| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 3000, | |
| "global_step": 88686, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.382721060821325e-05, | |
| "grad_norm": 1896.0, | |
| "learning_rate": 0.00029999661727893914, | |
| "loss": 32.25, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.10148163182463973, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.000289851836817536, | |
| "loss": 3.8074, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.10148163182463973, | |
| "eval_loss": 2.4561102390289307, | |
| "eval_runtime": 92.9065, | |
| "eval_samples_per_second": 1016.13, | |
| "eval_steps_per_second": 7.943, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 0.000279703673635072, | |
| "loss": 3.182, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.20296326364927947, | |
| "eval_loss": 2.3882644176483154, | |
| "eval_runtime": 92.9373, | |
| "eval_samples_per_second": 1015.792, | |
| "eval_steps_per_second": 7.941, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.30444489547391923, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 0.00026955551045260807, | |
| "loss": 3.1161, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.30444489547391923, | |
| "eval_loss": 2.365419626235962, | |
| "eval_runtime": 92.9113, | |
| "eval_samples_per_second": 1016.077, | |
| "eval_steps_per_second": 7.943, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.0002594073472701441, | |
| "loss": 3.0863, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.40592652729855894, | |
| "eval_loss": 2.339674472808838, | |
| "eval_runtime": 92.9289, | |
| "eval_samples_per_second": 1015.884, | |
| "eval_steps_per_second": 7.942, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.5074081591231987, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 0.0002492591840876801, | |
| "loss": 3.0643, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.5074081591231987, | |
| "eval_loss": 2.3348894119262695, | |
| "eval_runtime": 92.9352, | |
| "eval_samples_per_second": 1015.815, | |
| "eval_steps_per_second": 7.941, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "grad_norm": 1.25, | |
| "learning_rate": 0.0002391110209052161, | |
| "loss": 3.0593, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.6088897909478385, | |
| "eval_loss": 2.32326602935791, | |
| "eval_runtime": 92.9449, | |
| "eval_samples_per_second": 1015.709, | |
| "eval_steps_per_second": 7.94, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.7103714227724782, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.00022896285772275215, | |
| "loss": 3.0519, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.7103714227724782, | |
| "eval_loss": 2.325340986251831, | |
| "eval_runtime": 92.952, | |
| "eval_samples_per_second": 1015.632, | |
| "eval_steps_per_second": 7.94, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 0.0002188146945402882, | |
| "loss": 3.0464, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.8118530545971179, | |
| "eval_loss": 2.320810079574585, | |
| "eval_runtime": 92.965, | |
| "eval_samples_per_second": 1015.489, | |
| "eval_steps_per_second": 7.938, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.9133346864217576, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00020866653135782423, | |
| "loss": 3.0434, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.9133346864217576, | |
| "eval_loss": 2.3228955268859863, | |
| "eval_runtime": 93.1104, | |
| "eval_samples_per_second": 1013.904, | |
| "eval_steps_per_second": 7.926, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 0.00019851836817536025, | |
| "loss": 3.0389, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.0148163182463974, | |
| "eval_loss": 2.3193914890289307, | |
| "eval_runtime": 93.0867, | |
| "eval_samples_per_second": 1014.162, | |
| "eval_steps_per_second": 7.928, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.116297950071037, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.0001883702049928963, | |
| "loss": 3.0377, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.116297950071037, | |
| "eval_loss": 2.3154852390289307, | |
| "eval_runtime": 93.1029, | |
| "eval_samples_per_second": 1013.985, | |
| "eval_steps_per_second": 7.927, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.0001782220418104323, | |
| "loss": 3.0319, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.217779581895677, | |
| "eval_loss": 2.3132622241973877, | |
| "eval_runtime": 93.1111, | |
| "eval_samples_per_second": 1013.897, | |
| "eval_steps_per_second": 7.926, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.3192612137203166, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.00016807387862796832, | |
| "loss": 3.0328, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.3192612137203166, | |
| "eval_loss": 2.310361623764038, | |
| "eval_runtime": 93.0989, | |
| "eval_samples_per_second": 1014.03, | |
| "eval_steps_per_second": 7.927, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.00015792571544550436, | |
| "loss": 3.0314, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.4207428455449564, | |
| "eval_loss": 2.309779405593872, | |
| "eval_runtime": 93.116, | |
| "eval_samples_per_second": 1013.843, | |
| "eval_steps_per_second": 7.926, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.522224477369596, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 0.00014777755226304037, | |
| "loss": 3.0306, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.522224477369596, | |
| "eval_loss": 2.3095569610595703, | |
| "eval_runtime": 93.0822, | |
| "eval_samples_per_second": 1014.211, | |
| "eval_steps_per_second": 7.928, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 0.0001376293890805764, | |
| "loss": 3.0251, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.6237061091942357, | |
| "eval_loss": 2.314659595489502, | |
| "eval_runtime": 93.0784, | |
| "eval_samples_per_second": 1014.253, | |
| "eval_steps_per_second": 7.929, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.7251877410188756, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 0.00012748122589811243, | |
| "loss": 3.027, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.7251877410188756, | |
| "eval_loss": 2.3094723224639893, | |
| "eval_runtime": 92.9414, | |
| "eval_samples_per_second": 1015.748, | |
| "eval_steps_per_second": 7.94, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00011733306271564845, | |
| "loss": 3.0238, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.8266693728435153, | |
| "eval_loss": 2.3098957538604736, | |
| "eval_runtime": 92.9753, | |
| "eval_samples_per_second": 1015.377, | |
| "eval_steps_per_second": 7.938, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 1.928151004668155, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00010718489953318448, | |
| "loss": 3.0246, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 1.928151004668155, | |
| "eval_loss": 2.308159828186035, | |
| "eval_runtime": 92.9437, | |
| "eval_samples_per_second": 1015.722, | |
| "eval_steps_per_second": 7.94, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 9.703673635072052e-05, | |
| "loss": 3.0253, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.029632636492795, | |
| "eval_loss": 2.30761981010437, | |
| "eval_runtime": 92.9512, | |
| "eval_samples_per_second": 1015.641, | |
| "eval_steps_per_second": 7.94, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.1311142683174347, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 8.688857316825655e-05, | |
| "loss": 3.0261, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.1311142683174347, | |
| "eval_loss": 2.310319185256958, | |
| "eval_runtime": 92.926, | |
| "eval_samples_per_second": 1015.916, | |
| "eval_steps_per_second": 7.942, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 7.674040998579256e-05, | |
| "loss": 3.0242, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.232595900142074, | |
| "eval_loss": 2.3085408210754395, | |
| "eval_runtime": 92.9709, | |
| "eval_samples_per_second": 1015.425, | |
| "eval_steps_per_second": 7.938, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.334077531966714, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 6.659224680332859e-05, | |
| "loss": 3.0245, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.334077531966714, | |
| "eval_loss": 2.3078315258026123, | |
| "eval_runtime": 92.9373, | |
| "eval_samples_per_second": 1015.793, | |
| "eval_steps_per_second": 7.941, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 5.644408362086462e-05, | |
| "loss": 3.024, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.435559163791354, | |
| "eval_loss": 2.3096206188201904, | |
| "eval_runtime": 92.9124, | |
| "eval_samples_per_second": 1016.065, | |
| "eval_steps_per_second": 7.943, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.5370407956159937, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 4.629592043840065e-05, | |
| "loss": 3.0251, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 2.5370407956159937, | |
| "eval_loss": 2.30861496925354, | |
| "eval_runtime": 92.9585, | |
| "eval_samples_per_second": 1015.561, | |
| "eval_steps_per_second": 7.939, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 3.614775725593667e-05, | |
| "loss": 3.0257, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 2.638522427440633, | |
| "eval_loss": 2.308117389678955, | |
| "eval_runtime": 92.9722, | |
| "eval_samples_per_second": 1015.411, | |
| "eval_steps_per_second": 7.938, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 2.740004059265273, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 2.59995940734727e-05, | |
| "loss": 3.0224, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 2.740004059265273, | |
| "eval_loss": 2.308805465698242, | |
| "eval_runtime": 92.931, | |
| "eval_samples_per_second": 1015.861, | |
| "eval_steps_per_second": 7.941, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.5851430891008727e-05, | |
| "loss": 3.0267, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 2.841485691089913, | |
| "eval_loss": 2.3082549571990967, | |
| "eval_runtime": 92.9136, | |
| "eval_samples_per_second": 1016.051, | |
| "eval_steps_per_second": 7.943, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 2.9429673229145523, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 5.703267708544753e-06, | |
| "loss": 3.0243, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 2.9429673229145523, | |
| "eval_loss": 2.3083608150482178, | |
| "eval_runtime": 92.9434, | |
| "eval_samples_per_second": 1015.726, | |
| "eval_steps_per_second": 7.94, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 88686, | |
| "total_flos": 1.2889476521431695e+18, | |
| "train_loss": 3.0687490486097015, | |
| "train_runtime": 36823.0101, | |
| "train_samples_per_second": 308.271, | |
| "train_steps_per_second": 2.408 | |
| } | |
| ], | |
| "logging_steps": 3000, | |
| "max_steps": 88686, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2889476521431695e+18, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |