| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.05653710247349823, | |
| "eval_steps": 5, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0028268551236749115, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 7.6743, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0028268551236749115, | |
| "eval_loss": 7.129258632659912, | |
| "eval_runtime": 20.9973, | |
| "eval_samples_per_second": 28.385, | |
| "eval_steps_per_second": 3.572, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005653710247349823, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 7.3349, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008480565371024734, | |
| "grad_norm": 5.082520484924316, | |
| "learning_rate": 2e-05, | |
| "loss": 7.4669, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.011307420494699646, | |
| "grad_norm": 4.920047760009766, | |
| "learning_rate": 4e-05, | |
| "loss": 7.1521, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.014134275618374558, | |
| "grad_norm": NaN, | |
| "learning_rate": 4e-05, | |
| "loss": 7.7426, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014134275618374558, | |
| "eval_loss": 7.097348213195801, | |
| "eval_runtime": 20.8903, | |
| "eval_samples_per_second": 28.53, | |
| "eval_steps_per_second": 3.59, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01696113074204947, | |
| "grad_norm": 4.500086784362793, | |
| "learning_rate": 6e-05, | |
| "loss": 7.1859, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.019787985865724382, | |
| "grad_norm": 3.6024482250213623, | |
| "learning_rate": 8e-05, | |
| "loss": 6.9672, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.022614840989399292, | |
| "grad_norm": 4.331362724304199, | |
| "learning_rate": 0.0001, | |
| "loss": 7.2915, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.025441696113074206, | |
| "grad_norm": 3.071237564086914, | |
| "learning_rate": 0.00012, | |
| "loss": 7.2579, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.028268551236749116, | |
| "grad_norm": 3.218928813934326, | |
| "learning_rate": 0.00014, | |
| "loss": 7.0417, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.028268551236749116, | |
| "eval_loss": 6.332116603851318, | |
| "eval_runtime": 20.7973, | |
| "eval_samples_per_second": 28.658, | |
| "eval_steps_per_second": 3.606, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03109540636042403, | |
| "grad_norm": 2.1495420932769775, | |
| "learning_rate": 0.00016, | |
| "loss": 6.1519, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03392226148409894, | |
| "grad_norm": 2.033008337020874, | |
| "learning_rate": 0.00018, | |
| "loss": 6.6133, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03674911660777385, | |
| "grad_norm": 2.2604849338531494, | |
| "learning_rate": 0.0002, | |
| "loss": 6.1704, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.039575971731448764, | |
| "grad_norm": 1.981244683265686, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 6.0664, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04240282685512368, | |
| "grad_norm": 1.9168858528137207, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 5.7705, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04240282685512368, | |
| "eval_loss": 5.497771263122559, | |
| "eval_runtime": 20.7925, | |
| "eval_samples_per_second": 28.664, | |
| "eval_steps_per_second": 3.607, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.045229681978798585, | |
| "grad_norm": 1.6251788139343262, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 5.6184, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0480565371024735, | |
| "grad_norm": 1.4069232940673828, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 5.3642, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05088339222614841, | |
| "grad_norm": 1.9275850057601929, | |
| "learning_rate": 0.0001, | |
| "loss": 5.6553, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05371024734982332, | |
| "grad_norm": 2.3857240676879883, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 5.3992, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05653710247349823, | |
| "grad_norm": 1.5103741884231567, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 5.6954, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05653710247349823, | |
| "eval_loss": 5.104593276977539, | |
| "eval_runtime": 11.9721, | |
| "eval_samples_per_second": 49.783, | |
| "eval_steps_per_second": 6.265, | |
| "step": 20 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.638283774571315e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |