| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 2703, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05549389567147614, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 4.907510173880873e-05, | |
| "loss": 1.186, | |
| "mean_token_accuracy": 0.7049988703429699, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11098779134295228, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.8150203477617464e-05, | |
| "loss": 1.1463, | |
| "mean_token_accuracy": 0.7061308635771275, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16648168701442842, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 4.72253052164262e-05, | |
| "loss": 0.9759, | |
| "mean_token_accuracy": 0.7407757295668125, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.22197558268590456, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 4.6300406955234926e-05, | |
| "loss": 1.0052, | |
| "mean_token_accuracy": 0.7320078010857105, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27746947835738067, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.5375508694043653e-05, | |
| "loss": 1.0511, | |
| "mean_token_accuracy": 0.7330576995015144, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.33296337402885684, | |
| "grad_norm": 9.125, | |
| "learning_rate": 4.445061043285239e-05, | |
| "loss": 0.9694, | |
| "mean_token_accuracy": 0.7509057518839836, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.38845726970033295, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 4.352571217166112e-05, | |
| "loss": 1.029, | |
| "mean_token_accuracy": 0.7393789207935333, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4439511653718091, | |
| "grad_norm": 9.25, | |
| "learning_rate": 4.260081391046985e-05, | |
| "loss": 0.9199, | |
| "mean_token_accuracy": 0.760094360858202, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.49944506104328523, | |
| "grad_norm": 9.125, | |
| "learning_rate": 4.167591564927858e-05, | |
| "loss": 0.9908, | |
| "mean_token_accuracy": 0.7447640253603458, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5549389567147613, | |
| "grad_norm": 8.25, | |
| "learning_rate": 4.075101738808732e-05, | |
| "loss": 0.8984, | |
| "mean_token_accuracy": 0.7630760133266449, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6104328523862376, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 3.9826119126896046e-05, | |
| "loss": 0.9407, | |
| "mean_token_accuracy": 0.757173229753971, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6659267480577137, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 3.890122086570477e-05, | |
| "loss": 0.9575, | |
| "mean_token_accuracy": 0.7462678501009941, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7214206437291898, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 3.79763226045135e-05, | |
| "loss": 0.8748, | |
| "mean_token_accuracy": 0.7648530495911836, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7769145394006659, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.705142434332224e-05, | |
| "loss": 0.9121, | |
| "mean_token_accuracy": 0.7616761503368616, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.832408435072142, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 3.612652608213097e-05, | |
| "loss": 0.8764, | |
| "mean_token_accuracy": 0.7672346414625645, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8879023307436182, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 3.5201627820939697e-05, | |
| "loss": 0.8334, | |
| "mean_token_accuracy": 0.7711550496518612, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9433962264150944, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 3.4276729559748424e-05, | |
| "loss": 0.8035, | |
| "mean_token_accuracy": 0.7846515120565891, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9988901220865705, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 3.3351831298557165e-05, | |
| "loss": 0.784, | |
| "mean_token_accuracy": 0.7859167304635047, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.0543840177580466, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 3.242693303736589e-05, | |
| "loss": 0.4242, | |
| "mean_token_accuracy": 0.8643624025583267, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.1098779134295227, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 3.150203477617462e-05, | |
| "loss": 0.4209, | |
| "mean_token_accuracy": 0.8601007984578609, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1653718091009988, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 3.0577136514983354e-05, | |
| "loss": 0.4225, | |
| "mean_token_accuracy": 0.8565894016623496, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.220865704772475, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 2.9652238253792085e-05, | |
| "loss": 0.3897, | |
| "mean_token_accuracy": 0.8680702060461044, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2763596004439512, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.8727339992600816e-05, | |
| "loss": 0.43, | |
| "mean_token_accuracy": 0.8627835646271705, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.3318534961154274, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 2.7802441731409544e-05, | |
| "loss": 0.4254, | |
| "mean_token_accuracy": 0.8601828774809838, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.3873473917869035, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 2.6877543470218275e-05, | |
| "loss": 0.404, | |
| "mean_token_accuracy": 0.8627345436811447, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.4428412874583796, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 2.595264520902701e-05, | |
| "loss": 0.406, | |
| "mean_token_accuracy": 0.8670766687393189, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.4983351831298557, | |
| "grad_norm": 7.875, | |
| "learning_rate": 2.502774694783574e-05, | |
| "loss": 0.4123, | |
| "mean_token_accuracy": 0.8563414543867112, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.5538290788013318, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 2.410284868664447e-05, | |
| "loss": 0.3906, | |
| "mean_token_accuracy": 0.8690173977613449, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.609322974472808, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 2.31779504254532e-05, | |
| "loss": 0.3896, | |
| "mean_token_accuracy": 0.8703913018107414, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.6648168701442843, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.2253052164261932e-05, | |
| "loss": 0.3916, | |
| "mean_token_accuracy": 0.8693778309226036, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.7203107658157601, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 2.1328153903070663e-05, | |
| "loss": 0.3955, | |
| "mean_token_accuracy": 0.8697583265602589, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.7758046614872365, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 2.0403255641879394e-05, | |
| "loss": 0.4101, | |
| "mean_token_accuracy": 0.8629062753915787, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.8312985571587126, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 1.9478357380688125e-05, | |
| "loss": 0.3577, | |
| "mean_token_accuracy": 0.8781416714191437, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.8867924528301887, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.8553459119496856e-05, | |
| "loss": 0.3991, | |
| "mean_token_accuracy": 0.8602015821635723, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.9422863485016648, | |
| "grad_norm": 3.5, | |
| "learning_rate": 1.7628560858305587e-05, | |
| "loss": 0.3717, | |
| "mean_token_accuracy": 0.8744880908727646, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.997780244173141, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 1.670366259711432e-05, | |
| "loss": 0.3798, | |
| "mean_token_accuracy": 0.8699861750006675, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.0532741398446173, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 1.577876433592305e-05, | |
| "loss": 0.156, | |
| "mean_token_accuracy": 0.9473638749122619, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.108768035516093, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 1.4853866074731781e-05, | |
| "loss": 0.1442, | |
| "mean_token_accuracy": 0.9488543626666069, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.1642619311875695, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 1.392896781354051e-05, | |
| "loss": 0.1507, | |
| "mean_token_accuracy": 0.9458903276920319, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.2197558268590454, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 1.3004069552349243e-05, | |
| "loss": 0.1485, | |
| "mean_token_accuracy": 0.9459331405162811, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.2752497225305217, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 1.2079171291157974e-05, | |
| "loss": 0.137, | |
| "mean_token_accuracy": 0.946778584420681, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.3307436182019976, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 1.1154273029966705e-05, | |
| "loss": 0.1302, | |
| "mean_token_accuracy": 0.9501716086268425, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.386237513873474, | |
| "grad_norm": 3.875, | |
| "learning_rate": 1.0229374768775436e-05, | |
| "loss": 0.1487, | |
| "mean_token_accuracy": 0.9474012264609337, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.44173140954495, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 9.304476507584166e-06, | |
| "loss": 0.1371, | |
| "mean_token_accuracy": 0.9514653950929641, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.497225305216426, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 8.379578246392897e-06, | |
| "loss": 0.1384, | |
| "mean_token_accuracy": 0.9439902834594249, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.5527192008879025, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 7.454679985201627e-06, | |
| "loss": 0.1357, | |
| "mean_token_accuracy": 0.9507204574346543, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.6082130965593784, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 6.529781724010358e-06, | |
| "loss": 0.1284, | |
| "mean_token_accuracy": 0.9505923983454704, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.6637069922308547, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 5.60488346281909e-06, | |
| "loss": 0.1327, | |
| "mean_token_accuracy": 0.9514957949519157, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.7192008879023306, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 4.679985201627821e-06, | |
| "loss": 0.1422, | |
| "mean_token_accuracy": 0.950286613702774, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.774694783573807, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 3.7550869404365522e-06, | |
| "loss": 0.1351, | |
| "mean_token_accuracy": 0.9470736330747604, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.830188679245283, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 2.830188679245283e-06, | |
| "loss": 0.1247, | |
| "mean_token_accuracy": 0.9515706104040146, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.885682574916759, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 1.9052904180540142e-06, | |
| "loss": 0.1368, | |
| "mean_token_accuracy": 0.9486922469735145, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 9.80392156862745e-07, | |
| "loss": 0.1282, | |
| "mean_token_accuracy": 0.9518065723776817, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.9966703662597114, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 5.549389567147614e-08, | |
| "loss": 0.1399, | |
| "mean_token_accuracy": 0.9513461437821388, | |
| "step": 2700 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 2703, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.0206819147904e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |