| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 264, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07633587786259542, | |
| "grad_norm": 4.742532677896037, | |
| "learning_rate": 2.962962962962963e-06, | |
| "loss": 1.0457, | |
| "mean_token_accuracy": 0.783866959810257, | |
| "num_tokens": 787708.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.15267175572519084, | |
| "grad_norm": 1.9625388369532857, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.8897, | |
| "mean_token_accuracy": 0.7999136686325073, | |
| "num_tokens": 1584559.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22900763358778625, | |
| "grad_norm": 1.4592114838143702, | |
| "learning_rate": 1.037037037037037e-05, | |
| "loss": 0.6509, | |
| "mean_token_accuracy": 0.8498281538486481, | |
| "num_tokens": 2395528.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3053435114503817, | |
| "grad_norm": 0.5742660222521939, | |
| "learning_rate": 1.4074074074074075e-05, | |
| "loss": 0.5224, | |
| "mean_token_accuracy": 0.877668297290802, | |
| "num_tokens": 3193615.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3816793893129771, | |
| "grad_norm": 1.5539399744331717, | |
| "learning_rate": 1.7777777777777777e-05, | |
| "loss": 0.603, | |
| "mean_token_accuracy": 0.8636772155761718, | |
| "num_tokens": 4001905.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4580152671755725, | |
| "grad_norm": 1.383692795108315, | |
| "learning_rate": 1.9831223628691984e-05, | |
| "loss": 0.6026, | |
| "mean_token_accuracy": 0.8651778519153595, | |
| "num_tokens": 4805947.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5343511450381679, | |
| "grad_norm": 0.7212760668289794, | |
| "learning_rate": 1.9409282700421944e-05, | |
| "loss": 0.4854, | |
| "mean_token_accuracy": 0.8840490877628326, | |
| "num_tokens": 5613999.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6106870229007634, | |
| "grad_norm": 0.6119613892318467, | |
| "learning_rate": 1.89873417721519e-05, | |
| "loss": 0.5691, | |
| "mean_token_accuracy": 0.8745659828186035, | |
| "num_tokens": 6423827.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6870229007633588, | |
| "grad_norm": 0.47676818080470346, | |
| "learning_rate": 1.856540084388186e-05, | |
| "loss": 0.5074, | |
| "mean_token_accuracy": 0.8846244931221008, | |
| "num_tokens": 7239602.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7633587786259542, | |
| "grad_norm": 0.514784662228685, | |
| "learning_rate": 1.8143459915611816e-05, | |
| "loss": 0.4873, | |
| "mean_token_accuracy": 0.8858396947383881, | |
| "num_tokens": 8010548.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8396946564885496, | |
| "grad_norm": 0.5009616332970619, | |
| "learning_rate": 1.7721518987341772e-05, | |
| "loss": 0.5417, | |
| "mean_token_accuracy": 0.8772693455219269, | |
| "num_tokens": 8811142.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.916030534351145, | |
| "grad_norm": 0.6432299623723948, | |
| "learning_rate": 1.729957805907173e-05, | |
| "loss": 0.5506, | |
| "mean_token_accuracy": 0.8746115028858185, | |
| "num_tokens": 9606796.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9923664122137404, | |
| "grad_norm": 0.4603072170202572, | |
| "learning_rate": 1.687763713080169e-05, | |
| "loss": 0.4578, | |
| "mean_token_accuracy": 0.8920021951198578, | |
| "num_tokens": 10403148.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0610687022900764, | |
| "grad_norm": 3.3165369083686143, | |
| "learning_rate": 1.6455696202531647e-05, | |
| "loss": 0.4402, | |
| "mean_token_accuracy": 0.899945510758294, | |
| "num_tokens": 11048906.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1374045801526718, | |
| "grad_norm": 2.766201516331391, | |
| "learning_rate": 1.6033755274261603e-05, | |
| "loss": 0.4365, | |
| "mean_token_accuracy": 0.9022025763988495, | |
| "num_tokens": 11854043.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.213740458015267, | |
| "grad_norm": 0.48053334497719935, | |
| "learning_rate": 1.5611814345991563e-05, | |
| "loss": 0.3459, | |
| "mean_token_accuracy": 0.9143236577510834, | |
| "num_tokens": 12656567.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2900763358778626, | |
| "grad_norm": 0.4622047473779545, | |
| "learning_rate": 1.5189873417721521e-05, | |
| "loss": 0.3698, | |
| "mean_token_accuracy": 0.9105424880981445, | |
| "num_tokens": 13471671.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.366412213740458, | |
| "grad_norm": 0.4223954903184026, | |
| "learning_rate": 1.4767932489451477e-05, | |
| "loss": 0.3808, | |
| "mean_token_accuracy": 0.9093575775623322, | |
| "num_tokens": 14265714.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4427480916030535, | |
| "grad_norm": 0.43148362035876203, | |
| "learning_rate": 1.4345991561181437e-05, | |
| "loss": 0.4011, | |
| "mean_token_accuracy": 0.9054655790328979, | |
| "num_tokens": 15077565.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5190839694656488, | |
| "grad_norm": 0.40067549513568396, | |
| "learning_rate": 1.3924050632911395e-05, | |
| "loss": 0.3006, | |
| "mean_token_accuracy": 0.9241831779479981, | |
| "num_tokens": 15879506.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5954198473282442, | |
| "grad_norm": 0.46303246451803154, | |
| "learning_rate": 1.350210970464135e-05, | |
| "loss": 0.3233, | |
| "mean_token_accuracy": 0.9187310457229614, | |
| "num_tokens": 16679230.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.6717557251908397, | |
| "grad_norm": 0.4070661383743064, | |
| "learning_rate": 1.3080168776371309e-05, | |
| "loss": 0.378, | |
| "mean_token_accuracy": 0.90953728556633, | |
| "num_tokens": 17457542.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7480916030534353, | |
| "grad_norm": 0.769113494701985, | |
| "learning_rate": 1.2658227848101268e-05, | |
| "loss": 0.3829, | |
| "mean_token_accuracy": 0.9075210094451904, | |
| "num_tokens": 18274184.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8244274809160306, | |
| "grad_norm": 4.081490569449155, | |
| "learning_rate": 1.2236286919831224e-05, | |
| "loss": 0.4123, | |
| "mean_token_accuracy": 0.904869532585144, | |
| "num_tokens": 19083548.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.900763358778626, | |
| "grad_norm": 0.49471022422420063, | |
| "learning_rate": 1.1814345991561182e-05, | |
| "loss": 0.4281, | |
| "mean_token_accuracy": 0.8996932089328766, | |
| "num_tokens": 19876168.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.9770992366412212, | |
| "grad_norm": 0.40467517940465514, | |
| "learning_rate": 1.139240506329114e-05, | |
| "loss": 0.3043, | |
| "mean_token_accuracy": 0.9238722801208497, | |
| "num_tokens": 20673176.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.045801526717557, | |
| "grad_norm": 0.43951149160060266, | |
| "learning_rate": 1.0970464135021096e-05, | |
| "loss": 0.2862, | |
| "mean_token_accuracy": 0.9268535508049859, | |
| "num_tokens": 21344850.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.122137404580153, | |
| "grad_norm": 0.41447293210123126, | |
| "learning_rate": 1.0548523206751056e-05, | |
| "loss": 0.2231, | |
| "mean_token_accuracy": 0.9415154874324798, | |
| "num_tokens": 22160398.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.198473282442748, | |
| "grad_norm": 0.4452926170537663, | |
| "learning_rate": 1.0126582278481014e-05, | |
| "loss": 0.1973, | |
| "mean_token_accuracy": 0.9481983184814453, | |
| "num_tokens": 22965629.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.2748091603053435, | |
| "grad_norm": 0.4020318700510315, | |
| "learning_rate": 9.704641350210972e-06, | |
| "loss": 0.3415, | |
| "mean_token_accuracy": 0.9216605007648468, | |
| "num_tokens": 23741814.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.351145038167939, | |
| "grad_norm": 0.5262769275300023, | |
| "learning_rate": 9.28270042194093e-06, | |
| "loss": 0.3558, | |
| "mean_token_accuracy": 0.9182116627693176, | |
| "num_tokens": 24530627.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.427480916030534, | |
| "grad_norm": 0.3849592272345829, | |
| "learning_rate": 8.860759493670886e-06, | |
| "loss": 0.3288, | |
| "mean_token_accuracy": 0.9241446554660797, | |
| "num_tokens": 25322490.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5038167938931295, | |
| "grad_norm": 0.41304782030368414, | |
| "learning_rate": 8.438818565400846e-06, | |
| "loss": 0.1962, | |
| "mean_token_accuracy": 0.9485301196575164, | |
| "num_tokens": 26120386.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.5801526717557253, | |
| "grad_norm": 0.4001920507927691, | |
| "learning_rate": 8.016877637130802e-06, | |
| "loss": 0.2517, | |
| "mean_token_accuracy": 0.9370044708251953, | |
| "num_tokens": 26913645.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.6564885496183206, | |
| "grad_norm": 0.3240700971303689, | |
| "learning_rate": 7.5949367088607605e-06, | |
| "loss": 0.2148, | |
| "mean_token_accuracy": 0.9440955638885498, | |
| "num_tokens": 27700220.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.732824427480916, | |
| "grad_norm": 0.4462899971482025, | |
| "learning_rate": 7.172995780590718e-06, | |
| "loss": 0.2412, | |
| "mean_token_accuracy": 0.9370499432086945, | |
| "num_tokens": 28511433.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.8091603053435117, | |
| "grad_norm": 0.37405261169716236, | |
| "learning_rate": 6.751054852320675e-06, | |
| "loss": 0.1959, | |
| "mean_token_accuracy": 0.9486702799797058, | |
| "num_tokens": 29320345.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.885496183206107, | |
| "grad_norm": 0.5070713681058377, | |
| "learning_rate": 6.329113924050634e-06, | |
| "loss": 0.2469, | |
| "mean_token_accuracy": 0.937513917684555, | |
| "num_tokens": 30126410.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.9618320610687023, | |
| "grad_norm": 0.4518587773437182, | |
| "learning_rate": 5.907172995780591e-06, | |
| "loss": 0.2986, | |
| "mean_token_accuracy": 0.9325798094272614, | |
| "num_tokens": 30943204.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.030534351145038, | |
| "grad_norm": 0.6456773562151603, | |
| "learning_rate": 5.485232067510548e-06, | |
| "loss": 0.198, | |
| "mean_token_accuracy": 0.9480705327457852, | |
| "num_tokens": 31617040.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.1068702290076335, | |
| "grad_norm": 0.35354935160147144, | |
| "learning_rate": 5.063291139240507e-06, | |
| "loss": 0.1417, | |
| "mean_token_accuracy": 0.9633415877819062, | |
| "num_tokens": 32420901.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.183206106870229, | |
| "grad_norm": 0.43025067497911046, | |
| "learning_rate": 4.641350210970465e-06, | |
| "loss": 0.1633, | |
| "mean_token_accuracy": 0.9574208796024323, | |
| "num_tokens": 33221997.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.2595419847328246, | |
| "grad_norm": 0.44274438323121784, | |
| "learning_rate": 4.219409282700423e-06, | |
| "loss": 0.1728, | |
| "mean_token_accuracy": 0.9585293054580688, | |
| "num_tokens": 34035651.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.33587786259542, | |
| "grad_norm": 0.5041692104858938, | |
| "learning_rate": 3.7974683544303802e-06, | |
| "loss": 0.2769, | |
| "mean_token_accuracy": 0.940484470129013, | |
| "num_tokens": 34845591.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.4122137404580153, | |
| "grad_norm": 0.4135851021335505, | |
| "learning_rate": 3.3755274261603377e-06, | |
| "loss": 0.2343, | |
| "mean_token_accuracy": 0.9466347515583038, | |
| "num_tokens": 35646568.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.4885496183206106, | |
| "grad_norm": 0.3433633154751686, | |
| "learning_rate": 2.9535864978902956e-06, | |
| "loss": 0.1566, | |
| "mean_token_accuracy": 0.9608495891094208, | |
| "num_tokens": 36438898.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.564885496183206, | |
| "grad_norm": 0.3906070844239771, | |
| "learning_rate": 2.5316455696202535e-06, | |
| "loss": 0.23, | |
| "mean_token_accuracy": 0.9464579701423645, | |
| "num_tokens": 37250943.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.6412213740458013, | |
| "grad_norm": 0.38849266425243045, | |
| "learning_rate": 2.1097046413502114e-06, | |
| "loss": 0.134, | |
| "mean_token_accuracy": 0.9657551169395446, | |
| "num_tokens": 38038753.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.717557251908397, | |
| "grad_norm": 0.44426004445172707, | |
| "learning_rate": 1.6877637130801689e-06, | |
| "loss": 0.1854, | |
| "mean_token_accuracy": 0.9568316519260407, | |
| "num_tokens": 38839070.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.7938931297709924, | |
| "grad_norm": 0.46841967781432653, | |
| "learning_rate": 1.2658227848101267e-06, | |
| "loss": 0.1306, | |
| "mean_token_accuracy": 0.9657817900180816, | |
| "num_tokens": 39646744.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.8702290076335877, | |
| "grad_norm": 0.39182807996616936, | |
| "learning_rate": 8.438818565400844e-07, | |
| "loss": 0.1723, | |
| "mean_token_accuracy": 0.9591064333915711, | |
| "num_tokens": 40443210.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.9465648854961835, | |
| "grad_norm": 0.3836010955535366, | |
| "learning_rate": 4.219409282700422e-07, | |
| "loss": 0.139, | |
| "mean_token_accuracy": 0.9641590178012848, | |
| "num_tokens": 41234126.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "mean_token_accuracy": 0.952833524772099, | |
| "num_tokens": 41735472.0, | |
| "step": 264, | |
| "total_flos": 83711346343936.0, | |
| "train_loss": 0.35391224627241946, | |
| "train_runtime": 949.067, | |
| "train_samples_per_second": 4.396, | |
| "train_steps_per_second": 0.278 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 264, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 83711346343936.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |