| { | |
| "best_metric": 1.5939216613769531, | |
| "best_model_checkpoint": "opus-mt-zh-en-finetuned-zhen-checkpoints/checkpoint-25985", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 259850, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.777989149093628, | |
| "learning_rate": 1.9961593226861654e-05, | |
| "loss": 0.608, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.571350336074829, | |
| "learning_rate": 1.9923109486242066e-05, | |
| "loss": 0.6031, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.669718265533447, | |
| "learning_rate": 1.9884625745622478e-05, | |
| "loss": 0.5823, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.766803503036499, | |
| "learning_rate": 1.984614200500289e-05, | |
| "loss": 0.625, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.617687225341797, | |
| "learning_rate": 1.9807735231864538e-05, | |
| "loss": 0.6324, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 4.26535177230835, | |
| "learning_rate": 1.976932845872619e-05, | |
| "loss": 0.6184, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 5.035365581512451, | |
| "learning_rate": 1.9730844718106602e-05, | |
| "loss": 0.6277, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 6.626140594482422, | |
| "learning_rate": 1.9692360977487014e-05, | |
| "loss": 0.6557, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 3.718008279800415, | |
| "learning_rate": 1.9653877236867426e-05, | |
| "loss": 0.6623, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 4.26861047744751, | |
| "learning_rate": 1.9615470463729075e-05, | |
| "loss": 0.6719, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 5.266055107116699, | |
| "learning_rate": 1.9576986723109487e-05, | |
| "loss": 0.636, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 6.086512565612793, | |
| "learning_rate": 1.95385029824899e-05, | |
| "loss": 0.6585, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 4.0940327644348145, | |
| "learning_rate": 1.950001924187031e-05, | |
| "loss": 0.6341, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 7.52503776550293, | |
| "learning_rate": 1.9461535501250722e-05, | |
| "loss": 0.6589, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 3.5400032997131348, | |
| "learning_rate": 1.9423051760631134e-05, | |
| "loss": 0.6627, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 3.525592565536499, | |
| "learning_rate": 1.9384568020011546e-05, | |
| "loss": 0.639, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 3.1940386295318604, | |
| "learning_rate": 1.9346084279391958e-05, | |
| "loss": 0.6433, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 5.362303733825684, | |
| "learning_rate": 1.930767750625361e-05, | |
| "loss": 0.6714, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 4.885318279266357, | |
| "learning_rate": 1.9269193765634022e-05, | |
| "loss": 0.6334, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 7.104074001312256, | |
| "learning_rate": 1.9230710025014434e-05, | |
| "loss": 0.6496, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.086849689483643, | |
| "learning_rate": 1.9192226284394843e-05, | |
| "loss": 0.6647, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 10.493572235107422, | |
| "learning_rate": 1.9153819511256495e-05, | |
| "loss": 0.6609, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 10.206048011779785, | |
| "learning_rate": 1.9115335770636907e-05, | |
| "loss": 0.6806, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.489506721496582, | |
| "learning_rate": 1.907685203001732e-05, | |
| "loss": 0.641, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 4.525002479553223, | |
| "learning_rate": 1.903844525687897e-05, | |
| "loss": 0.6863, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.507730007171631, | |
| "learning_rate": 1.9000038483740623e-05, | |
| "loss": 0.6829, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.827548980712891, | |
| "learning_rate": 1.8961554743121035e-05, | |
| "loss": 0.6597, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.409604072570801, | |
| "learning_rate": 1.8923071002501447e-05, | |
| "loss": 0.6445, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 9.57291316986084, | |
| "learning_rate": 1.8884664229363095e-05, | |
| "loss": 0.6882, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 9.129105567932129, | |
| "learning_rate": 1.8846180488743507e-05, | |
| "loss": 0.6719, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 8.071757316589355, | |
| "learning_rate": 1.880769674812392e-05, | |
| "loss": 0.6668, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 3.636784791946411, | |
| "learning_rate": 1.876921300750433e-05, | |
| "loss": 0.6773, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 6.6433515548706055, | |
| "learning_rate": 1.8730729266884743e-05, | |
| "loss": 0.6776, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 3.0176546573638916, | |
| "learning_rate": 1.8692245526265155e-05, | |
| "loss": 0.6587, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 7.913615703582764, | |
| "learning_rate": 1.8653761785645567e-05, | |
| "loss": 0.6665, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 5.335203647613525, | |
| "learning_rate": 1.861527804502598e-05, | |
| "loss": 0.6684, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 5.51749324798584, | |
| "learning_rate": 1.857679430440639e-05, | |
| "loss": 0.6577, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.254025936126709, | |
| "learning_rate": 1.8538310563786802e-05, | |
| "loss": 0.6776, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.3869874477386475, | |
| "learning_rate": 1.8499826823167214e-05, | |
| "loss": 0.6351, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 12.704727172851562, | |
| "learning_rate": 1.8461343082547626e-05, | |
| "loss": 0.6904, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 3.478996515274048, | |
| "learning_rate": 1.8422859341928038e-05, | |
| "loss": 0.6829, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 6.513260841369629, | |
| "learning_rate": 1.838437560130845e-05, | |
| "loss": 0.6615, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 5.242546081542969, | |
| "learning_rate": 1.8345891860688862e-05, | |
| "loss": 0.6859, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 6.638239860534668, | |
| "learning_rate": 1.8307408120069274e-05, | |
| "loss": 0.6485, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 8.32187557220459, | |
| "learning_rate": 1.8269001346930923e-05, | |
| "loss": 0.6567, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 3.9929823875427246, | |
| "learning_rate": 1.8230517606311334e-05, | |
| "loss": 0.6996, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.426717281341553, | |
| "learning_rate": 1.8192110833172987e-05, | |
| "loss": 0.6529, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 7.328232288360596, | |
| "learning_rate": 1.81536270925534e-05, | |
| "loss": 0.6711, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 6.648090362548828, | |
| "learning_rate": 1.811514335193381e-05, | |
| "loss": 0.7026, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.047011375427246, | |
| "learning_rate": 1.8076659611314222e-05, | |
| "loss": 0.6867, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 10.354930877685547, | |
| "learning_rate": 1.803825283817587e-05, | |
| "loss": 0.6718, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.5939216613769531, | |
| "eval_runtime": 11.3331, | |
| "eval_samples_per_second": 351.271, | |
| "eval_steps_per_second": 87.884, | |
| "step": 25985 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.946779251098633, | |
| "learning_rate": 1.7999846065037523e-05, | |
| "loss": 0.6928, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 4.127852916717529, | |
| "learning_rate": 1.7961362324417935e-05, | |
| "loss": 0.5471, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 8.648958206176758, | |
| "learning_rate": 1.7922878583798347e-05, | |
| "loss": 0.5216, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 3.198063611984253, | |
| "learning_rate": 1.788439484317876e-05, | |
| "loss": 0.5649, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 7.709962368011475, | |
| "learning_rate": 1.784591110255917e-05, | |
| "loss": 0.5739, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.456827402114868, | |
| "learning_rate": 1.7807427361939583e-05, | |
| "loss": 0.549, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 5.343823432922363, | |
| "learning_rate": 1.7768943621319995e-05, | |
| "loss": 0.5624, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 6.258249759674072, | |
| "learning_rate": 1.7730459880700407e-05, | |
| "loss": 0.5497, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 3.7241768836975098, | |
| "learning_rate": 1.769197614008082e-05, | |
| "loss": 0.5422, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 4.565840244293213, | |
| "learning_rate": 1.765349239946123e-05, | |
| "loss": 0.5805, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 6.31953763961792, | |
| "learning_rate": 1.7615008658841642e-05, | |
| "loss": 0.5889, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.6492664813995361, | |
| "learning_rate": 1.7576524918222054e-05, | |
| "loss": 0.5732, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 6.656198978424072, | |
| "learning_rate": 1.7538041177602466e-05, | |
| "loss": 0.5499, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 4.865792274475098, | |
| "learning_rate": 1.7499557436982878e-05, | |
| "loss": 0.5519, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 3.8847908973693848, | |
| "learning_rate": 1.746107369636329e-05, | |
| "loss": 0.5695, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 9.86506462097168, | |
| "learning_rate": 1.74225899557437e-05, | |
| "loss": 0.5877, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 5.732994079589844, | |
| "learning_rate": 1.738418318260535e-05, | |
| "loss": 0.5629, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 2.4092957973480225, | |
| "learning_rate": 1.7345699441985762e-05, | |
| "loss": 0.5734, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 4.129636764526367, | |
| "learning_rate": 1.7307215701366174e-05, | |
| "loss": 0.5931, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 4.32247257232666, | |
| "learning_rate": 1.7268731960746586e-05, | |
| "loss": 0.5703, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 4.780704975128174, | |
| "learning_rate": 1.7230325187608238e-05, | |
| "loss": 0.5978, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 5.10893440246582, | |
| "learning_rate": 1.719184144698865e-05, | |
| "loss": 0.57, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 3.899763584136963, | |
| "learning_rate": 1.71534346738503e-05, | |
| "loss": 0.5778, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 6.0712480545043945, | |
| "learning_rate": 1.711495093323071e-05, | |
| "loss": 0.5744, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 9.512677192687988, | |
| "learning_rate": 1.7076467192611123e-05, | |
| "loss": 0.5825, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 5.049200534820557, | |
| "learning_rate": 1.7037983451991535e-05, | |
| "loss": 0.6039, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 5.021922588348389, | |
| "learning_rate": 1.6999499711371947e-05, | |
| "loss": 0.5708, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 3.546837568283081, | |
| "learning_rate": 1.696101597075236e-05, | |
| "loss": 0.5808, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 4.707235813140869, | |
| "learning_rate": 1.692253223013277e-05, | |
| "loss": 0.5894, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 8.543360710144043, | |
| "learning_rate": 1.688412545699442e-05, | |
| "loss": 0.5828, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 5.515990257263184, | |
| "learning_rate": 1.684564171637483e-05, | |
| "loss": 0.6009, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 8.573718070983887, | |
| "learning_rate": 1.6807234943236483e-05, | |
| "loss": 0.5796, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 3.3873181343078613, | |
| "learning_rate": 1.6768751202616895e-05, | |
| "loss": 0.6052, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 5.747949123382568, | |
| "learning_rate": 1.6730344429478547e-05, | |
| "loss": 0.5998, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 4.767616271972656, | |
| "learning_rate": 1.669186068885896e-05, | |
| "loss": 0.631, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 5.872443199157715, | |
| "learning_rate": 1.665337694823937e-05, | |
| "loss": 0.5903, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 5.379058837890625, | |
| "learning_rate": 1.6614893207619783e-05, | |
| "loss": 0.6141, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 7.023738384246826, | |
| "learning_rate": 1.6576409467000195e-05, | |
| "loss": 0.601, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 7.2762932777404785, | |
| "learning_rate": 1.6537925726380607e-05, | |
| "loss": 0.614, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 5.506533145904541, | |
| "learning_rate": 1.649944198576102e-05, | |
| "loss": 0.6156, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 4.782810688018799, | |
| "learning_rate": 1.646095824514143e-05, | |
| "loss": 0.5983, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 8.239398002624512, | |
| "learning_rate": 1.6422474504521842e-05, | |
| "loss": 0.6115, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 5.413235664367676, | |
| "learning_rate": 1.6383990763902254e-05, | |
| "loss": 0.6157, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 5.1087446212768555, | |
| "learning_rate": 1.6345507023282666e-05, | |
| "loss": 0.6225, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 10.661508560180664, | |
| "learning_rate": 1.6307023282663078e-05, | |
| "loss": 0.6185, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 5.0764312744140625, | |
| "learning_rate": 1.626853954204349e-05, | |
| "loss": 0.5878, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 6.266598224639893, | |
| "learning_rate": 1.6230055801423902e-05, | |
| "loss": 0.6057, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 5.597410678863525, | |
| "learning_rate": 1.6191572060804314e-05, | |
| "loss": 0.591, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 5.9798970222473145, | |
| "learning_rate": 1.6153088320184722e-05, | |
| "loss": 0.6114, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 5.151573181152344, | |
| "learning_rate": 1.6114604579565134e-05, | |
| "loss": 0.5875, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 3.2873849868774414, | |
| "learning_rate": 1.6076197806426786e-05, | |
| "loss": 0.6393, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 3.8436408042907715, | |
| "learning_rate": 1.6037714065807198e-05, | |
| "loss": 0.6202, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.6332776546478271, | |
| "eval_runtime": 11.3148, | |
| "eval_samples_per_second": 351.839, | |
| "eval_steps_per_second": 88.026, | |
| "step": 51970 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 11.219087600708008, | |
| "learning_rate": 1.599923032518761e-05, | |
| "loss": 0.6083, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 5.6387038230896, | |
| "learning_rate": 1.5960746584568022e-05, | |
| "loss": 0.5929, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 10.761116981506348, | |
| "learning_rate": 1.592241677891091e-05, | |
| "loss": 0.5913, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 4.182568073272705, | |
| "learning_rate": 1.5883933038291323e-05, | |
| "loss": 0.5983, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.4628002643585205, | |
| "learning_rate": 1.5845449297671735e-05, | |
| "loss": 0.5789, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 6.6901535987854, | |
| "learning_rate": 1.5806965557052147e-05, | |
| "loss": 0.5827, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 4.275672912597656, | |
| "learning_rate": 1.576848181643256e-05, | |
| "loss": 0.593, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 6.460977554321289, | |
| "learning_rate": 1.572999807581297e-05, | |
| "loss": 0.6012, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 6.66805362701416, | |
| "learning_rate": 1.5691514335193382e-05, | |
| "loss": 0.6148, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 9.386066436767578, | |
| "learning_rate": 1.5653030594573794e-05, | |
| "loss": 0.6147, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 4.505888938903809, | |
| "learning_rate": 1.5614700788916683e-05, | |
| "loss": 0.6, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 6.781282424926758, | |
| "learning_rate": 1.5576217048297095e-05, | |
| "loss": 0.5938, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 4.20095157623291, | |
| "learning_rate": 1.5537810275158747e-05, | |
| "loss": 0.6118, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 6.635080337524414, | |
| "learning_rate": 1.549932653453916e-05, | |
| "loss": 0.6375, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 5.1331329345703125, | |
| "learning_rate": 1.546084279391957e-05, | |
| "loss": 0.625, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 4.725217342376709, | |
| "learning_rate": 1.5422359053299983e-05, | |
| "loss": 0.6443, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 6.601204872131348, | |
| "learning_rate": 1.5383875312680395e-05, | |
| "loss": 0.6206, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 16.34359359741211, | |
| "learning_rate": 1.5345391572060807e-05, | |
| "loss": 0.6144, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 6.883621692657471, | |
| "learning_rate": 1.530690783144122e-05, | |
| "loss": 0.6112, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 5.42087984085083, | |
| "learning_rate": 1.526842409082163e-05, | |
| "loss": 0.6286, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 7.191906452178955, | |
| "learning_rate": 1.522994035020204e-05, | |
| "loss": 0.6103, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 8.114423751831055, | |
| "learning_rate": 1.5191456609582453e-05, | |
| "loss": 0.6464, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 6.188645839691162, | |
| "learning_rate": 1.5152972868962865e-05, | |
| "loss": 0.5874, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 5.572692394256592, | |
| "learning_rate": 1.5114489128343276e-05, | |
| "loss": 0.6235, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 8.618267059326172, | |
| "learning_rate": 1.5076082355204927e-05, | |
| "loss": 0.6144, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 6.580616474151611, | |
| "learning_rate": 1.5037598614585339e-05, | |
| "loss": 0.6544, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 7.99381160736084, | |
| "learning_rate": 1.499919184144699e-05, | |
| "loss": 0.6281, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 4.514256954193115, | |
| "learning_rate": 1.4960708100827401e-05, | |
| "loss": 0.6295, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 8.09423828125, | |
| "learning_rate": 1.4922224360207813e-05, | |
| "loss": 0.6332, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 5.654535293579102, | |
| "learning_rate": 1.4883740619588225e-05, | |
| "loss": 0.6573, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 8.867722511291504, | |
| "learning_rate": 1.4845256878968637e-05, | |
| "loss": 0.6331, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 11.65078353881836, | |
| "learning_rate": 1.4806773138349049e-05, | |
| "loss": 0.6522, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 4.735153675079346, | |
| "learning_rate": 1.476828939772946e-05, | |
| "loss": 0.628, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 10.033109664916992, | |
| "learning_rate": 1.4729805657109873e-05, | |
| "loss": 0.6588, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 5.765382289886475, | |
| "learning_rate": 1.4691398883971523e-05, | |
| "loss": 0.657, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 8.563407897949219, | |
| "learning_rate": 1.4652915143351935e-05, | |
| "loss": 0.6476, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 5.050264358520508, | |
| "learning_rate": 1.4614431402732347e-05, | |
| "loss": 0.643, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 4.829985618591309, | |
| "learning_rate": 1.4575947662112759e-05, | |
| "loss": 0.6637, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 10.406055450439453, | |
| "learning_rate": 1.453754088897441e-05, | |
| "loss": 0.6588, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 5.833386421203613, | |
| "learning_rate": 1.4499057148354821e-05, | |
| "loss": 0.6443, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 5.603066921234131, | |
| "learning_rate": 1.4460573407735233e-05, | |
| "loss": 0.6382, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 6.070736885070801, | |
| "learning_rate": 1.4422166634596883e-05, | |
| "loss": 0.654, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 4.568108081817627, | |
| "learning_rate": 1.4383682893977295e-05, | |
| "loss": 0.6496, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 5.071454048156738, | |
| "learning_rate": 1.4345199153357707e-05, | |
| "loss": 0.6412, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 9.910784721374512, | |
| "learning_rate": 1.4306715412738119e-05, | |
| "loss": 0.6329, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 4.771623134613037, | |
| "learning_rate": 1.4268231672118531e-05, | |
| "loss": 0.6661, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 12.325662612915039, | |
| "learning_rate": 1.4229747931498943e-05, | |
| "loss": 0.6389, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 4.688502311706543, | |
| "learning_rate": 1.4191341158360595e-05, | |
| "loss": 0.6475, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 7.340636730194092, | |
| "learning_rate": 1.4152857417741005e-05, | |
| "loss": 0.678, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 11.585808753967285, | |
| "learning_rate": 1.4114373677121417e-05, | |
| "loss": 0.6422, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 4.971041679382324, | |
| "learning_rate": 1.4075889936501829e-05, | |
| "loss": 0.6493, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 8.936264991760254, | |
| "learning_rate": 1.4037406195882241e-05, | |
| "loss": 0.6719, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.634682297706604, | |
| "eval_runtime": 11.4139, | |
| "eval_samples_per_second": 348.784, | |
| "eval_steps_per_second": 87.262, | |
| "step": 77955 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.788780212402344, | |
| "learning_rate": 1.3998922455262653e-05, | |
| "loss": 0.6236, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 8.352863311767578, | |
| "learning_rate": 1.3960438714643065e-05, | |
| "loss": 0.5519, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 9.03128433227539, | |
| "learning_rate": 1.3922031941504715e-05, | |
| "loss": 0.5617, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 8.761636734008789, | |
| "learning_rate": 1.3883548200885127e-05, | |
| "loss": 0.554, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 4.592207908630371, | |
| "learning_rate": 1.3845064460265539e-05, | |
| "loss": 0.5218, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 5.602917671203613, | |
| "learning_rate": 1.380658071964595e-05, | |
| "loss": 0.5432, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 10.146926879882812, | |
| "learning_rate": 1.3768096979026363e-05, | |
| "loss": 0.5332, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 7.7499494552612305, | |
| "learning_rate": 1.3729613238406773e-05, | |
| "loss": 0.5459, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 10.009146690368652, | |
| "learning_rate": 1.3691129497787185e-05, | |
| "loss": 0.5721, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 5.544896602630615, | |
| "learning_rate": 1.3652645757167597e-05, | |
| "loss": 0.5407, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 11.661575317382812, | |
| "learning_rate": 1.3614238984029249e-05, | |
| "loss": 0.5678, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 2.23122501373291, | |
| "learning_rate": 1.357575524340966e-05, | |
| "loss": 0.5577, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 2.9503915309906006, | |
| "learning_rate": 1.3537271502790073e-05, | |
| "loss": 0.568, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 5.630256175994873, | |
| "learning_rate": 1.3498787762170485e-05, | |
| "loss": 0.5524, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 5.076592922210693, | |
| "learning_rate": 1.3460304021550896e-05, | |
| "loss": 0.574, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 7.131458759307861, | |
| "learning_rate": 1.3421974215893786e-05, | |
| "loss": 0.5718, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 16.959537506103516, | |
| "learning_rate": 1.3383490475274197e-05, | |
| "loss": 0.5743, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 8.548674583435059, | |
| "learning_rate": 1.334500673465461e-05, | |
| "loss": 0.5747, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 16.18918800354004, | |
| "learning_rate": 1.3306522994035021e-05, | |
| "loss": 0.594, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 3.274156093597412, | |
| "learning_rate": 1.3268039253415433e-05, | |
| "loss": 0.5758, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 5.2989959716796875, | |
| "learning_rate": 1.3229555512795845e-05, | |
| "loss": 0.5658, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 3.634488821029663, | |
| "learning_rate": 1.3191148739657495e-05, | |
| "loss": 0.5572, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 4.335933208465576, | |
| "learning_rate": 1.3152664999037907e-05, | |
| "loss": 0.5654, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 7.001922607421875, | |
| "learning_rate": 1.311418125841832e-05, | |
| "loss": 0.5741, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 9.779496192932129, | |
| "learning_rate": 1.3075697517798731e-05, | |
| "loss": 0.5782, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 7.241322994232178, | |
| "learning_rate": 1.3037213777179143e-05, | |
| "loss": 0.585, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 10.485682487487793, | |
| "learning_rate": 1.2998730036559555e-05, | |
| "loss": 0.5743, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 5.369238376617432, | |
| "learning_rate": 1.2960246295939967e-05, | |
| "loss": 0.5743, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 2.732351064682007, | |
| "learning_rate": 1.2921762555320379e-05, | |
| "loss": 0.5753, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 5.905006408691406, | |
| "learning_rate": 1.288335578218203e-05, | |
| "loss": 0.5816, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 6.421899795532227, | |
| "learning_rate": 1.2844872041562441e-05, | |
| "loss": 0.5907, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 8.367085456848145, | |
| "learning_rate": 1.2806388300942853e-05, | |
| "loss": 0.6026, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 5.6347784996032715, | |
| "learning_rate": 1.2767904560323265e-05, | |
| "loss": 0.6028, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 4.885034084320068, | |
| "learning_rate": 1.2729497787184915e-05, | |
| "loss": 0.6062, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 9.47564697265625, | |
| "learning_rate": 1.2691014046565327e-05, | |
| "loss": 0.6116, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 4.893156051635742, | |
| "learning_rate": 1.2652530305945739e-05, | |
| "loss": 0.5769, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 7.7681074142456055, | |
| "learning_rate": 1.261404656532615e-05, | |
| "loss": 0.5922, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 9.778617858886719, | |
| "learning_rate": 1.2575639792187801e-05, | |
| "loss": 0.623, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 10.256739616394043, | |
| "learning_rate": 1.2537156051568213e-05, | |
| "loss": 0.5552, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 6.800660133361816, | |
| "learning_rate": 1.2498672310948625e-05, | |
| "loss": 0.5961, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 11.549945831298828, | |
| "learning_rate": 1.2460188570329037e-05, | |
| "loss": 0.5983, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 7.519701957702637, | |
| "learning_rate": 1.2421704829709449e-05, | |
| "loss": 0.5769, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 7.1441168785095215, | |
| "learning_rate": 1.23832980565711e-05, | |
| "loss": 0.6006, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 4.6727375984191895, | |
| "learning_rate": 1.2344814315951511e-05, | |
| "loss": 0.6096, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 10.61463737487793, | |
| "learning_rate": 1.2306407542813164e-05, | |
| "loss": 0.6088, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 11.35522747039795, | |
| "learning_rate": 1.2267923802193574e-05, | |
| "loss": 0.603, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 4.605931282043457, | |
| "learning_rate": 1.2229440061573986e-05, | |
| "loss": 0.5742, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 5.55601692199707, | |
| "learning_rate": 1.2190956320954398e-05, | |
| "loss": 0.6005, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 4.776731014251709, | |
| "learning_rate": 1.215247258033481e-05, | |
| "loss": 0.6167, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 6.725343227386475, | |
| "learning_rate": 1.2113988839715221e-05, | |
| "loss": 0.5845, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 2.1615917682647705, | |
| "learning_rate": 1.2075505099095633e-05, | |
| "loss": 0.6123, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 3.8668465614318848, | |
| "learning_rate": 1.2037021358476043e-05, | |
| "loss": 0.5945, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.6611100435256958, | |
| "eval_runtime": 11.217, | |
| "eval_samples_per_second": 354.908, | |
| "eval_steps_per_second": 88.794, | |
| "step": 103940 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.890031337738037, | |
| "learning_rate": 1.1998614585337696e-05, | |
| "loss": 0.5796, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 8.080361366271973, | |
| "learning_rate": 1.1960130844718107e-05, | |
| "loss": 0.5026, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 7.6675496101379395, | |
| "learning_rate": 1.192164710409852e-05, | |
| "loss": 0.522, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 6.038026332855225, | |
| "learning_rate": 1.1883163363478931e-05, | |
| "loss": 0.491, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 2.374932050704956, | |
| "learning_rate": 1.1844756590340582e-05, | |
| "loss": 0.5123, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 1.8449337482452393, | |
| "learning_rate": 1.1806272849720994e-05, | |
| "loss": 0.5104, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 7.582018852233887, | |
| "learning_rate": 1.1767789109101406e-05, | |
| "loss": 0.525, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 6.123476028442383, | |
| "learning_rate": 1.1729305368481817e-05, | |
| "loss": 0.5276, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 6.717319011688232, | |
| "learning_rate": 1.1690898595343468e-05, | |
| "loss": 0.5204, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 5.3222336769104, | |
| "learning_rate": 1.165241485472388e-05, | |
| "loss": 0.5039, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 7.88124942779541, | |
| "learning_rate": 1.1613931114104292e-05, | |
| "loss": 0.5177, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 13.466960906982422, | |
| "learning_rate": 1.1575447373484704e-05, | |
| "loss": 0.5076, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 3.5181362628936768, | |
| "learning_rate": 1.1536963632865115e-05, | |
| "loss": 0.5227, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 9.861627578735352, | |
| "learning_rate": 1.1498479892245527e-05, | |
| "loss": 0.5297, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 9.91634464263916, | |
| "learning_rate": 1.1460073119107178e-05, | |
| "loss": 0.5425, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 8.588738441467285, | |
| "learning_rate": 1.142158937848759e-05, | |
| "loss": 0.5298, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 4.605578899383545, | |
| "learning_rate": 1.1383105637868002e-05, | |
| "loss": 0.5202, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 4.7749481201171875, | |
| "learning_rate": 1.1344621897248413e-05, | |
| "loss": 0.5262, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 6.725896835327148, | |
| "learning_rate": 1.1306215124110066e-05, | |
| "loss": 0.5012, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 5.475027561187744, | |
| "learning_rate": 1.1267731383490476e-05, | |
| "loss": 0.5077, | |
| "step": 113500 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 6.180666923522949, | |
| "learning_rate": 1.1229247642870888e-05, | |
| "loss": 0.514, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 6.444517135620117, | |
| "learning_rate": 1.119084086973254e-05, | |
| "loss": 0.5538, | |
| "step": 114500 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 6.281551361083984, | |
| "learning_rate": 1.1152357129112952e-05, | |
| "loss": 0.5441, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 9.76223373413086, | |
| "learning_rate": 1.1113873388493362e-05, | |
| "loss": 0.5512, | |
| "step": 115500 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 5.260191440582275, | |
| "learning_rate": 1.1075389647873774e-05, | |
| "loss": 0.5178, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 7.131434917449951, | |
| "learning_rate": 1.1036905907254186e-05, | |
| "loss": 0.5017, | |
| "step": 116500 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 6.9372992515563965, | |
| "learning_rate": 1.0998422166634598e-05, | |
| "loss": 0.5323, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 3.6769461631774902, | |
| "learning_rate": 1.095993842601501e-05, | |
| "loss": 0.5472, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 4.933724403381348, | |
| "learning_rate": 1.0921454685395421e-05, | |
| "loss": 0.5275, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 8.011011123657227, | |
| "learning_rate": 1.0882970944775832e-05, | |
| "loss": 0.5754, | |
| "step": 118500 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 5.938791275024414, | |
| "learning_rate": 1.0844564171637484e-05, | |
| "loss": 0.5244, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 5.401825904846191, | |
| "learning_rate": 1.0806080431017896e-05, | |
| "loss": 0.5601, | |
| "step": 119500 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 3.8994340896606445, | |
| "learning_rate": 1.0767596690398308e-05, | |
| "loss": 0.5399, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 7.102384090423584, | |
| "learning_rate": 1.072911294977872e-05, | |
| "loss": 0.5426, | |
| "step": 120500 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 4.785587787628174, | |
| "learning_rate": 1.069070617664037e-05, | |
| "loss": 0.541, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 3.6323719024658203, | |
| "learning_rate": 1.0652222436020782e-05, | |
| "loss": 0.5327, | |
| "step": 121500 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 3.2487778663635254, | |
| "learning_rate": 1.0613738695401194e-05, | |
| "loss": 0.5437, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 5.607531547546387, | |
| "learning_rate": 1.0575254954781606e-05, | |
| "loss": 0.5691, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 4.990538597106934, | |
| "learning_rate": 1.0536848181643256e-05, | |
| "loss": 0.5544, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 11.409914016723633, | |
| "learning_rate": 1.0498364441023668e-05, | |
| "loss": 0.5229, | |
| "step": 123500 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 5.91348123550415, | |
| "learning_rate": 1.045988070040408e-05, | |
| "loss": 0.5609, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 5.885928630828857, | |
| "learning_rate": 1.0421396959784492e-05, | |
| "loss": 0.5503, | |
| "step": 124500 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 3.7525486946105957, | |
| "learning_rate": 1.0382913219164904e-05, | |
| "loss": 0.5452, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 4.8420305252075195, | |
| "learning_rate": 1.0344506446026554e-05, | |
| "loss": 0.566, | |
| "step": 125500 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 5.222369194030762, | |
| "learning_rate": 1.0306022705406966e-05, | |
| "loss": 0.5481, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 6.775496006011963, | |
| "learning_rate": 1.0267538964787378e-05, | |
| "loss": 0.5353, | |
| "step": 126500 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 13.353194236755371, | |
| "learning_rate": 1.022905522416779e-05, | |
| "loss": 0.5574, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 6.453900337219238, | |
| "learning_rate": 1.0190648451029442e-05, | |
| "loss": 0.5715, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 9.484274864196777, | |
| "learning_rate": 1.0152241677891092e-05, | |
| "loss": 0.5678, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 5.4346160888671875, | |
| "learning_rate": 1.0113757937271504e-05, | |
| "loss": 0.5661, | |
| "step": 128500 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 7.368044376373291, | |
| "learning_rate": 1.0075274196651916e-05, | |
| "loss": 0.5524, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 5.8485941886901855, | |
| "learning_rate": 1.0036790456032328e-05, | |
| "loss": 0.5472, | |
| "step": 129500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.6937280893325806, | |
| "eval_runtime": 11.3508, | |
| "eval_samples_per_second": 350.726, | |
| "eval_steps_per_second": 87.747, | |
| "step": 129925 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 3.9238245487213135, | |
| "learning_rate": 9.998306715412738e-06, | |
| "loss": 0.5324, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 3.3814291954040527, | |
| "learning_rate": 9.95982297479315e-06, | |
| "loss": 0.4778, | |
| "step": 130500 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "grad_norm": 3.6909124851226807, | |
| "learning_rate": 9.921339234173562e-06, | |
| "loss": 0.4872, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 3.4291248321533203, | |
| "learning_rate": 9.882855493553974e-06, | |
| "loss": 0.4572, | |
| "step": 131500 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "grad_norm": 7.226670742034912, | |
| "learning_rate": 9.844448720415625e-06, | |
| "loss": 0.4887, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 3.1274802684783936, | |
| "learning_rate": 9.805964979796036e-06, | |
| "loss": 0.4687, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "grad_norm": 6.210909366607666, | |
| "learning_rate": 9.767558206657689e-06, | |
| "loss": 0.4695, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "grad_norm": 4.919899940490723, | |
| "learning_rate": 9.7290744660381e-06, | |
| "loss": 0.4697, | |
| "step": 133500 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 5.735771179199219, | |
| "learning_rate": 9.690590725418512e-06, | |
| "loss": 0.4663, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 4.188859462738037, | |
| "learning_rate": 9.652106984798924e-06, | |
| "loss": 0.4997, | |
| "step": 134500 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 6.6499786376953125, | |
| "learning_rate": 9.613623244179336e-06, | |
| "loss": 0.4713, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 5.279160499572754, | |
| "learning_rate": 9.575139503559748e-06, | |
| "loss": 0.4742, | |
| "step": 135500 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "grad_norm": 19.4254093170166, | |
| "learning_rate": 9.536655762940158e-06, | |
| "loss": 0.4906, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "grad_norm": 4.88319206237793, | |
| "learning_rate": 9.49817202232057e-06, | |
| "loss": 0.466, | |
| "step": 136500 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "grad_norm": 3.700117826461792, | |
| "learning_rate": 9.459688281700982e-06, | |
| "loss": 0.5211, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 5.820217609405518, | |
| "learning_rate": 9.421281508562632e-06, | |
| "loss": 0.5012, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 4.963629245758057, | |
| "learning_rate": 9.382797767943044e-06, | |
| "loss": 0.4893, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 3.2575323581695557, | |
| "learning_rate": 9.344314027323456e-06, | |
| "loss": 0.4881, | |
| "step": 138500 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "grad_norm": 4.5853424072265625, | |
| "learning_rate": 9.305830286703868e-06, | |
| "loss": 0.497, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "grad_norm": 9.991304397583008, | |
| "learning_rate": 9.26734654608428e-06, | |
| "loss": 0.5029, | |
| "step": 139500 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "grad_norm": 7.186006546020508, | |
| "learning_rate": 9.228862805464692e-06, | |
| "loss": 0.5146, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "grad_norm": 4.827492713928223, | |
| "learning_rate": 9.190456032326342e-06, | |
| "loss": 0.5155, | |
| "step": 140500 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 6.840179443359375, | |
| "learning_rate": 9.151972291706754e-06, | |
| "loss": 0.5204, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 3.26387357711792, | |
| "learning_rate": 9.113488551087166e-06, | |
| "loss": 0.4909, | |
| "step": 141500 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "grad_norm": 8.286077499389648, | |
| "learning_rate": 9.075004810467578e-06, | |
| "loss": 0.4973, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "grad_norm": 6.248400688171387, | |
| "learning_rate": 9.03659803732923e-06, | |
| "loss": 0.5213, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 4.121056079864502, | |
| "learning_rate": 8.998114296709642e-06, | |
| "loss": 0.5176, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 5.463682651519775, | |
| "learning_rate": 8.959630556090052e-06, | |
| "loss": 0.4707, | |
| "step": 143500 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 7.710265159606934, | |
| "learning_rate": 8.921146815470464e-06, | |
| "loss": 0.4728, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "grad_norm": 6.872325420379639, | |
| "learning_rate": 8.882740042332115e-06, | |
| "loss": 0.5183, | |
| "step": 144500 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "grad_norm": 7.6797614097595215, | |
| "learning_rate": 8.844256301712527e-06, | |
| "loss": 0.5065, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 4.729648590087891, | |
| "learning_rate": 8.805772561092938e-06, | |
| "loss": 0.5117, | |
| "step": 145500 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 6.701267719268799, | |
| "learning_rate": 8.76728882047335e-06, | |
| "loss": 0.5427, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "grad_norm": 8.45148754119873, | |
| "learning_rate": 8.728805079853762e-06, | |
| "loss": 0.4782, | |
| "step": 146500 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "grad_norm": 4.278749465942383, | |
| "learning_rate": 8.690398306715413e-06, | |
| "loss": 0.5093, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "grad_norm": 6.3357343673706055, | |
| "learning_rate": 8.651914566095825e-06, | |
| "loss": 0.5328, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 9.23112678527832, | |
| "learning_rate": 8.613430825476237e-06, | |
| "loss": 0.526, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "grad_norm": 2.7398366928100586, | |
| "learning_rate": 8.574947084856648e-06, | |
| "loss": 0.5004, | |
| "step": 148500 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 7.174339771270752, | |
| "learning_rate": 8.5365403117183e-06, | |
| "loss": 0.4947, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 6.2081427574157715, | |
| "learning_rate": 8.498056571098712e-06, | |
| "loss": 0.4961, | |
| "step": 149500 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "grad_norm": 5.464582443237305, | |
| "learning_rate": 8.459572830479124e-06, | |
| "loss": 0.4984, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "grad_norm": 4.864925384521484, | |
| "learning_rate": 8.421089089859536e-06, | |
| "loss": 0.5101, | |
| "step": 150500 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "grad_norm": 10.367297172546387, | |
| "learning_rate": 8.382682316721185e-06, | |
| "loss": 0.5004, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 4.001099109649658, | |
| "learning_rate": 8.344198576101597e-06, | |
| "loss": 0.5109, | |
| "step": 151500 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "grad_norm": 8.188924789428711, | |
| "learning_rate": 8.305714835482009e-06, | |
| "loss": 0.4988, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "grad_norm": 5.423083782196045, | |
| "learning_rate": 8.26723109486242e-06, | |
| "loss": 0.4957, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "grad_norm": 5.513723373413086, | |
| "learning_rate": 8.228824321724073e-06, | |
| "loss": 0.5113, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "grad_norm": 5.753915309906006, | |
| "learning_rate": 8.190340581104485e-06, | |
| "loss": 0.5088, | |
| "step": 153500 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "grad_norm": 5.672900676727295, | |
| "learning_rate": 8.151933807966135e-06, | |
| "loss": 0.5117, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 5.95, | |
| "grad_norm": 6.9963202476501465, | |
| "learning_rate": 8.113450067346547e-06, | |
| "loss": 0.5474, | |
| "step": 154500 | |
| }, | |
| { | |
| "epoch": 5.96, | |
| "grad_norm": 6.640045166015625, | |
| "learning_rate": 8.074966326726959e-06, | |
| "loss": 0.5135, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 5.076375961303711, | |
| "learning_rate": 8.036482586107371e-06, | |
| "loss": 0.5147, | |
| "step": 155500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.7142382860183716, | |
| "eval_runtime": 11.2855, | |
| "eval_samples_per_second": 352.753, | |
| "eval_steps_per_second": 88.255, | |
| "step": 155910 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 6.40647554397583, | |
| "learning_rate": 7.997998845487783e-06, | |
| "loss": 0.467, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 3.021810293197632, | |
| "learning_rate": 7.959515104868195e-06, | |
| "loss": 0.4311, | |
| "step": 156500 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 11.23317813873291, | |
| "learning_rate": 7.921031364248607e-06, | |
| "loss": 0.4569, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "grad_norm": 4.276322841644287, | |
| "learning_rate": 7.882547623629019e-06, | |
| "loss": 0.4523, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "grad_norm": 5.726410388946533, | |
| "learning_rate": 7.844140850490667e-06, | |
| "loss": 0.4601, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 8.776612281799316, | |
| "learning_rate": 7.80565710987108e-06, | |
| "loss": 0.4254, | |
| "step": 158500 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "grad_norm": 9.005372047424316, | |
| "learning_rate": 7.767173369251491e-06, | |
| "loss": 0.4328, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 6.14, | |
| "grad_norm": 9.345296859741211, | |
| "learning_rate": 7.728689628631903e-06, | |
| "loss": 0.4619, | |
| "step": 159500 | |
| }, | |
| { | |
| "epoch": 6.16, | |
| "grad_norm": 5.089489936828613, | |
| "learning_rate": 7.690282855493555e-06, | |
| "loss": 0.4546, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "grad_norm": 5.524288177490234, | |
| "learning_rate": 7.651799114873967e-06, | |
| "loss": 0.4444, | |
| "step": 160500 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 5.413343906402588, | |
| "learning_rate": 7.613315374254379e-06, | |
| "loss": 0.4672, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "grad_norm": 2.5175459384918213, | |
| "learning_rate": 7.5749086011160285e-06, | |
| "loss": 0.4534, | |
| "step": 161500 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "grad_norm": 9.147507667541504, | |
| "learning_rate": 7.5364248604964404e-06, | |
| "loss": 0.4642, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 4.600451946258545, | |
| "learning_rate": 7.497941119876852e-06, | |
| "loss": 0.4448, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 6.27, | |
| "grad_norm": 5.7864556312561035, | |
| "learning_rate": 7.459457379257264e-06, | |
| "loss": 0.4463, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "grad_norm": 6.185606002807617, | |
| "learning_rate": 7.420973638637676e-06, | |
| "loss": 0.4608, | |
| "step": 163500 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "grad_norm": 4.961802959442139, | |
| "learning_rate": 7.382489898018088e-06, | |
| "loss": 0.4609, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "grad_norm": 8.758057594299316, | |
| "learning_rate": 7.3440061573985e-06, | |
| "loss": 0.459, | |
| "step": 164500 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "grad_norm": 13.342981338500977, | |
| "learning_rate": 7.305522416778912e-06, | |
| "loss": 0.4684, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "grad_norm": 3.678190231323242, | |
| "learning_rate": 7.267038676159324e-06, | |
| "loss": 0.4482, | |
| "step": 165500 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "grad_norm": 5.763678550720215, | |
| "learning_rate": 7.2285549355397356e-06, | |
| "loss": 0.4828, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "grad_norm": 2.4365394115448, | |
| "learning_rate": 7.1900711949201475e-06, | |
| "loss": 0.4519, | |
| "step": 166500 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "grad_norm": 7.325679779052734, | |
| "learning_rate": 7.1515874543005585e-06, | |
| "loss": 0.4725, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 6.45, | |
| "grad_norm": 7.1001811027526855, | |
| "learning_rate": 7.11310371368097e-06, | |
| "loss": 0.4552, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "grad_norm": 4.369767189025879, | |
| "learning_rate": 7.0746199730613814e-06, | |
| "loss": 0.4615, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 6.48, | |
| "grad_norm": 5.81917667388916, | |
| "learning_rate": 7.036213199923034e-06, | |
| "loss": 0.4518, | |
| "step": 168500 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 4.713387966156006, | |
| "learning_rate": 6.997729459303445e-06, | |
| "loss": 0.473, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "grad_norm": 7.274464130401611, | |
| "learning_rate": 6.9592457186838565e-06, | |
| "loss": 0.4729, | |
| "step": 169500 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "grad_norm": 8.085463523864746, | |
| "learning_rate": 6.9207619780642684e-06, | |
| "loss": 0.4723, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "grad_norm": 9.212237358093262, | |
| "learning_rate": 6.88227823744468e-06, | |
| "loss": 0.4571, | |
| "step": 170500 | |
| }, | |
| { | |
| "epoch": 6.58, | |
| "grad_norm": 3.6301112174987793, | |
| "learning_rate": 6.843871464306331e-06, | |
| "loss": 0.4726, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 5.698800563812256, | |
| "learning_rate": 6.805464691167982e-06, | |
| "loss": 0.4831, | |
| "step": 171500 | |
| }, | |
| { | |
| "epoch": 6.62, | |
| "grad_norm": 5.532044887542725, | |
| "learning_rate": 6.766980950548394e-06, | |
| "loss": 0.4708, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "grad_norm": 7.939385414123535, | |
| "learning_rate": 6.728497209928806e-06, | |
| "loss": 0.468, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 6.66, | |
| "grad_norm": 3.2672829627990723, | |
| "learning_rate": 6.690013469309218e-06, | |
| "loss": 0.4935, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "grad_norm": 11.197509765625, | |
| "learning_rate": 6.65152972868963e-06, | |
| "loss": 0.4777, | |
| "step": 173500 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 2.946389675140381, | |
| "learning_rate": 6.6130459880700416e-06, | |
| "loss": 0.4786, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "grad_norm": 1.725356936454773, | |
| "learning_rate": 6.574562247450453e-06, | |
| "loss": 0.4471, | |
| "step": 174500 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "grad_norm": 3.4278807640075684, | |
| "learning_rate": 6.5360785068308645e-06, | |
| "loss": 0.472, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 6.75, | |
| "grad_norm": 7.619284629821777, | |
| "learning_rate": 6.4975947662112756e-06, | |
| "loss": 0.4566, | |
| "step": 175500 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "grad_norm": 3.8815529346466064, | |
| "learning_rate": 6.459187993072928e-06, | |
| "loss": 0.4738, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "grad_norm": 4.85827112197876, | |
| "learning_rate": 6.420704252453339e-06, | |
| "loss": 0.4637, | |
| "step": 176500 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "grad_norm": 5.672398090362549, | |
| "learning_rate": 6.382220511833751e-06, | |
| "loss": 0.4646, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 6.83, | |
| "grad_norm": 3.5175821781158447, | |
| "learning_rate": 6.3437367712141626e-06, | |
| "loss": 0.4949, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "grad_norm": 12.251681327819824, | |
| "learning_rate": 6.3052530305945744e-06, | |
| "loss": 0.4764, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 6.87, | |
| "grad_norm": 7.721076488494873, | |
| "learning_rate": 6.266846257456225e-06, | |
| "loss": 0.4982, | |
| "step": 178500 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "grad_norm": 3.9712600708007812, | |
| "learning_rate": 6.228362516836637e-06, | |
| "loss": 0.4859, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 6.91, | |
| "grad_norm": 5.2524213790893555, | |
| "learning_rate": 6.189878776217049e-06, | |
| "loss": 0.4697, | |
| "step": 179500 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "grad_norm": 4.3481340408325195, | |
| "learning_rate": 6.151395035597461e-06, | |
| "loss": 0.4916, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 6.95, | |
| "grad_norm": 5.635584354400635, | |
| "learning_rate": 6.1129112949778725e-06, | |
| "loss": 0.476, | |
| "step": 180500 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "grad_norm": 4.938673973083496, | |
| "learning_rate": 6.074504521839524e-06, | |
| "loss": 0.4842, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 6.98, | |
| "grad_norm": 3.648951530456543, | |
| "learning_rate": 6.036097748701174e-06, | |
| "loss": 0.4961, | |
| "step": 181500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 1.7367790937423706, | |
| "eval_runtime": 11.3431, | |
| "eval_samples_per_second": 350.961, | |
| "eval_steps_per_second": 87.806, | |
| "step": 181895 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.272878646850586, | |
| "learning_rate": 5.997614008081586e-06, | |
| "loss": 0.4793, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "grad_norm": 8.145177841186523, | |
| "learning_rate": 5.959130267461998e-06, | |
| "loss": 0.4003, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "grad_norm": 3.661898374557495, | |
| "learning_rate": 5.92064652684241e-06, | |
| "loss": 0.4182, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "grad_norm": 4.206645488739014, | |
| "learning_rate": 5.882162786222822e-06, | |
| "loss": 0.4278, | |
| "step": 183500 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "grad_norm": 2.098565101623535, | |
| "learning_rate": 5.843679045603233e-06, | |
| "loss": 0.4281, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 5.96292781829834, | |
| "learning_rate": 5.805195304983645e-06, | |
| "loss": 0.4374, | |
| "step": 184500 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "grad_norm": 2.972966194152832, | |
| "learning_rate": 5.766711564364057e-06, | |
| "loss": 0.4143, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "grad_norm": 9.22352123260498, | |
| "learning_rate": 5.7282278237444686e-06, | |
| "loss": 0.401, | |
| "step": 185500 | |
| }, | |
| { | |
| "epoch": 7.16, | |
| "grad_norm": 8.188728332519531, | |
| "learning_rate": 5.689821050606119e-06, | |
| "loss": 0.4134, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "grad_norm": 8.370118141174316, | |
| "learning_rate": 5.65141427746777e-06, | |
| "loss": 0.445, | |
| "step": 186500 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 4.1989922523498535, | |
| "learning_rate": 5.612930536848182e-06, | |
| "loss": 0.4428, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "grad_norm": 8.339373588562012, | |
| "learning_rate": 5.574446796228594e-06, | |
| "loss": 0.4091, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "grad_norm": 5.580036640167236, | |
| "learning_rate": 5.535963055609006e-06, | |
| "loss": 0.4169, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "grad_norm": 5.285128116607666, | |
| "learning_rate": 5.4975562824706565e-06, | |
| "loss": 0.4282, | |
| "step": 188500 | |
| }, | |
| { | |
| "epoch": 7.27, | |
| "grad_norm": 4.649808406829834, | |
| "learning_rate": 5.459072541851068e-06, | |
| "loss": 0.4199, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 7.711951732635498, | |
| "learning_rate": 5.42058880123148e-06, | |
| "loss": 0.4331, | |
| "step": 189500 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "grad_norm": 3.5178699493408203, | |
| "learning_rate": 5.382105060611892e-06, | |
| "loss": 0.4362, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "grad_norm": 6.855963230133057, | |
| "learning_rate": 5.343621319992304e-06, | |
| "loss": 0.449, | |
| "step": 190500 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "grad_norm": 5.831274509429932, | |
| "learning_rate": 5.305137579372716e-06, | |
| "loss": 0.4286, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 5.696512699127197, | |
| "learning_rate": 5.266653838753127e-06, | |
| "loss": 0.4388, | |
| "step": 191500 | |
| }, | |
| { | |
| "epoch": 7.39, | |
| "grad_norm": 6.7039055824279785, | |
| "learning_rate": 5.228170098133539e-06, | |
| "loss": 0.4395, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "grad_norm": 7.356657028198242, | |
| "learning_rate": 5.189686357513951e-06, | |
| "loss": 0.4488, | |
| "step": 192500 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "grad_norm": 4.289228439331055, | |
| "learning_rate": 5.151279584375601e-06, | |
| "loss": 0.4367, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "grad_norm": 6.5351948738098145, | |
| "learning_rate": 5.112795843756013e-06, | |
| "loss": 0.438, | |
| "step": 193500 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "grad_norm": 6.228538990020752, | |
| "learning_rate": 5.0743890706176645e-06, | |
| "loss": 0.4419, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 7.49, | |
| "grad_norm": 8.515239715576172, | |
| "learning_rate": 5.035905329998076e-06, | |
| "loss": 0.4501, | |
| "step": 194500 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 6.8445725440979, | |
| "learning_rate": 4.9974215893784874e-06, | |
| "loss": 0.4257, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 7.52, | |
| "grad_norm": 9.162801742553711, | |
| "learning_rate": 4.958937848758899e-06, | |
| "loss": 0.4349, | |
| "step": 195500 | |
| }, | |
| { | |
| "epoch": 7.54, | |
| "grad_norm": 10.447943687438965, | |
| "learning_rate": 4.920454108139311e-06, | |
| "loss": 0.4364, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 7.56, | |
| "grad_norm": 6.493594169616699, | |
| "learning_rate": 4.881970367519723e-06, | |
| "loss": 0.4436, | |
| "step": 196500 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "grad_norm": 6.790309429168701, | |
| "learning_rate": 4.843486626900135e-06, | |
| "loss": 0.4535, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 4.3611860275268555, | |
| "learning_rate": 4.805002886280547e-06, | |
| "loss": 0.443, | |
| "step": 197500 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "grad_norm": 7.639214992523193, | |
| "learning_rate": 4.766519145660959e-06, | |
| "loss": 0.4433, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "grad_norm": 4.556090831756592, | |
| "learning_rate": 4.72811237252261e-06, | |
| "loss": 0.4433, | |
| "step": 198500 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "grad_norm": 6.17908239364624, | |
| "learning_rate": 4.689628631903021e-06, | |
| "loss": 0.4389, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "grad_norm": 5.435972213745117, | |
| "learning_rate": 4.651144891283433e-06, | |
| "loss": 0.4413, | |
| "step": 199500 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 8.422209739685059, | |
| "learning_rate": 4.612661150663845e-06, | |
| "loss": 0.4444, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "grad_norm": 9.431715965270996, | |
| "learning_rate": 4.574254377525496e-06, | |
| "loss": 0.4495, | |
| "step": 200500 | |
| }, | |
| { | |
| "epoch": 7.74, | |
| "grad_norm": 2.9533908367156982, | |
| "learning_rate": 4.535847604387147e-06, | |
| "loss": 0.4417, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "grad_norm": 4.222201824188232, | |
| "learning_rate": 4.497363863767559e-06, | |
| "loss": 0.4456, | |
| "step": 201500 | |
| }, | |
| { | |
| "epoch": 7.77, | |
| "grad_norm": 8.314460754394531, | |
| "learning_rate": 4.4588801231479705e-06, | |
| "loss": 0.4297, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "grad_norm": 8.237698554992676, | |
| "learning_rate": 4.4203963825283815e-06, | |
| "loss": 0.4637, | |
| "step": 202500 | |
| }, | |
| { | |
| "epoch": 7.81, | |
| "grad_norm": 7.0175933837890625, | |
| "learning_rate": 4.3819126419087934e-06, | |
| "loss": 0.4253, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "grad_norm": 9.906578063964844, | |
| "learning_rate": 4.343428901289205e-06, | |
| "loss": 0.4261, | |
| "step": 203500 | |
| }, | |
| { | |
| "epoch": 7.85, | |
| "grad_norm": 2.473310947418213, | |
| "learning_rate": 4.304945160669617e-06, | |
| "loss": 0.4576, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 7.87, | |
| "grad_norm": 7.461137771606445, | |
| "learning_rate": 4.266461420050029e-06, | |
| "loss": 0.458, | |
| "step": 204500 | |
| }, | |
| { | |
| "epoch": 7.89, | |
| "grad_norm": 4.642632007598877, | |
| "learning_rate": 4.227977679430441e-06, | |
| "loss": 0.4477, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "grad_norm": 4.437539100646973, | |
| "learning_rate": 4.189570906292092e-06, | |
| "loss": 0.4441, | |
| "step": 205500 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "grad_norm": 6.345538139343262, | |
| "learning_rate": 4.151087165672504e-06, | |
| "loss": 0.4549, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "grad_norm": 5.00540828704834, | |
| "learning_rate": 4.112603425052915e-06, | |
| "loss": 0.4563, | |
| "step": 206500 | |
| }, | |
| { | |
| "epoch": 7.97, | |
| "grad_norm": 2.3221421241760254, | |
| "learning_rate": 4.074119684433327e-06, | |
| "loss": 0.4462, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 7.99, | |
| "grad_norm": 4.091787815093994, | |
| "learning_rate": 4.0357129112949785e-06, | |
| "loss": 0.4588, | |
| "step": 207500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.752930998802185, | |
| "eval_runtime": 11.2624, | |
| "eval_samples_per_second": 353.477, | |
| "eval_steps_per_second": 88.436, | |
| "step": 207880 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 10.46442985534668, | |
| "learning_rate": 3.99722917067539e-06, | |
| "loss": 0.4772, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 2.7908215522766113, | |
| "learning_rate": 3.958745430055801e-06, | |
| "loss": 0.4165, | |
| "step": 208500 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "grad_norm": 7.39349365234375, | |
| "learning_rate": 3.920261689436213e-06, | |
| "loss": 0.4127, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "grad_norm": 7.740577697753906, | |
| "learning_rate": 3.881777948816625e-06, | |
| "loss": 0.4235, | |
| "step": 209500 | |
| }, | |
| { | |
| "epoch": 8.08, | |
| "grad_norm": 3.9219789505004883, | |
| "learning_rate": 3.843294208197037e-06, | |
| "loss": 0.4272, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 6.718163967132568, | |
| "learning_rate": 3.804887435058688e-06, | |
| "loss": 0.4062, | |
| "step": 210500 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "grad_norm": 4.927985191345215, | |
| "learning_rate": 3.7664036944391e-06, | |
| "loss": 0.4, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 8.14, | |
| "grad_norm": 6.605733871459961, | |
| "learning_rate": 3.7279199538195113e-06, | |
| "loss": 0.4224, | |
| "step": 211500 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "grad_norm": 5.900667190551758, | |
| "learning_rate": 3.6894362131999232e-06, | |
| "loss": 0.401, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 8.18, | |
| "grad_norm": 3.047395706176758, | |
| "learning_rate": 3.651029440061574e-06, | |
| "loss": 0.4099, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 10.504358291625977, | |
| "learning_rate": 3.612545699441986e-06, | |
| "loss": 0.4094, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "grad_norm": 2.9013960361480713, | |
| "learning_rate": 3.574061958822398e-06, | |
| "loss": 0.4225, | |
| "step": 213500 | |
| }, | |
| { | |
| "epoch": 8.24, | |
| "grad_norm": 1.0684086084365845, | |
| "learning_rate": 3.53557821820281e-06, | |
| "loss": 0.4333, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "grad_norm": 1.7722007036209106, | |
| "learning_rate": 3.4970944775832217e-06, | |
| "loss": 0.419, | |
| "step": 214500 | |
| }, | |
| { | |
| "epoch": 8.27, | |
| "grad_norm": 5.168225288391113, | |
| "learning_rate": 3.4586107369636327e-06, | |
| "loss": 0.429, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "grad_norm": 9.540655136108398, | |
| "learning_rate": 3.4201269963440446e-06, | |
| "loss": 0.4024, | |
| "step": 215500 | |
| }, | |
| { | |
| "epoch": 8.31, | |
| "grad_norm": 7.926085472106934, | |
| "learning_rate": 3.3817202232056955e-06, | |
| "loss": 0.4081, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 3.9807188510894775, | |
| "learning_rate": 3.3432364825861074e-06, | |
| "loss": 0.4098, | |
| "step": 216500 | |
| }, | |
| { | |
| "epoch": 8.35, | |
| "grad_norm": 4.447396755218506, | |
| "learning_rate": 3.3047527419665193e-06, | |
| "loss": 0.4045, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 8.37, | |
| "grad_norm": 5.585531711578369, | |
| "learning_rate": 3.266269001346931e-06, | |
| "loss": 0.409, | |
| "step": 217500 | |
| }, | |
| { | |
| "epoch": 8.39, | |
| "grad_norm": 4.316734790802002, | |
| "learning_rate": 3.227862228208582e-06, | |
| "loss": 0.4239, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "grad_norm": 7.370926380157471, | |
| "learning_rate": 3.189378487588994e-06, | |
| "loss": 0.4104, | |
| "step": 218500 | |
| }, | |
| { | |
| "epoch": 8.43, | |
| "grad_norm": 6.9890522956848145, | |
| "learning_rate": 3.1508947469694055e-06, | |
| "loss": 0.4024, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 8.45, | |
| "grad_norm": 5.685979843139648, | |
| "learning_rate": 3.1124110063498174e-06, | |
| "loss": 0.4167, | |
| "step": 219500 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "grad_norm": 5.407013416290283, | |
| "learning_rate": 3.0739272657302293e-06, | |
| "loss": 0.3967, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "grad_norm": 3.3302626609802246, | |
| "learning_rate": 3.03552049259188e-06, | |
| "loss": 0.4303, | |
| "step": 220500 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 9.727193832397461, | |
| "learning_rate": 2.997113719453531e-06, | |
| "loss": 0.4108, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 8.52, | |
| "grad_norm": 5.004329681396484, | |
| "learning_rate": 2.958629978833943e-06, | |
| "loss": 0.4135, | |
| "step": 221500 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "grad_norm": 5.899045467376709, | |
| "learning_rate": 2.920146238214355e-06, | |
| "loss": 0.4214, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 8.56, | |
| "grad_norm": 9.472413063049316, | |
| "learning_rate": 2.8816624975947667e-06, | |
| "loss": 0.4063, | |
| "step": 222500 | |
| }, | |
| { | |
| "epoch": 8.58, | |
| "grad_norm": 3.2998526096343994, | |
| "learning_rate": 2.8431787569751786e-06, | |
| "loss": 0.4152, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 9.45439338684082, | |
| "learning_rate": 2.8046950163555897e-06, | |
| "loss": 0.4156, | |
| "step": 223500 | |
| }, | |
| { | |
| "epoch": 8.62, | |
| "grad_norm": 10.916293144226074, | |
| "learning_rate": 2.7662112757360016e-06, | |
| "loss": 0.4177, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 8.64, | |
| "grad_norm": 9.067635536193848, | |
| "learning_rate": 2.7277275351164134e-06, | |
| "loss": 0.4107, | |
| "step": 224500 | |
| }, | |
| { | |
| "epoch": 8.66, | |
| "grad_norm": 8.393898010253906, | |
| "learning_rate": 2.6892437944968253e-06, | |
| "loss": 0.4167, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 8.68, | |
| "grad_norm": 4.151984691619873, | |
| "learning_rate": 2.6508370213584762e-06, | |
| "loss": 0.415, | |
| "step": 225500 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 5.011842250823975, | |
| "learning_rate": 2.612353280738888e-06, | |
| "loss": 0.4358, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 8.72, | |
| "grad_norm": 2.5369720458984375, | |
| "learning_rate": 2.5738695401192996e-06, | |
| "loss": 0.4114, | |
| "step": 226500 | |
| }, | |
| { | |
| "epoch": 8.74, | |
| "grad_norm": 6.167263507843018, | |
| "learning_rate": 2.5353857994997115e-06, | |
| "loss": 0.4117, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "grad_norm": 3.8160266876220703, | |
| "learning_rate": 2.4969020588801234e-06, | |
| "loss": 0.4195, | |
| "step": 227500 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "grad_norm": 6.264139175415039, | |
| "learning_rate": 2.4584952857417743e-06, | |
| "loss": 0.4182, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "grad_norm": 2.6473159790039062, | |
| "learning_rate": 2.420011545122186e-06, | |
| "loss": 0.4109, | |
| "step": 228500 | |
| }, | |
| { | |
| "epoch": 8.81, | |
| "grad_norm": 5.942705154418945, | |
| "learning_rate": 2.3815278045025976e-06, | |
| "loss": 0.4385, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 8.83, | |
| "grad_norm": 6.846287250518799, | |
| "learning_rate": 2.3430440638830095e-06, | |
| "loss": 0.4406, | |
| "step": 229500 | |
| }, | |
| { | |
| "epoch": 8.85, | |
| "grad_norm": 3.9785611629486084, | |
| "learning_rate": 2.3045603232634214e-06, | |
| "loss": 0.4154, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 8.87, | |
| "grad_norm": 4.567147254943848, | |
| "learning_rate": 2.2661535501250723e-06, | |
| "loss": 0.4214, | |
| "step": 230500 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "grad_norm": 7.9429240226745605, | |
| "learning_rate": 2.227669809505484e-06, | |
| "loss": 0.446, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 8.91, | |
| "grad_norm": 6.396302223205566, | |
| "learning_rate": 2.189186068885896e-06, | |
| "loss": 0.4306, | |
| "step": 231500 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "grad_norm": 7.975338935852051, | |
| "learning_rate": 2.1507023282663076e-06, | |
| "loss": 0.419, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 8.95, | |
| "grad_norm": 6.531688690185547, | |
| "learning_rate": 2.1122955551279585e-06, | |
| "loss": 0.4144, | |
| "step": 232500 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "grad_norm": 4.969785213470459, | |
| "learning_rate": 2.0738118145083704e-06, | |
| "loss": 0.4283, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 8.99, | |
| "grad_norm": 7.7668538093566895, | |
| "learning_rate": 2.0353280738887822e-06, | |
| "loss": 0.4025, | |
| "step": 233500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 1.7583389282226562, | |
| "eval_runtime": 11.3764, | |
| "eval_samples_per_second": 349.935, | |
| "eval_steps_per_second": 87.55, | |
| "step": 233865 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 6.1113409996032715, | |
| "learning_rate": 1.9968443332691937e-06, | |
| "loss": 0.4198, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "grad_norm": 6.071191787719727, | |
| "learning_rate": 1.9583605926496056e-06, | |
| "loss": 0.3948, | |
| "step": 234500 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "grad_norm": 2.7634339332580566, | |
| "learning_rate": 1.9198768520300175e-06, | |
| "loss": 0.3819, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 9.06, | |
| "grad_norm": 5.456170082092285, | |
| "learning_rate": 1.8813931114104292e-06, | |
| "loss": 0.3905, | |
| "step": 235500 | |
| }, | |
| { | |
| "epoch": 9.08, | |
| "grad_norm": 6.395251750946045, | |
| "learning_rate": 1.842909370790841e-06, | |
| "loss": 0.3947, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 4.872887134552002, | |
| "learning_rate": 1.804425630171253e-06, | |
| "loss": 0.3775, | |
| "step": 236500 | |
| }, | |
| { | |
| "epoch": 9.12, | |
| "grad_norm": 7.3761162757873535, | |
| "learning_rate": 1.7660188570329036e-06, | |
| "loss": 0.3947, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "grad_norm": 4.171831130981445, | |
| "learning_rate": 1.7275351164133155e-06, | |
| "loss": 0.4119, | |
| "step": 237500 | |
| }, | |
| { | |
| "epoch": 9.16, | |
| "grad_norm": 2.4952399730682373, | |
| "learning_rate": 1.6890513757937274e-06, | |
| "loss": 0.394, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "grad_norm": 6.803173542022705, | |
| "learning_rate": 1.650567635174139e-06, | |
| "loss": 0.4007, | |
| "step": 238500 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 3.5348174571990967, | |
| "learning_rate": 1.6120838945545508e-06, | |
| "loss": 0.4027, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "grad_norm": 5.090949058532715, | |
| "learning_rate": 1.5736771214162017e-06, | |
| "loss": 0.3902, | |
| "step": 239500 | |
| }, | |
| { | |
| "epoch": 9.24, | |
| "grad_norm": 12.865615844726562, | |
| "learning_rate": 1.5351933807966136e-06, | |
| "loss": 0.4002, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 9.26, | |
| "grad_norm": 8.621195793151855, | |
| "learning_rate": 1.4967096401770253e-06, | |
| "loss": 0.4197, | |
| "step": 240500 | |
| }, | |
| { | |
| "epoch": 9.27, | |
| "grad_norm": 6.878337383270264, | |
| "learning_rate": 1.4582258995574371e-06, | |
| "loss": 0.4024, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "grad_norm": 6.962910175323486, | |
| "learning_rate": 1.4197421589378488e-06, | |
| "loss": 0.4079, | |
| "step": 241500 | |
| }, | |
| { | |
| "epoch": 9.31, | |
| "grad_norm": 4.2871174812316895, | |
| "learning_rate": 1.3813353857995e-06, | |
| "loss": 0.3943, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 9.33, | |
| "grad_norm": 3.1228811740875244, | |
| "learning_rate": 1.3428516451799118e-06, | |
| "loss": 0.3982, | |
| "step": 242500 | |
| }, | |
| { | |
| "epoch": 9.35, | |
| "grad_norm": 4.895651340484619, | |
| "learning_rate": 1.3043679045603233e-06, | |
| "loss": 0.3939, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 9.37, | |
| "grad_norm": 3.8321521282196045, | |
| "learning_rate": 1.2658841639407352e-06, | |
| "loss": 0.4048, | |
| "step": 243500 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "grad_norm": 4.628911972045898, | |
| "learning_rate": 1.2274004233211469e-06, | |
| "loss": 0.4099, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 9.41, | |
| "grad_norm": 4.80955696105957, | |
| "learning_rate": 1.1889166827015588e-06, | |
| "loss": 0.3891, | |
| "step": 244500 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "grad_norm": 7.214938163757324, | |
| "learning_rate": 1.1504329420819704e-06, | |
| "loss": 0.4, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 9.45, | |
| "grad_norm": 5.537662982940674, | |
| "learning_rate": 1.1120261689436213e-06, | |
| "loss": 0.4052, | |
| "step": 245500 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "grad_norm": 4.672789573669434, | |
| "learning_rate": 1.0736193958052725e-06, | |
| "loss": 0.4121, | |
| "step": 246000 | |
| }, | |
| { | |
| "epoch": 9.49, | |
| "grad_norm": 7.284801959991455, | |
| "learning_rate": 1.0351356551856841e-06, | |
| "loss": 0.4168, | |
| "step": 246500 | |
| }, | |
| { | |
| "epoch": 9.51, | |
| "grad_norm": 8.112405776977539, | |
| "learning_rate": 9.96651914566096e-07, | |
| "loss": 0.3951, | |
| "step": 247000 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "grad_norm": 5.421095848083496, | |
| "learning_rate": 9.581681739465077e-07, | |
| "loss": 0.384, | |
| "step": 247500 | |
| }, | |
| { | |
| "epoch": 9.54, | |
| "grad_norm": 6.66250467300415, | |
| "learning_rate": 9.196844333269195e-07, | |
| "loss": 0.4037, | |
| "step": 248000 | |
| }, | |
| { | |
| "epoch": 9.56, | |
| "grad_norm": 7.84215784072876, | |
| "learning_rate": 8.812006927073312e-07, | |
| "loss": 0.4025, | |
| "step": 248500 | |
| }, | |
| { | |
| "epoch": 9.58, | |
| "grad_norm": 4.539148807525635, | |
| "learning_rate": 8.427169520877431e-07, | |
| "loss": 0.4072, | |
| "step": 249000 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 4.91790246963501, | |
| "learning_rate": 8.042332114681547e-07, | |
| "loss": 0.3853, | |
| "step": 249500 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "grad_norm": 7.577627658843994, | |
| "learning_rate": 7.658264383298057e-07, | |
| "loss": 0.3997, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 9.64, | |
| "grad_norm": 14.857439994812012, | |
| "learning_rate": 7.273426977102174e-07, | |
| "loss": 0.3852, | |
| "step": 250500 | |
| }, | |
| { | |
| "epoch": 9.66, | |
| "grad_norm": 3.0629687309265137, | |
| "learning_rate": 6.888589570906293e-07, | |
| "loss": 0.3887, | |
| "step": 251000 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "grad_norm": 3.039710760116577, | |
| "learning_rate": 6.50375216471041e-07, | |
| "loss": 0.4124, | |
| "step": 251500 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 7.001966953277588, | |
| "learning_rate": 6.11968443332692e-07, | |
| "loss": 0.4102, | |
| "step": 252000 | |
| }, | |
| { | |
| "epoch": 9.72, | |
| "grad_norm": 3.581230878829956, | |
| "learning_rate": 5.734847027131038e-07, | |
| "loss": 0.3916, | |
| "step": 252500 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "grad_norm": 3.673539638519287, | |
| "learning_rate": 5.350009620935155e-07, | |
| "loss": 0.3973, | |
| "step": 253000 | |
| }, | |
| { | |
| "epoch": 9.76, | |
| "grad_norm": 9.442609786987305, | |
| "learning_rate": 4.965172214739272e-07, | |
| "loss": 0.3913, | |
| "step": 253500 | |
| }, | |
| { | |
| "epoch": 9.77, | |
| "grad_norm": 4.602294921875, | |
| "learning_rate": 4.5811044833557826e-07, | |
| "loss": 0.3989, | |
| "step": 254000 | |
| }, | |
| { | |
| "epoch": 9.79, | |
| "grad_norm": 2.9956846237182617, | |
| "learning_rate": 4.1962670771599004e-07, | |
| "loss": 0.4018, | |
| "step": 254500 | |
| }, | |
| { | |
| "epoch": 9.81, | |
| "grad_norm": 5.884944438934326, | |
| "learning_rate": 3.811429670964018e-07, | |
| "loss": 0.3974, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 9.83, | |
| "grad_norm": 3.6451961994171143, | |
| "learning_rate": 3.426592264768136e-07, | |
| "loss": 0.3981, | |
| "step": 255500 | |
| }, | |
| { | |
| "epoch": 9.85, | |
| "grad_norm": 5.324589729309082, | |
| "learning_rate": 3.0417548585722534e-07, | |
| "loss": 0.3885, | |
| "step": 256000 | |
| }, | |
| { | |
| "epoch": 9.87, | |
| "grad_norm": 5.142848968505859, | |
| "learning_rate": 2.6569174523763713e-07, | |
| "loss": 0.4074, | |
| "step": 256500 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "grad_norm": 2.519761085510254, | |
| "learning_rate": 2.272080046180489e-07, | |
| "loss": 0.416, | |
| "step": 257000 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "grad_norm": 8.230846405029297, | |
| "learning_rate": 1.8880123147969984e-07, | |
| "loss": 0.4173, | |
| "step": 257500 | |
| }, | |
| { | |
| "epoch": 9.93, | |
| "grad_norm": 8.473825454711914, | |
| "learning_rate": 1.5031749086011162e-07, | |
| "loss": 0.3753, | |
| "step": 258000 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "grad_norm": 9.58931827545166, | |
| "learning_rate": 1.1183375024052338e-07, | |
| "loss": 0.4112, | |
| "step": 258500 | |
| }, | |
| { | |
| "epoch": 9.97, | |
| "grad_norm": 7.562945365905762, | |
| "learning_rate": 7.335000962093516e-08, | |
| "loss": 0.3932, | |
| "step": 259000 | |
| }, | |
| { | |
| "epoch": 9.99, | |
| "grad_norm": 3.519716501235962, | |
| "learning_rate": 3.4866269001346934e-08, | |
| "loss": 0.4157, | |
| "step": 259500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.769174575805664, | |
| "eval_runtime": 11.3755, | |
| "eval_samples_per_second": 349.964, | |
| "eval_steps_per_second": 87.557, | |
| "step": 259850 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 259850, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.0298513413767168e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |