| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 100, |
| "global_step": 1875, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 1.3540814772248269, |
| "epoch": 0.16, |
| "grad_norm": 0.29296875, |
| "learning_rate": 0.00019973673694024, |
| "loss": 1.4633206176757811, |
| "mean_token_accuracy": 0.6500224708765745, |
| "num_tokens": 1238177.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_entropy": 1.0408895498230344, |
| "eval_loss": 1.0171571969985962, |
| "eval_mean_token_accuracy": 0.7146944498258924, |
| "eval_num_tokens": 1238177.0, |
| "eval_runtime": 78.8867, |
| "eval_samples_per_second": 6.338, |
| "eval_steps_per_second": 0.799, |
| "step": 100 |
| }, |
| { |
| "entropy": 0.9953572849929333, |
| "epoch": 0.32, |
| "grad_norm": 0.283203125, |
| "learning_rate": 0.00019700443730801413, |
| "loss": 0.9830435943603516, |
| "mean_token_accuracy": 0.7228847779333591, |
| "num_tokens": 2525507.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_entropy": 1.0137588391228327, |
| "eval_loss": 0.9846630692481995, |
| "eval_mean_token_accuracy": 0.7225528132347834, |
| "eval_num_tokens": 2525507.0, |
| "eval_runtime": 78.7428, |
| "eval_samples_per_second": 6.35, |
| "eval_steps_per_second": 0.8, |
| "step": 200 |
| }, |
| { |
| "entropy": 0.998912007957697, |
| "epoch": 0.48, |
| "grad_norm": 0.392578125, |
| "learning_rate": 0.00019138263836827288, |
| "loss": 0.9803644561767578, |
| "mean_token_accuracy": 0.7225449255108833, |
| "num_tokens": 3772872.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_entropy": 0.9792825617487468, |
| "eval_loss": 0.9684324860572815, |
| "eval_mean_token_accuracy": 0.7254447255815778, |
| "eval_num_tokens": 3772872.0, |
| "eval_runtime": 78.7266, |
| "eval_samples_per_second": 6.351, |
| "eval_steps_per_second": 0.8, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.9658883933722973, |
| "epoch": 0.64, |
| "grad_norm": 0.314453125, |
| "learning_rate": 0.00018303879827647975, |
| "loss": 0.9450923919677734, |
| "mean_token_accuracy": 0.7311204792559147, |
| "num_tokens": 4984433.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_entropy": 0.9519493419026571, |
| "eval_loss": 0.9587358832359314, |
| "eval_mean_token_accuracy": 0.7281927579925174, |
| "eval_num_tokens": 4984433.0, |
| "eval_runtime": 78.7548, |
| "eval_samples_per_second": 6.349, |
| "eval_steps_per_second": 0.8, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.9506746862828731, |
| "epoch": 0.8, |
| "grad_norm": 0.287109375, |
| "learning_rate": 0.00017222145741734626, |
| "loss": 0.9352159118652343, |
| "mean_token_accuracy": 0.7334290930628776, |
| "num_tokens": 6278212.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_entropy": 0.9630418930734906, |
| "eval_loss": 0.9499430060386658, |
| "eval_mean_token_accuracy": 0.7301400247074309, |
| "eval_num_tokens": 6278212.0, |
| "eval_runtime": 78.8947, |
| "eval_samples_per_second": 6.338, |
| "eval_steps_per_second": 0.799, |
| "step": 500 |
| }, |
| { |
| "entropy": 0.9404085071384907, |
| "epoch": 0.96, |
| "grad_norm": 0.36328125, |
| "learning_rate": 0.0001592528350603103, |
| "loss": 0.9235909271240235, |
| "mean_token_accuracy": 0.7360797208547593, |
| "num_tokens": 7515414.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_entropy": 0.9610084844014001, |
| "eval_loss": 0.9445458054542542, |
| "eval_mean_token_accuracy": 0.7312900613224695, |
| "eval_num_tokens": 7515414.0, |
| "eval_runtime": 78.809, |
| "eval_samples_per_second": 6.344, |
| "eval_steps_per_second": 0.799, |
| "step": 600 |
| }, |
| { |
| "entropy": 0.8863665771484375, |
| "epoch": 1.12, |
| "grad_norm": 0.33984375, |
| "learning_rate": 0.0001445192313207067, |
| "loss": 0.8664443969726563, |
| "mean_token_accuracy": 0.751860016733408, |
| "num_tokens": 8791049.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_entropy": 0.8678532384690785, |
| "eval_loss": 0.9490408897399902, |
| "eval_mean_token_accuracy": 0.7310751146740384, |
| "eval_num_tokens": 8791049.0, |
| "eval_runtime": 78.889, |
| "eval_samples_per_second": 6.338, |
| "eval_steps_per_second": 0.799, |
| "step": 700 |
| }, |
| { |
| "entropy": 0.8653571680933237, |
| "epoch": 1.28, |
| "grad_norm": 0.341796875, |
| "learning_rate": 0.0001284595203261965, |
| "loss": 0.8425308227539062, |
| "mean_token_accuracy": 0.7570811548829078, |
| "num_tokens": 10023730.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_entropy": 0.8799236341128274, |
| "eval_loss": 0.9464381337165833, |
| "eval_mean_token_accuracy": 0.731743007425278, |
| "eval_num_tokens": 10023730.0, |
| "eval_runtime": 78.7704, |
| "eval_samples_per_second": 6.348, |
| "eval_steps_per_second": 0.8, |
| "step": 800 |
| }, |
| { |
| "entropy": 0.8525189873576164, |
| "epoch": 1.44, |
| "grad_norm": 0.326171875, |
| "learning_rate": 0.00011155207734584263, |
| "loss": 0.8311178588867187, |
| "mean_token_accuracy": 0.758762137889862, |
| "num_tokens": 11272481.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_entropy": 0.8737099454516456, |
| "eval_loss": 0.9428508281707764, |
| "eval_mean_token_accuracy": 0.7328412476040068, |
| "eval_num_tokens": 11272481.0, |
| "eval_runtime": 78.9808, |
| "eval_samples_per_second": 6.331, |
| "eval_steps_per_second": 0.798, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.8523794415593148, |
| "epoch": 1.6, |
| "grad_norm": 0.322265625, |
| "learning_rate": 9.430052928723153e-05, |
| "loss": 0.8334156799316407, |
| "mean_token_accuracy": 0.7607765494287014, |
| "num_tokens": 12541228.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_entropy": 0.8634761429968334, |
| "eval_loss": 0.939626932144165, |
| "eval_mean_token_accuracy": 0.7337213830342368, |
| "eval_num_tokens": 12541228.0, |
| "eval_runtime": 78.7473, |
| "eval_samples_per_second": 6.349, |
| "eval_steps_per_second": 0.8, |
| "step": 1000 |
| }, |
| { |
| "entropy": 0.848904173374176, |
| "epoch": 1.76, |
| "grad_norm": 0.296875, |
| "learning_rate": 7.721875301571359e-05, |
| "loss": 0.8279781341552734, |
| "mean_token_accuracy": 0.761102753430605, |
| "num_tokens": 13787594.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_entropy": 0.8816202557276166, |
| "eval_loss": 0.9332679510116577, |
| "eval_mean_token_accuracy": 0.7345179215310111, |
| "eval_num_tokens": 13787594.0, |
| "eval_runtime": 78.8468, |
| "eval_samples_per_second": 6.341, |
| "eval_steps_per_second": 0.799, |
| "step": 1100 |
| }, |
| { |
| "entropy": 0.8487374657392501, |
| "epoch": 1.92, |
| "grad_norm": 0.361328125, |
| "learning_rate": 6.0815568355179556e-05, |
| "loss": 0.8284202575683594, |
| "mean_token_accuracy": 0.7601838798820972, |
| "num_tokens": 15024952.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_entropy": 0.8655259221319168, |
| "eval_loss": 0.9299662113189697, |
| "eval_mean_token_accuracy": 0.7359667079789298, |
| "eval_num_tokens": 15024952.0, |
| "eval_runtime": 78.81, |
| "eval_samples_per_second": 6.344, |
| "eval_steps_per_second": 0.799, |
| "step": 1200 |
| }, |
| { |
| "entropy": 0.7959265542775392, |
| "epoch": 2.08, |
| "grad_norm": 0.39453125, |
| "learning_rate": 4.5579581724397255e-05, |
| "loss": 0.765469741821289, |
| "mean_token_accuracy": 0.7783753572404385, |
| "num_tokens": 16289561.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_entropy": 0.7987193597687615, |
| "eval_loss": 0.9503117799758911, |
| "eval_mean_token_accuracy": 0.73391526464432, |
| "eval_num_tokens": 16289561.0, |
| "eval_runtime": 78.8942, |
| "eval_samples_per_second": 6.338, |
| "eval_steps_per_second": 0.799, |
| "step": 1300 |
| }, |
| { |
| "entropy": 0.7321610506623983, |
| "epoch": 2.24, |
| "grad_norm": 0.392578125, |
| "learning_rate": 3.196463187590929e-05, |
| "loss": 0.6977394866943359, |
| "mean_token_accuracy": 0.7958511321246624, |
| "num_tokens": 17566179.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_entropy": 0.7957971247415694, |
| "eval_loss": 0.9516386985778809, |
| "eval_mean_token_accuracy": 0.7339397157941546, |
| "eval_num_tokens": 17566179.0, |
| "eval_runtime": 78.9187, |
| "eval_samples_per_second": 6.336, |
| "eval_steps_per_second": 0.798, |
| "step": 1400 |
| }, |
| { |
| "entropy": 0.7484312203526496, |
| "epoch": 2.4, |
| "grad_norm": 0.36328125, |
| "learning_rate": 2.0376271269487514e-05, |
| "loss": 0.7148262023925781, |
| "mean_token_accuracy": 0.7925588050484658, |
| "num_tokens": 18823409.0, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_entropy": 0.7901531098380922, |
| "eval_loss": 0.952657163143158, |
| "eval_mean_token_accuracy": 0.7342138498548477, |
| "eval_num_tokens": 18823409.0, |
| "eval_runtime": 78.7246, |
| "eval_samples_per_second": 6.351, |
| "eval_steps_per_second": 0.8, |
| "step": 1500 |
| }, |
| { |
| "entropy": 0.7353997033834457, |
| "epoch": 2.56, |
| "grad_norm": 0.404296875, |
| "learning_rate": 1.1159685763395111e-05, |
| "loss": 0.6989010620117188, |
| "mean_token_accuracy": 0.7955636675655842, |
| "num_tokens": 20029991.0, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_entropy": 0.7921941649346125, |
| "eval_loss": 0.9510936141014099, |
| "eval_mean_token_accuracy": 0.7344856574421837, |
| "eval_num_tokens": 20029991.0, |
| "eval_runtime": 78.6039, |
| "eval_samples_per_second": 6.361, |
| "eval_steps_per_second": 0.801, |
| "step": 1600 |
| }, |
| { |
| "entropy": 0.7453224293142557, |
| "epoch": 2.7199999999999998, |
| "grad_norm": 0.376953125, |
| "learning_rate": 4.58941246311464e-06, |
| "loss": 0.712043685913086, |
| "mean_token_accuracy": 0.7924758359789849, |
| "num_tokens": 21285971.0, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_entropy": 0.7924881426114885, |
| "eval_loss": 0.9510090351104736, |
| "eval_mean_token_accuracy": 0.7344099529205806, |
| "eval_num_tokens": 21285971.0, |
| "eval_runtime": 78.7793, |
| "eval_samples_per_second": 6.347, |
| "eval_steps_per_second": 0.8, |
| "step": 1700 |
| }, |
| { |
| "entropy": 0.7430764560401439, |
| "epoch": 2.88, |
| "grad_norm": 0.404296875, |
| "learning_rate": 8.611620049653879e-07, |
| "loss": 0.7093801879882813, |
| "mean_token_accuracy": 0.7932735744118691, |
| "num_tokens": 22529694.0, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_entropy": 0.7928767279973106, |
| "eval_loss": 0.9508963823318481, |
| "eval_mean_token_accuracy": 0.7343375957201398, |
| "eval_num_tokens": 22529694.0, |
| "eval_runtime": 78.867, |
| "eval_samples_per_second": 6.34, |
| "eval_steps_per_second": 0.799, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_entropy": 0.7928942896070934, |
| "eval_loss": 0.9508863687515259, |
| "eval_mean_token_accuracy": 0.734485215610928, |
| "eval_num_tokens": 23462628.0, |
| "eval_runtime": 78.8215, |
| "eval_samples_per_second": 6.343, |
| "eval_steps_per_second": 0.799, |
| "step": 1875 |
| } |
| ], |
| "logging_steps": 100, |
| "max_steps": 1875, |
| "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": 1.6703999720219935e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
