| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 9.0, |
| "eval_steps": 500, |
| "global_step": 936, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.24096385542168675, |
| "grad_norm": 0.17040389776229858, |
| "learning_rate": 0.000511, |
| "loss": 0.4303, |
| "mean_token_accuracy": 0.8983636182546616, |
| "num_tokens": 569446.0, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.4819277108433735, |
| "grad_norm": 0.19750550389289856, |
| "learning_rate": 0.000511, |
| "loss": 0.2953, |
| "mean_token_accuracy": 0.9258013522624969, |
| "num_tokens": 1138403.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.7228915662650602, |
| "grad_norm": 0.2441304326057434, |
| "learning_rate": 0.000511, |
| "loss": 0.2179, |
| "mean_token_accuracy": 0.9427212655544281, |
| "num_tokens": 1705606.0, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.963855421686747, |
| "grad_norm": 0.19773538410663605, |
| "learning_rate": 0.000511, |
| "loss": 0.194, |
| "mean_token_accuracy": 0.9479873812198639, |
| "num_tokens": 2271060.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 0.1995953470468521, |
| "eval_mean_token_accuracy": 0.9460804703387808, |
| "eval_num_tokens": 2345411.0, |
| "eval_runtime": 4.3258, |
| "eval_samples_per_second": 85.303, |
| "eval_steps_per_second": 10.865, |
| "step": 104 |
| }, |
| { |
| "epoch": 1.202409638554217, |
| "grad_norm": 0.0789426863193512, |
| "learning_rate": 0.000511, |
| "loss": 0.16, |
| "mean_token_accuracy": 0.9584937607399141, |
| "num_tokens": 2836512.0, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.4433734939759035, |
| "grad_norm": 0.10643763095140457, |
| "learning_rate": 0.000511, |
| "loss": 0.1408, |
| "mean_token_accuracy": 0.9620217531919479, |
| "num_tokens": 3403395.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.6843373493975904, |
| "grad_norm": 0.08403506129980087, |
| "learning_rate": 0.000511, |
| "loss": 0.1314, |
| "mean_token_accuracy": 0.9630346685647965, |
| "num_tokens": 3970344.0, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.9253012048192772, |
| "grad_norm": 0.10631190985441208, |
| "learning_rate": 0.000511, |
| "loss": 0.1272, |
| "mean_token_accuracy": 0.9646087974309921, |
| "num_tokens": 4538034.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_loss": 0.14445580542087555, |
| "eval_mean_token_accuracy": 0.9612192435467497, |
| "eval_num_tokens": 4690349.0, |
| "eval_runtime": 4.2683, |
| "eval_samples_per_second": 86.451, |
| "eval_steps_per_second": 11.011, |
| "step": 208 |
| }, |
| { |
| "epoch": 2.163855421686747, |
| "grad_norm": 0.08164115995168686, |
| "learning_rate": 0.000511, |
| "loss": 0.1067, |
| "mean_token_accuracy": 0.9689010110768405, |
| "num_tokens": 5091219.0, |
| "step": 225 |
| }, |
| { |
| "epoch": 2.404819277108434, |
| "grad_norm": 0.08637778460979462, |
| "learning_rate": 0.000511, |
| "loss": 0.0968, |
| "mean_token_accuracy": 0.9721185141801834, |
| "num_tokens": 5658328.0, |
| "step": 250 |
| }, |
| { |
| "epoch": 2.6457831325301204, |
| "grad_norm": 0.0902683362364769, |
| "learning_rate": 0.000511, |
| "loss": 0.091, |
| "mean_token_accuracy": 0.9744718617200852, |
| "num_tokens": 6225744.0, |
| "step": 275 |
| }, |
| { |
| "epoch": 2.886746987951807, |
| "grad_norm": 0.09357521682977676, |
| "learning_rate": 0.000511, |
| "loss": 0.0929, |
| "mean_token_accuracy": 0.9727102434635162, |
| "num_tokens": 6794402.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_loss": 0.14705069363117218, |
| "eval_mean_token_accuracy": 0.9624257531571896, |
| "eval_num_tokens": 7035273.0, |
| "eval_runtime": 4.2634, |
| "eval_samples_per_second": 86.55, |
| "eval_steps_per_second": 11.024, |
| "step": 312 |
| }, |
| { |
| "epoch": 3.125301204819277, |
| "grad_norm": 0.14760874211788177, |
| "learning_rate": 0.000511, |
| "loss": 0.0842, |
| "mean_token_accuracy": 0.9764150320881545, |
| "num_tokens": 7334495.0, |
| "step": 325 |
| }, |
| { |
| "epoch": 3.3662650602409636, |
| "grad_norm": 0.09895172715187073, |
| "learning_rate": 0.000511, |
| "loss": 0.0777, |
| "mean_token_accuracy": 0.9774631917476654, |
| "num_tokens": 7903478.0, |
| "step": 350 |
| }, |
| { |
| "epoch": 3.6072289156626507, |
| "grad_norm": 0.10538128763437271, |
| "learning_rate": 0.000511, |
| "loss": 0.0742, |
| "mean_token_accuracy": 0.9783486902713776, |
| "num_tokens": 8469479.0, |
| "step": 375 |
| }, |
| { |
| "epoch": 3.8481927710843373, |
| "grad_norm": 0.09741026163101196, |
| "learning_rate": 0.000511, |
| "loss": 0.0679, |
| "mean_token_accuracy": 0.9803410685062408, |
| "num_tokens": 9036138.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_loss": 0.14550796151161194, |
| "eval_mean_token_accuracy": 0.9638357936067784, |
| "eval_num_tokens": 9380804.0, |
| "eval_runtime": 4.2387, |
| "eval_samples_per_second": 87.055, |
| "eval_steps_per_second": 11.088, |
| "step": 416 |
| }, |
| { |
| "epoch": 4.086746987951807, |
| "grad_norm": 0.10616449266672134, |
| "learning_rate": 0.000511, |
| "loss": 0.0625, |
| "mean_token_accuracy": 0.9811063475079007, |
| "num_tokens": 9598059.0, |
| "step": 425 |
| }, |
| { |
| "epoch": 4.327710843373494, |
| "grad_norm": 0.09445353597402573, |
| "learning_rate": 0.000511, |
| "loss": 0.0545, |
| "mean_token_accuracy": 0.9833864039182663, |
| "num_tokens": 10165089.0, |
| "step": 450 |
| }, |
| { |
| "epoch": 4.5686746987951805, |
| "grad_norm": 0.07407805323600769, |
| "learning_rate": 0.000511, |
| "loss": 0.0545, |
| "mean_token_accuracy": 0.984256454706192, |
| "num_tokens": 10732931.0, |
| "step": 475 |
| }, |
| { |
| "epoch": 4.809638554216868, |
| "grad_norm": 0.07322381436824799, |
| "learning_rate": 0.000511, |
| "loss": 0.0462, |
| "mean_token_accuracy": 0.9861221539974213, |
| "num_tokens": 11301466.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_loss": 0.14242176711559296, |
| "eval_mean_token_accuracy": 0.9691625554510888, |
| "eval_num_tokens": 11726407.0, |
| "eval_runtime": 4.2397, |
| "eval_samples_per_second": 87.035, |
| "eval_steps_per_second": 11.086, |
| "step": 520 |
| }, |
| { |
| "epoch": 5.048192771084337, |
| "grad_norm": 0.06890378147363663, |
| "learning_rate": 0.000511, |
| "loss": 0.0538, |
| "mean_token_accuracy": 0.9846178467827614, |
| "num_tokens": 11856936.0, |
| "step": 525 |
| }, |
| { |
| "epoch": 5.289156626506024, |
| "grad_norm": 0.05453705042600632, |
| "learning_rate": 0.000511, |
| "loss": 0.0485, |
| "mean_token_accuracy": 0.9858993107080459, |
| "num_tokens": 12423275.0, |
| "step": 550 |
| }, |
| { |
| "epoch": 5.530120481927711, |
| "grad_norm": 0.0743594691157341, |
| "learning_rate": 0.000511, |
| "loss": 0.0455, |
| "mean_token_accuracy": 0.9857969325780869, |
| "num_tokens": 12992045.0, |
| "step": 575 |
| }, |
| { |
| "epoch": 5.771084337349397, |
| "grad_norm": 0.06587184965610504, |
| "learning_rate": 0.000511, |
| "loss": 0.0446, |
| "mean_token_accuracy": 0.9862634456157684, |
| "num_tokens": 13560037.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_loss": 0.1241711750626564, |
| "eval_mean_token_accuracy": 0.9710005204728309, |
| "eval_num_tokens": 14071932.0, |
| "eval_runtime": 4.2228, |
| "eval_samples_per_second": 87.383, |
| "eval_steps_per_second": 11.13, |
| "step": 624 |
| }, |
| { |
| "epoch": 6.009638554216868, |
| "grad_norm": 0.06124307960271835, |
| "learning_rate": 0.000511, |
| "loss": 0.0368, |
| "mean_token_accuracy": 0.9886124525407348, |
| "num_tokens": 14102915.0, |
| "step": 625 |
| }, |
| { |
| "epoch": 6.250602409638554, |
| "grad_norm": 0.10462699830532074, |
| "learning_rate": 0.000511, |
| "loss": 0.0378, |
| "mean_token_accuracy": 0.9887062352895737, |
| "num_tokens": 14669309.0, |
| "step": 650 |
| }, |
| { |
| "epoch": 6.491566265060241, |
| "grad_norm": 0.09343062341213226, |
| "learning_rate": 0.000511, |
| "loss": 0.0353, |
| "mean_token_accuracy": 0.9892213380336762, |
| "num_tokens": 15237353.0, |
| "step": 675 |
| }, |
| { |
| "epoch": 6.732530120481927, |
| "grad_norm": 0.08443740010261536, |
| "learning_rate": 0.000511, |
| "loss": 0.0364, |
| "mean_token_accuracy": 0.9892494148015976, |
| "num_tokens": 15804441.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 6.973493975903614, |
| "grad_norm": 0.07635796070098877, |
| "learning_rate": 0.000511, |
| "loss": 0.0397, |
| "mean_token_accuracy": 0.9886371964216232, |
| "num_tokens": 16368577.0, |
| "step": 725 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_loss": 0.17054519057273865, |
| "eval_mean_token_accuracy": 0.9649876089806252, |
| "eval_num_tokens": 16416958.0, |
| "eval_runtime": 4.2384, |
| "eval_samples_per_second": 87.061, |
| "eval_steps_per_second": 11.089, |
| "step": 728 |
| }, |
| { |
| "epoch": 7.212048192771085, |
| "grad_norm": 0.0668734461069107, |
| "learning_rate": 0.000511, |
| "loss": 0.0333, |
| "mean_token_accuracy": 0.9895464802029157, |
| "num_tokens": 16927490.0, |
| "step": 750 |
| }, |
| { |
| "epoch": 7.453012048192771, |
| "grad_norm": 0.06639474630355835, |
| "learning_rate": 0.000511, |
| "loss": 0.0325, |
| "mean_token_accuracy": 0.9908391135931015, |
| "num_tokens": 17494143.0, |
| "step": 775 |
| }, |
| { |
| "epoch": 7.693975903614458, |
| "grad_norm": 0.10801058262586594, |
| "learning_rate": 0.000511, |
| "loss": 0.029, |
| "mean_token_accuracy": 0.9909009468555451, |
| "num_tokens": 18063423.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 7.934939759036144, |
| "grad_norm": 0.048982683569192886, |
| "learning_rate": 0.000511, |
| "loss": 0.0313, |
| "mean_token_accuracy": 0.9908118671178818, |
| "num_tokens": 18630905.0, |
| "step": 825 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_loss": 0.1113305315375328, |
| "eval_mean_token_accuracy": 0.9737493814306056, |
| "eval_num_tokens": 18762023.0, |
| "eval_runtime": 4.2275, |
| "eval_samples_per_second": 87.285, |
| "eval_steps_per_second": 11.118, |
| "step": 832 |
| }, |
| { |
| "epoch": 8.173493975903614, |
| "grad_norm": 0.05220003426074982, |
| "learning_rate": 0.000511, |
| "loss": 0.0277, |
| "mean_token_accuracy": 0.9916808304160533, |
| "num_tokens": 19184427.0, |
| "step": 850 |
| }, |
| { |
| "epoch": 8.4144578313253, |
| "grad_norm": 0.08605129271745682, |
| "learning_rate": 0.000511, |
| "loss": 0.0273, |
| "mean_token_accuracy": 0.9915985196828843, |
| "num_tokens": 19754417.0, |
| "step": 875 |
| }, |
| { |
| "epoch": 8.655421686746989, |
| "grad_norm": 0.0517394132912159, |
| "learning_rate": 0.000511, |
| "loss": 0.0258, |
| "mean_token_accuracy": 0.9922689855098724, |
| "num_tokens": 20322925.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 8.896385542168675, |
| "grad_norm": 0.059128183871507645, |
| "learning_rate": 0.000511, |
| "loss": 0.0256, |
| "mean_token_accuracy": 0.9923817366361618, |
| "num_tokens": 20889889.0, |
| "step": 925 |
| }, |
| { |
| "epoch": 9.0, |
| "eval_loss": 0.1412837952375412, |
| "eval_mean_token_accuracy": 0.97177672893443, |
| "eval_num_tokens": 21107006.0, |
| "eval_runtime": 4.2426, |
| "eval_samples_per_second": 86.974, |
| "eval_steps_per_second": 11.078, |
| "step": 936 |
| } |
| ], |
| "logging_steps": 25, |
| "max_steps": 1248, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 12, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 9.230033655148052e+17, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
