| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.1988619347109912, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02395926924228811, | |
| "grad_norm": 16.852388381958008, | |
| "learning_rate": 1.1904761904761906e-06, | |
| "loss": 1.3705, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04791853848457622, | |
| "grad_norm": 4.8825883865356445, | |
| "learning_rate": 2.380952380952381e-06, | |
| "loss": 0.9316, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07187780772686433, | |
| "grad_norm": 2.1795146465301514, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.5955, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09583707696915245, | |
| "grad_norm": 1.2667920589447021, | |
| "learning_rate": 4.761904761904762e-06, | |
| "loss": 0.4746, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11979634621144056, | |
| "grad_norm": 1.1073658466339111, | |
| "learning_rate": 5.9523809523809525e-06, | |
| "loss": 0.421, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14375561545372867, | |
| "grad_norm": 1.2127255201339722, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.3877, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16771488469601678, | |
| "grad_norm": 1.0557327270507812, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.3716, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1916741539383049, | |
| "grad_norm": 1.04291570186615, | |
| "learning_rate": 9.523809523809525e-06, | |
| "loss": 0.3575, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.215633423180593, | |
| "grad_norm": 1.1732438802719116, | |
| "learning_rate": 9.9984209464165e-06, | |
| "loss": 0.3494, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.23959269242288112, | |
| "grad_norm": 1.0016248226165771, | |
| "learning_rate": 9.988774786134235e-06, | |
| "loss": 0.343, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2635519616651692, | |
| "grad_norm": 1.0416539907455444, | |
| "learning_rate": 9.970376619680024e-06, | |
| "loss": 0.3379, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.28751123090745734, | |
| "grad_norm": 0.9761713743209839, | |
| "learning_rate": 9.94325872368957e-06, | |
| "loss": 0.3297, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3114705001497454, | |
| "grad_norm": 1.0316894054412842, | |
| "learning_rate": 9.907468672167165e-06, | |
| "loss": 0.3317, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.33542976939203356, | |
| "grad_norm": 0.9167680740356445, | |
| "learning_rate": 9.863069253024719e-06, | |
| "loss": 0.3275, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.35938903863432164, | |
| "grad_norm": 0.8503734469413757, | |
| "learning_rate": 9.81013835793043e-06, | |
| "loss": 0.3237, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3833483078766098, | |
| "grad_norm": 0.8351507186889648, | |
| "learning_rate": 9.748768845660335e-06, | |
| "loss": 0.3213, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.40730757711889787, | |
| "grad_norm": 0.8993275165557861, | |
| "learning_rate": 9.679068379192455e-06, | |
| "loss": 0.3178, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.431266846361186, | |
| "grad_norm": 0.8393663167953491, | |
| "learning_rate": 9.601159236829353e-06, | |
| "loss": 0.3162, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4552261156034741, | |
| "grad_norm": 0.7886475920677185, | |
| "learning_rate": 9.515178097680437e-06, | |
| "loss": 0.3165, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.47918538484576223, | |
| "grad_norm": 0.8455172777175903, | |
| "learning_rate": 9.421275801880363e-06, | |
| "loss": 0.3118, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5031446540880503, | |
| "grad_norm": 0.8376420140266418, | |
| "learning_rate": 9.319617085964177e-06, | |
| "loss": 0.3102, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5271039233303384, | |
| "grad_norm": 0.8927198052406311, | |
| "learning_rate": 9.210380293863462e-06, | |
| "loss": 0.311, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5510631925726265, | |
| "grad_norm": 0.7255620360374451, | |
| "learning_rate": 9.093757064030473e-06, | |
| "loss": 0.3086, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5750224618149147, | |
| "grad_norm": 0.7063446640968323, | |
| "learning_rate": 8.969951993239177e-06, | |
| "loss": 0.3053, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5989817310572028, | |
| "grad_norm": 0.7932593822479248, | |
| "learning_rate": 8.83918227765299e-06, | |
| "loss": 0.304, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6229410002994908, | |
| "grad_norm": 0.7496311664581299, | |
| "learning_rate": 8.701677331788891e-06, | |
| "loss": 0.3022, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6469002695417789, | |
| "grad_norm": 0.7745276689529419, | |
| "learning_rate": 8.557678386046429e-06, | |
| "loss": 0.3033, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6708595387840671, | |
| "grad_norm": 0.68818598985672, | |
| "learning_rate": 8.4074380635076e-06, | |
| "loss": 0.3017, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6948188080263552, | |
| "grad_norm": 0.8030797243118286, | |
| "learning_rate": 8.251219936750145e-06, | |
| "loss": 0.3049, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7187780772686433, | |
| "grad_norm": 0.7876457571983337, | |
| "learning_rate": 8.089298065451673e-06, | |
| "loss": 0.297, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7427373465109314, | |
| "grad_norm": 0.693367600440979, | |
| "learning_rate": 7.921956515595861e-06, | |
| "loss": 0.2955, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7666966157532196, | |
| "grad_norm": 0.7320050001144409, | |
| "learning_rate": 7.7494888611242e-06, | |
| "loss": 0.2955, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7906558849955077, | |
| "grad_norm": 0.7215656042098999, | |
| "learning_rate": 7.572197668907533e-06, | |
| "loss": 0.2936, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8146151542377957, | |
| "grad_norm": 0.7600423693656921, | |
| "learning_rate": 7.390393967940962e-06, | |
| "loss": 0.2942, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8385744234800838, | |
| "grad_norm": 0.7513334155082703, | |
| "learning_rate": 7.2043967036932935e-06, | |
| "loss": 0.2938, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.862533692722372, | |
| "grad_norm": 0.6746324300765991, | |
| "learning_rate": 7.014532178568314e-06, | |
| "loss": 0.29, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8864929619646601, | |
| "grad_norm": 0.6711202263832092, | |
| "learning_rate": 6.821133479459492e-06, | |
| "loss": 0.2922, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9104522312069482, | |
| "grad_norm": 0.657655656337738, | |
| "learning_rate": 6.624539893402383e-06, | |
| "loss": 0.2867, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9344115004492363, | |
| "grad_norm": 0.762762188911438, | |
| "learning_rate": 6.425096312349881e-06, | |
| "loss": 0.2908, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9583707696915245, | |
| "grad_norm": 0.6959704756736755, | |
| "learning_rate": 6.223152628114537e-06, | |
| "loss": 0.2871, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9823300389338125, | |
| "grad_norm": 0.7447584271430969, | |
| "learning_rate": 6.019063118539425e-06, | |
| "loss": 0.2898, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0071877807726863, | |
| "grad_norm": 0.6954111456871033, | |
| "learning_rate": 5.813185825974419e-06, | |
| "loss": 0.2862, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0311470500149746, | |
| "grad_norm": 0.5891335606575012, | |
| "learning_rate": 5.605881929148254e-06, | |
| "loss": 0.2651, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0551063192572627, | |
| "grad_norm": 0.6528168320655823, | |
| "learning_rate": 5.3975151095383e-06, | |
| "loss": 0.2648, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0790655884995508, | |
| "grad_norm": 0.6022618412971497, | |
| "learning_rate": 5.188450913349674e-06, | |
| "loss": 0.2634, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.103024857741839, | |
| "grad_norm": 0.6223345994949341, | |
| "learning_rate": 4.979056110222982e-06, | |
| "loss": 0.2613, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 0.618979275226593, | |
| "learning_rate": 4.769698049795739e-06, | |
| "loss": 0.2626, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.150943396226415, | |
| "grad_norm": 0.5937119126319885, | |
| "learning_rate": 4.560744017246284e-06, | |
| "loss": 0.2597, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1749026654687031, | |
| "grad_norm": 0.6550348401069641, | |
| "learning_rate": 4.352560588950766e-06, | |
| "loss": 0.263, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.1988619347109912, | |
| "grad_norm": 0.5479949116706848, | |
| "learning_rate": 4.145512989383618e-06, | |
| "loss": 0.2599, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 834, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0159083954242585e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |