| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 739, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0013531799729364006, | |
| "grad_norm": 22.625, | |
| "learning_rate": 1.3513513513513515e-07, | |
| "loss": 2.1989, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013531799729364006, | |
| "grad_norm": 22.5, | |
| "learning_rate": 1.3513513513513515e-06, | |
| "loss": 2.2213, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02706359945872801, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 2.2125, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04059539918809202, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 4.0540540540540545e-06, | |
| "loss": 2.1298, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05412719891745602, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 2.0609, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06765899864682003, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 2.0108, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08119079837618404, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 1.918, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09472259810554803, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 9.45945945945946e-06, | |
| "loss": 1.8292, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10825439783491204, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 9.99799150901764e-06, | |
| "loss": 1.8037, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12178619756427606, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 9.985723240513795e-06, | |
| "loss": 1.7704, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.13531799729364005, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 9.962329873928743e-06, | |
| "loss": 1.7527, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14884979702300405, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 9.92786360901067e-06, | |
| "loss": 1.7479, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16238159675236807, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 9.882401353473711e-06, | |
| "loss": 1.7169, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17591339648173207, | |
| "grad_norm": 2.625, | |
| "learning_rate": 9.826044551386743e-06, | |
| "loss": 1.7196, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18944519621109607, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 9.758918956812024e-06, | |
| "loss": 1.7238, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2029769959404601, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 1.7291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2165087956698241, | |
| "grad_norm": 2.625, | |
| "learning_rate": 9.59298421915731e-06, | |
| "loss": 1.7071, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.23004059539918809, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 9.494545341361291e-06, | |
| "loss": 1.715, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2435723951285521, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 9.386077375438848e-06, | |
| "loss": 1.7119, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2571041948579161, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 9.267822355835402e-06, | |
| "loss": 1.7079, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2706359945872801, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 9.140044155740102e-06, | |
| "loss": 1.6912, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.28416779431664413, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 9.003027898281551e-06, | |
| "loss": 1.7097, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2976995940460081, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 8.85707932030663e-06, | |
| "loss": 1.6883, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3112313937753721, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 8.702524090162023e-06, | |
| "loss": 1.6917, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.32476319350473615, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 8.539707081000808e-06, | |
| "loss": 1.6951, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3382949932341001, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 8.36899160123559e-06, | |
| "loss": 1.7043, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.35182679296346414, | |
| "grad_norm": 2.5, | |
| "learning_rate": 8.190758583855379e-06, | |
| "loss": 1.6851, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.36535859269282817, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 8.005405736415127e-06, | |
| "loss": 1.6912, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.37889039242219213, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 7.813346653594667e-06, | |
| "loss": 1.7011, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.39242219215155616, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 7.615009894307263e-06, | |
| "loss": 1.6955, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4059539918809202, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 7.410838025417083e-06, | |
| "loss": 1.6848, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.41948579161028415, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 7.201286634199484e-06, | |
| "loss": 1.6749, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4330175913396482, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 6.986823311747652e-06, | |
| "loss": 1.6797, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4465493910690122, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 6.767926609594032e-06, | |
| "loss": 1.6744, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.46008119079837617, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 1.6636, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4736129905277402, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 6.3187956454196885e-06, | |
| "loss": 1.6949, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4871447902571042, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 6.0895635702004985e-06, | |
| "loss": 1.6903, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5006765899864682, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 5.857900252611959e-06, | |
| "loss": 1.6602, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5142083897158322, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 5.624322624101255e-06, | |
| "loss": 1.6888, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5277401894451962, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 5.3893518876917795e-06, | |
| "loss": 1.6876, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5412719891745602, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 5.153512354975388e-06, | |
| "loss": 1.656, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5548037889039242, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 4.917330276168208e-06, | |
| "loss": 1.6737, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5683355886332883, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.681332665840647e-06, | |
| "loss": 1.6856, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5818673883626523, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 4.446046126941801e-06, | |
| "loss": 1.6917, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5953991880920162, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 4.211995675742358e-06, | |
| "loss": 1.6728, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6089309878213802, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 3.979703570318017e-06, | |
| "loss": 1.6919, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6224627875507442, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 3.749688145187497e-06, | |
| "loss": 1.6789, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6359945872801083, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 3.5224626547055463e-06, | |
| "loss": 1.687, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6495263870094723, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 3.298534127791785e-06, | |
| "loss": 1.678, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6630581867388363, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 3.078402236550926e-06, | |
| "loss": 1.6756, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6765899864682002, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 2.86255818130892e-06, | |
| "loss": 1.6895, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6901217861975643, | |
| "grad_norm": 2.5, | |
| "learning_rate": 2.6514835945529706e-06, | |
| "loss": 1.685, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7036535859269283, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 2.4456494662211082e-06, | |
| "loss": 1.6879, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7171853856562923, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 1.6812, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7307171853856563, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 2.0515270521524562e-06, | |
| "loss": 1.6727, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7442489851150202, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.864118207632315e-06, | |
| "loss": 1.6839, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7577807848443843, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 1.683706741592327e-06, | |
| "loss": 1.6745, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7713125845737483, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 1.5106952225582312e-06, | |
| "loss": 1.677, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7848443843031123, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.3454697068804434e-06, | |
| "loss": 1.6677, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7983761840324763, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 1.1883988772913924e-06, | |
| "loss": 1.6725, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8119079837618404, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.0398332202301708e-06, | |
| "loss": 1.669, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8254397834912043, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 9.001042437702468e-07, | |
| "loss": 1.6775, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8389715832205683, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 7.695237378953224e-07, | |
| "loss": 1.6914, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8525033829499323, | |
| "grad_norm": 2.5, | |
| "learning_rate": 6.483830787739659e-07, | |
| "loss": 1.6743, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8660351826792964, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 5.369525785854368e-07, | |
| "loss": 1.6784, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8795669824086604, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 4.3548088234752814e-07, | |
| "loss": 1.687, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8930987821380244, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 3.4419441309229587e-07, | |
| "loss": 1.6729, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9066305818673883, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 2.6329686662774247e-07, | |
| "loss": 1.6775, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.9201623815967523, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 1.9296875701281858e-07, | |
| "loss": 1.6813, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9336941813261164, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 1.333670137599713e-07, | |
| "loss": 1.6944, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9472259810554804, | |
| "grad_norm": 2.75, | |
| "learning_rate": 8.462463166403978e-08, | |
| "loss": 1.6899, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9607577807848444, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 4.685037403886483e-08, | |
| "loss": 1.6669, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9742895805142084, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 1.6881, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9878213802435724, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.51872650170937e-09, | |
| "loss": 1.6671, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.7291793823242188, | |
| "eval_runtime": 27.3164, | |
| "eval_samples_per_second": 15.924, | |
| "eval_steps_per_second": 2.013, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 739, | |
| "total_flos": 7.51411531916247e+16, | |
| "train_loss": 1.726003464891075, | |
| "train_runtime": 1754.8001, | |
| "train_samples_per_second": 3.368, | |
| "train_steps_per_second": 0.421 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 739, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.51411531916247e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |