| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.021108179419525065, | |
| "eval_steps": 4, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00042216358839050134, | |
| "grad_norm": 6.515402317047119, | |
| "learning_rate": 2e-05, | |
| "loss": 6.2351, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00042216358839050134, | |
| "eval_loss": 7.163845539093018, | |
| "eval_runtime": 59.6773, | |
| "eval_samples_per_second": 8.362, | |
| "eval_steps_per_second": 8.362, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008443271767810027, | |
| "grad_norm": 6.853526592254639, | |
| "learning_rate": 4e-05, | |
| "loss": 6.2997, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.001266490765171504, | |
| "grad_norm": 7.778280735015869, | |
| "learning_rate": 6e-05, | |
| "loss": 7.021, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0016886543535620053, | |
| "grad_norm": 8.638328552246094, | |
| "learning_rate": 8e-05, | |
| "loss": 7.51, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0016886543535620053, | |
| "eval_loss": 6.921504974365234, | |
| "eval_runtime": 59.6612, | |
| "eval_samples_per_second": 8.364, | |
| "eval_steps_per_second": 8.364, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0021108179419525065, | |
| "grad_norm": 8.384474754333496, | |
| "learning_rate": 0.0001, | |
| "loss": 6.6173, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.002532981530343008, | |
| "grad_norm": 6.948892116546631, | |
| "learning_rate": 0.00012, | |
| "loss": 5.3785, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0029551451187335093, | |
| "grad_norm": 7.122462749481201, | |
| "learning_rate": 0.00014, | |
| "loss": 5.7491, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0033773087071240107, | |
| "grad_norm": 8.084318161010742, | |
| "learning_rate": 0.00016, | |
| "loss": 4.7822, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0033773087071240107, | |
| "eval_loss": 4.463127613067627, | |
| "eval_runtime": 59.3591, | |
| "eval_samples_per_second": 8.406, | |
| "eval_steps_per_second": 8.406, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.003799472295514512, | |
| "grad_norm": 7.751031875610352, | |
| "learning_rate": 0.00018, | |
| "loss": 4.0501, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.004221635883905013, | |
| "grad_norm": 9.506431579589844, | |
| "learning_rate": 0.0002, | |
| "loss": 4.0302, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004643799472295515, | |
| "grad_norm": 7.7947001457214355, | |
| "learning_rate": 0.0001996917333733128, | |
| "loss": 2.6523, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.005065963060686016, | |
| "grad_norm": 7.803070068359375, | |
| "learning_rate": 0.00019876883405951377, | |
| "loss": 1.5597, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005065963060686016, | |
| "eval_loss": 1.5647698640823364, | |
| "eval_runtime": 59.7333, | |
| "eval_samples_per_second": 8.354, | |
| "eval_steps_per_second": 8.354, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005488126649076517, | |
| "grad_norm": 7.336414813995361, | |
| "learning_rate": 0.00019723699203976766, | |
| "loss": 1.2889, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.005910290237467019, | |
| "grad_norm": 7.312011241912842, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 1.2851, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0063324538258575196, | |
| "grad_norm": 11.156982421875, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 0.9687, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.006754617414248021, | |
| "grad_norm": 8.215885162353516, | |
| "learning_rate": 0.0001891006524188368, | |
| "loss": 0.481, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.006754617414248021, | |
| "eval_loss": 0.6843006014823914, | |
| "eval_runtime": 59.5333, | |
| "eval_samples_per_second": 8.382, | |
| "eval_steps_per_second": 8.382, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.007176781002638522, | |
| "grad_norm": 4.760833263397217, | |
| "learning_rate": 0.00018526401643540922, | |
| "loss": 0.4751, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.007598944591029024, | |
| "grad_norm": 7.501655578613281, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.6609, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.008021108179419525, | |
| "grad_norm": 6.412684917449951, | |
| "learning_rate": 0.0001760405965600031, | |
| "loss": 0.6547, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.008443271767810026, | |
| "grad_norm": 5.393802642822266, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.3143, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008443271767810026, | |
| "eval_loss": 0.530745804309845, | |
| "eval_runtime": 59.4057, | |
| "eval_samples_per_second": 8.4, | |
| "eval_steps_per_second": 8.4, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008865435356200527, | |
| "grad_norm": 6.753420352935791, | |
| "learning_rate": 0.00016494480483301836, | |
| "loss": 0.7354, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.00928759894459103, | |
| "grad_norm": 3.744786500930786, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.3945, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.00970976253298153, | |
| "grad_norm": 3.2725045680999756, | |
| "learning_rate": 0.0001522498564715949, | |
| "loss": 0.2881, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.010131926121372032, | |
| "grad_norm": 2.0320169925689697, | |
| "learning_rate": 0.00014539904997395468, | |
| "loss": 0.1861, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.010131926121372032, | |
| "eval_loss": 0.4393819272518158, | |
| "eval_runtime": 59.4549, | |
| "eval_samples_per_second": 8.393, | |
| "eval_steps_per_second": 8.393, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.010554089709762533, | |
| "grad_norm": 5.322549343109131, | |
| "learning_rate": 0.000138268343236509, | |
| "loss": 0.7912, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.010976253298153034, | |
| "grad_norm": 1.55814790725708, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.2046, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.011398416886543536, | |
| "grad_norm": 2.2971291542053223, | |
| "learning_rate": 0.00012334453638559057, | |
| "loss": 0.2717, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.011820580474934037, | |
| "grad_norm": 1.6769708395004272, | |
| "learning_rate": 0.0001156434465040231, | |
| "loss": 0.1889, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.011820580474934037, | |
| "eval_loss": 0.43992501497268677, | |
| "eval_runtime": 59.4329, | |
| "eval_samples_per_second": 8.396, | |
| "eval_steps_per_second": 8.396, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.012242744063324538, | |
| "grad_norm": 4.204702854156494, | |
| "learning_rate": 0.0001078459095727845, | |
| "loss": 0.4801, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.012664907651715039, | |
| "grad_norm": 3.246248483657837, | |
| "learning_rate": 0.0001, | |
| "loss": 0.425, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01308707124010554, | |
| "grad_norm": 2.989017963409424, | |
| "learning_rate": 9.215409042721552e-05, | |
| "loss": 0.46, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.013509234828496043, | |
| "grad_norm": 2.7211666107177734, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 0.2487, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.013509234828496043, | |
| "eval_loss": 0.3397166132926941, | |
| "eval_runtime": 59.4567, | |
| "eval_samples_per_second": 8.393, | |
| "eval_steps_per_second": 8.393, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.013931398416886544, | |
| "grad_norm": 1.7053347826004028, | |
| "learning_rate": 7.66554636144095e-05, | |
| "loss": 0.251, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.014353562005277045, | |
| "grad_norm": 2.717104434967041, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.5596, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.014775725593667546, | |
| "grad_norm": 1.9112799167633057, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 0.2708, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.015197889182058048, | |
| "grad_norm": 2.3562815189361572, | |
| "learning_rate": 5.4600950026045326e-05, | |
| "loss": 0.4158, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.015197889182058048, | |
| "eval_loss": 0.31511208415031433, | |
| "eval_runtime": 59.4134, | |
| "eval_samples_per_second": 8.399, | |
| "eval_steps_per_second": 8.399, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01562005277044855, | |
| "grad_norm": 2.5325310230255127, | |
| "learning_rate": 4.7750143528405126e-05, | |
| "loss": 0.415, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01604221635883905, | |
| "grad_norm": 1.722407579421997, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.1969, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.016464379947229553, | |
| "grad_norm": 2.104142189025879, | |
| "learning_rate": 3.5055195166981645e-05, | |
| "loss": 0.2823, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.016886543535620052, | |
| "grad_norm": 2.8751745223999023, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.3524, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.016886543535620052, | |
| "eval_loss": 0.2879149615764618, | |
| "eval_runtime": 59.4767, | |
| "eval_samples_per_second": 8.39, | |
| "eval_steps_per_second": 8.39, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.017308707124010555, | |
| "grad_norm": 3.010507345199585, | |
| "learning_rate": 2.3959403439996907e-05, | |
| "loss": 0.3272, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.017730870712401054, | |
| "grad_norm": 2.1966700553894043, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.3737, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.018153034300791557, | |
| "grad_norm": 1.8249499797821045, | |
| "learning_rate": 1.4735983564590783e-05, | |
| "loss": 0.236, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01857519788918206, | |
| "grad_norm": 2.0696208477020264, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 0.2272, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01857519788918206, | |
| "eval_loss": 0.27388355135917664, | |
| "eval_runtime": 59.4525, | |
| "eval_samples_per_second": 8.393, | |
| "eval_steps_per_second": 8.393, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01899736147757256, | |
| "grad_norm": 3.1134109497070312, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 0.4219, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.01941952506596306, | |
| "grad_norm": 2.3598380088806152, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.3093, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.01984168865435356, | |
| "grad_norm": 2.2470579147338867, | |
| "learning_rate": 2.7630079602323442e-06, | |
| "loss": 0.3218, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.020263852242744063, | |
| "grad_norm": 2.2850100994110107, | |
| "learning_rate": 1.231165940486234e-06, | |
| "loss": 0.3537, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.020263852242744063, | |
| "eval_loss": 0.26765426993370056, | |
| "eval_runtime": 60.0554, | |
| "eval_samples_per_second": 8.309, | |
| "eval_steps_per_second": 8.309, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.020686015831134566, | |
| "grad_norm": 2.18385910987854, | |
| "learning_rate": 3.0826662668720364e-07, | |
| "loss": 0.1944, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.021108179419525065, | |
| "grad_norm": 2.7979819774627686, | |
| "learning_rate": 0.0, | |
| "loss": 0.3756, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 4, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.85471820890112e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |