| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.929787234042553, | |
| "eval_steps": 5, | |
| "global_step": 87, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9849, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 2.0810964107513428, | |
| "eval_runtime": 6.2278, | |
| "eval_samples_per_second": 44.96, | |
| "eval_steps_per_second": 5.62, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 4e-05, | |
| "loss": 1.8931, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 6e-05, | |
| "loss": 1.763, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 8e-05, | |
| "loss": 1.6688, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3107, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_loss": 1.1991726160049438, | |
| "eval_runtime": 6.3173, | |
| "eval_samples_per_second": 44.322, | |
| "eval_steps_per_second": 5.54, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 0.00012, | |
| "loss": 1.1899, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 0.00014, | |
| "loss": 1.041, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 0.00016, | |
| "loss": 0.9779, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.00018, | |
| "loss": 0.7671, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6399, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_loss": 0.6359391212463379, | |
| "eval_runtime": 6.2721, | |
| "eval_samples_per_second": 44.642, | |
| "eval_steps_per_second": 5.58, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 0.0001999167799344583, | |
| "loss": 0.6777, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.00019966725824941932, | |
| "loss": 0.492, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.00019925185024910277, | |
| "loss": 0.3781, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 0.0001986712473391289, | |
| "loss": 0.4503, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.00019792641587574212, | |
| "loss": 0.2779, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_loss": 0.28619876503944397, | |
| "eval_runtime": 6.3023, | |
| "eval_samples_per_second": 44.428, | |
| "eval_steps_per_second": 5.554, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 0.00019701859555740648, | |
| "loss": 0.2658, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 0.2282, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 0.000194720301029191, | |
| "loss": 0.2149, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 0.0001933336521037367, | |
| "loss": 0.1792, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 0.00019179165852537596, | |
| "loss": 0.1807, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_loss": 0.1634436696767807, | |
| "eval_runtime": 6.2724, | |
| "eval_samples_per_second": 44.64, | |
| "eval_steps_per_second": 5.58, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 0.1425, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 0.00018825215767863214, | |
| "loss": 0.1377, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.00018626054156009806, | |
| "loss": 0.1173, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 0.1338, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 0.00018185014665785936, | |
| "loss": 0.1256, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_loss": 0.1177145466208458, | |
| "eval_runtime": 6.2558, | |
| "eval_samples_per_second": 44.758, | |
| "eval_steps_per_second": 5.595, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 0.00017943870854121124, | |
| "loss": 0.1234, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 0.0001768950525339362, | |
| "loss": 0.1164, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 0.000174223412300427, | |
| "loss": 0.1077, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 0.00017142823452219038, | |
| "loss": 0.0877, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.00016851417149678444, | |
| "loss": 0.097, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.08908110111951828, | |
| "eval_runtime": 6.274, | |
| "eval_samples_per_second": 44.628, | |
| "eval_steps_per_second": 5.579, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 0.0741, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 0.0811, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.00015910811325286768, | |
| "loss": 0.0743, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 0.00015576886669868296, | |
| "loss": 0.0799, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 0.00015233679836966122, | |
| "loss": 0.1063, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_loss": 0.07344743609428406, | |
| "eval_runtime": 6.3055, | |
| "eval_samples_per_second": 44.406, | |
| "eval_steps_per_second": 5.551, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 0.00014881762060482814, | |
| "loss": 0.0835, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 0.00014521719072826858, | |
| "loss": 0.0661, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.0637, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 0.00013779667014289065, | |
| "loss": 0.0642, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 0.00013398893015826167, | |
| "loss": 0.0769, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "eval_loss": 0.07225509732961655, | |
| "eval_runtime": 6.2831, | |
| "eval_samples_per_second": 44.564, | |
| "eval_steps_per_second": 5.571, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 0.00013012461895372344, | |
| "loss": 0.0746, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 0.00012621016829391022, | |
| "loss": 0.0592, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 0.078, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 0.00011825698208393619, | |
| "loss": 0.088, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 0.0694, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_loss": 0.06333146244287491, | |
| "eval_runtime": 6.2803, | |
| "eval_samples_per_second": 44.584, | |
| "eval_steps_per_second": 5.573, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 0.00011018229867038356, | |
| "loss": 0.0591, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 0.00010611616608218429, | |
| "loss": 0.0646, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 0.00010203985373914056, | |
| "loss": 0.068, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 9.79601462608595e-05, | |
| "loss": 0.055, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 9.388383391781575e-05, | |
| "loss": 0.0687, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "eval_loss": 0.06235363706946373, | |
| "eval_runtime": 6.2849, | |
| "eval_samples_per_second": 44.551, | |
| "eval_steps_per_second": 5.569, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 8.981770132961649e-05, | |
| "loss": 0.0643, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 0.0589, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 8.174301791606385e-05, | |
| "loss": 0.0729, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 0.0553, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 7.378983170608982e-05, | |
| "loss": 0.0575, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "eval_loss": 0.062243152409791946, | |
| "eval_runtime": 6.2854, | |
| "eval_samples_per_second": 44.548, | |
| "eval_steps_per_second": 5.568, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 6.98753810462766e-05, | |
| "loss": 0.0675, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 6.601106984173835e-05, | |
| "loss": 0.068, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 6.220332985710936e-05, | |
| "loss": 0.0601, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 0.0764, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 5.478280927173145e-05, | |
| "loss": 0.0516, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "eval_loss": 0.060868892818689346, | |
| "eval_runtime": 5.9359, | |
| "eval_samples_per_second": 47.171, | |
| "eval_steps_per_second": 5.896, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 5.11823793951719e-05, | |
| "loss": 0.0516, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 4.7663201630338816e-05, | |
| "loss": 0.0543, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 4.423113330131707e-05, | |
| "loss": 0.0519, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 4.089188674713236e-05, | |
| "loss": 0.0602, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 0.0582, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "eval_loss": 0.060279980301856995, | |
| "eval_runtime": 5.9625, | |
| "eval_samples_per_second": 46.96, | |
| "eval_steps_per_second": 5.87, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 0.055, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 3.1485828503215585e-05, | |
| "loss": 0.0515, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2.8571765477809643e-05, | |
| "loss": 0.0515, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 2.5776587699573006e-05, | |
| "loss": 0.0497, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 2.3104947466063787e-05, | |
| "loss": 0.0611, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_loss": 0.05999369919300079, | |
| "eval_runtime": 5.9591, | |
| "eval_samples_per_second": 46.987, | |
| "eval_steps_per_second": 5.873, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 2.0561291458788733e-05, | |
| "loss": 0.0564, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 1.8149853342140645e-05, | |
| "loss": 0.0498, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 0.0503, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 1.373945843990192e-05, | |
| "loss": 0.0494, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 1.1747842321367886e-05, | |
| "loss": 0.0515, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "eval_loss": 0.05982312187552452, | |
| "eval_runtime": 5.856, | |
| "eval_samples_per_second": 47.814, | |
| "eval_steps_per_second": 5.977, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 9.903113209758096e-06, | |
| "loss": 0.0508, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 8.208341474624071e-06, | |
| "loss": 0.0596, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 6.666347896263325e-06, | |
| "loss": 0.0606, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 5.27969897080901e-06, | |
| "loss": 0.0484, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 0.0704, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "eval_loss": 0.05977718159556389, | |
| "eval_runtime": 5.9454, | |
| "eval_samples_per_second": 47.095, | |
| "eval_steps_per_second": 5.887, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2.9814044425935606e-06, | |
| "loss": 0.0534, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 2.073584124257899e-06, | |
| "loss": 0.0765, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 1.3287526608711131e-06, | |
| "loss": 0.0557, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 7.481497508972312e-07, | |
| "loss": 0.056, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 3.3274175058067846e-07, | |
| "loss": 0.0525, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_loss": 0.05975683405995369, | |
| "eval_runtime": 5.9096, | |
| "eval_samples_per_second": 47.381, | |
| "eval_steps_per_second": 5.923, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 8.322006554171146e-08, | |
| "loss": 0.055, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 0.0, | |
| "loss": 0.0625, | |
| "step": 87 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 87, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 5.5952097199364506e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |