "loss": 0.1129, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.18327402135231316, | |
| "grad_norm": 2.168172486956003, | |
| "learning_rate": 9.794230957881785e-06, | |
| "loss": 0.1551, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.18505338078291814, | |
| "grad_norm": 2.316161236824659, | |
| "learning_rate": 9.79024435456893e-06, | |
| "loss": 0.1562, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.18683274021352314, | |
| "grad_norm": 1.9367865674126532, | |
| "learning_rate": 9.786220329470334e-06, | |
| "loss": 0.1252, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.18861209964412812, | |
| "grad_norm": 2.0310727901530865, | |
| "learning_rate": 9.782158914022011e-06, | |
| "loss": 0.1209, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1903914590747331, | |
| "grad_norm": 1.816237044194044, | |
| "learning_rate": 9.778060139952075e-06, | |
| "loss": 0.1263, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.19217081850533807, | |
| "grad_norm": 2.10183625403862, | |
| "learning_rate": 9.773924039280488e-06, | |
| "loss": 0.1494, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.19395017793594305, | |
| "grad_norm": 1.7096421758593112, | |
| "learning_rate": 9.769750644318814e-06, | |
| "loss": 0.1129, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.19572953736654805, | |
| "grad_norm": 2.1246164917614117, | |
| "learning_rate": 9.765539987669956e-06, | |
| "loss": 0.1309, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.19750889679715303, | |
| "grad_norm": 2.0518698201546512, | |
| "learning_rate": 9.761292102227917e-06, | |
| "loss": 0.15, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.199288256227758, | |
| "grad_norm": 2.4973895272401343, | |
| "learning_rate": 9.757007021177529e-06, | |
| "loss": 0.1826, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.20106761565836298, | |
| "grad_norm": 1.4352488862690622, | |
| "learning_rate": 9.752684777994197e-06, | |
| "loss": 0.1111, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20284697508896798, | |
| "grad_norm": 2.1493876954569187, | |
| "learning_rate": 9.748325406443647e-06, | |
| "loss": 0.1351, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.20462633451957296, | |
| "grad_norm": 2.167583031870829, | |
| "learning_rate": 9.743928940581646e-06, | |
| "loss": 0.1776, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.20640569395017794, | |
| "grad_norm": 2.079333452495255, | |
| "learning_rate": 9.739495414753754e-06, | |
| "loss": 0.14, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.20818505338078291, | |
| "grad_norm": 1.7159999220045428, | |
| "learning_rate": 9.73502486359504e-06, | |
| "loss": 0.1131, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2099644128113879, | |
| "grad_norm": 2.2668620362303913, | |
| "learning_rate": 9.73051732202982e-06, | |
| "loss": 0.1365, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2117437722419929, | |
| "grad_norm": 1.6400652392684887, | |
| "learning_rate": 9.725972825271381e-06, | |
| "loss": 0.1256, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21352313167259787, | |
| "grad_norm": 1.925694871610653, | |
| "learning_rate": 9.721391408821713e-06, | |
| "loss": 0.1446, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21530249110320285, | |
| "grad_norm": 1.9054453544294028, | |
| "learning_rate": 9.716773108471213e-06, | |
| "loss": 0.1467, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.21708185053380782, | |
| "grad_norm": 1.8696234738312318, | |
| "learning_rate": 9.712117960298433e-06, | |
| "loss": 0.1553, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2188612099644128, | |
| "grad_norm": 1.868895945451807, | |
| "learning_rate": 9.707426000669773e-06, | |
| "loss": 0.1273, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2206405693950178, | |
| "grad_norm": 2.1368463371939472, | |
| "learning_rate": 9.702697266239211e-06, | |
| "loss": 0.1668, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22241992882562278, | |
| "grad_norm": 1.9181341318535814, | |
| "learning_rate": 9.697931793948012e-06, | |
| "loss": 0.1342, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.22419928825622776, | |
| "grad_norm": 1.823980847621837, | |
| "learning_rate": 9.693129621024441e-06, | |
| "loss": 0.1431, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.22597864768683273, | |
| "grad_norm": 1.778588368481193, | |
| "learning_rate": 9.68829078498347e-06, | |
| "loss": 0.1225, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2277580071174377, | |
| "grad_norm": 1.9748265914521732, | |
| "learning_rate": 9.683415323626487e-06, | |
| "loss": 0.1235, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.22953736654804271, | |
| "grad_norm": 2.5225336683948645, | |
| "learning_rate": 9.678503275040997e-06, | |
| "loss": 0.1539, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2313167259786477, | |
| "grad_norm": 1.6657395600962392, | |
| "learning_rate": 9.673554677600336e-06, | |
| "loss": 0.1008, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23309608540925267, | |
| "grad_norm": 2.004877807731837, | |
| "learning_rate": 9.668569569963355e-06, | |
| "loss": 0.1406, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.23487544483985764, | |
| "grad_norm": 1.5795309646696507, | |
| "learning_rate": 9.663547991074129e-06, | |
| "loss": 0.1232, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.23665480427046262, | |
| "grad_norm": 2.418881186183738, | |
| "learning_rate": 9.658489980161643e-06, | |
| "loss": 0.1715, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.23843416370106763, | |
| "grad_norm": 2.122915577586358, | |
| "learning_rate": 9.653395576739504e-06, | |
| "loss": 0.1459, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2402135231316726, | |
| "grad_norm": 2.2930306286148756, | |
| "learning_rate": 9.648264820605611e-06, | |
| "loss": 0.1367, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24199288256227758, | |
| "grad_norm": 1.9425312023359438, | |
| "learning_rate": 9.643097751841854e-06, | |
| "loss": 0.1486, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.24377224199288255, | |
| "grad_norm": 1.8014157699113282, | |
| "learning_rate": 9.637894410813803e-06, | |
| "loss": 0.1446, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.24555160142348753, | |
| "grad_norm": 2.3065350616778386, | |
| "learning_rate": 9.632654838170393e-06, | |
| "loss": 0.1373, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.24733096085409254, | |
| "grad_norm": 1.9513606068606788, | |
| "learning_rate": 9.627379074843595e-06, | |
| "loss": 0.1295, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2491103202846975, | |
| "grad_norm": 1.868700657316523, | |
| "learning_rate": 9.622067162048111e-06, | |
| "loss": 0.1324, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2508896797153025, | |
| "grad_norm": 1.5891391169781761, | |
| "learning_rate": 9.616719141281044e-06, | |
| "loss": 0.1256, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2526690391459075, | |
| "grad_norm": 2.3232588286286857, | |
| "learning_rate": 9.611335054321576e-06, | |
| "loss": 0.1819, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.25444839857651247, | |
| "grad_norm": 2.084404875185492, | |
| "learning_rate": 9.605914943230637e-06, | |
| "loss": 0.1477, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.25622775800711745, | |
| "grad_norm": 1.7616509473707371, | |
| "learning_rate": 9.600458850350588e-06, | |
| "loss": 0.1286, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2580071174377224, | |
| "grad_norm": 2.1039965764072552, | |
| "learning_rate": 9.594966818304875e-06, | |
| "loss": 0.1221, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2597864768683274, | |
| "grad_norm": 1.6838574453504735, | |
| "learning_rate": 9.589438889997712e-06, | |
| "loss": 0.1184, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2615658362989324, | |
| "grad_norm": 1.729332872590852, | |
| "learning_rate": 9.583875108613727e-06, | |
| "loss": 0.1, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.26334519572953735, | |
| "grad_norm": 2.062850629278771, | |
| "learning_rate": 9.578275517617646e-06, | |
| "loss": 0.1482, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.26512455516014233, | |
| "grad_norm": 1.802132208562917, | |
| "learning_rate": 9.572640160753936e-06, | |
| "loss": 0.1369, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.2669039145907473, | |
| "grad_norm": 1.763637055498114, | |
| "learning_rate": 9.566969082046471e-06, | |
| "loss": 0.1288, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.26868327402135234, | |
| "grad_norm": 1.619884850861297, | |
| "learning_rate": 9.561262325798188e-06, | |
| "loss": 0.123, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2704626334519573, | |
| "grad_norm": 1.6964950867807882, | |
| "learning_rate": 9.555519936590739e-06, | |
| "loss": 0.1206, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2722419928825623, | |
| "grad_norm": 1.8265151082224544, | |
| "learning_rate": 9.549741959284147e-06, | |
| "loss": 0.1584, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.27402135231316727, | |
| "grad_norm": 1.8707693790088105, | |
| "learning_rate": 9.543928439016445e-06, | |
| "loss": 0.1251, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.27580071174377224, | |
| "grad_norm": 1.7602122164021248, | |
| "learning_rate": 9.538079421203339e-06, | |
| "loss": 0.1304, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2775800711743772, | |
| "grad_norm": 1.6522595410691336, | |
| "learning_rate": 9.532194951537838e-06, | |
| "loss": 0.1169, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2793594306049822, | |
| "grad_norm": 1.5262924119470684, | |
| "learning_rate": 9.52627507598991e-06, | |
| "loss": 0.1031, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.28113879003558717, | |
| "grad_norm": 1.440983776296039, | |
| "learning_rate": 9.52031984080611e-06, | |
| "loss": 0.0937, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.28291814946619215, | |
| "grad_norm": 1.828843274034301, | |
| "learning_rate": 9.514329292509227e-06, | |
| "loss": 0.1266, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2846975088967972, | |
| "grad_norm": 2.450461512701518, | |
| "learning_rate": 9.508303477897925e-06, | |
| "loss": 0.1472, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.28647686832740216, | |
| "grad_norm": 2.111322718355945, | |
| "learning_rate": 9.502242444046365e-06, | |
| "loss": 0.1413, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.28825622775800713, | |
| "grad_norm": 1.8118749341906069, | |
| "learning_rate": 9.496146238303846e-06, | |
| "loss": 0.1319, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2900355871886121, | |
| "grad_norm": 2.3238041012569584, | |
| "learning_rate": 9.49001490829443e-06, | |
| "loss": 0.164, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2918149466192171, | |
| "grad_norm": 2.274046842610906, | |
| "learning_rate": 9.483848501916578e-06, | |
| "loss": 0.1864, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.29359430604982206, | |
| "grad_norm": 2.0005303780222827, | |
| "learning_rate": 9.477647067342766e-06, | |
| "loss": 0.1383, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.29537366548042704, | |
| "grad_norm": 1.7614172003829343, | |
| "learning_rate": 9.471410653019115e-06, | |
| "loss": 0.1088, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.297153024911032, | |
| "grad_norm": 2.061466220289302, | |
| "learning_rate": 9.46513930766501e-06, | |
| "loss": 0.1423, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.298932384341637, | |
| "grad_norm": 1.7544620131908184, | |
| "learning_rate": 9.458833080272723e-06, | |
| "loss": 0.1306, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.30071174377224197, | |
| "grad_norm": 1.6969146095169154, | |
| "learning_rate": 9.45249202010702e-06, | |
| "loss": 0.1085, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.302491103202847, | |
| "grad_norm": 2.3018345693609215, | |
| "learning_rate": 9.446116176704791e-06, | |
| "loss": 0.1477, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.304270462633452, | |
| "grad_norm": 1.8144854241543134, | |
| "learning_rate": 9.439705599874653e-06, | |
| "loss": 0.1185, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.30604982206405695, | |
| "grad_norm": 1.93924924281574, | |
| "learning_rate": 9.433260339696564e-06, | |
| "loss": 0.1335, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.30782918149466193, | |
| "grad_norm": 2.1445317733091005, | |
| "learning_rate": 9.426780446521429e-06, | |
| "loss": 0.1562, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3096085409252669, | |
| "grad_norm": 1.7525833158554183, | |
| "learning_rate": 9.42026597097071e-06, | |
| "loss": 0.1519, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3113879003558719, | |
| "grad_norm": 1.8745009592234907, | |
| "learning_rate": 9.413716963936033e-06, | |
| "loss": 0.1297, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.31316725978647686, | |
| "grad_norm": 1.7905240646112448, | |
| "learning_rate": 9.407133476578778e-06, | |
| "loss": 0.1211, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.31494661921708184, | |
| "grad_norm": 2.620598102368107, | |
| "learning_rate": 9.400515560329698e-06, | |
| "loss": 0.1861, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3167259786476868, | |
| "grad_norm": 1.9199663344653912, | |
| "learning_rate": 9.393863266888501e-06, | |
| "loss": 0.1121, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3185053380782918, | |
| "grad_norm": 2.303835436255933, | |
| "learning_rate": 9.387176648223457e-06, | |
| "loss": 0.1577, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3202846975088968, | |
| "grad_norm": 2.5725211804297796, | |
| "learning_rate": 9.38045575657098e-06, | |
| "loss": 0.1527, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3220640569395018, | |
| "grad_norm": 2.2059523103434544, | |
| "learning_rate": 9.37370064443524e-06, | |
| "loss": 0.1232, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3238434163701068, | |
| "grad_norm": 1.4612683730316718, | |
| "learning_rate": 9.366911364587726e-06, | |
| "loss": 0.0981, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.32562277580071175, | |
| "grad_norm": 1.9675096444201836, | |
| "learning_rate": 9.360087970066854e-06, | |
| "loss": 0.1293, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3274021352313167, | |
| "grad_norm": 1.8012652789722516, | |
| "learning_rate": 9.353230514177553e-06, | |
| "loss": 0.113, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3291814946619217, | |
| "grad_norm": 1.8708566598483425, | |
| "learning_rate": 9.346339050490832e-06, | |
| "loss": 0.1271, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3309608540925267, | |
| "grad_norm": 2.858076575009163, | |
| "learning_rate": 9.33941363284338e-06, | |
| "loss": 0.1394, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.33274021352313166, | |
| "grad_norm": 1.9800884762903848, | |
| "learning_rate": 9.332454315337129e-06, | |
| "loss": 0.1377, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.33451957295373663, | |
| "grad_norm": 1.8017351708994598, | |
| "learning_rate": 9.325461152338846e-06, | |
| "loss": 0.1302, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.33629893238434166, | |
| "grad_norm": 1.8627645531758872, | |
| "learning_rate": 9.3184341984797e-06, | |
| "loss": 0.139, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.33807829181494664, | |
| "grad_norm": 1.6087334485632374, | |
| "learning_rate": 9.311373508654838e-06, | |
| "loss": 0.1028, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3398576512455516, | |
| "grad_norm": 2.283610679799863, | |
| "learning_rate": 9.30427913802295e-06, | |
| "loss": 0.1574, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.3416370106761566, | |
| "grad_norm": 2.2180935828628603, | |
| "learning_rate": 9.297151142005852e-06, | |
| "loss": 0.1391, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.34341637010676157, | |
| "grad_norm": 1.9145714261243696, | |
| "learning_rate": 9.289989576288035e-06, | |
| "loss": 0.1411, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.34519572953736655, | |
| "grad_norm": 2.0455326768937176, | |
| "learning_rate": 9.282794496816244e-06, | |
| "loss": 0.1752, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3469750889679715, | |
| "grad_norm": 2.296487446186895, | |
| "learning_rate": 9.27556595979904e-06, | |
| "loss": 0.142, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3487544483985765, | |
| "grad_norm": 1.9045424282228405, | |
| "learning_rate": 9.26830402170635e-06, | |
| "loss": 0.1265, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3505338078291815, | |
| "grad_norm": 1.7648265263183143, | |
| "learning_rate": 9.261008739269035e-06, | |
| "loss": 0.1168, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.35231316725978645, | |
| "grad_norm": 1.6660434367157462, | |
| "learning_rate": 9.253680169478448e-06, | |
| "loss": 0.115, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3540925266903915, | |
| "grad_norm": 2.0098103858739504, | |
| "learning_rate": 9.246318369585983e-06, | |
| "loss": 0.1374, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.35587188612099646, | |
| "grad_norm": 2.1062139217216402, | |
| "learning_rate": 9.238923397102629e-06, | |
| "loss": 0.1523, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.35587188612099646, | |
| "eval_loss": 0.1489383727312088, | |
| "eval_runtime": 1.9393, | |
| "eval_samples_per_second": 23.72, | |
| "eval_steps_per_second": 6.188, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.35765124555160144, | |
| "grad_norm": 1.82737657240112, | |
| "learning_rate": 9.231495309798525e-06, | |
| "loss": 0.1293, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.3594306049822064, | |
| "grad_norm": 1.968888809117369, | |
| "learning_rate": 9.224034165702506e-06, | |
| "loss": 0.1339, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.3612099644128114, | |
| "grad_norm": 1.3379121552988635, | |
| "learning_rate": 9.216540023101646e-06, | |
| "loss": 0.1007, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.36298932384341637, | |
| "grad_norm": 2.117061114365621, | |
| "learning_rate": 9.209012940540806e-06, | |
| "loss": 0.1543, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.36476868327402134, | |
| "grad_norm": 1.5774503058121385, | |
| "learning_rate": 9.20145297682218e-06, | |
| "loss": 0.118, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3665480427046263, | |
| "grad_norm": 2.508006304283157, | |
| "learning_rate": 9.193860191004833e-06, | |
| "loss": 0.1939, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3683274021352313, | |
| "grad_norm": 2.2709734282118568, | |
| "learning_rate": 9.186234642404234e-06, | |
| "loss": 0.1602, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3701067615658363, | |
| "grad_norm": 1.980296254541841, | |
| "learning_rate": 9.178576390591803e-06, | |
| "loss": 0.1319, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3718861209964413, | |
| "grad_norm": 1.627166091801382, | |
| "learning_rate": 9.170885495394435e-06, | |
| "loss": 0.1137, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3736654804270463, | |
| "grad_norm": 2.0535176330016522, | |
| "learning_rate": 9.16316201689404e-06, | |
| "loss": 0.1382, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.37544483985765126, | |
| "grad_norm": 2.2026299899166415, | |
| "learning_rate": 9.155406015427076e-06, | |
| "loss": 0.1643, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.37722419928825623, | |
| "grad_norm": 1.9827206570141291, | |
| "learning_rate": 9.147617551584066e-06, | |
| "loss": 0.1261, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.3790035587188612, | |
| "grad_norm": 1.812586017156392, | |
| "learning_rate": 9.139796686209135e-06, | |
| "loss": 0.1784, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3807829181494662, | |
| "grad_norm": 1.4698103648824048, | |
| "learning_rate": 9.131943480399531e-06, | |
| "loss": 0.1065, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.38256227758007116, | |
| "grad_norm": 1.6311538504504814, | |
| "learning_rate": 9.124057995505148e-06, | |
| "loss": 0.1276, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.38434163701067614, | |
| "grad_norm": 2.696981921866431, | |
| "learning_rate": 9.11614029312805e-06, | |
| "loss": 0.2113, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3861209964412811, | |
| "grad_norm": 2.2984569011103493, | |
| "learning_rate": 9.108190435121982e-06, | |
| "loss": 0.1447, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3879003558718861, | |
| "grad_norm": 2.350433595803923, | |
| "learning_rate": 9.100208483591892e-06, | |
| "loss": 0.1542, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3896797153024911, | |
| "grad_norm": 2.24136204397637, | |
| "learning_rate": 9.092194500893448e-06, | |
| "loss": 0.1787, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3914590747330961, | |
| "grad_norm": 1.7354438133326449, | |
| "learning_rate": 9.084148549632547e-06, | |
| "loss": 0.1086, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3932384341637011, | |
| "grad_norm": 2.16791568368727, | |
| "learning_rate": 9.076070692664827e-06, | |
| "loss": 0.1538, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.39501779359430605, | |
| "grad_norm": 1.8399439607049857, | |
| "learning_rate": 9.067960993095176e-06, | |
| "loss": 0.1256, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.39679715302491103, | |
| "grad_norm": 1.8146529775663895, | |
| "learning_rate": 9.059819514277238e-06, | |
| "loss": 0.115, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.398576512455516, | |
| "grad_norm": 1.6446584525984949, | |
| "learning_rate": 9.05164631981292e-06, | |
| "loss": 0.1272, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.400355871886121, | |
| "grad_norm": 1.724807538404155, | |
| "learning_rate": 9.043441473551893e-06, | |
| "loss": 0.1302, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.40213523131672596, | |
| "grad_norm": 2.0480014377554525, | |
| "learning_rate": 9.035205039591099e-06, | |
| "loss": 0.1363, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.40391459074733094, | |
| "grad_norm": 2.411552392849718, | |
| "learning_rate": 9.02693708227424e-06, | |
| "loss": 0.1792, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.40569395017793597, | |
| "grad_norm": 1.8820734760015407, | |
| "learning_rate": 9.018637666191284e-06, | |
| "loss": 0.1694, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.40747330960854095, | |
| "grad_norm": 2.0736111763540483, | |
| "learning_rate": 9.010306856177958e-06, | |
| "loss": 0.1485, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.4092526690391459, | |
| "grad_norm": 2.0774985553312675, | |
| "learning_rate": 9.001944717315236e-06, | |
| "loss": 0.1723, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4110320284697509, | |
| "grad_norm": 2.153019546210872, | |
| "learning_rate": 8.993551314928846e-06, | |
| "loss": 0.172, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.4128113879003559, | |
| "grad_norm": 1.7596660656703789, | |
| "learning_rate": 8.985126714588739e-06, | |
| "loss": 0.1064, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.41459074733096085, | |
| "grad_norm": 2.050212279106589, | |
| "learning_rate": 8.976670982108591e-06, | |
| "loss": 0.1615, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.41637010676156583, | |
| "grad_norm": 2.0199767485731472, | |
| "learning_rate": 8.968184183545285e-06, | |
| "loss": 0.1595, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.4181494661921708, | |
| "grad_norm": 2.2903354374900364, | |
| "learning_rate": 8.959666385198396e-06, | |
| "loss": 0.16, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.4199288256227758, | |
| "grad_norm": 1.7311816344743154, | |
| "learning_rate": 8.951117653609666e-06, | |
| "loss": 0.132, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.42170818505338076, | |
| "grad_norm": 2.282726355585462, | |
| "learning_rate": 8.9425380555625e-06, | |
| "loss": 0.143, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.4234875444839858, | |
| "grad_norm": 1.3548124067386482, | |
| "learning_rate": 8.933927658081423e-06, | |
| "loss": 0.1001, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.42526690391459077, | |
| "grad_norm": 1.6678129486487536, | |
| "learning_rate": 8.925286528431578e-06, | |
| "loss": 0.126, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.42704626334519574, | |
| "grad_norm": 2.0240582937958314, | |
| "learning_rate": 8.916614734118184e-06, | |
| "loss": 0.1155, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4288256227758007, | |
| "grad_norm": 1.5154199196672344, | |
| "learning_rate": 8.907912342886016e-06, | |
| "loss": 0.1079, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4306049822064057, | |
| "grad_norm": 1.9345248241941613, | |
| "learning_rate": 8.899179422718877e-06, | |
| "loss": 0.1339, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.43238434163701067, | |
| "grad_norm": 2.0198358190111216, | |
| "learning_rate": 8.890416041839061e-06, | |
| "loss": 0.1472, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.43416370106761565, | |
| "grad_norm": 1.9432122003698364, | |
| "learning_rate": 8.881622268706825e-06, | |
| "loss": 0.1383, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.4359430604982206, | |
| "grad_norm": 1.7110962716287574, | |
| "learning_rate": 8.872798172019856e-06, | |
| "loss": 0.1341, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4377224199288256, | |
| "grad_norm": 1.7978527419615087, | |
| "learning_rate": 8.863943820712726e-06, | |
| "loss": 0.1287, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4395017793594306, | |
| "grad_norm": 1.7770100710251968, | |
| "learning_rate": 8.855059283956363e-06, | |
| "loss": 0.1587, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4412811387900356, | |
| "grad_norm": 2.2199555218586897, | |
| "learning_rate": 8.8461446311575e-06, | |
| "loss": 0.151, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4430604982206406, | |
| "grad_norm": 2.1465622201874757, | |
| "learning_rate": 8.837199931958147e-06, | |
| "loss": 0.1328, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.44483985765124556, | |
| "grad_norm": 1.8944999197678942, | |
| "learning_rate": 8.828225256235035e-06, | |
| "loss": 0.1487, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.44661921708185054, | |
| "grad_norm": 1.9278885564480934, | |
| "learning_rate": 8.819220674099074e-06, | |
| "loss": 0.1246, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.4483985765124555, | |
| "grad_norm": 1.655052551788192, | |
| "learning_rate": 8.810186255894804e-06, | |
| "loss": 0.1215, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.4501779359430605, | |
| "grad_norm": 1.9974700534312668, | |
| "learning_rate": 8.801122072199848e-06, | |
| "loss": 0.1287, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.45195729537366547, | |
| "grad_norm": 2.111237827175182, | |
| "learning_rate": 8.792028193824364e-06, | |
| "loss": 0.1329, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.45373665480427045, | |
| "grad_norm": 1.655033231401923, | |
| "learning_rate": 8.782904691810478e-06, | |
| "loss": 0.1138, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.4555160142348754, | |
| "grad_norm": 2.0011081929994914, | |
| "learning_rate": 8.77375163743175e-06, | |
| "loss": 0.1214, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.45729537366548045, | |
| "grad_norm": 1.9329829382424308, | |
| "learning_rate": 8.764569102192593e-06, | |
| "loss": 0.1623, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.45907473309608543, | |
| "grad_norm": 1.9358747271549825, | |
| "learning_rate": 8.755357157827735e-06, | |
| "loss": 0.1216, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.4608540925266904, | |
| "grad_norm": 2.3238350928952753, | |
| "learning_rate": 8.746115876301651e-06, | |
| "loss": 0.1829, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.4626334519572954, | |
| "grad_norm": 1.8512659728956768, | |
| "learning_rate": 8.736845329807994e-06, | |
| "loss": 0.1268, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.46441281138790036, | |
| "grad_norm": 2.362421839618523, | |
| "learning_rate": 8.727545590769044e-06, | |
| "loss": 0.1787, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.46619217081850534, | |
| "grad_norm": 1.8750109633950303, | |
| "learning_rate": 8.718216731835131e-06, | |
| "loss": 0.1371, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.4679715302491103, | |
| "grad_norm": 1.828984033541546, | |
| "learning_rate": 8.708858825884075e-06, | |
| "loss": 0.1528, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.4697508896797153, | |
| "grad_norm": 2.028100053651681, | |
| "learning_rate": 8.699471946020612e-06, | |
| "loss": 0.1316, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.47153024911032027, | |
| "grad_norm": 1.9913127243144877, | |
| "learning_rate": 8.690056165575825e-06, | |
| "loss": 0.1269, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.47330960854092524, | |
| "grad_norm": 1.733281028495838, | |
| "learning_rate": 8.680611558106571e-06, | |
| "loss": 0.1175, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.4750889679715303, | |
| "grad_norm": 1.7202175780024154, | |
| "learning_rate": 8.671138197394907e-06, | |
| "loss": 0.1069, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.47686832740213525, | |
| "grad_norm": 1.4414730224242311, | |
| "learning_rate": 8.661636157447511e-06, | |
| "loss": 0.1071, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.4786476868327402, | |
| "grad_norm": 2.2644269503266745, | |
| "learning_rate": 8.652105512495106e-06, | |
| "loss": 0.1408, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.4804270462633452, | |
| "grad_norm": 1.764548619184198, | |
| "learning_rate": 8.64254633699188e-06, | |
| "loss": 0.1179, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4822064056939502, | |
| "grad_norm": 1.817289657456145, | |
| "learning_rate": 8.632958705614905e-06, | |
| "loss": 0.1213, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.48398576512455516, | |
| "grad_norm": 1.9622502159273811, | |
| "learning_rate": 8.623342693263549e-06, | |
| "loss": 0.1306, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.48576512455516013, | |
| "grad_norm": 2.060841295682792, | |
| "learning_rate": 8.6136983750589e-06, | |
| "loss": 0.1545, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.4875444839857651, | |
| "grad_norm": 2.1843036948227676, | |
| "learning_rate": 8.604025826343167e-06, | |
| "loss": 0.1657, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4893238434163701, | |
| "grad_norm": 1.8859677469712872, | |
| "learning_rate": 8.594325122679107e-06, | |
| "loss": 0.1342, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.49110320284697506, | |
| "grad_norm": 1.6957528694808495, | |
| "learning_rate": 8.584596339849419e-06, | |
| "loss": 0.1382, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.4928825622775801, | |
| "grad_norm": 1.6179145710673422, | |
| "learning_rate": 8.574839553856157e-06, | |
| "loss": 0.1142, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.49466192170818507, | |
| "grad_norm": 1.9839374607959928, | |
| "learning_rate": 8.565054840920145e-06, | |
| "loss": 0.1531, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.49644128113879005, | |
| "grad_norm": 1.6282069337641418, | |
| "learning_rate": 8.55524227748037e-06, | |
| "loss": 0.1015, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.498220640569395, | |
| "grad_norm": 1.7162499529242055, | |
| "learning_rate": 8.545401940193392e-06, | |
| "loss": 0.1111, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.073586158325745, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.1528, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.501779359430605, | |
| "grad_norm": 2.3997390328097934, | |
| "learning_rate": 8.525638251788312e-06, | |
| "loss": 0.1638, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.50355871886121, | |
| "grad_norm": 1.8013202784408173, | |
| "learning_rate": 8.515715055065783e-06, | |
| "loss": 0.1467, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.505338078291815, | |
| "grad_norm": 1.9094506792871493, | |
| "learning_rate": 8.505764393285985e-06, | |
| "loss": 0.1559, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5071174377224199, | |
| "grad_norm": 1.8208572422943416, | |
| "learning_rate": 8.495786344184314e-06, | |
| "loss": 0.1218, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5088967971530249, | |
| "grad_norm": 1.8505921616187488, | |
| "learning_rate": 8.485780985710113e-06, | |
| "loss": 0.1367, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5106761565836299, | |
| "grad_norm": 1.6355200769783111, | |
| "learning_rate": 8.475748396026074e-06, | |
| "loss": 0.1216, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.5124555160142349, | |
| "grad_norm": 2.291343144883803, | |
| "learning_rate": 8.46568865350762e-06, | |
| "loss": 0.1594, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.5142348754448398, | |
| "grad_norm": 2.157466048249321, | |
| "learning_rate": 8.45560183674229e-06, | |
| "loss": 0.1606, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5160142348754448, | |
| "grad_norm": 1.9155449280684218, | |
| "learning_rate": 8.445488024529133e-06, | |
| "loss": 0.1589, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5177935943060499, | |
| "grad_norm": 1.8253569578023725, | |
| "learning_rate": 8.435347295878087e-06, | |
| "loss": 0.1428, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5195729537366548, | |
| "grad_norm": 1.4256609096704345, | |
| "learning_rate": 8.425179730009368e-06, | |
| "loss": 0.092, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5213523131672598, | |
| "grad_norm": 1.4393090847164607, | |
| "learning_rate": 8.41498540635284e-06, | |
| "loss": 0.1185, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5231316725978647, | |
| "grad_norm": 1.8605876882786099, | |
| "learning_rate": 8.404764404547404e-06, | |
| "loss": 0.1383, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5249110320284698, | |
| "grad_norm": 2.7236835096290255, | |
| "learning_rate": 8.394516804440374e-06, | |
| "loss": 0.1623, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5266903914590747, | |
| "grad_norm": 2.082228620521577, | |
| "learning_rate": 8.384242686086848e-06, | |
| "loss": 0.1663, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.5284697508896797, | |
| "grad_norm": 1.9045451072672848, | |
| "learning_rate": 8.373942129749094e-06, | |
| "loss": 0.1694, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5302491103202847, | |
| "grad_norm": 1.6459830475666937, | |
| "learning_rate": 8.363615215895908e-06, | |
| "loss": 0.1256, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5320284697508897, | |
| "grad_norm": 2.3594720417593567, | |
| "learning_rate": 8.353262025202e-06, | |
| "loss": 0.1391, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5338078291814946, | |
| "grad_norm": 2.0479536405322043, | |
| "learning_rate": 8.342882638547351e-06, | |
| "loss": 0.1181, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5355871886120996, | |
| "grad_norm": 1.6532555614688147, | |
| "learning_rate": 8.332477137016587e-06, | |
| "loss": 0.1068, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.5373665480427047, | |
| "grad_norm": 1.572478660703893, | |
| "learning_rate": 8.322045601898354e-06, | |
| "loss": 0.1088, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.5391459074733096, | |
| "grad_norm": 1.7526979074796074, | |
| "learning_rate": 8.311588114684665e-06, | |
| "loss": 0.1254, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.5409252669039146, | |
| "grad_norm": 2.3339793040734627, | |
| "learning_rate": 8.301104757070276e-06, | |
| "loss": 0.2202, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.5427046263345195, | |
| "grad_norm": 1.6503766819932704, | |
| "learning_rate": 8.290595610952045e-06, | |
| "loss": 0.1105, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5444839857651246, | |
| "grad_norm": 2.1102082160281044, | |
| "learning_rate": 8.280060758428294e-06, | |
| "loss": 0.155, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5462633451957295, | |
| "grad_norm": 1.5925313645061512, | |
| "learning_rate": 8.269500281798164e-06, | |
| "loss": 0.1084, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5480427046263345, | |
| "grad_norm": 2.1012482402590473, | |
| "learning_rate": 8.258914263560971e-06, | |
| "loss": 0.138, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5498220640569395, | |
| "grad_norm": 1.8286671524351013, | |
| "learning_rate": 8.248302786415567e-06, | |
| "loss": 0.1291, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5516014234875445, | |
| "grad_norm": 2.3990254797946857, | |
| "learning_rate": 8.237665933259693e-06, | |
| "loss": 0.1578, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5533807829181495, | |
| "grad_norm": 1.8431391591979345, | |
| "learning_rate": 8.227003787189323e-06, | |
| "loss": 0.1306, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5551601423487544, | |
| "grad_norm": 1.708077799553025, | |
| "learning_rate": 8.216316431498028e-06, | |
| "loss": 0.1049, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5569395017793595, | |
| "grad_norm": 2.188711795640406, | |
| "learning_rate": 8.205603949676317e-06, | |
| "loss": 0.1939, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.5587188612099644, | |
| "grad_norm": 1.7811414760534494, | |
| "learning_rate": 8.194866425410984e-06, | |
| "loss": 0.1294, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.5604982206405694, | |
| "grad_norm": 2.3527324164520294, | |
| "learning_rate": 8.184103942584456e-06, | |
| "loss": 0.153, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5622775800711743, | |
| "grad_norm": 1.7593461939366073, | |
| "learning_rate": 8.173316585274144e-06, | |
| "loss": 0.1296, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.5640569395017794, | |
| "grad_norm": 1.9849456630473954, | |
| "learning_rate": 8.162504437751775e-06, | |
| "loss": 0.1313, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.5658362989323843, | |
| "grad_norm": 1.978305370006857, | |
| "learning_rate": 8.151667584482742e-06, | |
| "loss": 0.1413, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.5676156583629893, | |
| "grad_norm": 2.3795317923485633, | |
| "learning_rate": 8.140806110125442e-06, | |
| "loss": 0.1761, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.5693950177935944, | |
| "grad_norm": 1.7630768195954098, | |
| "learning_rate": 8.129920099530608e-06, | |
| "loss": 0.135, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5711743772241993, | |
| "grad_norm": 1.4466369034653457, | |
| "learning_rate": 8.119009637740663e-06, | |
| "loss": 0.102, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.5729537366548043, | |
| "grad_norm": 1.2877298671473556, | |
| "learning_rate": 8.108074809989032e-06, | |
| "loss": 0.1003, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.5747330960854092, | |
| "grad_norm": 1.7791989709267704, | |
| "learning_rate": 8.097115701699498e-06, | |
| "loss": 0.1006, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.5765124555160143, | |
| "grad_norm": 1.9110862710555443, | |
| "learning_rate": 8.086132398485525e-06, | |
| "loss": 0.1579, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.5782918149466192, | |
| "grad_norm": 1.3057712667626018, | |
| "learning_rate": 8.075124986149583e-06, | |
| "loss": 0.1077, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5800711743772242, | |
| "grad_norm": 2.022062455619658, | |
| "learning_rate": 8.064093550682494e-06, | |
| "loss": 0.1322, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.5818505338078291, | |
| "grad_norm": 1.7155645040960275, | |
| "learning_rate": 8.053038178262742e-06, | |
| "loss": 0.1276, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.5836298932384342, | |
| "grad_norm": 1.7635481897670906, | |
| "learning_rate": 8.041958955255815e-06, | |
| "loss": 0.1342, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.5854092526690391, | |
| "grad_norm": 2.132120001139485, | |
| "learning_rate": 8.030855968213518e-06, | |
| "loss": 0.1413, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.5871886120996441, | |
| "grad_norm": 2.1408351595723, | |
| "learning_rate": 8.019729303873307e-06, | |
| "loss": 0.1456, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5889679715302492, | |
| "grad_norm": 1.5143764295524942, | |
| "learning_rate": 8.008579049157607e-06, | |
| "loss": 0.1123, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.5907473309608541, | |
| "grad_norm": 1.5080192518811768, | |
| "learning_rate": 7.99740529117313e-06, | |
| "loss": 0.1241, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.5925266903914591, | |
| "grad_norm": 1.6522216644823624, | |
| "learning_rate": 7.986208117210198e-06, | |
| "loss": 0.1042, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.594306049822064, | |
| "grad_norm": 1.8663923371659379, | |
| "learning_rate": 7.974987614742066e-06, | |
| "loss": 0.1486, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.5960854092526691, | |
| "grad_norm": 1.9061371176255044, | |
| "learning_rate": 7.963743871424224e-06, | |
| "loss": 0.1299, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.597864768683274, | |
| "grad_norm": 1.9899625896217472, | |
| "learning_rate": 7.952476975093729e-06, | |
| "loss": 0.1371, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.599644128113879, | |
| "grad_norm": 1.929091823609846, | |
| "learning_rate": 7.941187013768508e-06, | |
| "loss": 0.1316, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6014234875444839, | |
| "grad_norm": 1.9559812782740347, | |
| "learning_rate": 7.929874075646673e-06, | |
| "loss": 0.16, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.603202846975089, | |
| "grad_norm": 2.2540141804309157, | |
| "learning_rate": 7.918538249105835e-06, | |
| "loss": 0.1429, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.604982206405694, | |
| "grad_norm": 2.0138608355619216, | |
| "learning_rate": 7.907179622702409e-06, | |
| "loss": 0.1393, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6067615658362989, | |
| "grad_norm": 1.6115079296495123, | |
| "learning_rate": 7.895798285170927e-06, | |
| "loss": 0.127, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.608540925266904, | |
| "grad_norm": 1.696170485085791, | |
| "learning_rate": 7.88439432542334e-06, | |
| "loss": 0.1092, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6103202846975089, | |
| "grad_norm": 1.50992348645248, | |
| "learning_rate": 7.872967832548327e-06, | |
| "loss": 0.1238, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6120996441281139, | |
| "grad_norm": 1.8901436218002905, | |
| "learning_rate": 7.861518895810597e-06, | |
| "loss": 0.1624, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6138790035587188, | |
| "grad_norm": 1.6846632419632614, | |
| "learning_rate": 7.850047604650188e-06, | |
| "loss": 0.1292, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6156583629893239, | |
| "grad_norm": 2.067291303798022, | |
| "learning_rate": 7.838554048681783e-06, | |
| "loss": 0.1649, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6174377224199288, | |
| "grad_norm": 2.2669296385513458, | |
| "learning_rate": 7.827038317693988e-06, | |
| "loss": 0.1903, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6192170818505338, | |
| "grad_norm": 1.7136340957303902, | |
| "learning_rate": 7.815500501648654e-06, | |
| "loss": 0.1116, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.6209964412811388, | |
| "grad_norm": 2.0093891267306803, | |
| "learning_rate": 7.80394069068015e-06, | |
| "loss": 0.1522, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.6227758007117438, | |
| "grad_norm": 1.701346621062405, | |
| "learning_rate": 7.79235897509468e-06, | |
| "loss": 0.1296, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6245551601423488, | |
| "grad_norm": 1.9447268966552496, | |
| "learning_rate": 7.780755445369563e-06, | |
| "loss": 0.1456, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.6263345195729537, | |
| "grad_norm": 2.575529464529574, | |
| "learning_rate": 7.769130192152538e-06, | |
| "loss": 0.1778, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.6281138790035588, | |
| "grad_norm": 1.9049494022890523, | |
| "learning_rate": 7.757483306261042e-06, | |
| "loss": 0.1396, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.6298932384341637, | |
| "grad_norm": 2.372324099813749, | |
| "learning_rate": 7.745814878681516e-06, | |
| "loss": 0.1526, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.6316725978647687, | |
| "grad_norm": 1.793671292504791, | |
| "learning_rate": 7.734125000568684e-06, | |
| "loss": 0.1488, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6334519572953736, | |
| "grad_norm": 1.8116907864007974, | |
| "learning_rate": 7.722413763244837e-06, | |
| "loss": 0.1253, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.6352313167259787, | |
| "grad_norm": 1.391854247170714, | |
| "learning_rate": 7.710681258199136e-06, | |
| "loss": 0.112, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.6370106761565836, | |
| "grad_norm": 1.9494524743226096, | |
| "learning_rate": 7.69892757708688e-06, | |
| "loss": 0.121, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.6387900355871886, | |
| "grad_norm": 1.944172924841546, | |
| "learning_rate": 7.687152811728799e-06, | |
| "loss": 0.1356, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.6405693950177936, | |
| "grad_norm": 1.8677741590142636, | |
| "learning_rate": 7.675357054110337e-06, | |
| "loss": 0.1313, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6423487544483986, | |
| "grad_norm": 1.851724904866175, | |
| "learning_rate": 7.663540396380931e-06, | |
| "loss": 0.1307, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.6441281138790036, | |
| "grad_norm": 1.8447237118960993, | |
| "learning_rate": 7.651702930853287e-06, | |
| "loss": 0.1255, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.6459074733096085, | |
| "grad_norm": 2.1126688352246044, | |
| "learning_rate": 7.639844750002668e-06, | |
| "loss": 0.1583, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.6476868327402135, | |
| "grad_norm": 1.7988465637286415, | |
| "learning_rate": 7.627965946466167e-06, | |
| "loss": 0.1389, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.6494661921708185, | |
| "grad_norm": 1.9875699577813057, | |
| "learning_rate": 7.616066613041977e-06, | |
| "loss": 0.1495, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6512455516014235, | |
| "grad_norm": 2.01549944312167, | |
| "learning_rate": 7.6041468426886785e-06, | |
| "loss": 0.1383, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.6530249110320284, | |
| "grad_norm": 1.3587805725771642, | |
| "learning_rate": 7.592206728524507e-06, | |
| "loss": 0.0944, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.6548042704626335, | |
| "grad_norm": 2.1684785389669616, | |
| "learning_rate": 7.580246363826621e-06, | |
| "loss": 0.1735, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6565836298932385, | |
| "grad_norm": 2.3732748545326294, | |
| "learning_rate": 7.568265842030381e-06, | |
| "loss": 0.2111, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.6583629893238434, | |
| "grad_norm": 1.7358977964208648, | |
| "learning_rate": 7.556265256728618e-06, | |
| "loss": 0.1248, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6601423487544484, | |
| "grad_norm": 1.9236737054433828, | |
| "learning_rate": 7.544244701670894e-06, | |
| "loss": 0.1374, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.6619217081850534, | |
| "grad_norm": 1.6459222170012857, | |
| "learning_rate": 7.532204270762786e-06, | |
| "loss": 0.1205, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6637010676156584, | |
| "grad_norm": 1.6843969341131255, | |
| "learning_rate": 7.520144058065133e-06, | |
| "loss": 0.119, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6654804270462633, | |
| "grad_norm": 1.7742584158125871, | |
| "learning_rate": 7.50806415779332e-06, | |
| "loss": 0.1334, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.6672597864768683, | |
| "grad_norm": 1.8414246976104107, | |
| "learning_rate": 7.495964664316525e-06, | |
| "loss": 0.16, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6690391459074733, | |
| "grad_norm": 2.1004998093194933, | |
| "learning_rate": 7.4838456721569975e-06, | |
| "loss": 0.1609, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.6708185053380783, | |
| "grad_norm": 1.6846814490738962, | |
| "learning_rate": 7.471707275989304e-06, | |
| "loss": 0.1422, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.6725978647686833, | |
| "grad_norm": 2.04809946063895, | |
| "learning_rate": 7.459549570639602e-06, | |
| "loss": 0.1403, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.6743772241992882, | |
| "grad_norm": 1.8961854273266294, | |
| "learning_rate": 7.447372651084896e-06, | |
| "loss": 0.1326, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.6761565836298933, | |
| "grad_norm": 1.5502259966791825, | |
| "learning_rate": 7.435176612452286e-06, | |
| "loss": 0.0943, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6779359430604982, | |
| "grad_norm": 1.662497454905377, | |
| "learning_rate": 7.4229615500182396e-06, | |
| "loss": 0.1261, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.6797153024911032, | |
| "grad_norm": 1.4624470878899167, | |
| "learning_rate": 7.4107275592078345e-06, | |
| "loss": 0.11, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.6814946619217082, | |
| "grad_norm": 1.6777774873063425, | |
| "learning_rate": 7.398474735594022e-06, | |
| "loss": 0.1283, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.6832740213523132, | |
| "grad_norm": 2.293855702135554, | |
| "learning_rate": 7.386203174896872e-06, | |
| "loss": 0.1701, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.6850533807829181, | |
| "grad_norm": 1.8089331971909137, | |
| "learning_rate": 7.373912972982838e-06, | |
| "loss": 0.1138, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6868327402135231, | |
| "grad_norm": 1.483804789664669, | |
| "learning_rate": 7.361604225863992e-06, | |
| "loss": 0.1217, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.6886120996441281, | |
| "grad_norm": 2.075552114959099, | |
| "learning_rate": 7.349277029697287e-06, | |
| "loss": 0.1426, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.6903914590747331, | |
| "grad_norm": 1.7820665155676334, | |
| "learning_rate": 7.336931480783801e-06, | |
| "loss": 0.1176, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.6921708185053381, | |
| "grad_norm": 1.8678908684908941, | |
| "learning_rate": 7.3245676755679854e-06, | |
| "loss": 0.1337, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.693950177935943, | |
| "grad_norm": 2.0690006388488116, | |
| "learning_rate": 7.312185710636911e-06, | |
| "loss": 0.1519, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6957295373665481, | |
| "grad_norm": 1.671363112955066, | |
| "learning_rate": 7.299785682719512e-06, | |
| "loss": 0.1267, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.697508896797153, | |
| "grad_norm": 2.045022613885222, | |
| "learning_rate": 7.287367688685835e-06, | |
| "loss": 0.1585, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.699288256227758, | |
| "grad_norm": 2.030325281815586, | |
| "learning_rate": 7.274931825546279e-06, | |
| "loss": 0.1318, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.701067615658363, | |
| "grad_norm": 1.8312469008330514, | |
| "learning_rate": 7.262478190450834e-06, | |
| "loss": 0.122, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.702846975088968, | |
| "grad_norm": 2.3297532583102636, | |
| "learning_rate": 7.250006880688332e-06, | |
| "loss": 0.2023, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7046263345195729, | |
| "grad_norm": 1.890394181087404, | |
| "learning_rate": 7.2375179936856775e-06, | |
| "loss": 0.1232, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7064056939501779, | |
| "grad_norm": 1.6317140462424817, | |
| "learning_rate": 7.22501162700709e-06, | |
| "loss": 0.1104, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.708185053380783, | |
| "grad_norm": 1.6184272530048542, | |
| "learning_rate": 7.21248787835334e-06, | |
| "loss": 0.1009, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7099644128113879, | |
| "grad_norm": 1.6073658229476548, | |
| "learning_rate": 7.199946845560994e-06, | |
| "loss": 0.0967, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7117437722419929, | |
| "grad_norm": 2.025455744120962, | |
| "learning_rate": 7.1873886266016365e-06, | |
| "loss": 0.1438, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7117437722419929, | |
| "eval_loss": 0.1424187570810318, | |
| "eval_runtime": 1.8886, | |
| "eval_samples_per_second": 24.357, | |
| "eval_steps_per_second": 6.354, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7135231316725978, | |
| "grad_norm": 1.654858585601299, | |
| "learning_rate": 7.174813319581115e-06, | |
| "loss": 0.1238, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.7153024911032029, | |
| "grad_norm": 2.6468631957702846, | |
| "learning_rate": 7.162221022738768e-06, | |
| "loss": 0.1489, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.7170818505338078, | |
| "grad_norm": 1.7192257443494814, | |
| "learning_rate": 7.149611834446664e-06, | |
| "loss": 0.1301, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.7188612099644128, | |
| "grad_norm": 1.9419047478527747, | |
| "learning_rate": 7.136985853208824e-06, | |
| "loss": 0.1412, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.7206405693950177, | |
| "grad_norm": 1.8196636599509193, | |
| "learning_rate": 7.124343177660462e-06, | |
| "loss": 0.1057, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.7224199288256228, | |
| "grad_norm": 2.7705565717009186, | |
| "learning_rate": 7.111683906567206e-06, | |
| "loss": 0.1711, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.7241992882562278, | |
| "grad_norm": 2.0996857968821234, | |
| "learning_rate": 7.099008138824329e-06, | |
| "loss": 0.1294, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.7259786476868327, | |
| "grad_norm": 1.739217458452229, | |
| "learning_rate": 7.086315973455982e-06, | |
| "loss": 0.1282, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.7277580071174378, | |
| "grad_norm": 1.626935071009314, | |
| "learning_rate": 7.0736075096144084e-06, | |
| "loss": 0.1078, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.7295373665480427, | |
| "grad_norm": 1.6241702284618669, | |
| "learning_rate": 7.060882846579182e-06, | |
| "loss": 0.111, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7313167259786477, | |
| "grad_norm": 1.4584817108770307, | |
| "learning_rate": 7.048142083756427e-06, | |
| "loss": 0.1134, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.7330960854092526, | |
| "grad_norm": 1.886165158080951, | |
| "learning_rate": 7.035385320678035e-06, | |
| "loss": 0.132, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.7348754448398577, | |
| "grad_norm": 1.924305122482625, | |
| "learning_rate": 7.022612657000898e-06, | |
| "loss": 0.1306, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.7366548042704626, | |
| "grad_norm": 1.6899391798601864, | |
| "learning_rate": 7.0098241925061215e-06, | |
| "loss": 0.136, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.7384341637010676, | |
| "grad_norm": 1.9004468299507355, | |
| "learning_rate": 6.997020027098249e-06, | |
| "loss": 0.1371, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7402135231316725, | |
| "grad_norm": 1.8537357538861918, | |
| "learning_rate": 6.9842002608044844e-06, | |
| "loss": 0.1399, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.7419928825622776, | |
| "grad_norm": 1.7386577519917863, | |
| "learning_rate": 6.971364993773901e-06, | |
| "loss": 0.1233, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.7437722419928826, | |
| "grad_norm": 1.7258723753634884, | |
| "learning_rate": 6.958514326276669e-06, | |
| "loss": 0.1144, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.7455516014234875, | |
| "grad_norm": 1.7851255306210316, | |
| "learning_rate": 6.945648358703269e-06, | |
| "loss": 0.1256, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.7473309608540926, | |
| "grad_norm": 1.8140412762502862, | |
| "learning_rate": 6.932767191563703e-06, | |
| "loss": 0.1501, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7491103202846975, | |
| "grad_norm": 1.7349143274674739, | |
| "learning_rate": 6.919870925486718e-06, | |
| "loss": 0.1291, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.7508896797153025, | |
| "grad_norm": 1.5420328047064338, | |
| "learning_rate": 6.906959661219011e-06, | |
| "loss": 0.1061, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.7526690391459074, | |
| "grad_norm": 1.817881333519245, | |
| "learning_rate": 6.8940334996244505e-06, | |
| "loss": 0.1187, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.7544483985765125, | |
| "grad_norm": 1.812941189274749, | |
| "learning_rate": 6.881092541683279e-06, | |
| "loss": 0.113, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.7562277580071174, | |
| "grad_norm": 1.8799423931861792, | |
| "learning_rate": 6.8681368884913345e-06, | |
| "loss": 0.1187, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7580071174377224, | |
| "grad_norm": 2.0813782426274594, | |
| "learning_rate": 6.855166641259252e-06, | |
| "loss": 0.1766, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.7597864768683275, | |
| "grad_norm": 1.3767432345074846, | |
| "learning_rate": 6.8421819013116766e-06, | |
| "loss": 0.0991, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.7615658362989324, | |
| "grad_norm": 2.175448653302719, | |
| "learning_rate": 6.829182770086474e-06, | |
| "loss": 0.1916, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.7633451957295374, | |
| "grad_norm": 1.7816622393469639, | |
| "learning_rate": 6.816169349133934e-06, | |
| "loss": 0.1279, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.7651245551601423, | |
| "grad_norm": 1.812634116602351, | |
| "learning_rate": 6.803141740115979e-06, | |
| "loss": 0.096, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7669039145907474, | |
| "grad_norm": 2.774707819248449, | |
| "learning_rate": 6.7901000448053676e-06, | |
| "loss": 0.1768, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.7686832740213523, | |
| "grad_norm": 1.47482803191325, | |
| "learning_rate": 6.777044365084907e-06, | |
| "loss": 0.0974, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.7704626334519573, | |
| "grad_norm": 1.1824454013998207, | |
| "learning_rate": 6.763974802946649e-06, | |
| "loss": 0.0803, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.7722419928825622, | |
| "grad_norm": 2.0471031782609015, | |
| "learning_rate": 6.750891460491093e-06, | |
| "loss": 0.1549, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.7740213523131673, | |
| "grad_norm": 2.878679834821763, | |
| "learning_rate": 6.737794439926395e-06, | |
| "loss": 0.2033, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.7758007117437722, | |
| "grad_norm": 1.759868970556609, | |
| "learning_rate": 6.724683843567567e-06, | |
| "loss": 0.1188, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.7775800711743772, | |
| "grad_norm": 2.1487990086170057, | |
| "learning_rate": 6.711559773835672e-06, | |
| "loss": 0.1404, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.7793594306049823, | |
| "grad_norm": 1.7793012057738797, | |
| "learning_rate": 6.69842233325703e-06, | |
| "loss": 0.122, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.7811387900355872, | |
| "grad_norm": 2.1171807946525893, | |
| "learning_rate": 6.685271624462416e-06, | |
| "loss": 0.1588, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.7829181494661922, | |
| "grad_norm": 1.5619480971657118, | |
| "learning_rate": 6.672107750186255e-06, | |
| "loss": 0.1171, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7846975088967971, | |
| "grad_norm": 1.275620084824659, | |
| "learning_rate": 6.658930813265825e-06, | |
| "loss": 0.0839, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.7864768683274022, | |
| "grad_norm": 1.957516608717714, | |
| "learning_rate": 6.645740916640449e-06, | |
| "loss": 0.1297, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.7882562277580071, | |
| "grad_norm": 1.4690390038104089, | |
| "learning_rate": 6.63253816335069e-06, | |
| "loss": 0.1014, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.7900355871886121, | |
| "grad_norm": 1.6731926564114215, | |
| "learning_rate": 6.619322656537552e-06, | |
| "loss": 0.1339, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.791814946619217, | |
| "grad_norm": 1.7492778854960134, | |
| "learning_rate": 6.606094499441671e-06, | |
| "loss": 0.1286, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.7935943060498221, | |
| "grad_norm": 1.8684021925137513, | |
| "learning_rate": 6.592853795402502e-06, | |
| "loss": 0.1297, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.7953736654804271, | |
| "grad_norm": 2.0094291684456107, | |
| "learning_rate": 6.579600647857525e-06, | |
| "loss": 0.1339, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.797153024911032, | |
| "grad_norm": 2.1491520913379443, | |
| "learning_rate": 6.566335160341425e-06, | |
| "loss": 0.1421, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.798932384341637, | |
| "grad_norm": 1.5242968614554042, | |
| "learning_rate": 6.553057436485289e-06, | |
| "loss": 0.1145, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.800711743772242, | |
| "grad_norm": 1.8086736236273908, | |
| "learning_rate": 6.539767580015799e-06, | |
| "loss": 0.1699, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.802491103202847, | |
| "grad_norm": 1.897446523525275, | |
| "learning_rate": 6.52646569475441e-06, | |
| "loss": 0.1147, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8042704626334519, | |
| "grad_norm": 1.9964778745258163, | |
| "learning_rate": 6.513151884616556e-06, | |
| "loss": 0.1737, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.806049822064057, | |
| "grad_norm": 2.2307496861100122, | |
| "learning_rate": 6.499826253610823e-06, | |
| "loss": 0.139, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.8078291814946619, | |
| "grad_norm": 1.9657630711570364, | |
| "learning_rate": 6.486488905838143e-06, | |
| "loss": 0.1292, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.8096085409252669, | |
| "grad_norm": 1.5828740024604897, | |
| "learning_rate": 6.473139945490984e-06, | |
| "loss": 0.0885, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8113879003558719, | |
| "grad_norm": 2.0124219401931076, | |
| "learning_rate": 6.459779476852528e-06, | |
| "loss": 0.139, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.8131672597864769, | |
| "grad_norm": 1.7520306057433719, | |
| "learning_rate": 6.446407604295863e-06, | |
| "loss": 0.1222, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.8149466192170819, | |
| "grad_norm": 1.6425374064606146, | |
| "learning_rate": 6.433024432283169e-06, | |
| "loss": 0.1102, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.8167259786476868, | |
| "grad_norm": 1.8216450355659084, | |
| "learning_rate": 6.41963006536489e-06, | |
| "loss": 0.1265, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.8185053380782918, | |
| "grad_norm": 1.8071396471172743, | |
| "learning_rate": 6.4062246081789316e-06, | |
| "loss": 0.1198, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8202846975088968, | |
| "grad_norm": 1.587261685008709, | |
| "learning_rate": 6.392808165449836e-06, | |
| "loss": 0.0921, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.8220640569395018, | |
| "grad_norm": 1.8199731253981306, | |
| "learning_rate": 6.379380841987965e-06, | |
| "loss": 0.1389, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.8238434163701067, | |
| "grad_norm": 1.7735946726314031, | |
| "learning_rate": 6.365942742688684e-06, | |
| "loss": 0.1052, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.8256227758007118, | |
| "grad_norm": 2.1256334716644023, | |
| "learning_rate": 6.352493972531535e-06, | |
| "loss": 0.1422, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.8274021352313167, | |
| "grad_norm": 2.349989054143199, | |
| "learning_rate": 6.339034636579425e-06, | |
| "loss": 0.1628, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.8291814946619217, | |
| "grad_norm": 1.7494465771888674, | |
| "learning_rate": 6.325564839977802e-06, | |
| "loss": 0.1025, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.8309608540925267, | |
| "grad_norm": 1.5734154602497274, | |
| "learning_rate": 6.312084687953835e-06, | |
| "loss": 0.1033, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.8327402135231317, | |
| "grad_norm": 2.0202867863557517, | |
| "learning_rate": 6.298594285815585e-06, | |
| "loss": 0.133, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.8345195729537367, | |
| "grad_norm": 1.9315641072085195, | |
| "learning_rate": 6.2850937389511936e-06, | |
| "loss": 0.1422, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.8362989323843416, | |
| "grad_norm": 1.9169847918305984, | |
| "learning_rate": 6.271583152828049e-06, | |
| "loss": 0.1228, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8380782918149466, | |
| "grad_norm": 1.9633986099592902, | |
| "learning_rate": 6.258062632991972e-06, | |
| "loss": 0.113, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.8398576512455516, | |
| "grad_norm": 1.544011691783178, | |
| "learning_rate": 6.244532285066382e-06, | |
| "loss": 0.0937, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.8416370106761566, | |
| "grad_norm": 1.3671751767869913, | |
| "learning_rate": 6.2309922147514775e-06, | |
| "loss": 0.1026, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.8434163701067615, | |
| "grad_norm": 1.964688483955758, | |
| "learning_rate": 6.2174425278234115e-06, | |
| "loss": 0.202, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.8451957295373665, | |
| "grad_norm": 1.793389760945486, | |
| "learning_rate": 6.20388333013346e-06, | |
| "loss": 0.1093, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.8469750889679716, | |
| "grad_norm": 1.8510148387980616, | |
| "learning_rate": 6.190314727607196e-06, | |
| "loss": 0.1414, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.8487544483985765, | |
| "grad_norm": 1.8136361826464784, | |
| "learning_rate": 6.176736826243671e-06, | |
| "loss": 0.1268, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.8505338078291815, | |
| "grad_norm": 1.9408467264668174, | |
| "learning_rate": 6.163149732114571e-06, | |
| "loss": 0.1555, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.8523131672597865, | |
| "grad_norm": 2.1590934070859817, | |
| "learning_rate": 6.149553551363404e-06, | |
| "loss": 0.1581, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.8540925266903915, | |
| "grad_norm": 1.7794767472045578, | |
| "learning_rate": 6.1359483902046605e-06, | |
| "loss": 0.1289, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8558718861209964, | |
| "grad_norm": 1.5790841082886007, | |
| "learning_rate": 6.122334354922984e-06, | |
| "loss": 0.1075, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.8576512455516014, | |
| "grad_norm": 1.5192547846542448, | |
| "learning_rate": 6.108711551872347e-06, | |
| "loss": 0.1193, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.8594306049822064, | |
| "grad_norm": 2.065710210287708, | |
| "learning_rate": 6.095080087475218e-06, | |
| "loss": 0.1235, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.8612099644128114, | |
| "grad_norm": 1.8659984411908528, | |
| "learning_rate": 6.0814400682217236e-06, | |
| "loss": 0.1339, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.8629893238434164, | |
| "grad_norm": 2.1247580623026074, | |
| "learning_rate": 6.067791600668823e-06, | |
| "loss": 0.1036, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.8647686832740213, | |
| "grad_norm": 1.8002971823510578, | |
| "learning_rate": 6.054134791439479e-06, | |
| "loss": 0.1162, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.8665480427046264, | |
| "grad_norm": 1.9218462333574762, | |
| "learning_rate": 6.040469747221815e-06, | |
| "loss": 0.1283, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.8683274021352313, | |
| "grad_norm": 1.9344519373131381, | |
| "learning_rate": 6.026796574768288e-06, | |
| "loss": 0.1111, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.8701067615658363, | |
| "grad_norm": 1.6850369570713912, | |
| "learning_rate": 6.013115380894854e-06, | |
| "loss": 0.1096, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.8718861209964412, | |
| "grad_norm": 1.3403244072720497, | |
| "learning_rate": 5.999426272480133e-06, | |
| "loss": 0.0915, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8736654804270463, | |
| "grad_norm": 1.931107221060752, | |
| "learning_rate": 5.985729356464575e-06, | |
| "loss": 0.1381, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.8754448398576512, | |
| "grad_norm": 1.3531376310672671, | |
| "learning_rate": 5.972024739849622e-06, | |
| "loss": 0.0924, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.8772241992882562, | |
| "grad_norm": 1.4952652847106684, | |
| "learning_rate": 5.958312529696874e-06, | |
| "loss": 0.0992, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.8790035587188612, | |
| "grad_norm": 1.8736450376031155, | |
| "learning_rate": 5.944592833127253e-06, | |
| "loss": 0.1501, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.8807829181494662, | |
| "grad_norm": 1.6065912150333321, | |
| "learning_rate": 5.9308657573201645e-06, | |
| "loss": 0.1303, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.8825622775800712, | |
| "grad_norm": 1.9698435717769076, | |
| "learning_rate": 5.917131409512663e-06, | |
| "loss": 0.1432, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.8843416370106761, | |
| "grad_norm": 1.7909064939454904, | |
| "learning_rate": 5.903389896998611e-06, | |
| "loss": 0.1151, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.8861209964412812, | |
| "grad_norm": 2.0208481295236425, | |
| "learning_rate": 5.889641327127843e-06, | |
| "loss": 0.142, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.8879003558718861, | |
| "grad_norm": 1.6014140709610698, | |
| "learning_rate": 5.875885807305326e-06, | |
| "loss": 0.1146, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.8896797153024911, | |
| "grad_norm": 1.5491263033771028, | |
| "learning_rate": 5.862123444990319e-06, | |
| "loss": 0.103, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.891459074733096, | |
| "grad_norm": 1.5779584955402308, | |
| "learning_rate": 5.848354347695537e-06, | |
| "loss": 0.1455, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.8932384341637011, | |
| "grad_norm": 2.024656879331989, | |
| "learning_rate": 5.83457862298631e-06, | |
| "loss": 0.1625, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.895017793594306, | |
| "grad_norm": 2.0655908536614547, | |
| "learning_rate": 5.8207963784797396e-06, | |
| "loss": 0.1338, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.896797153024911, | |
| "grad_norm": 1.897107770753346, | |
| "learning_rate": 5.807007721843862e-06, | |
| "loss": 0.1427, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.8985765124555161, | |
| "grad_norm": 1.8140619397780933, | |
| "learning_rate": 5.793212760796804e-06, | |
| "loss": 0.1221, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.900355871886121, | |
| "grad_norm": 1.862615782913275, | |
| "learning_rate": 5.779411603105947e-06, | |
| "loss": 0.1271, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.902135231316726, | |
| "grad_norm": 2.0068902895955145, | |
| "learning_rate": 5.765604356587076e-06, | |
| "loss": 0.1327, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.9039145907473309, | |
| "grad_norm": 1.8545626603171097, | |
| "learning_rate": 5.751791129103545e-06, | |
| "loss": 0.1272, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.905693950177936, | |
| "grad_norm": 1.9644698218131842, | |
| "learning_rate": 5.737972028565431e-06, | |
| "loss": 0.1549, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.9074733096085409, | |
| "grad_norm": 1.9170250532825894, | |
| "learning_rate": 5.7241471629286934e-06, | |
| "loss": 0.1351, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9092526690391459, | |
| "grad_norm": 2.3418282100368777, | |
| "learning_rate": 5.7103166401943276e-06, | |
| "loss": 0.1494, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.9110320284697508, | |
| "grad_norm": 1.9680860973541718, | |
| "learning_rate": 5.696480568407523e-06, | |
| "loss": 0.138, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.9128113879003559, | |
| "grad_norm": 2.2377315061579455, | |
| "learning_rate": 5.682639055656817e-06, | |
| "loss": 0.1881, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.9145907473309609, | |
| "grad_norm": 2.149426355121831, | |
| "learning_rate": 5.668792210073255e-06, | |
| "loss": 0.1478, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.9163701067615658, | |
| "grad_norm": 2.008100193861197, | |
| "learning_rate": 5.654940139829544e-06, | |
| "loss": 0.1384, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9181494661921709, | |
| "grad_norm": 1.9899017641586887, | |
| "learning_rate": 5.641082953139201e-06, | |
| "loss": 0.1262, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.9199288256227758, | |
| "grad_norm": 1.8456682064322725, | |
| "learning_rate": 5.6272207582557195e-06, | |
| "loss": 0.141, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.9217081850533808, | |
| "grad_norm": 1.8161361879527038, | |
| "learning_rate": 5.61335366347171e-06, | |
| "loss": 0.1162, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.9234875444839857, | |
| "grad_norm": 1.8648811434300516, | |
| "learning_rate": 5.599481777118071e-06, | |
| "loss": 0.1179, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.9252669039145908, | |
| "grad_norm": 1.8221935180859308, | |
| "learning_rate": 5.585605207563124e-06, | |
| "loss": 0.1106, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9270462633451957, | |
| "grad_norm": 1.88519467317523, | |
| "learning_rate": 5.571724063211782e-06, | |
| "loss": 0.1349, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.9288256227758007, | |
| "grad_norm": 1.5988518529631495, | |
| "learning_rate": 5.557838452504692e-06, | |
| "loss": 0.0924, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.9306049822064056, | |
| "grad_norm": 2.499901743444125, | |
| "learning_rate": 5.5439484839173996e-06, | |
| "loss": 0.1525, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.9323843416370107, | |
| "grad_norm": 1.5889241083232466, | |
| "learning_rate": 5.530054265959486e-06, | |
| "loss": 0.1094, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.9341637010676157, | |
| "grad_norm": 2.2743815077379965, | |
| "learning_rate": 5.516155907173735e-06, | |
| "loss": 0.177, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9359430604982206, | |
| "grad_norm": 1.6028126813151364, | |
| "learning_rate": 5.5022535161352764e-06, | |
| "loss": 0.1305, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.9377224199288257, | |
| "grad_norm": 2.0619042742903892, | |
| "learning_rate": 5.488347201450741e-06, | |
| "loss": 0.1261, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.9395017793594306, | |
| "grad_norm": 1.8109585809901696, | |
| "learning_rate": 5.47443707175741e-06, | |
| "loss": 0.1266, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.9412811387900356, | |
| "grad_norm": 1.6694460949039043, | |
| "learning_rate": 5.46052323572237e-06, | |
| "loss": 0.1076, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.9430604982206405, | |
| "grad_norm": 1.8207248504643856, | |
| "learning_rate": 5.446605802041662e-06, | |
| "loss": 0.1395, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9448398576512456, | |
| "grad_norm": 1.638010398392601, | |
| "learning_rate": 5.432684879439428e-06, | |
| "loss": 0.1102, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.9466192170818505, | |
| "grad_norm": 1.8517353169199433, | |
| "learning_rate": 5.418760576667071e-06, | |
| "loss": 0.0976, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.9483985765124555, | |
| "grad_norm": 1.8584270347802287, | |
| "learning_rate": 5.404833002502398e-06, | |
| "loss": 0.136, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.9501779359430605, | |
| "grad_norm": 2.1826881413410333, | |
| "learning_rate": 5.39090226574877e-06, | |
| "loss": 0.1343, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.9519572953736655, | |
| "grad_norm": 2.1686115740018397, | |
| "learning_rate": 5.376968475234258e-06, | |
| "loss": 0.1414, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.9537366548042705, | |
| "grad_norm": 1.2567577661315184, | |
| "learning_rate": 5.363031739810787e-06, | |
| "loss": 0.0886, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.9555160142348754, | |
| "grad_norm": 1.5653978311138095, | |
| "learning_rate": 5.349092168353291e-06, | |
| "loss": 0.1114, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.9572953736654805, | |
| "grad_norm": 1.693470690655894, | |
| "learning_rate": 5.335149869758855e-06, | |
| "loss": 0.1151, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.9590747330960854, | |
| "grad_norm": 1.657767374446765, | |
| "learning_rate": 5.32120495294587e-06, | |
| "loss": 0.1278, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.9608540925266904, | |
| "grad_norm": 1.5085226078036016, | |
| "learning_rate": 5.3072575268531835e-06, | |
| "loss": 0.1069, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9626334519572953, | |
| "grad_norm": 2.007314818730687, | |
| "learning_rate": 5.293307700439242e-06, | |
| "loss": 0.1453, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.9644128113879004, | |
| "grad_norm": 2.2141218412006616, | |
| "learning_rate": 5.2793555826812456e-06, | |
| "loss": 0.1299, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.9661921708185054, | |
| "grad_norm": 1.4462746575417975, | |
| "learning_rate": 5.265401282574294e-06, | |
| "loss": 0.0836, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.9679715302491103, | |
| "grad_norm": 1.8420941410743605, | |
| "learning_rate": 5.2514449091305375e-06, | |
| "loss": 0.1201, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.9697508896797153, | |
| "grad_norm": 1.7117205243442641, | |
| "learning_rate": 5.237486571378317e-06, | |
| "loss": 0.1031, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.9715302491103203, | |
| "grad_norm": 1.5672473235367783, | |
| "learning_rate": 5.22352637836133e-06, | |
| "loss": 0.116, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.9733096085409253, | |
| "grad_norm": 1.4999906557566207, | |
| "learning_rate": 5.209564439137755e-06, | |
| "loss": 0.0967, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.9750889679715302, | |
| "grad_norm": 2.0274070851590844, | |
| "learning_rate": 5.195600862779421e-06, | |
| "loss": 0.1381, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.9768683274021353, | |
| "grad_norm": 2.0541241199712856, | |
| "learning_rate": 5.181635758370942e-06, | |
| "loss": 0.1507, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.9786476868327402, | |
| "grad_norm": 1.845466287692656, | |
| "learning_rate": 5.167669235008871e-06, | |
| "loss": 0.1455, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9804270462633452, | |
| "grad_norm": 1.8830229823267266, | |
| "learning_rate": 5.153701401800845e-06, | |
| "loss": 0.1415, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.9822064056939501, | |
| "grad_norm": 2.0250847801432084, | |
| "learning_rate": 5.139732367864736e-06, | |
| "loss": 0.1316, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.9839857651245552, | |
| "grad_norm": 1.6346619638641466, | |
| "learning_rate": 5.1257622423277934e-06, | |
| "loss": 0.1039, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.9857651245551602, | |
| "grad_norm": 1.8208190403030007, | |
| "learning_rate": 5.111791134325793e-06, | |
| "loss": 0.1359, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.9875444839857651, | |
| "grad_norm": 1.7342489526060136, | |
| "learning_rate": 5.097819153002192e-06, | |
| "loss": 0.1294, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.9893238434163701, | |
| "grad_norm": 1.477894160456886, | |
| "learning_rate": 5.083846407507263e-06, | |
| "loss": 0.0897, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.9911032028469751, | |
| "grad_norm": 2.010061332837563, | |
| "learning_rate": 5.0698730069972535e-06, | |
| "loss": 0.1466, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.9928825622775801, | |
| "grad_norm": 1.561393228267325, | |
| "learning_rate": 5.055899060633524e-06, | |
| "loss": 0.1083, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.994661921708185, | |
| "grad_norm": 1.6360898249569782, | |
| "learning_rate": 5.041924677581702e-06, | |
| "loss": 0.1213, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.99644128113879, | |
| "grad_norm": 1.682319578302086, | |
| "learning_rate": 5.0279499670108245e-06, | |
| "loss": 0.1187, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.998220640569395, | |
| "grad_norm": 1.522252041891195, | |
| "learning_rate": 5.013975038092491e-06, | |
| "loss": 0.0979, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.5739077053411878, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0762, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.001779359430605, | |
| "grad_norm": 1.353305669111731, | |
| "learning_rate": 4.98602496190751e-06, | |
| "loss": 0.0743, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.00355871886121, | |
| "grad_norm": 0.9771132127089612, | |
| "learning_rate": 4.9720500329891755e-06, | |
| "loss": 0.0577, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.0053380782918149, | |
| "grad_norm": 1.7729222484901388, | |
| "learning_rate": 4.9580753224183005e-06, | |
| "loss": 0.0818, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.00711743772242, | |
| "grad_norm": 1.1922552423414674, | |
| "learning_rate": 4.944100939366478e-06, | |
| "loss": 0.0602, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.008896797153025, | |
| "grad_norm": 1.506665684383127, | |
| "learning_rate": 4.930126993002748e-06, | |
| "loss": 0.0787, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.01067615658363, | |
| "grad_norm": 0.9236703755605269, | |
| "learning_rate": 4.9161535924927375e-06, | |
| "loss": 0.0461, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.0124555160142348, | |
| "grad_norm": 1.091874665301268, | |
| "learning_rate": 4.90218084699781e-06, | |
| "loss": 0.045, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.0142348754448398, | |
| "grad_norm": 1.4995026492683319, | |
| "learning_rate": 4.888208865674208e-06, | |
| "loss": 0.0666, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.0160142348754448, | |
| "grad_norm": 1.3272975250731802, | |
| "learning_rate": 4.874237757672209e-06, | |
| "loss": 0.0671, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.0177935943060499, | |
| "grad_norm": 1.4194886009096468, | |
| "learning_rate": 4.8602676321352646e-06, | |
| "loss": 0.0677, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.019572953736655, | |
| "grad_norm": 1.3177112726308065, | |
| "learning_rate": 4.846298598199155e-06, | |
| "loss": 0.0461, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.0213523131672597, | |
| "grad_norm": 1.191373489417224, | |
| "learning_rate": 4.832330764991131e-06, | |
| "loss": 0.0465, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.0231316725978647, | |
| "grad_norm": 1.6896241894367776, | |
| "learning_rate": 4.81836424162906e-06, | |
| "loss": 0.0601, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.0249110320284698, | |
| "grad_norm": 1.4206512850264201, | |
| "learning_rate": 4.80439913722058e-06, | |
| "loss": 0.0529, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.0266903914590748, | |
| "grad_norm": 1.6627160559626963, | |
| "learning_rate": 4.790435560862247e-06, | |
| "loss": 0.063, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.0284697508896796, | |
| "grad_norm": 1.5467567210729698, | |
| "learning_rate": 4.776473621638673e-06, | |
| "loss": 0.0406, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.0302491103202847, | |
| "grad_norm": 1.454971428638148, | |
| "learning_rate": 4.762513428621684e-06, | |
| "loss": 0.0521, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.0320284697508897, | |
| "grad_norm": 1.694597382683876, | |
| "learning_rate": 4.748555090869464e-06, | |
| "loss": 0.0531, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.0338078291814947, | |
| "grad_norm": 1.6481918057013685, | |
| "learning_rate": 4.734598717425706e-06, | |
| "loss": 0.0577, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.0355871886120998, | |
| "grad_norm": 1.5020755779197987, | |
| "learning_rate": 4.720644417318755e-06, | |
| "loss": 0.0478, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.0373665480427046, | |
| "grad_norm": 2.1611885166305465, | |
| "learning_rate": 4.70669229956076e-06, | |
| "loss": 0.0511, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.0391459074733096, | |
| "grad_norm": 1.746488095064038, | |
| "learning_rate": 4.692742473146818e-06, | |
| "loss": 0.0526, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.0409252669039146, | |
| "grad_norm": 2.1842420874625965, | |
| "learning_rate": 4.678795047054131e-06, | |
| "loss": 0.0702, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.0427046263345197, | |
| "grad_norm": 1.549401712307886, | |
| "learning_rate": 4.664850130241146e-06, | |
| "loss": 0.0377, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.0444839857651245, | |
| "grad_norm": 1.4520333627527706, | |
| "learning_rate": 4.650907831646711e-06, | |
| "loss": 0.0338, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.0462633451957295, | |
| "grad_norm": 1.673567804182864, | |
| "learning_rate": 4.636968260189214e-06, | |
| "loss": 0.0602, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.0480427046263345, | |
| "grad_norm": 1.8029858503170495, | |
| "learning_rate": 4.623031524765744e-06, | |
| "loss": 0.0446, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.0498220640569396, | |
| "grad_norm": 1.8263418979021693, | |
| "learning_rate": 4.609097734251231e-06, | |
| "loss": 0.0623, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.0516014234875444, | |
| "grad_norm": 2.027208397144687, | |
| "learning_rate": 4.595166997497605e-06, | |
| "loss": 0.0408, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.0533807829181494, | |
| "grad_norm": 1.8688709469112337, | |
| "learning_rate": 4.58123942333293e-06, | |
| "loss": 0.0627, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.0551601423487544, | |
| "grad_norm": 1.5932237670597085, | |
| "learning_rate": 4.567315120560573e-06, | |
| "loss": 0.0523, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.0569395017793595, | |
| "grad_norm": 1.834307159468834, | |
| "learning_rate": 4.553394197958339e-06, | |
| "loss": 0.0588, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.0587188612099645, | |
| "grad_norm": 2.074271474803056, | |
| "learning_rate": 4.539476764277631e-06, | |
| "loss": 0.0574, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.0604982206405693, | |
| "grad_norm": 1.3641185794855366, | |
| "learning_rate": 4.525562928242592e-06, | |
| "loss": 0.0462, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.0622775800711743, | |
| "grad_norm": 2.06801481914098, | |
| "learning_rate": 4.511652798549261e-06, | |
| "loss": 0.0555, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.0640569395017794, | |
| "grad_norm": 1.548455930321502, | |
| "learning_rate": 4.497746483864725e-06, | |
| "loss": 0.0456, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.0658362989323844, | |
| "grad_norm": 1.5777176853351251, | |
| "learning_rate": 4.483844092826267e-06, | |
| "loss": 0.0518, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.0676156583629894, | |
| "grad_norm": 1.4825371391235314, | |
| "learning_rate": 4.469945734040516e-06, | |
| "loss": 0.044, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0676156583629894, | |
| "eval_loss": 0.15300770103931427, | |
| "eval_runtime": 1.8893, | |
| "eval_samples_per_second": 24.348, | |
| "eval_steps_per_second": 6.352, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0693950177935942, | |
| "grad_norm": 1.5022650785366567, | |
| "learning_rate": 4.456051516082603e-06, | |
| "loss": 0.0432, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.0711743772241993, | |
| "grad_norm": 1.751718962151422, | |
| "learning_rate": 4.442161547495309e-06, | |
| "loss": 0.0476, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.0729537366548043, | |
| "grad_norm": 1.7916838347261523, | |
| "learning_rate": 4.42827593678822e-06, | |
| "loss": 0.0427, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.0747330960854093, | |
| "grad_norm": 1.9605205102269285, | |
| "learning_rate": 4.414394792436877e-06, | |
| "loss": 0.0561, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.0765124555160142, | |
| "grad_norm": 1.7311117633763786, | |
| "learning_rate": 4.400518222881931e-06, | |
| "loss": 0.0604, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.0782918149466192, | |
| "grad_norm": 1.514679280035021, | |
| "learning_rate": 4.386646336528291e-06, | |
| "loss": 0.0501, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.0800711743772242, | |
| "grad_norm": 1.5877766075224946, | |
| "learning_rate": 4.372779241744282e-06, | |
| "loss": 0.0487, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.0818505338078293, | |
| "grad_norm": 1.3547349784532614, | |
| "learning_rate": 4.358917046860799e-06, | |
| "loss": 0.0344, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.083629893238434, | |
| "grad_norm": 1.4709838816765004, | |
| "learning_rate": 4.345059860170458e-06, | |
| "loss": 0.05, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.085409252669039, | |
| "grad_norm": 1.4062342533082566, | |
| "learning_rate": 4.331207789926746e-06, | |
| "loss": 0.0382, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.0871886120996441, | |
| "grad_norm": 1.376615369689365, | |
| "learning_rate": 4.317360944343184e-06, | |
| "loss": 0.0444, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.0889679715302492, | |
| "grad_norm": 1.2530903941266593, | |
| "learning_rate": 4.303519431592479e-06, | |
| "loss": 0.0326, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.0907473309608542, | |
| "grad_norm": 1.966665300492479, | |
| "learning_rate": 4.289683359805673e-06, | |
| "loss": 0.0555, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.092526690391459, | |
| "grad_norm": 1.7776975232760048, | |
| "learning_rate": 4.275852837071309e-06, | |
| "loss": 0.0521, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.094306049822064, | |
| "grad_norm": 1.3253857365992163, | |
| "learning_rate": 4.26202797143457e-06, | |
| "loss": 0.0417, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.096085409252669, | |
| "grad_norm": 1.608646246279028, | |
| "learning_rate": 4.248208870896456e-06, | |
| "loss": 0.0529, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.097864768683274, | |
| "grad_norm": 1.498852590365668, | |
| "learning_rate": 4.234395643412925e-06, | |
| "loss": 0.0508, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.099644128113879, | |
| "grad_norm": 1.554019647281601, | |
| "learning_rate": 4.220588396894055e-06, | |
| "loss": 0.0483, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.101423487544484, | |
| "grad_norm": 1.4033454122163562, | |
| "learning_rate": 4.2067872392031965e-06, | |
| "loss": 0.0449, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.103202846975089, | |
| "grad_norm": 2.216035737647856, | |
| "learning_rate": 4.192992278156141e-06, | |
| "loss": 0.0783, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.104982206405694, | |
| "grad_norm": 1.4344741984855178, | |
| "learning_rate": 4.179203621520262e-06, | |
| "loss": 0.0505, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.106761565836299, | |
| "grad_norm": 1.4539588129398393, | |
| "learning_rate": 4.165421377013691e-06, | |
| "loss": 0.0401, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.1085409252669038, | |
| "grad_norm": 1.6620274068540544, | |
| "learning_rate": 4.151645652304465e-06, | |
| "loss": 0.0411, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.1103202846975089, | |
| "grad_norm": 1.7790092582107746, | |
| "learning_rate": 4.137876555009684e-06, | |
| "loss": 0.0548, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.112099644128114, | |
| "grad_norm": 3.294165679530209, | |
| "learning_rate": 4.124114192694676e-06, | |
| "loss": 0.0863, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.113879003558719, | |
| "grad_norm": 1.8095446452236716, | |
| "learning_rate": 4.110358672872158e-06, | |
| "loss": 0.0622, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.1156583629893237, | |
| "grad_norm": 1.5950313163633831, | |
| "learning_rate": 4.0966101030013915e-06, | |
| "loss": 0.0563, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.1174377224199288, | |
| "grad_norm": 1.5832352283341409, | |
| "learning_rate": 4.082868590487339e-06, | |
| "loss": 0.0563, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.1192170818505338, | |
| "grad_norm": 1.4508858125121225, | |
| "learning_rate": 4.069134242679837e-06, | |
| "loss": 0.0478, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.1209964412811388, | |
| "grad_norm": 1.90992314964518, | |
| "learning_rate": 4.055407166872748e-06, | |
| "loss": 0.0499, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.1227758007117439, | |
| "grad_norm": 1.6111231192359678, | |
| "learning_rate": 4.041687470303127e-06, | |
| "loss": 0.0614, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.1245551601423487, | |
| "grad_norm": 1.195876755915428, | |
| "learning_rate": 4.02797526015038e-06, | |
| "loss": 0.0417, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.1263345195729537, | |
| "grad_norm": 1.564074489891255, | |
| "learning_rate": 4.014270643535427e-06, | |
| "loss": 0.0488, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.1281138790035588, | |
| "grad_norm": 1.6805727566522493, | |
| "learning_rate": 4.000573727519868e-06, | |
| "loss": 0.0468, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.1298932384341638, | |
| "grad_norm": 1.6359854851157571, | |
| "learning_rate": 3.9868846191051465e-06, | |
| "loss": 0.0513, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.1316725978647686, | |
| "grad_norm": 1.64390851673991, | |
| "learning_rate": 3.973203425231715e-06, | |
| "loss": 0.0596, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.1334519572953736, | |
| "grad_norm": 1.7432998478143718, | |
| "learning_rate": 3.959530252778187e-06, | |
| "loss": 0.0634, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.1352313167259787, | |
| "grad_norm": 1.7972165464741237, | |
| "learning_rate": 3.945865208560522e-06, | |
| "loss": 0.0516, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.1370106761565837, | |
| "grad_norm": 1.6937542949760003, | |
| "learning_rate": 3.932208399331177e-06, | |
| "loss": 0.0621, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.1387900355871885, | |
| "grad_norm": 1.7963169569269741, | |
| "learning_rate": 3.918559931778277e-06, | |
| "loss": 0.0561, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.1405693950177935, | |
| "grad_norm": 1.893183406017728, | |
| "learning_rate": 3.904919912524784e-06, | |
| "loss": 0.0453, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.1423487544483986, | |
| "grad_norm": 1.5798109776387077, | |
| "learning_rate": 3.891288448127654e-06, | |
| "loss": 0.0527, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.1441281138790036, | |
| "grad_norm": 1.7950833489968363, | |
| "learning_rate": 3.877665645077017e-06, | |
| "loss": 0.0574, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.1459074733096086, | |
| "grad_norm": 1.4260088179722528, | |
| "learning_rate": 3.86405160979534e-06, | |
| "loss": 0.0492, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.1476868327402134, | |
| "grad_norm": 1.5553889730658728, | |
| "learning_rate": 3.850446448636597e-06, | |
| "loss": 0.0525, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.1494661921708185, | |
| "grad_norm": 1.2949647947678804, | |
| "learning_rate": 3.8368502678854296e-06, | |
| "loss": 0.0377, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.1512455516014235, | |
| "grad_norm": 1.2232774191586702, | |
| "learning_rate": 3.8232631737563306e-06, | |
| "loss": 0.0476, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.1530249110320285, | |
| "grad_norm": 2.1538248000464955, | |
| "learning_rate": 3.809685272392804e-06, | |
| "loss": 0.0581, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.1548042704626336, | |
| "grad_norm": 2.0019780165715497, | |
| "learning_rate": 3.796116669866543e-06, | |
| "loss": 0.0723, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.1565836298932384, | |
| "grad_norm": 1.5894388284049505, | |
| "learning_rate": 3.78255747217659e-06, | |
| "loss": 0.0467, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.1583629893238434, | |
| "grad_norm": 1.8009735436404846, | |
| "learning_rate": 3.769007785248523e-06, | |
| "loss": 0.0545, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.1601423487544484, | |
| "grad_norm": 1.5749947756629827, | |
| "learning_rate": 3.7554677149336186e-06, | |
| "loss": 0.0459, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.1619217081850535, | |
| "grad_norm": 1.728667661831068, | |
| "learning_rate": 3.7419373670080284e-06, | |
| "loss": 0.0597, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.1637010676156583, | |
| "grad_norm": 1.696381828983317, | |
| "learning_rate": 3.7284168471719527e-06, | |
| "loss": 0.0496, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.1654804270462633, | |
| "grad_norm": 1.2695639362981899, | |
| "learning_rate": 3.7149062610488085e-06, | |
| "loss": 0.0497, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.1672597864768683, | |
| "grad_norm": 1.6215659340990989, | |
| "learning_rate": 3.701405714184416e-06, | |
| "loss": 0.0441, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.1690391459074734, | |
| "grad_norm": 1.4452519796604475, | |
| "learning_rate": 3.687915312046166e-06, | |
| "loss": 0.0495, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.1708185053380782, | |
| "grad_norm": 1.8573603137973824, | |
| "learning_rate": 3.6744351600221994e-06, | |
| "loss": 0.0502, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.1725978647686832, | |
| "grad_norm": 1.845051179319759, | |
| "learning_rate": 3.6609653634205773e-06, | |
| "loss": 0.0763, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.1743772241992882, | |
| "grad_norm": 1.4299135110080567, | |
| "learning_rate": 3.647506027468467e-06, | |
| "loss": 0.0332, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.1761565836298933, | |
| "grad_norm": 1.611081621365616, | |
| "learning_rate": 3.6340572573113176e-06, | |
| "loss": 0.0511, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.1779359430604983, | |
| "grad_norm": 1.2934608746173604, | |
| "learning_rate": 3.6206191580120346e-06, | |
| "loss": 0.0418, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.1797153024911031, | |
| "grad_norm": 2.44640726451789, | |
| "learning_rate": 3.6071918345501655e-06, | |
| "loss": 0.0691, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.1814946619217082, | |
| "grad_norm": 1.7188001038795346, | |
| "learning_rate": 3.5937753918210705e-06, | |
| "loss": 0.0559, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.1832740213523132, | |
| "grad_norm": 1.5395475545311244, | |
| "learning_rate": 3.5803699346351117e-06, | |
| "loss": 0.044, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.1850533807829182, | |
| "grad_norm": 1.2562045401570148, | |
| "learning_rate": 3.566975567716833e-06, | |
| "loss": 0.042, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.1868327402135233, | |
| "grad_norm": 1.4815867644230685, | |
| "learning_rate": 3.5535923957041374e-06, | |
| "loss": 0.0408, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.188612099644128, | |
| "grad_norm": 1.3254374693352022, | |
| "learning_rate": 3.540220523147474e-06, | |
| "loss": 0.0483, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.190391459074733, | |
| "grad_norm": 1.514616367629608, | |
| "learning_rate": 3.5268600545090183e-06, | |
| "loss": 0.0447, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.1921708185053381, | |
| "grad_norm": 1.3190680523508571, | |
| "learning_rate": 3.513511094161858e-06, | |
| "loss": 0.0416, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.193950177935943, | |
| "grad_norm": 1.7575480551432359, | |
| "learning_rate": 3.5001737463891793e-06, | |
| "loss": 0.0574, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.195729537366548, | |
| "grad_norm": 1.9651893418619537, | |
| "learning_rate": 3.4868481153834454e-06, | |
| "loss": 0.0617, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.197508896797153, | |
| "grad_norm": 1.4195721792388443, | |
| "learning_rate": 3.4735343052455905e-06, | |
| "loss": 0.0361, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.199288256227758, | |
| "grad_norm": 1.896892315030391, | |
| "learning_rate": 3.4602324199842026e-06, | |
| "loss": 0.0547, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.201067615658363, | |
| "grad_norm": 1.9438845971613863, | |
| "learning_rate": 3.446942563514711e-06, | |
| "loss": 0.0638, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.2028469750889679, | |
| "grad_norm": 1.5695181560606608, | |
| "learning_rate": 3.4336648396585777e-06, | |
| "loss": 0.0422, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.204626334519573, | |
| "grad_norm": 1.4637566899325707, | |
| "learning_rate": 3.4203993521424774e-06, | |
| "loss": 0.0485, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.206405693950178, | |
| "grad_norm": 1.2159962956603534, | |
| "learning_rate": 3.407146204597499e-06, | |
| "loss": 0.0326, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.208185053380783, | |
| "grad_norm": 1.3240485056169564, | |
| "learning_rate": 3.3939055005583305e-06, | |
| "loss": 0.0403, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.209964412811388, | |
| "grad_norm": 1.6147845007160504, | |
| "learning_rate": 3.3806773434624475e-06, | |
| "loss": 0.0508, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.2117437722419928, | |
| "grad_norm": 1.7710821612722727, | |
| "learning_rate": 3.3674618366493117e-06, | |
| "loss": 0.0606, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.2135231316725978, | |
| "grad_norm": 1.580296546142346, | |
| "learning_rate": 3.3542590833595533e-06, | |
| "loss": 0.0595, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.2153024911032029, | |
| "grad_norm": 1.6147197629529721, | |
| "learning_rate": 3.341069186734176e-06, | |
| "loss": 0.0504, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.217081850533808, | |
| "grad_norm": 1.3994027119598413, | |
| "learning_rate": 3.3278922498137455e-06, | |
| "loss": 0.0425, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.2188612099644127, | |
| "grad_norm": 1.6087729721334214, | |
| "learning_rate": 3.314728375537587e-06, | |
| "loss": 0.0496, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.2206405693950177, | |
| "grad_norm": 1.5070051694347988, | |
| "learning_rate": 3.3015776667429724e-06, | |
| "loss": 0.0659, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.2224199288256228, | |
| "grad_norm": 1.6418262653771072, | |
| "learning_rate": 3.2884402261643296e-06, | |
| "loss": 0.0419, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.2241992882562278, | |
| "grad_norm": 1.758600658516154, | |
| "learning_rate": 3.2753161564324344e-06, | |
| "loss": 0.0528, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.2259786476868326, | |
| "grad_norm": 1.919217455802356, | |
| "learning_rate": 3.262205560073605e-06, | |
| "loss": 0.0476, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.2277580071174377, | |
| "grad_norm": 1.1902688178138883, | |
| "learning_rate": 3.249108539508909e-06, | |
| "loss": 0.0309, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.2295373665480427, | |
| "grad_norm": 1.7886361974643792, | |
| "learning_rate": 3.2360251970533527e-06, | |
| "loss": 0.0673, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.2313167259786477, | |
| "grad_norm": 1.2908012752728788, | |
| "learning_rate": 3.2229556349150947e-06, | |
| "loss": 0.0419, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.2330960854092528, | |
| "grad_norm": 1.7251442608848604, | |
| "learning_rate": 3.2098999551946337e-06, | |
| "loss": 0.0571, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.2348754448398576, | |
| "grad_norm": 1.652831109922192, | |
| "learning_rate": 3.1968582598840234e-06, | |
| "loss": 0.0513, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.2366548042704626, | |
| "grad_norm": 1.68735981770456, | |
| "learning_rate": 3.183830650866068e-06, | |
| "loss": 0.0442, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.2384341637010676, | |
| "grad_norm": 1.6297274428543973, | |
| "learning_rate": 3.1708172299135266e-06, | |
| "loss": 0.0516, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.2402135231316727, | |
| "grad_norm": 2.058364125352861, | |
| "learning_rate": 3.1578180986883234e-06, | |
| "loss": 0.0627, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.2419928825622777, | |
| "grad_norm": 1.5427608115598344, | |
| "learning_rate": 3.1448333587407486e-06, | |
| "loss": 0.0451, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.2437722419928825, | |
| "grad_norm": 1.3500871996309878, | |
| "learning_rate": 3.131863111508667e-06, | |
| "loss": 0.046, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.2455516014234875, | |
| "grad_norm": 1.6563451474609625, | |
| "learning_rate": 3.118907458316722e-06, | |
| "loss": 0.046, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2473309608540926, | |
| "grad_norm": 1.4085506319458705, | |
| "learning_rate": 3.105966500375551e-06, | |
| "loss": 0.0341, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.2491103202846976, | |
| "grad_norm": 1.9173810033113514, | |
| "learning_rate": 3.0930403387809892e-06, | |
| "loss": 0.0601, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.2508896797153026, | |
| "grad_norm": 1.6597342870329217, | |
| "learning_rate": 3.080129074513285e-06, | |
| "loss": 0.0554, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.2526690391459074, | |
| "grad_norm": 2.0744117124963655, | |
| "learning_rate": 3.067232808436299e-06, | |
| "loss": 0.0547, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.2544483985765125, | |
| "grad_norm": 2.1879636267069595, | |
| "learning_rate": 3.0543516412967327e-06, | |
| "loss": 0.0581, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.2562277580071175, | |
| "grad_norm": 1.4189505243048706, | |
| "learning_rate": 3.041485673723331e-06, | |
| "loss": 0.0361, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.2580071174377223, | |
| "grad_norm": 1.5369331785753853, | |
| "learning_rate": 3.0286350062261017e-06, | |
| "loss": 0.046, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.2597864768683273, | |
| "grad_norm": 1.5822949045048327, | |
| "learning_rate": 3.0157997391955172e-06, | |
| "loss": 0.0517, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.2615658362989324, | |
| "grad_norm": 1.815927206897285, | |
| "learning_rate": 3.0029799729017518e-06, | |
| "loss": 0.0438, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.2633451957295374, | |
| "grad_norm": 1.2811984379238093, | |
| "learning_rate": 2.9901758074938797e-06, | |
| "loss": 0.0463, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.2651245551601424, | |
| "grad_norm": 1.760919593951551, | |
| "learning_rate": 2.977387342999103e-06, | |
| "loss": 0.0531, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.2669039145907472, | |
| "grad_norm": 1.8971619003456028, | |
| "learning_rate": 2.964614679321966e-06, | |
| "loss": 0.0708, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.2686832740213523, | |
| "grad_norm": 1.4697028898412405, | |
| "learning_rate": 2.951857916243574e-06, | |
| "loss": 0.0422, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.2704626334519573, | |
| "grad_norm": 1.8248835447625493, | |
| "learning_rate": 2.9391171534208185e-06, | |
| "loss": 0.0642, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.2722419928825623, | |
| "grad_norm": 1.904513312593701, | |
| "learning_rate": 2.9263924903855932e-06, | |
| "loss": 0.0567, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.2740213523131674, | |
| "grad_norm": 1.3600928772265222, | |
| "learning_rate": 2.9136840265440213e-06, | |
| "loss": 0.0403, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.2758007117437722, | |
| "grad_norm": 1.5135170300211649, | |
| "learning_rate": 2.9009918611756732e-06, | |
| "loss": 0.0474, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.2775800711743772, | |
| "grad_norm": 1.6611502256756587, | |
| "learning_rate": 2.8883160934327968e-06, | |
| "loss": 0.043, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.2793594306049823, | |
| "grad_norm": 2.042543430815399, | |
| "learning_rate": 2.8756568223395396e-06, | |
| "loss": 0.0616, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.281138790035587, | |
| "grad_norm": 1.5135168345472616, | |
| "learning_rate": 2.8630141467911777e-06, | |
| "loss": 0.0482, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.282918149466192, | |
| "grad_norm": 1.4809668197696582, | |
| "learning_rate": 2.8503881655533395e-06, | |
| "loss": 0.0486, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.2846975088967971, | |
| "grad_norm": 1.3897156561593598, | |
| "learning_rate": 2.837778977261235e-06, | |
| "loss": 0.0406, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.2864768683274022, | |
| "grad_norm": 1.927920030346448, | |
| "learning_rate": 2.8251866804188875e-06, | |
| "loss": 0.0456, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.2882562277580072, | |
| "grad_norm": 1.4906557561551745, | |
| "learning_rate": 2.812611373398365e-06, | |
| "loss": 0.0558, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.290035587188612, | |
| "grad_norm": 1.6355323195989593, | |
| "learning_rate": 2.8000531544390064e-06, | |
| "loss": 0.0438, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.291814946619217, | |
| "grad_norm": 1.7562587577622542, | |
| "learning_rate": 2.7875121216466595e-06, | |
| "loss": 0.06, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.293594306049822, | |
| "grad_norm": 1.6777571708779444, | |
| "learning_rate": 2.7749883729929105e-06, | |
| "loss": 0.0434, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.295373665480427, | |
| "grad_norm": 1.6355279469681712, | |
| "learning_rate": 2.762482006314324e-06, | |
| "loss": 0.0548, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.2971530249110321, | |
| "grad_norm": 1.9270186736270023, | |
| "learning_rate": 2.7499931193116692e-06, | |
| "loss": 0.0631, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.298932384341637, | |
| "grad_norm": 1.6864200228057962, | |
| "learning_rate": 2.737521809549167e-06, | |
| "loss": 0.0443, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.300711743772242, | |
| "grad_norm": 1.5660972690339905, | |
| "learning_rate": 2.725068174453722e-06, | |
| "loss": 0.0562, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.302491103202847, | |
| "grad_norm": 1.5336928863358907, | |
| "learning_rate": 2.712632311314165e-06, | |
| "loss": 0.0416, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.304270462633452, | |
| "grad_norm": 2.0410788318606814, | |
| "learning_rate": 2.7002143172804875e-06, | |
| "loss": 0.0638, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.306049822064057, | |
| "grad_norm": 2.10762851325803, | |
| "learning_rate": 2.6878142893630904e-06, | |
| "loss": 0.0581, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.3078291814946619, | |
| "grad_norm": 1.5034193254396262, | |
| "learning_rate": 2.6754323244320154e-06, | |
| "loss": 0.051, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.309608540925267, | |
| "grad_norm": 2.0287864679373593, | |
| "learning_rate": 2.6630685192161995e-06, | |
| "loss": 0.0582, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.311387900355872, | |
| "grad_norm": 1.798472771264233, | |
| "learning_rate": 2.650722970302714e-06, | |
| "loss": 0.0453, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.3131672597864767, | |
| "grad_norm": 2.112499963297033, | |
| "learning_rate": 2.638395774136009e-06, | |
| "loss": 0.0482, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.3149466192170818, | |
| "grad_norm": 1.6401046497106775, | |
| "learning_rate": 2.6260870270171645e-06, | |
| "loss": 0.044, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.3167259786476868, | |
| "grad_norm": 2.0852398847183053, | |
| "learning_rate": 2.613796825103129e-06, | |
| "loss": 0.0659, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.3185053380782918, | |
| "grad_norm": 1.5247014319411996, | |
| "learning_rate": 2.60152526440598e-06, | |
| "loss": 0.0329, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.3202846975088969, | |
| "grad_norm": 1.529414833928814, | |
| "learning_rate": 2.5892724407921667e-06, | |
| "loss": 0.0405, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.3220640569395017, | |
| "grad_norm": 2.102332443066376, | |
| "learning_rate": 2.577038449981763e-06, | |
| "loss": 0.0685, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.3238434163701067, | |
| "grad_norm": 1.8310627272949573, | |
| "learning_rate": 2.564823387547716e-06, | |
| "loss": 0.0523, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.3256227758007118, | |
| "grad_norm": 1.447883644708352, | |
| "learning_rate": 2.552627348915106e-06, | |
| "loss": 0.0496, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.3274021352313168, | |
| "grad_norm": 2.4642343978710337, | |
| "learning_rate": 2.5404504293603983e-06, | |
| "loss": 0.0634, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.3291814946619218, | |
| "grad_norm": 1.7886969180563022, | |
| "learning_rate": 2.528292724010697e-06, | |
| "loss": 0.0628, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.3309608540925266, | |
| "grad_norm": 1.6104244874898914, | |
| "learning_rate": 2.5161543278430055e-06, | |
| "loss": 0.0588, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.3327402135231317, | |
| "grad_norm": 1.636627371567464, | |
| "learning_rate": 2.5040353356834756e-06, | |
| "loss": 0.0429, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.3345195729537367, | |
| "grad_norm": 1.6247286470880693, | |
| "learning_rate": 2.4919358422066816e-06, | |
| "loss": 0.0508, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.3362989323843417, | |
| "grad_norm": 1.8654528914333375, | |
| "learning_rate": 2.4798559419348672e-06, | |
| "loss": 0.0479, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.3380782918149468, | |
| "grad_norm": 1.3881278472352054, | |
| "learning_rate": 2.4677957292372166e-06, | |
| "loss": 0.0431, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.3398576512455516, | |
| "grad_norm": 1.7467417958380345, | |
| "learning_rate": 2.455755298329107e-06, | |
| "loss": 0.0604, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.3416370106761566, | |
| "grad_norm": 1.49022623069703, | |
| "learning_rate": 2.4437347432713838e-06, | |
| "loss": 0.0384, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.3434163701067616, | |
| "grad_norm": 1.738799702285033, | |
| "learning_rate": 2.431734157969619e-06, | |
| "loss": 0.0592, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.3451957295373664, | |
| "grad_norm": 1.9083474964605713, | |
| "learning_rate": 2.4197536361733792e-06, | |
| "loss": 0.066, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.3469750889679715, | |
| "grad_norm": 1.4312368525037908, | |
| "learning_rate": 2.407793271475495e-06, | |
| "loss": 0.0504, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.3487544483985765, | |
| "grad_norm": 1.3937632917287868, | |
| "learning_rate": 2.3958531573113223e-06, | |
| "loss": 0.0601, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.3505338078291815, | |
| "grad_norm": 1.4790336087653384, | |
| "learning_rate": 2.3839333869580243e-06, | |
| "loss": 0.0433, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.3523131672597866, | |
| "grad_norm": 1.405969216008383, | |
| "learning_rate": 2.372034053533835e-06, | |
| "loss": 0.0356, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.3540925266903914, | |
| "grad_norm": 1.338831551620881, | |
| "learning_rate": 2.360155249997334e-06, | |
| "loss": 0.0434, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.3558718861209964, | |
| "grad_norm": 1.4852392169464352, | |
| "learning_rate": 2.348297069146715e-06, | |
| "loss": 0.0394, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.3576512455516014, | |
| "grad_norm": 1.286955298306548, | |
| "learning_rate": 2.3364596036190706e-06, | |
| "loss": 0.0317, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.3594306049822065, | |
| "grad_norm": 1.8218290396765193, | |
| "learning_rate": 2.3246429458896637e-06, | |
| "loss": 0.0592, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.3612099644128115, | |
| "grad_norm": 1.6847700004072366, | |
| "learning_rate": 2.312847188271203e-06, | |
| "loss": 0.0562, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.3629893238434163, | |
| "grad_norm": 1.4514400222356858, | |
| "learning_rate": 2.301072422913123e-06, | |
| "loss": 0.0373, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.3647686832740213, | |
| "grad_norm": 1.3991491705973067, | |
| "learning_rate": 2.2893187418008666e-06, | |
| "loss": 0.044, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.3665480427046264, | |
| "grad_norm": 1.6347719352259367, | |
| "learning_rate": 2.2775862367551642e-06, | |
| "loss": 0.0435, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.3683274021352312, | |
| "grad_norm": 1.4012138726711725, | |
| "learning_rate": 2.265874999431318e-06, | |
| "loss": 0.0366, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.3701067615658362, | |
| "grad_norm": 1.8241247381341326, | |
| "learning_rate": 2.254185121318484e-06, | |
| "loss": 0.0561, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.3718861209964412, | |
| "grad_norm": 1.5662115285203324, | |
| "learning_rate": 2.2425166937389596e-06, | |
| "loss": 0.0497, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.3736654804270463, | |
| "grad_norm": 1.3227843145360212, | |
| "learning_rate": 2.2308698078474645e-06, | |
| "loss": 0.0397, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.3754448398576513, | |
| "grad_norm": 1.693109557402715, | |
| "learning_rate": 2.219244554630438e-06, | |
| "loss": 0.0502, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.3772241992882561, | |
| "grad_norm": 1.4935851648348395, | |
| "learning_rate": 2.207641024905322e-06, | |
| "loss": 0.0455, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.3790035587188612, | |
| "grad_norm": 1.3187260632148614, | |
| "learning_rate": 2.1960593093198508e-06, | |
| "loss": 0.0339, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.3807829181494662, | |
| "grad_norm": 1.3378224243172256, | |
| "learning_rate": 2.184499498351347e-06, | |
| "loss": 0.0343, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.3825622775800712, | |
| "grad_norm": 1.48550825281536, | |
| "learning_rate": 2.172961682306011e-06, | |
| "loss": 0.0478, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.3843416370106763, | |
| "grad_norm": 1.5061146542775, | |
| "learning_rate": 2.1614459513182173e-06, | |
| "loss": 0.0408, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.386120996441281, | |
| "grad_norm": 1.5605749725879396, | |
| "learning_rate": 2.149952395349813e-06, | |
| "loss": 0.04, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.387900355871886, | |
| "grad_norm": 2.3475214433666367, | |
| "learning_rate": 2.1384811041894055e-06, | |
| "loss": 0.0726, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.3896797153024911, | |
| "grad_norm": 1.6250187454178806, | |
| "learning_rate": 2.1270321674516736e-06, | |
| "loss": 0.0387, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.3914590747330962, | |
| "grad_norm": 1.7200311423438022, | |
| "learning_rate": 2.1156056745766593e-06, | |
| "loss": 0.0416, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.3932384341637012, | |
| "grad_norm": 1.5987701792483358, | |
| "learning_rate": 2.104201714829074e-06, | |
| "loss": 0.0515, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.395017793594306, | |
| "grad_norm": 1.8830750383152088, | |
| "learning_rate": 2.0928203772975917e-06, | |
| "loss": 0.0605, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.396797153024911, | |
| "grad_norm": 1.6205998701999682, | |
| "learning_rate": 2.081461750894166e-06, | |
| "loss": 0.051, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.398576512455516, | |
| "grad_norm": 1.5130056274071655, | |
| "learning_rate": 2.070125924353328e-06, | |
| "loss": 0.0407, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.4003558718861209, | |
| "grad_norm": 1.4350157269999744, | |
| "learning_rate": 2.058812986231493e-06, | |
| "loss": 0.0315, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.402135231316726, | |
| "grad_norm": 1.9223090539296386, | |
| "learning_rate": 2.0475230249062727e-06, | |
| "loss": 0.0619, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.403914590747331, | |
| "grad_norm": 1.6155893694812649, | |
| "learning_rate": 2.0362561285757766e-06, | |
| "loss": 0.0508, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.405693950177936, | |
| "grad_norm": 1.9301184560518097, | |
| "learning_rate": 2.0250123852579347e-06, | |
| "loss": 0.058, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.407473309608541, | |
| "grad_norm": 1.5353566007659045, | |
| "learning_rate": 2.013791882789801e-06, | |
| "loss": 0.0469, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.4092526690391458, | |
| "grad_norm": 1.3681548446723928, | |
| "learning_rate": 2.0025947088268714e-06, | |
| "loss": 0.0351, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.4110320284697508, | |
| "grad_norm": 1.6803430043809926, | |
| "learning_rate": 1.9914209508423943e-06, | |
| "loss": 0.0605, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.4128113879003559, | |
| "grad_norm": 2.1981628298390636, | |
| "learning_rate": 1.9802706961266936e-06, | |
| "loss": 0.0544, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.414590747330961, | |
| "grad_norm": 1.796806214364753, | |
| "learning_rate": 1.969144031786483e-06, | |
| "loss": 0.0587, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.416370106761566, | |
| "grad_norm": 1.6404472508002965, | |
| "learning_rate": 1.958041044744186e-06, | |
| "loss": 0.0567, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.4181494661921707, | |
| "grad_norm": 1.5898717386945593, | |
| "learning_rate": 1.94696182173726e-06, | |
| "loss": 0.04, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.4199288256227758, | |
| "grad_norm": 1.662695910080773, | |
| "learning_rate": 1.9359064493175077e-06, | |
| "loss": 0.0462, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.4217081850533808, | |
| "grad_norm": 1.4257775056257276, | |
| "learning_rate": 1.9248750138504176e-06, | |
| "loss": 0.0443, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.4234875444839858, | |
| "grad_norm": 1.4320094357335684, | |
| "learning_rate": 1.9138676015144765e-06, | |
| "loss": 0.0352, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4234875444839858, | |
| "eval_loss": 0.14595071971416473, | |
| "eval_runtime": 1.8889, | |
| "eval_samples_per_second": 24.352, | |
| "eval_steps_per_second": 6.353, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4252669039145909, | |
| "grad_norm": 1.542305943500346, | |
| "learning_rate": 1.9028842983005036e-06, | |
| "loss": 0.0578, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.4270462633451957, | |
| "grad_norm": 1.8648944594226182, | |
| "learning_rate": 1.8919251900109697e-06, | |
| "loss": 0.0686, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.4288256227758007, | |
| "grad_norm": 1.408539141876749, | |
| "learning_rate": 1.8809903622593395e-06, | |
| "loss": 0.0363, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.4306049822064058, | |
| "grad_norm": 1.5777799487776285, | |
| "learning_rate": 1.870079900469392e-06, | |
| "loss": 0.0451, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.4323843416370106, | |
| "grad_norm": 1.2300626956615635, | |
| "learning_rate": 1.8591938898745593e-06, | |
| "loss": 0.0373, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.4341637010676156, | |
| "grad_norm": 1.4155111194882426, | |
| "learning_rate": 1.8483324155172594e-06, | |
| "loss": 0.032, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.4359430604982206, | |
| "grad_norm": 1.9545802196259987, | |
| "learning_rate": 1.837495562248226e-06, | |
| "loss": 0.0676, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.4377224199288257, | |
| "grad_norm": 1.5426528914977735, | |
| "learning_rate": 1.8266834147258577e-06, | |
| "loss": 0.0446, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.4395017793594307, | |
| "grad_norm": 1.3706696059691963, | |
| "learning_rate": 1.8158960574155455e-06, | |
| "loss": 0.048, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.4412811387900355, | |
| "grad_norm": 1.923852821053572, | |
| "learning_rate": 1.8051335745890196e-06, | |
| "loss": 0.0485, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.4430604982206405, | |
| "grad_norm": 1.538454678357912, | |
| "learning_rate": 1.7943960503236856e-06, | |
| "loss": 0.0511, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.4448398576512456, | |
| "grad_norm": 1.2699215475852161, | |
| "learning_rate": 1.7836835685019732e-06, | |
| "loss": 0.0415, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.4466192170818506, | |
| "grad_norm": 1.5079641838195619, | |
| "learning_rate": 1.7729962128106787e-06, | |
| "loss": 0.0464, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.4483985765124556, | |
| "grad_norm": 1.3396635005681008, | |
| "learning_rate": 1.7623340667403089e-06, | |
| "loss": 0.036, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.4501779359430604, | |
| "grad_norm": 1.944032103288467, | |
| "learning_rate": 1.7516972135844352e-06, | |
| "loss": 0.0481, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.4519572953736655, | |
| "grad_norm": 1.7584288747521668, | |
| "learning_rate": 1.741085736439031e-06, | |
| "loss": 0.054, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.4537366548042705, | |
| "grad_norm": 1.8784996399254477, | |
| "learning_rate": 1.730499718201838e-06, | |
| "loss": 0.0581, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.4555160142348753, | |
| "grad_norm": 1.585623365884921, | |
| "learning_rate": 1.7199392415717064e-06, | |
| "loss": 0.047, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.4572953736654806, | |
| "grad_norm": 1.0908992660187677, | |
| "learning_rate": 1.7094043890479557e-06, | |
| "loss": 0.0353, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.4590747330960854, | |
| "grad_norm": 1.5480233871791025, | |
| "learning_rate": 1.698895242929725e-06, | |
| "loss": 0.0494, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.4608540925266904, | |
| "grad_norm": 1.7995348335643881, | |
| "learning_rate": 1.6884118853153358e-06, | |
| "loss": 0.0514, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.4626334519572954, | |
| "grad_norm": 1.814776913660295, | |
| "learning_rate": 1.6779543981016478e-06, | |
| "loss": 0.0459, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.4644128113879002, | |
| "grad_norm": 1.6112682309489867, | |
| "learning_rate": 1.6675228629834133e-06, | |
| "loss": 0.0517, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.4661921708185053, | |
| "grad_norm": 1.3671655412017953, | |
| "learning_rate": 1.657117361452651e-06, | |
| "loss": 0.0407, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.4679715302491103, | |
| "grad_norm": 1.1351235798355912, | |
| "learning_rate": 1.6467379747980011e-06, | |
| "loss": 0.0366, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.4697508896797153, | |
| "grad_norm": 1.2601068370284656, | |
| "learning_rate": 1.6363847841040914e-06, | |
| "loss": 0.0394, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.4715302491103204, | |
| "grad_norm": 1.5781445493139832, | |
| "learning_rate": 1.626057870250906e-06, | |
| "loss": 0.0407, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.4733096085409252, | |
| "grad_norm": 1.4192275215482895, | |
| "learning_rate": 1.6157573139131527e-06, | |
| "loss": 0.0372, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.4750889679715302, | |
| "grad_norm": 1.3040976042337626, | |
| "learning_rate": 1.605483195559628e-06, | |
| "loss": 0.0483, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.4768683274021353, | |
| "grad_norm": 1.4808992385309687, | |
| "learning_rate": 1.5952355954525966e-06, | |
| "loss": 0.0375, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.4786476868327403, | |
| "grad_norm": 1.5434824292086615, | |
| "learning_rate": 1.5850145936471607e-06, | |
| "loss": 0.0464, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.4804270462633453, | |
| "grad_norm": 2.0181781279521593, | |
| "learning_rate": 1.5748202699906335e-06, | |
| "loss": 0.0571, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.4822064056939501, | |
| "grad_norm": 1.8018946073991524, | |
| "learning_rate": 1.5646527041219128e-06, | |
| "loss": 0.0498, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.4839857651245552, | |
| "grad_norm": 1.47467339069657, | |
| "learning_rate": 1.5545119754708682e-06, | |
| "loss": 0.039, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.4857651245551602, | |
| "grad_norm": 1.7111292361011305, | |
| "learning_rate": 1.544398163257711e-06, | |
| "loss": 0.05, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.487544483985765, | |
| "grad_norm": 1.5753387872671925, | |
| "learning_rate": 1.5343113464923808e-06, | |
| "loss": 0.0494, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.48932384341637, | |
| "grad_norm": 1.497662835665088, | |
| "learning_rate": 1.524251603973927e-06, | |
| "loss": 0.0493, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.491103202846975, | |
| "grad_norm": 1.6141445517392699, | |
| "learning_rate": 1.5142190142898883e-06, | |
| "loss": 0.0469, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.49288256227758, | |
| "grad_norm": 1.6437524147273326, | |
| "learning_rate": 1.5042136558156883e-06, | |
| "loss": 0.0544, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.4946619217081851, | |
| "grad_norm": 1.4570589734380823, | |
| "learning_rate": 1.4942356067140162e-06, | |
| "loss": 0.042, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.49644128113879, | |
| "grad_norm": 1.17670273551654, | |
| "learning_rate": 1.4842849449342195e-06, | |
| "loss": 0.0367, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.498220640569395, | |
| "grad_norm": 1.3292605130554214, | |
| "learning_rate": 1.4743617482116896e-06, | |
| "loss": 0.0376, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.8950906601916515, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.047, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.501779359430605, | |
| "grad_norm": 1.5503491262034121, | |
| "learning_rate": 1.454598059806609e-06, | |
| "loss": 0.0485, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.50355871886121, | |
| "grad_norm": 1.7854432270440623, | |
| "learning_rate": 1.4447577225196296e-06, | |
| "loss": 0.0504, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.5053380782918149, | |
| "grad_norm": 1.6010720830304457, | |
| "learning_rate": 1.4349451590798564e-06, | |
| "loss": 0.0566, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.50711743772242, | |
| "grad_norm": 1.585467916335836, | |
| "learning_rate": 1.4251604461438444e-06, | |
| "loss": 0.0512, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.508896797153025, | |
| "grad_norm": 1.2602219314937615, | |
| "learning_rate": 1.4154036601505834e-06, | |
| "loss": 0.0331, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.5106761565836297, | |
| "grad_norm": 1.4216698283865359, | |
| "learning_rate": 1.4056748773208933e-06, | |
| "loss": 0.0382, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.512455516014235, | |
| "grad_norm": 1.8972846300824147, | |
| "learning_rate": 1.3959741736568339e-06, | |
| "loss": 0.0621, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.5142348754448398, | |
| "grad_norm": 1.4745012600105427, | |
| "learning_rate": 1.3863016249411027e-06, | |
| "loss": 0.041, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.5160142348754448, | |
| "grad_norm": 1.4439525302505538, | |
| "learning_rate": 1.376657306736453e-06, | |
| "loss": 0.0432, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.5177935943060499, | |
| "grad_norm": 1.6129015850929418, | |
| "learning_rate": 1.3670412943850975e-06, | |
| "loss": 0.0485, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.5195729537366547, | |
| "grad_norm": 1.4636499139086063, | |
| "learning_rate": 1.3574536630081208e-06, | |
| "loss": 0.0393, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.52135231316726, | |
| "grad_norm": 1.766669994257703, | |
| "learning_rate": 1.347894487504896e-06, | |
| "loss": 0.0541, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.5231316725978647, | |
| "grad_norm": 1.2833048017170003, | |
| "learning_rate": 1.3383638425524909e-06, | |
| "loss": 0.0339, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.5249110320284698, | |
| "grad_norm": 1.431768026011943, | |
| "learning_rate": 1.3288618026050943e-06, | |
| "loss": 0.0405, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.5266903914590748, | |
| "grad_norm": 2.007033223429606, | |
| "learning_rate": 1.31938844189343e-06, | |
| "loss": 0.0649, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.5284697508896796, | |
| "grad_norm": 2.029919770627406, | |
| "learning_rate": 1.3099438344241777e-06, | |
| "loss": 0.0608, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.5302491103202847, | |
| "grad_norm": 1.6218815496312928, | |
| "learning_rate": 1.3005280539793908e-06, | |
| "loss": 0.0507, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.5320284697508897, | |
| "grad_norm": 1.583831100550564, | |
| "learning_rate": 1.2911411741159273e-06, | |
| "loss": 0.0381, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.5338078291814945, | |
| "grad_norm": 1.4120420932331246, | |
| "learning_rate": 1.2817832681648712e-06, | |
| "loss": 0.0474, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.5355871886120998, | |
| "grad_norm": 1.2751903940001499, | |
| "learning_rate": 1.2724544092309581e-06, | |
| "loss": 0.0346, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.5373665480427046, | |
| "grad_norm": 1.2386946124213627, | |
| "learning_rate": 1.2631546701920073e-06, | |
| "loss": 0.0471, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.5391459074733096, | |
| "grad_norm": 1.7962767249122955, | |
| "learning_rate": 1.2538841236983519e-06, | |
| "loss": 0.0496, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.5409252669039146, | |
| "grad_norm": 1.6058636158515285, | |
| "learning_rate": 1.244642842172266e-06, | |
| "loss": 0.0369, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.5427046263345194, | |
| "grad_norm": 1.4172154429302468, | |
| "learning_rate": 1.2354308978074088e-06, | |
| "loss": 0.0399, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.5444839857651247, | |
| "grad_norm": 1.5028770290011404, | |
| "learning_rate": 1.2262483625682514e-06, | |
| "loss": 0.0512, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.5462633451957295, | |
| "grad_norm": 1.4722689968669584, | |
| "learning_rate": 1.2170953081895214e-06, | |
| "loss": 0.0394, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.5480427046263345, | |
| "grad_norm": 1.8104076677352459, | |
| "learning_rate": 1.2079718061756369e-06, | |
| "loss": 0.0525, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.5498220640569396, | |
| "grad_norm": 1.317580245578167, | |
| "learning_rate": 1.1988779278001517e-06, | |
| "loss": 0.0377, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.5516014234875444, | |
| "grad_norm": 1.9284146589556046, | |
| "learning_rate": 1.1898137441051982e-06, | |
| "loss": 0.0631, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.5533807829181496, | |
| "grad_norm": 1.6843574765305045, | |
| "learning_rate": 1.1807793259009282e-06, | |
| "loss": 0.0531, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.5551601423487544, | |
| "grad_norm": 1.4092482984988974, | |
| "learning_rate": 1.1717747437649657e-06, | |
| "loss": 0.039, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.5569395017793595, | |
| "grad_norm": 1.3697495828604107, | |
| "learning_rate": 1.1628000680418533e-06, | |
| "loss": 0.0359, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.5587188612099645, | |
| "grad_norm": 1.2962713717025165, | |
| "learning_rate": 1.1538553688425002e-06, | |
| "loss": 0.0297, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.5604982206405693, | |
| "grad_norm": 1.7485510128107082, | |
| "learning_rate": 1.14494071604364e-06, | |
| "loss": 0.0398, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.5622775800711743, | |
| "grad_norm": 1.6008116762807254, | |
| "learning_rate": 1.1360561792872754e-06, | |
| "loss": 0.0474, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.5640569395017794, | |
| "grad_norm": 2.4210865256005825, | |
| "learning_rate": 1.127201827980145e-06, | |
| "loss": 0.0643, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.5658362989323842, | |
| "grad_norm": 1.3381633595504001, | |
| "learning_rate": 1.1183777312931748e-06, | |
| "loss": 0.034, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.5676156583629894, | |
| "grad_norm": 1.4566811241631359, | |
| "learning_rate": 1.1095839581609407e-06, | |
| "loss": 0.043, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.5693950177935942, | |
| "grad_norm": 1.9994171530686031, | |
| "learning_rate": 1.1008205772811248e-06, | |
| "loss": 0.0505, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.5711743772241993, | |
| "grad_norm": 1.8048729396737153, | |
| "learning_rate": 1.0920876571139843e-06, | |
| "loss": 0.0616, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.5729537366548043, | |
| "grad_norm": 1.3161748092556311, | |
| "learning_rate": 1.0833852658818167e-06, | |
| "loss": 0.0368, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.5747330960854091, | |
| "grad_norm": 1.3148941673186614, | |
| "learning_rate": 1.0747134715684221e-06, | |
| "loss": 0.0301, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.5765124555160144, | |
| "grad_norm": 1.9825464522594134, | |
| "learning_rate": 1.0660723419185776e-06, | |
| "loss": 0.0535, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.5782918149466192, | |
| "grad_norm": 1.5185782757706232, | |
| "learning_rate": 1.0574619444375017e-06, | |
| "loss": 0.044, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.5800711743772242, | |
| "grad_norm": 1.132409434790568, | |
| "learning_rate": 1.0488823463903341e-06, | |
| "loss": 0.0332, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.5818505338078293, | |
| "grad_norm": 1.5991028704907984, | |
| "learning_rate": 1.0403336148016053e-06, | |
| "loss": 0.0507, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.583629893238434, | |
| "grad_norm": 1.3477345790247133, | |
| "learning_rate": 1.0318158164547159e-06, | |
| "loss": 0.0447, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.585409252669039, | |
| "grad_norm": 1.6751225009529225, | |
| "learning_rate": 1.0233290178914096e-06, | |
| "loss": 0.0369, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.5871886120996441, | |
| "grad_norm": 1.3728336702006203, | |
| "learning_rate": 1.014873285411262e-06, | |
| "loss": 0.0478, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.5889679715302492, | |
| "grad_norm": 1.8411034892207845, | |
| "learning_rate": 1.006448685071154e-06, | |
| "loss": 0.0645, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.5907473309608542, | |
| "grad_norm": 1.9138440322307035, | |
| "learning_rate": 9.980552826847635e-07, | |
| "loss": 0.053, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.592526690391459, | |
| "grad_norm": 1.5915492619322404, | |
| "learning_rate": 9.896931438220453e-07, | |
| "loss": 0.0535, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.594306049822064, | |
| "grad_norm": 1.613903218644777, | |
| "learning_rate": 9.813623338087181e-07, | |
| "loss": 0.0432, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.596085409252669, | |
| "grad_norm": 2.0961848428547505, | |
| "learning_rate": 9.730629177257623e-07, | |
| "loss": 0.047, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.5978647686832739, | |
| "grad_norm": 1.4830440521961352, | |
| "learning_rate": 9.64794960408903e-07, | |
| "loss": 0.0382, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.5996441281138791, | |
| "grad_norm": 1.230528563404261, | |
| "learning_rate": 9.565585264481092e-07, | |
| "loss": 0.0386, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.601423487544484, | |
| "grad_norm": 1.3567558897360548, | |
| "learning_rate": 9.483536801870835e-07, | |
| "loss": 0.0359, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.603202846975089, | |
| "grad_norm": 1.4751351647789543, | |
| "learning_rate": 9.401804857227648e-07, | |
| "loss": 0.0324, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.604982206405694, | |
| "grad_norm": 1.7510649490555823, | |
| "learning_rate": 9.320390069048258e-07, | |
| "loss": 0.0652, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.6067615658362988, | |
| "grad_norm": 1.2747184871148731, | |
| "learning_rate": 9.239293073351735e-07, | |
| "loss": 0.0377, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.608540925266904, | |
| "grad_norm": 1.6869616869045594, | |
| "learning_rate": 9.158514503674543e-07, | |
| "loss": 0.0461, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.6103202846975089, | |
| "grad_norm": 1.305958732210241, | |
| "learning_rate": 9.078054991065532e-07, | |
| "loss": 0.0406, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.612099644128114, | |
| "grad_norm": 1.4224733071510518, | |
| "learning_rate": 8.997915164081095e-07, | |
| "loss": 0.042, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.613879003558719, | |
| "grad_norm": 1.3550883460121246, | |
| "learning_rate": 8.918095648780195e-07, | |
| "loss": 0.0354, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.6156583629893237, | |
| "grad_norm": 1.5611002986773197, | |
| "learning_rate": 8.838597068719518e-07, | |
| "loss": 0.0425, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.6174377224199288, | |
| "grad_norm": 1.7764766275708408, | |
| "learning_rate": 8.75942004494853e-07, | |
| "loss": 0.0388, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.6192170818505338, | |
| "grad_norm": 1.7337801110275877, | |
| "learning_rate": 8.680565196004704e-07, | |
| "loss": 0.0482, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.6209964412811388, | |
| "grad_norm": 1.5985962495391328, | |
| "learning_rate": 8.602033137908666e-07, | |
| "loss": 0.0365, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.6227758007117439, | |
| "grad_norm": 1.543492343723246, | |
| "learning_rate": 8.523824484159348e-07, | |
| "loss": 0.0441, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.6245551601423487, | |
| "grad_norm": 1.1832595248208693, | |
| "learning_rate": 8.445939845729245e-07, | |
| "loss": 0.0321, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.6263345195729537, | |
| "grad_norm": 1.303130905507554, | |
| "learning_rate": 8.368379831059592e-07, | |
| "loss": 0.0385, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.6281138790035588, | |
| "grad_norm": 1.2050174662192676, | |
| "learning_rate": 8.29114504605566e-07, | |
| "loss": 0.0383, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.6298932384341636, | |
| "grad_norm": 1.3458303658271265, | |
| "learning_rate": 8.21423609408199e-07, | |
| "loss": 0.0422, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.6316725978647688, | |
| "grad_norm": 1.7137814763276404, | |
| "learning_rate": 8.137653575957666e-07, | |
| "loss": 0.0393, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.6334519572953736, | |
| "grad_norm": 1.7271114528108913, | |
| "learning_rate": 8.061398089951678e-07, | |
| "loss": 0.0558, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.6352313167259787, | |
| "grad_norm": 1.348840499427854, | |
| "learning_rate": 7.985470231778203e-07, | |
| "loss": 0.0325, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.6370106761565837, | |
| "grad_norm": 1.682962691501169, | |
| "learning_rate": 7.909870594591951e-07, | |
| "loss": 0.0422, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.6387900355871885, | |
| "grad_norm": 1.7036539888543565, | |
| "learning_rate": 7.834599768983553e-07, | |
| "loss": 0.0373, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.6405693950177938, | |
| "grad_norm": 1.5621321963767334, | |
| "learning_rate": 7.759658342974951e-07, | |
| "loss": 0.0409, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.6423487544483986, | |
| "grad_norm": 1.7947883183129458, | |
| "learning_rate": 7.685046902014747e-07, | |
| "loss": 0.0622, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.6441281138790036, | |
| "grad_norm": 1.7635057073853968, | |
| "learning_rate": 7.61076602897371e-07, | |
| "loss": 0.0426, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.6459074733096086, | |
| "grad_norm": 1.549021065145144, | |
| "learning_rate": 7.536816304140177e-07, | |
| "loss": 0.0432, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.6476868327402134, | |
| "grad_norm": 1.5804084914709269, | |
| "learning_rate": 7.46319830521553e-07, | |
| "loss": 0.0441, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.6494661921708185, | |
| "grad_norm": 1.2244876138852487, | |
| "learning_rate": 7.389912607309662e-07, | |
| "loss": 0.0333, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.6512455516014235, | |
| "grad_norm": 3.5111212805642436, | |
| "learning_rate": 7.316959782936516e-07, | |
| "loss": 0.0677, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.6530249110320283, | |
| "grad_norm": 1.308507354078122, | |
| "learning_rate": 7.244340402009608e-07, | |
| "loss": 0.044, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.6548042704626336, | |
| "grad_norm": 1.9423259867437728, | |
| "learning_rate": 7.172055031837572e-07, | |
| "loss": 0.0527, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.6565836298932384, | |
| "grad_norm": 1.5781335015475328, | |
| "learning_rate": 7.100104237119676e-07, | |
| "loss": 0.0408, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.6583629893238434, | |
| "grad_norm": 1.2203804340189908, | |
| "learning_rate": 7.028488579941506e-07, | |
| "loss": 0.0407, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.6601423487544484, | |
| "grad_norm": 1.9772794732536039, | |
| "learning_rate": 6.957208619770505e-07, | |
| "loss": 0.0459, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.6619217081850532, | |
| "grad_norm": 1.7546367312553026, | |
| "learning_rate": 6.886264913451635e-07, | |
| "loss": 0.0403, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.6637010676156585, | |
| "grad_norm": 1.823588410398394, | |
| "learning_rate": 6.815658015203014e-07, | |
| "loss": 0.0394, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.6654804270462633, | |
| "grad_norm": 1.346251285157429, | |
| "learning_rate": 6.745388476611553e-07, | |
| "loss": 0.04, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.6672597864768683, | |
| "grad_norm": 1.516411584900659, | |
| "learning_rate": 6.67545684662873e-07, | |
| "loss": 0.0423, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.6690391459074734, | |
| "grad_norm": 1.5385602152144275, | |
| "learning_rate": 6.605863671566221e-07, | |
| "loss": 0.0476, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.6708185053380782, | |
| "grad_norm": 2.0200801763181215, | |
| "learning_rate": 6.536609495091695e-07, | |
| "loss": 0.046, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.6725978647686834, | |
| "grad_norm": 1.9829192253954908, | |
| "learning_rate": 6.467694858224488e-07, | |
| "loss": 0.0436, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.6743772241992882, | |
| "grad_norm": 1.721581487774052, | |
| "learning_rate": 6.399120299331468e-07, | |
| "loss": 0.0398, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.6761565836298933, | |
| "grad_norm": 1.4020425861909211, | |
| "learning_rate": 6.330886354122768e-07, | |
| "loss": 0.0511, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.6779359430604983, | |
| "grad_norm": 1.396580862521342, | |
| "learning_rate": 6.262993555647617e-07, | |
| "loss": 0.0433, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.6797153024911031, | |
| "grad_norm": 1.5517704121801634, | |
| "learning_rate": 6.1954424342902e-07, | |
| "loss": 0.0446, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.6814946619217082, | |
| "grad_norm": 2.1155752156225067, | |
| "learning_rate": 6.128233517765448e-07, | |
| "loss": 0.0642, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.6832740213523132, | |
| "grad_norm": 1.6660999431118102, | |
| "learning_rate": 6.061367331114992e-07, | |
| "loss": 0.0471, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.685053380782918, | |
| "grad_norm": 1.258438120691111, | |
| "learning_rate": 5.994844396703025e-07, | |
| "loss": 0.0372, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.6868327402135233, | |
| "grad_norm": 1.2814530611157964, | |
| "learning_rate": 5.928665234212233e-07, | |
| "loss": 0.0392, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.688612099644128, | |
| "grad_norm": 1.3533114025454038, | |
| "learning_rate": 5.862830360639698e-07, | |
| "loss": 0.0353, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.690391459074733, | |
| "grad_norm": 1.5411665340001826, | |
| "learning_rate": 5.797340290292907e-07, | |
| "loss": 0.0375, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.6921708185053381, | |
| "grad_norm": 1.6681942263190166, | |
| "learning_rate": 5.732195534785723e-07, | |
| "loss": 0.0435, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.693950177935943, | |
| "grad_norm": 1.3363002138561992, | |
| "learning_rate": 5.667396603034369e-07, | |
| "loss": 0.0367, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.6957295373665482, | |
| "grad_norm": 1.5148617436116476, | |
| "learning_rate": 5.602944001253486e-07, | |
| "loss": 0.0379, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.697508896797153, | |
| "grad_norm": 1.448788818734238, | |
| "learning_rate": 5.538838232952104e-07, | |
| "loss": 0.0391, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.699288256227758, | |
| "grad_norm": 1.56583381557676, | |
| "learning_rate": 5.475079798929816e-07, | |
| "loss": 0.0515, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.701067615658363, | |
| "grad_norm": 1.4174363400085765, | |
| "learning_rate": 5.411669197272795e-07, | |
| "loss": 0.032, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.7028469750889679, | |
| "grad_norm": 1.5204352676241315, | |
| "learning_rate": 5.348606923349903e-07, | |
| "loss": 0.0409, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.704626334519573, | |
| "grad_norm": 1.23194145988896, | |
| "learning_rate": 5.285893469808855e-07, | |
| "loss": 0.031, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.706405693950178, | |
| "grad_norm": 1.4689389452086652, | |
| "learning_rate": 5.223529326572352e-07, | |
| "loss": 0.0389, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.708185053380783, | |
| "grad_norm": 1.5532010962171097, | |
| "learning_rate": 5.161514980834232e-07, | |
| "loss": 0.0366, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.709964412811388, | |
| "grad_norm": 1.4962434596925702, | |
| "learning_rate": 5.099850917055709e-07, | |
| "loss": 0.0298, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.7117437722419928, | |
| "grad_norm": 1.6643086101682487, | |
| "learning_rate": 5.038537616961559e-07, | |
| "loss": 0.0492, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.7135231316725978, | |
| "grad_norm": 1.3089303328933728, | |
| "learning_rate": 4.977575559536358e-07, | |
| "loss": 0.0299, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.7153024911032029, | |
| "grad_norm": 1.608175117123677, | |
| "learning_rate": 4.916965221020753e-07, | |
| "loss": 0.0431, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.7170818505338077, | |
| "grad_norm": 1.4102691798736315, | |
| "learning_rate": 4.856707074907729e-07, | |
| "loss": 0.0398, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.718861209964413, | |
| "grad_norm": 1.9933302609537924, | |
| "learning_rate": 4.796801591938922e-07, | |
| "loss": 0.061, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.7206405693950177, | |
| "grad_norm": 1.688352460479466, | |
| "learning_rate": 4.737249240100911e-07, | |
| "loss": 0.0488, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.7224199288256228, | |
| "grad_norm": 1.7612512059358014, | |
| "learning_rate": 4.6780504846216155e-07, | |
| "loss": 0.0372, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.7241992882562278, | |
| "grad_norm": 1.411829654149764, | |
| "learning_rate": 4.619205787966613e-07, | |
| "loss": 0.0379, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.7259786476868326, | |
| "grad_norm": 1.1325365642245606, | |
| "learning_rate": 4.560715609835548e-07, | |
| "loss": 0.0321, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.7277580071174379, | |
| "grad_norm": 1.605416264323445, | |
| "learning_rate": 4.5025804071585464e-07, | |
| "loss": 0.052, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.7295373665480427, | |
| "grad_norm": 1.593284079402667, | |
| "learning_rate": 4.4448006340926163e-07, | |
| "loss": 0.0416, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.7313167259786477, | |
| "grad_norm": 1.5253350519078592, | |
| "learning_rate": 4.3873767420181344e-07, | |
| "loss": 0.0466, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.7330960854092528, | |
| "grad_norm": 1.589287409685416, | |
| "learning_rate": 4.3303091795353024e-07, | |
| "loss": 0.043, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.7348754448398576, | |
| "grad_norm": 1.3485162191396995, | |
| "learning_rate": 4.2735983924606596e-07, | |
| "loss": 0.0295, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.7366548042704626, | |
| "grad_norm": 1.6709173335097418, | |
| "learning_rate": 4.2172448238235464e-07, | |
| "loss": 0.0367, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.7384341637010676, | |
| "grad_norm": 1.5968733757492934, | |
| "learning_rate": 4.161248913862731e-07, | |
| "loss": 0.0447, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.7402135231316724, | |
| "grad_norm": 1.393351838125729, | |
| "learning_rate": 4.1056111000228937e-07, | |
| "loss": 0.0293, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.7419928825622777, | |
| "grad_norm": 1.6256800448978685, | |
| "learning_rate": 4.0503318169512417e-07, | |
| "loss": 0.0473, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.7437722419928825, | |
| "grad_norm": 1.3344117346476265, | |
| "learning_rate": 3.9954114964941336e-07, | |
| "loss": 0.0309, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.7455516014234875, | |
| "grad_norm": 1.5741412144751346, | |
| "learning_rate": 3.9408505676936327e-07, | |
| "loss": 0.0414, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.7473309608540926, | |
| "grad_norm": 1.7420665698274114, | |
| "learning_rate": 3.886649456784253e-07, | |
| "loss": 0.05, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.7491103202846974, | |
| "grad_norm": 1.5310990117639014, | |
| "learning_rate": 3.8328085871895624e-07, | |
| "loss": 0.0442, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.7508896797153026, | |
| "grad_norm": 1.5359607037507725, | |
| "learning_rate": 3.779328379518898e-07, | |
| "loss": 0.0532, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.7526690391459074, | |
| "grad_norm": 1.3029445443316392, | |
| "learning_rate": 3.7262092515640556e-07, | |
| "loss": 0.0304, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.7544483985765125, | |
| "grad_norm": 1.3061999155259794, | |
| "learning_rate": 3.673451618296081e-07, | |
| "loss": 0.0341, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.7562277580071175, | |
| "grad_norm": 1.441777236465274, | |
| "learning_rate": 3.621055891861963e-07, | |
| "loss": 0.0389, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.7580071174377223, | |
| "grad_norm": 1.8150746345067803, | |
| "learning_rate": 3.56902248158148e-07, | |
| "loss": 0.0389, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.7597864768683276, | |
| "grad_norm": 1.3762404788368663, | |
| "learning_rate": 3.517351793943913e-07, | |
| "loss": 0.0339, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.7615658362989324, | |
| "grad_norm": 1.6633356829038823, | |
| "learning_rate": 3.4660442326049704e-07, | |
| "loss": 0.0314, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.7633451957295374, | |
| "grad_norm": 1.373782144418225, | |
| "learning_rate": 3.4151001983835696e-07, | |
| "loss": 0.0486, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.7651245551601424, | |
| "grad_norm": 1.3663382307897525, | |
| "learning_rate": 3.364520089258727e-07, | |
| "loss": 0.0341, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.7669039145907472, | |
| "grad_norm": 1.7893823256003312, | |
| "learning_rate": 3.314304300366461e-07, | |
| "loss": 0.044, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.7686832740213523, | |
| "grad_norm": 1.866056791876848, | |
| "learning_rate": 3.2644532239966444e-07, | |
| "loss": 0.0505, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.7704626334519573, | |
| "grad_norm": 1.219021007258483, | |
| "learning_rate": 3.2149672495900286e-07, | |
| "loss": 0.034, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.7722419928825621, | |
| "grad_norm": 2.0300011505812865, | |
| "learning_rate": 3.165846763735153e-07, | |
| "loss": 0.0586, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.7740213523131674, | |
| "grad_norm": 1.5781286566972308, | |
| "learning_rate": 3.117092150165324e-07, | |
| "loss": 0.0485, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.7758007117437722, | |
| "grad_norm": 1.5551448290853336, | |
| "learning_rate": 3.068703789755606e-07, | |
| "loss": 0.0436, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.7775800711743772, | |
| "grad_norm": 1.6019665127129685, | |
| "learning_rate": 3.020682060519886e-07, | |
| "loss": 0.0405, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.7793594306049823, | |
| "grad_norm": 1.4781099663151758, | |
| "learning_rate": 2.9730273376078923e-07, | |
| "loss": 0.0432, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7793594306049823, | |
| "eval_loss": 0.14510178565979004, | |
| "eval_runtime": 1.8874, | |
| "eval_samples_per_second": 24.372, | |
| "eval_steps_per_second": 6.358, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.781138790035587, | |
| "grad_norm": 2.2736654799135962, | |
| "learning_rate": 2.9257399933022737e-07, | |
| "loss": 0.0529, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.7829181494661923, | |
| "grad_norm": 1.336929375340167, | |
| "learning_rate": 2.8788203970156805e-07, | |
| "loss": 0.0386, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.7846975088967971, | |
| "grad_norm": 1.527181971436809, | |
| "learning_rate": 2.832268915287878e-07, | |
| "loss": 0.0487, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.7864768683274022, | |
| "grad_norm": 1.7140840449195633, | |
| "learning_rate": 2.7860859117828985e-07, | |
| "loss": 0.0485, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.7882562277580072, | |
| "grad_norm": 1.8572849776902922, | |
| "learning_rate": 2.740271747286194e-07, | |
| "loss": 0.0449, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.790035587188612, | |
| "grad_norm": 1.8161164332548432, | |
| "learning_rate": 2.6948267797018145e-07, | |
| "loss": 0.0451, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 1.791814946619217, | |
| "grad_norm": 1.6794821656093764, | |
| "learning_rate": 2.649751364049613e-07, | |
| "loss": 0.0344, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 1.793594306049822, | |
| "grad_norm": 1.5333015954147242, | |
| "learning_rate": 2.6050458524624735e-07, | |
| "loss": 0.0418, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 1.795373665480427, | |
| "grad_norm": 1.399998276036899, | |
| "learning_rate": 2.560710594183552e-07, | |
| "loss": 0.0302, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 1.7971530249110321, | |
| "grad_norm": 1.4324461620505187, | |
| "learning_rate": 2.5167459355635524e-07, | |
| "loss": 0.0378, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.798932384341637, | |
| "grad_norm": 1.783897390942567, | |
| "learning_rate": 2.473152220058039e-07, | |
| "loss": 0.0397, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 1.800711743772242, | |
| "grad_norm": 1.6412342224434973, | |
| "learning_rate": 2.429929788224722e-07, | |
| "loss": 0.0362, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 1.802491103202847, | |
| "grad_norm": 1.68923877739915, | |
| "learning_rate": 2.38707897772083e-07, | |
| "loss": 0.0586, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.8042704626334518, | |
| "grad_norm": 1.9356223167399953, | |
| "learning_rate": 2.3446001233004333e-07, | |
| "loss": 0.0651, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 1.806049822064057, | |
| "grad_norm": 1.5902613291829113, | |
| "learning_rate": 2.3024935568118745e-07, | |
| "loss": 0.0393, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.8078291814946619, | |
| "grad_norm": 1.7747641255537248, | |
| "learning_rate": 2.2607596071951288e-07, | |
| "loss": 0.0475, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 1.809608540925267, | |
| "grad_norm": 1.3675793118529163, | |
| "learning_rate": 2.2193986004792667e-07, | |
| "loss": 0.0439, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 1.811387900355872, | |
| "grad_norm": 1.316286402064375, | |
| "learning_rate": 2.1784108597799058e-07, | |
| "loss": 0.0292, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 1.8131672597864767, | |
| "grad_norm": 1.2331395521319486, | |
| "learning_rate": 2.1377967052966685e-07, | |
| "loss": 0.0276, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 1.814946619217082, | |
| "grad_norm": 1.1809571187950396, | |
| "learning_rate": 2.0975564543107007e-07, | |
| "loss": 0.0355, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.8167259786476868, | |
| "grad_norm": 1.229103334152263, | |
| "learning_rate": 2.057690421182168e-07, | |
| "loss": 0.0277, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 1.8185053380782918, | |
| "grad_norm": 1.2080584468363225, | |
| "learning_rate": 2.01819891734783e-07, | |
| "loss": 0.039, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 1.8202846975088969, | |
| "grad_norm": 1.265595129436382, | |
| "learning_rate": 1.979082251318576e-07, | |
| "loss": 0.0282, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 1.8220640569395017, | |
| "grad_norm": 1.4153250679538338, | |
| "learning_rate": 1.9403407286770592e-07, | |
| "loss": 0.0369, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 1.8238434163701067, | |
| "grad_norm": 1.3781818620563666, | |
| "learning_rate": 1.9019746520752502e-07, | |
| "loss": 0.0381, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.8256227758007118, | |
| "grad_norm": 1.718512200773955, | |
| "learning_rate": 1.8639843212321206e-07, | |
| "loss": 0.0403, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 1.8274021352313166, | |
| "grad_norm": 1.299058506626459, | |
| "learning_rate": 1.826370032931285e-07, | |
| "loss": 0.036, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 1.8291814946619218, | |
| "grad_norm": 1.6276872788797643, | |
| "learning_rate": 1.789132081018674e-07, | |
| "loss": 0.0537, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 1.8309608540925266, | |
| "grad_norm": 1.3298178653795178, | |
| "learning_rate": 1.7522707564002706e-07, | |
| "loss": 0.0323, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 1.8327402135231317, | |
| "grad_norm": 1.855773721049355, | |
| "learning_rate": 1.7157863470397718e-07, | |
| "loss": 0.0317, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.8345195729537367, | |
| "grad_norm": 1.4343753664257393, | |
| "learning_rate": 1.6796791379564138e-07, | |
| "loss": 0.0309, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 1.8362989323843415, | |
| "grad_norm": 1.5438212727690983, | |
| "learning_rate": 1.6439494112227173e-07, | |
| "loss": 0.0372, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 1.8380782918149468, | |
| "grad_norm": 1.720044473115758, | |
| "learning_rate": 1.6085974459622567e-07, | |
| "loss": 0.0465, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 1.8398576512455516, | |
| "grad_norm": 1.3342590877467573, | |
| "learning_rate": 1.573623518347517e-07, | |
| "loss": 0.0336, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 1.8416370106761566, | |
| "grad_norm": 1.4515084770098712, | |
| "learning_rate": 1.5390279015977117e-07, | |
| "loss": 0.0407, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.8434163701067616, | |
| "grad_norm": 1.3336644568591236, | |
| "learning_rate": 1.5048108659766693e-07, | |
| "loss": 0.0379, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 1.8451957295373664, | |
| "grad_norm": 1.5172439501449795, | |
| "learning_rate": 1.470972678790711e-07, | |
| "loss": 0.0449, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 1.8469750889679717, | |
| "grad_norm": 1.5581006598794573, | |
| "learning_rate": 1.437513604386559e-07, | |
| "loss": 0.0396, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 1.8487544483985765, | |
| "grad_norm": 1.530422392615402, | |
| "learning_rate": 1.404433904149266e-07, | |
| "loss": 0.0385, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.8505338078291815, | |
| "grad_norm": 1.48114575580019, | |
| "learning_rate": 1.3717338365001943e-07, | |
| "loss": 0.0389, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.8523131672597866, | |
| "grad_norm": 1.742749223954352, | |
| "learning_rate": 1.3394136568949834e-07, | |
| "loss": 0.0503, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 1.8540925266903914, | |
| "grad_norm": 1.569008405149613, | |
| "learning_rate": 1.307473617821553e-07, | |
| "loss": 0.0507, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 1.8558718861209964, | |
| "grad_norm": 2.065604662397892, | |
| "learning_rate": 1.275913968798137e-07, | |
| "loss": 0.0529, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 1.8576512455516014, | |
| "grad_norm": 1.4589821077191103, | |
| "learning_rate": 1.2447349563713186e-07, | |
| "loss": 0.0407, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.8594306049822062, | |
| "grad_norm": 1.4219623281496945, | |
| "learning_rate": 1.213936824114137e-07, | |
| "loss": 0.0308, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.8612099644128115, | |
| "grad_norm": 1.9874070729938271, | |
| "learning_rate": 1.1835198126241509e-07, | |
| "loss": 0.0445, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 1.8629893238434163, | |
| "grad_norm": 2.122007557527852, | |
| "learning_rate": 1.1534841595215617e-07, | |
| "loss": 0.0442, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 1.8647686832740213, | |
| "grad_norm": 1.1931204369385333, | |
| "learning_rate": 1.1238300994473983e-07, | |
| "loss": 0.0302, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 1.8665480427046264, | |
| "grad_norm": 1.513122014631756, | |
| "learning_rate": 1.0945578640616183e-07, | |
| "loss": 0.0525, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 1.8683274021352312, | |
| "grad_norm": 1.5519865700437034, | |
| "learning_rate": 1.0656676820413603e-07, | |
| "loss": 0.0358, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.8701067615658364, | |
| "grad_norm": 1.4271183827174334, | |
| "learning_rate": 1.0371597790791166e-07, | |
| "loss": 0.0409, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 1.8718861209964412, | |
| "grad_norm": 1.643965589040426, | |
| "learning_rate": 1.0090343778809908e-07, | |
| "loss": 0.0491, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 1.8736654804270463, | |
| "grad_norm": 1.8310312707555436, | |
| "learning_rate": 9.812916981649433e-08, | |
| "loss": 0.0401, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 1.8754448398576513, | |
| "grad_norm": 1.6578569407330417, | |
| "learning_rate": 9.539319566590766e-08, | |
| "loss": 0.0398, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 1.8772241992882561, | |
| "grad_norm": 1.7392343103257353, | |
| "learning_rate": 9.269553670999743e-08, | |
| "loss": 0.0424, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.8790035587188612, | |
| "grad_norm": 1.7454675157902426, | |
| "learning_rate": 9.003621402309815e-08, | |
| "loss": 0.0386, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 1.8807829181494662, | |
| "grad_norm": 1.803814566407453, | |
| "learning_rate": 8.741524838005888e-08, | |
| "loss": 0.038, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 1.8825622775800712, | |
| "grad_norm": 1.6575111639928062, | |
| "learning_rate": 8.483266025608061e-08, | |
| "loss": 0.0482, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 1.8843416370106763, | |
| "grad_norm": 1.2224749902025194, | |
| "learning_rate": 8.228846982655525e-08, | |
| "loss": 0.0297, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 1.886120996441281, | |
| "grad_norm": 2.00204445186587, | |
| "learning_rate": 7.978269696691021e-08, | |
| "loss": 0.0523, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.887900355871886, | |
| "grad_norm": 1.7306623324895363, | |
| "learning_rate": 7.731536125244965e-08, | |
| "loss": 0.0421, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 1.8896797153024911, | |
| "grad_norm": 1.7131523945031208, | |
| "learning_rate": 7.488648195820513e-08, | |
| "loss": 0.0526, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 1.891459074733096, | |
| "grad_norm": 1.7238410801029755, | |
| "learning_rate": 7.249607805878245e-08, | |
| "loss": 0.0325, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 1.8932384341637012, | |
| "grad_norm": 1.3678173526850477, | |
| "learning_rate": 7.014416822821557e-08, | |
| "loss": 0.0248, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 1.895017793594306, | |
| "grad_norm": 1.27011801830683, | |
| "learning_rate": 6.783077083981793e-08, | |
| "loss": 0.0314, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.896797153024911, | |
| "grad_norm": 1.8805708313276215, | |
| "learning_rate": 6.55559039660425e-08, | |
| "loss": 0.0537, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 1.898576512455516, | |
| "grad_norm": 1.7469676732724069, | |
| "learning_rate": 6.331958537833693e-08, | |
| "loss": 0.0364, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 1.9003558718861209, | |
| "grad_norm": 1.2633528618797971, | |
| "learning_rate": 6.112183254700866e-08, | |
| "loss": 0.032, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 1.9021352313167261, | |
| "grad_norm": 1.2127577777024772, | |
| "learning_rate": 5.8962662641083856e-08, | |
| "loss": 0.0237, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 1.903914590747331, | |
| "grad_norm": 1.4129750870511029, | |
| "learning_rate": 5.6842092528176516e-08, | |
| "loss": 0.0397, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.905693950177936, | |
| "grad_norm": 1.372404896515472, | |
| "learning_rate": 5.476013877435626e-08, | |
| "loss": 0.0355, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 1.907473309608541, | |
| "grad_norm": 1.6343034786652952, | |
| "learning_rate": 5.271681764401848e-08, | |
| "loss": 0.0563, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 1.9092526690391458, | |
| "grad_norm": 1.3618732836689813, | |
| "learning_rate": 5.071214509975775e-08, | |
| "loss": 0.0417, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 1.9110320284697508, | |
| "grad_norm": 1.8449529172109376, | |
| "learning_rate": 4.8746136802240716e-08, | |
| "loss": 0.0568, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 1.9128113879003559, | |
| "grad_norm": 1.8807573324899285, | |
| "learning_rate": 4.6818808110087875e-08, | |
| "loss": 0.0453, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.914590747330961, | |
| "grad_norm": 1.714253162855563, | |
| "learning_rate": 4.493017407975087e-08, | |
| "loss": 0.0632, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 1.916370106761566, | |
| "grad_norm": 1.8957743709260997, | |
| "learning_rate": 4.308024946539424e-08, | |
| "loss": 0.0444, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 1.9181494661921707, | |
| "grad_norm": 1.6567956649623439, | |
| "learning_rate": 4.1269048718783344e-08, | |
| "loss": 0.0362, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 1.9199288256227758, | |
| "grad_norm": 1.5167388864095845, | |
| "learning_rate": 3.9496585989167726e-08, | |
| "loss": 0.0376, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 1.9217081850533808, | |
| "grad_norm": 1.528227497972969, | |
| "learning_rate": 3.776287512317345e-08, | |
| "loss": 0.0469, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.9234875444839856, | |
| "grad_norm": 1.5847079828922794, | |
| "learning_rate": 3.606792966469375e-08, | |
| "loss": 0.031, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 1.9252669039145909, | |
| "grad_norm": 1.2317473860643373, | |
| "learning_rate": 3.4411762854782426e-08, | |
| "loss": 0.0367, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 1.9270462633451957, | |
| "grad_norm": 1.5927918914675636, | |
| "learning_rate": 3.279438763155174e-08, | |
| "loss": 0.0331, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 1.9288256227758007, | |
| "grad_norm": 1.4951441317298986, | |
| "learning_rate": 3.121581663007134e-08, | |
| "loss": 0.043, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 1.9306049822064058, | |
| "grad_norm": 1.1080721001577736, | |
| "learning_rate": 2.967606218226837e-08, | |
| "loss": 0.0312, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.9323843416370106, | |
| "grad_norm": 1.3676350931908878, | |
| "learning_rate": 2.8175136316832e-08, | |
| "loss": 0.0405, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 1.9341637010676158, | |
| "grad_norm": 1.6079602037518028, | |
| "learning_rate": 2.6713050759120117e-08, | |
| "loss": 0.0411, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 1.9359430604982206, | |
| "grad_norm": 1.7339251064517849, | |
| "learning_rate": 2.528981693106558e-08, | |
| "loss": 0.0466, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 1.9377224199288257, | |
| "grad_norm": 1.818344682500453, | |
| "learning_rate": 2.3905445951089013e-08, | |
| "loss": 0.0584, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 1.9395017793594307, | |
| "grad_norm": 1.744417521321875, | |
| "learning_rate": 2.2559948634011673e-08, | |
| "loss": 0.0437, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.9412811387900355, | |
| "grad_norm": 1.5382364425251263, | |
| "learning_rate": 2.125333549096942e-08, | |
| "loss": 0.0364, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 1.9430604982206405, | |
| "grad_norm": 1.4302711260083396, | |
| "learning_rate": 1.9985616729332747e-08, | |
| "loss": 0.0381, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 1.9448398576512456, | |
| "grad_norm": 1.3977794866496016, | |
| "learning_rate": 1.8756802252625773e-08, | |
| "loss": 0.0332, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 1.9466192170818504, | |
| "grad_norm": 1.6475591860935994, | |
| "learning_rate": 1.75669016604485e-08, | |
| "loss": 0.0524, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 1.9483985765124556, | |
| "grad_norm": 1.347916794977455, | |
| "learning_rate": 1.6415924248403547e-08, | |
| "loss": 0.0317, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.9501779359430604, | |
| "grad_norm": 1.4159572041864439, | |
| "learning_rate": 1.5303879008021773e-08, | |
| "loss": 0.04, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.9519572953736655, | |
| "grad_norm": 1.8267451048695906, | |
| "learning_rate": 1.4230774626691756e-08, | |
| "loss": 0.0347, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 1.9537366548042705, | |
| "grad_norm": 1.4403126185026325, | |
| "learning_rate": 1.3196619487594875e-08, | |
| "loss": 0.0329, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 1.9555160142348753, | |
| "grad_norm": 1.4349197249214471, | |
| "learning_rate": 1.2201421669636448e-08, | |
| "loss": 0.035, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 1.9572953736654806, | |
| "grad_norm": 2.0370873490560344, | |
| "learning_rate": 1.1245188947384133e-08, | |
| "loss": 0.045, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.9590747330960854, | |
| "grad_norm": 1.420798190137548, | |
| "learning_rate": 1.0327928791006858e-08, | |
| "loss": 0.0368, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 1.9608540925266904, | |
| "grad_norm": 1.449672057318401, | |
| "learning_rate": 9.449648366217645e-09, | |
| "loss": 0.0498, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 1.9626334519572954, | |
| "grad_norm": 1.4076625197907682, | |
| "learning_rate": 8.61035453421588e-09, | |
| "loss": 0.0355, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 1.9644128113879002, | |
| "grad_norm": 1.4685046640726878, | |
| "learning_rate": 7.81005385163458e-09, | |
| "loss": 0.0404, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 1.9661921708185055, | |
| "grad_norm": 1.371933027522928, | |
| "learning_rate": 7.048752570488205e-09, | |
| "loss": 0.0409, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.9679715302491103, | |
| "grad_norm": 2.016497856941647, | |
| "learning_rate": 6.326456638125478e-09, | |
| "loss": 0.063, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 1.9697508896797153, | |
| "grad_norm": 1.576290521878247, | |
| "learning_rate": 5.643171697183314e-09, | |
| "loss": 0.0407, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 1.9715302491103204, | |
| "grad_norm": 1.7466033853572596, | |
| "learning_rate": 4.998903085539075e-09, | |
| "loss": 0.0375, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 1.9733096085409252, | |
| "grad_norm": 1.3476478092234006, | |
| "learning_rate": 4.393655836272825e-09, | |
| "loss": 0.0338, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 1.9750889679715302, | |
| "grad_norm": 1.666024727327656, | |
| "learning_rate": 3.8274346776262514e-09, | |
| "loss": 0.0406, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.9768683274021353, | |
| "grad_norm": 1.4827254876172484, | |
| "learning_rate": 3.300244032966582e-09, | |
| "loss": 0.0349, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 1.97864768683274, | |
| "grad_norm": 1.5513686345362376, | |
| "learning_rate": 2.8120880207493928e-09, | |
| "loss": 0.0387, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 1.9804270462633453, | |
| "grad_norm": 1.4113814355549943, | |
| "learning_rate": 2.362970454491409e-09, | |
| "loss": 0.0306, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 1.9822064056939501, | |
| "grad_norm": 1.8372886903366759, | |
| "learning_rate": 1.952894842735531e-09, | |
| "loss": 0.0448, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 1.9839857651245552, | |
| "grad_norm": 0.9186204717079001, | |
| "learning_rate": 1.5818643890258555e-09, | |
| "loss": 0.0282, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.9857651245551602, | |
| "grad_norm": 1.6321989839260644, | |
| "learning_rate": 1.2498819918843609e-09, | |
| "loss": 0.0358, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 1.987544483985765, | |
| "grad_norm": 1.8770605870017814, | |
| "learning_rate": 9.569502447837053e-10, | |
| "loss": 0.0454, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 1.9893238434163703, | |
| "grad_norm": 1.719347137159773, | |
| "learning_rate": 7.03071436131686e-10, | |
| "loss": 0.0466, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 1.991103202846975, | |
| "grad_norm": 1.1223209242509902, | |
| "learning_rate": 4.882475492506977e-10, | |
| "loss": 0.0296, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 1.99288256227758, | |
| "grad_norm": 2.0212017603962753, | |
| "learning_rate": 3.124802623627465e-10, | |
| "loss": 0.0564, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.9946619217081851, | |
| "grad_norm": 1.6838975415280004, | |
| "learning_rate": 1.7577094857557097e-10, | |
| "loss": 0.0504, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 1.99644128113879, | |
| "grad_norm": 1.7985556873188533, | |
| "learning_rate": 7.812067587487093e-11, | |
| "loss": 0.0513, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 1.998220640569395, | |
| "grad_norm": 1.2838644811176287, | |
| "learning_rate": 1.9530207111539967e-11, | |
| "loss": 0.0333, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0960112206337198, | |
| "learning_rate": 0.0, | |
| "loss": 0.0323, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 1124, | |
| "total_flos": 4574545231872.0, | |
| "train_loss": 0.09024254273079681, | |
| "train_runtime": 852.4204, | |
| "train_samples_per_second": 10.542, | |
| "train_steps_per_second": 1.319 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1124, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4574545231872.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |