| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 20, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 0.8752671869306899, | |
| "learning_rate": 9.692307692307693e-06, | |
| "loss": 0.2364, | |
| "mean_token_accuracy": 0.927388173341751, | |
| "num_tokens": 654484.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.5912352927421961, | |
| "learning_rate": 9.307692307692308e-06, | |
| "loss": 0.2175, | |
| "mean_token_accuracy": 0.9301090151071548, | |
| "num_tokens": 1308735.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 0.5167232252277533, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 0.2043, | |
| "mean_token_accuracy": 0.933148954808712, | |
| "num_tokens": 1963205.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.43810472204253587, | |
| "learning_rate": 8.53846153846154e-06, | |
| "loss": 0.2092, | |
| "mean_token_accuracy": 0.9336670354008675, | |
| "num_tokens": 2617407.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "eval_loss": 0.7984817028045654, | |
| "eval_mean_token_accuracy": 0.7897448325157166, | |
| "eval_num_tokens": 2617407.0, | |
| "eval_runtime": 207.2592, | |
| "eval_samples_per_second": 1.911, | |
| "eval_steps_per_second": 0.121, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 0.4229067328272521, | |
| "learning_rate": 8.153846153846154e-06, | |
| "loss": 0.1993, | |
| "mean_token_accuracy": 0.9341381952166558, | |
| "num_tokens": 3271788.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.4038605805362403, | |
| "learning_rate": 7.76923076923077e-06, | |
| "loss": 0.1782, | |
| "mean_token_accuracy": 0.9413852587342262, | |
| "num_tokens": 3927148.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "grad_norm": 0.44386495794019754, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 0.1716, | |
| "mean_token_accuracy": 0.9436837241053582, | |
| "num_tokens": 4581399.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.4248743921174713, | |
| "learning_rate": 7e-06, | |
| "loss": 0.169, | |
| "mean_token_accuracy": 0.9443423271179199, | |
| "num_tokens": 5236759.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_loss": 0.8438917994499207, | |
| "eval_mean_token_accuracy": 0.7881988358497619, | |
| "eval_num_tokens": 5236759.0, | |
| "eval_runtime": 206.3838, | |
| "eval_samples_per_second": 1.919, | |
| "eval_steps_per_second": 0.121, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "grad_norm": 0.39016509839724534, | |
| "learning_rate": 6.615384615384616e-06, | |
| "loss": 0.1703, | |
| "mean_token_accuracy": 0.94403627961874, | |
| "num_tokens": 5890523.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.3919683518651608, | |
| "learning_rate": 6.230769230769231e-06, | |
| "loss": 0.1723, | |
| "mean_token_accuracy": 0.9430388778448104, | |
| "num_tokens": 6543649.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.1153846153846154, | |
| "grad_norm": 0.3777794605646071, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 0.1578, | |
| "mean_token_accuracy": 0.948506236076355, | |
| "num_tokens": 7198863.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.41041127040715936, | |
| "learning_rate": 5.461538461538461e-06, | |
| "loss": 0.15, | |
| "mean_token_accuracy": 0.9507905378937721, | |
| "num_tokens": 7853347.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "eval_loss": 0.8952463269233704, | |
| "eval_mean_token_accuracy": 0.7861712884902954, | |
| "eval_num_tokens": 7853347.0, | |
| "eval_runtime": 205.7768, | |
| "eval_samples_per_second": 1.924, | |
| "eval_steps_per_second": 0.121, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.3964654882595507, | |
| "learning_rate": 5.076923076923077e-06, | |
| "loss": 0.1497, | |
| "mean_token_accuracy": 0.9505507811903954, | |
| "num_tokens": 8506440.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.3768825306789265, | |
| "learning_rate": 4.692307692307693e-06, | |
| "loss": 0.1503, | |
| "mean_token_accuracy": 0.9506417199969291, | |
| "num_tokens": 9160442.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8846153846153846, | |
| "grad_norm": 0.398264112857023, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 0.1488, | |
| "mean_token_accuracy": 0.9506732374429703, | |
| "num_tokens": 9815802.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.39276972642065144, | |
| "learning_rate": 3.923076923076923e-06, | |
| "loss": 0.1394, | |
| "mean_token_accuracy": 0.9544374644756317, | |
| "num_tokens": 10470724.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_loss": 0.8906534910202026, | |
| "eval_mean_token_accuracy": 0.786395034790039, | |
| "eval_num_tokens": 10470724.0, | |
| "eval_runtime": 206.1358, | |
| "eval_samples_per_second": 1.921, | |
| "eval_steps_per_second": 0.121, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.269230769230769, | |
| "grad_norm": 0.3730257614019165, | |
| "learning_rate": 3.538461538461539e-06, | |
| "loss": 0.1378, | |
| "mean_token_accuracy": 0.9548449605703354, | |
| "num_tokens": 11125135.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.3683238050739676, | |
| "learning_rate": 3.153846153846154e-06, | |
| "loss": 0.1314, | |
| "mean_token_accuracy": 0.9570518404245376, | |
| "num_tokens": 11780057.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.6538461538461537, | |
| "grad_norm": 0.38257552430474834, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.1313, | |
| "mean_token_accuracy": 0.9570926904678345, | |
| "num_tokens": 12435417.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.36203400285990167, | |
| "learning_rate": 2.384615384615385e-06, | |
| "loss": 0.135, | |
| "mean_token_accuracy": 0.9562257811427116, | |
| "num_tokens": 13089419.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.9335225224494934, | |
| "eval_mean_token_accuracy": 0.7845649313926697, | |
| "eval_num_tokens": 13089419.0, | |
| "eval_runtime": 205.9656, | |
| "eval_samples_per_second": 1.923, | |
| "eval_steps_per_second": 0.121, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.038461538461538, | |
| "grad_norm": 0.35419211664728983, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.1285, | |
| "mean_token_accuracy": 0.9583714559674263, | |
| "num_tokens": 13742512.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.34375461778603095, | |
| "learning_rate": 1.6153846153846157e-06, | |
| "loss": 0.1242, | |
| "mean_token_accuracy": 0.9601028740406037, | |
| "num_tokens": 14395386.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.423076923076923, | |
| "grad_norm": 0.3525911653900964, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.1217, | |
| "mean_token_accuracy": 0.9608346611261368, | |
| "num_tokens": 15049637.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.3677261206987421, | |
| "learning_rate": 8.461538461538463e-07, | |
| "loss": 0.1245, | |
| "mean_token_accuracy": 0.9600024372339249, | |
| "num_tokens": 15704997.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "eval_loss": 0.9684593081474304, | |
| "eval_mean_token_accuracy": 0.7830931186676026, | |
| "eval_num_tokens": 15704997.0, | |
| "eval_runtime": 205.7113, | |
| "eval_samples_per_second": 1.925, | |
| "eval_steps_per_second": 0.122, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.8076923076923075, | |
| "grad_norm": 0.35206380647203034, | |
| "learning_rate": 4.615384615384616e-07, | |
| "loss": 0.1222, | |
| "mean_token_accuracy": 0.9602848619222641, | |
| "num_tokens": 16360357.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.3554888857605734, | |
| "learning_rate": 7.692307692307694e-08, | |
| "loss": 0.1239, | |
| "mean_token_accuracy": 0.959795056283474, | |
| "num_tokens": 17014300.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 130, | |
| "total_flos": 14799366782976.0, | |
| "train_loss": 0.15786109383289632, | |
| "train_runtime": 17510.8371, | |
| "train_samples_per_second": 0.474, | |
| "train_steps_per_second": 0.007 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 14799366782976.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |