| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 5490, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.757542133331299, | |
| "learning_rate": 1.817850637522769e-05, | |
| "loss": 1.1409, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 140.5492, | |
| "eval_loss": 0.6597685217857361, | |
| "eval_rouge-1": 46.5406, | |
| "eval_rouge-2": 22.9769, | |
| "eval_rouge-l": 32.9451, | |
| "eval_runtime": 2725.7664, | |
| "eval_samples_per_second": 0.403, | |
| "eval_steps_per_second": 0.067, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "predict_gen_len": 141.1755, | |
| "predict_loss": 0.6476863622665405, | |
| "predict_rouge-1": 47.1078, | |
| "predict_rouge-2": 23.5074, | |
| "predict_rouge-l": 33.3516, | |
| "predict_runtime": 2766.9583, | |
| "predict_samples_per_second": 0.398, | |
| "predict_steps_per_second": 0.066 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 3.1357340812683105, | |
| "learning_rate": 1.6357012750455374e-05, | |
| "loss": 0.6419, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 196.4153, | |
| "eval_loss": 0.5546132922172546, | |
| "eval_rouge-1": 48.0276, | |
| "eval_rouge-2": 22.0229, | |
| "eval_rouge-l": 32.1636, | |
| "eval_runtime": 3772.7644, | |
| "eval_samples_per_second": 0.291, | |
| "eval_steps_per_second": 0.049, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "predict_gen_len": 196.2973, | |
| "predict_loss": 0.5356236100196838, | |
| "predict_rouge-1": 48.5987, | |
| "predict_rouge-2": 22.7122, | |
| "predict_rouge-l": 32.8889, | |
| "predict_runtime": 3794.5656, | |
| "predict_samples_per_second": 0.29, | |
| "predict_steps_per_second": 0.048 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 1.975117802619934, | |
| "learning_rate": 1.4535519125683062e-05, | |
| "loss": 0.5212, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 159.2368, | |
| "eval_loss": 0.5185256004333496, | |
| "eval_rouge-1": 50.9653, | |
| "eval_rouge-2": 25.9311, | |
| "eval_rouge-l": 35.8033, | |
| "eval_runtime": 2838.4663, | |
| "eval_samples_per_second": 0.387, | |
| "eval_steps_per_second": 0.064, | |
| "step": 1647 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "predict_gen_len": 160.9509, | |
| "predict_loss": 0.49497663974761963, | |
| "predict_rouge-1": 51.167, | |
| "predict_rouge-2": 26.1709, | |
| "predict_rouge-l": 36.1567, | |
| "predict_runtime": 2882.8728, | |
| "predict_samples_per_second": 0.382, | |
| "predict_steps_per_second": 0.064 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 2.144341468811035, | |
| "learning_rate": 1.2714025500910747e-05, | |
| "loss": 0.4477, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 168.326, | |
| "eval_loss": 0.49395063519477844, | |
| "eval_rouge-1": 48.9221, | |
| "eval_rouge-2": 23.9901, | |
| "eval_rouge-l": 33.7623, | |
| "eval_runtime": 3077.747, | |
| "eval_samples_per_second": 0.357, | |
| "eval_steps_per_second": 0.059, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "predict_gen_len": 168.0445, | |
| "predict_loss": 0.4722643792629242, | |
| "predict_rouge-1": 49.6399, | |
| "predict_rouge-2": 24.5059, | |
| "predict_rouge-l": 34.4684, | |
| "predict_runtime": 3035.2896, | |
| "predict_samples_per_second": 0.362, | |
| "predict_steps_per_second": 0.061 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 2.392031669616699, | |
| "learning_rate": 1.0892531876138435e-05, | |
| "loss": 0.3979, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 175.8224, | |
| "eval_loss": 0.4893759787082672, | |
| "eval_rouge-1": 50.0387, | |
| "eval_rouge-2": 24.3981, | |
| "eval_rouge-l": 34.4437, | |
| "eval_runtime": 3577.8144, | |
| "eval_samples_per_second": 0.307, | |
| "eval_steps_per_second": 0.051, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "predict_gen_len": 176.2991, | |
| "predict_loss": 0.4671793282032013, | |
| "predict_rouge-1": 50.3641, | |
| "predict_rouge-2": 24.5994, | |
| "predict_rouge-l": 34.933, | |
| "predict_runtime": 3579.3858, | |
| "predict_samples_per_second": 0.307, | |
| "predict_steps_per_second": 0.051 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "grad_norm": 2.4226653575897217, | |
| "learning_rate": 9.071038251366122e-06, | |
| "loss": 0.3643, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_gen_len": 164.6248, | |
| "eval_loss": 0.48532548546791077, | |
| "eval_rouge-1": 49.8422, | |
| "eval_rouge-2": 25.0516, | |
| "eval_rouge-l": 34.9932, | |
| "eval_runtime": 3032.1092, | |
| "eval_samples_per_second": 0.362, | |
| "eval_steps_per_second": 0.06, | |
| "step": 3294 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "predict_gen_len": 165.0855, | |
| "predict_loss": 0.4631520211696625, | |
| "predict_rouge-1": 50.5523, | |
| "predict_rouge-2": 25.7796, | |
| "predict_rouge-l": 35.8814, | |
| "predict_runtime": 2810.2495, | |
| "predict_samples_per_second": 0.391, | |
| "predict_steps_per_second": 0.065 | |
| }, | |
| { | |
| "epoch": 6.38, | |
| "grad_norm": 2.65415620803833, | |
| "learning_rate": 7.249544626593807e-06, | |
| "loss": 0.3238, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_gen_len": 160.2577, | |
| "eval_loss": 0.4873065650463104, | |
| "eval_rouge-1": 50.8821, | |
| "eval_rouge-2": 26.3218, | |
| "eval_rouge-l": 36.3449, | |
| "eval_runtime": 2730.9734, | |
| "eval_samples_per_second": 0.402, | |
| "eval_steps_per_second": 0.067, | |
| "step": 3843 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "predict_gen_len": 163.3909, | |
| "predict_loss": 0.46200141310691833, | |
| "predict_rouge-1": 51.2034, | |
| "predict_rouge-2": 26.3222, | |
| "predict_rouge-l": 36.5416, | |
| "predict_runtime": 2814.6597, | |
| "predict_samples_per_second": 0.391, | |
| "predict_steps_per_second": 0.065 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 3.1916089057922363, | |
| "learning_rate": 5.428051001821493e-06, | |
| "loss": 0.2993, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 166.0501, | |
| "eval_loss": 0.4905695617198944, | |
| "eval_rouge-1": 50.4851, | |
| "eval_rouge-2": 25.7187, | |
| "eval_rouge-l": 35.9106, | |
| "eval_runtime": 2922.2897, | |
| "eval_samples_per_second": 0.376, | |
| "eval_steps_per_second": 0.063, | |
| "step": 4392 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "predict_gen_len": 166.8727, | |
| "predict_loss": 0.4648807942867279, | |
| "predict_rouge-1": 50.8481, | |
| "predict_rouge-2": 25.9305, | |
| "predict_rouge-l": 36.2178, | |
| "predict_runtime": 2955.6147, | |
| "predict_samples_per_second": 0.372, | |
| "predict_steps_per_second": 0.062 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 2.4203264713287354, | |
| "learning_rate": 3.6065573770491806e-06, | |
| "loss": 0.2735, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_gen_len": 167.5301, | |
| "eval_loss": 0.4907337725162506, | |
| "eval_rouge-1": 51.017, | |
| "eval_rouge-2": 26.0933, | |
| "eval_rouge-l": 36.1259, | |
| "eval_runtime": 3054.2577, | |
| "eval_samples_per_second": 0.359, | |
| "eval_steps_per_second": 0.06, | |
| "step": 4941 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "predict_gen_len": 168.3891, | |
| "predict_loss": 0.46499770879745483, | |
| "predict_rouge-1": 51.8088, | |
| "predict_rouge-2": 27.0355, | |
| "predict_rouge-l": 37.2662, | |
| "predict_runtime": 3070.5212, | |
| "predict_samples_per_second": 0.358, | |
| "predict_steps_per_second": 0.06 | |
| }, | |
| { | |
| "epoch": 9.11, | |
| "grad_norm": 3.681400775909424, | |
| "learning_rate": 1.7850637522768672e-06, | |
| "loss": 0.2645, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_gen_len": 163.3725, | |
| "eval_loss": 0.4916660189628601, | |
| "eval_rouge-1": 51.2775, | |
| "eval_rouge-2": 26.6234, | |
| "eval_rouge-l": 36.7381, | |
| "eval_runtime": 2893.3367, | |
| "eval_samples_per_second": 0.379, | |
| "eval_steps_per_second": 0.063, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "predict_gen_len": 164.1664, | |
| "predict_loss": 0.4661196768283844, | |
| "predict_rouge-1": 51.6636, | |
| "predict_rouge-2": 26.831, | |
| "predict_rouge-l": 37.0015, | |
| "predict_runtime": 2923.1866, | |
| "predict_samples_per_second": 0.376, | |
| "predict_steps_per_second": 0.063 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 5490, | |
| "total_flos": 3.718809382109184e+16, | |
| "train_loss": 0.4479398911550831, | |
| "train_runtime": 64358.151, | |
| "train_samples_per_second": 0.511, | |
| "train_steps_per_second": 0.085 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 5490, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 3.718809382109184e+16, | |
| "train_batch_size": 6, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |