| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 200, | |
| "global_step": 280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 33.699462890625, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 12.5831, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 32.89402770996094, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 12.2228, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 29.5042781829834, | |
| "learning_rate": 6e-06, | |
| "loss": 11.3684, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 29.208715438842773, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 9.8704, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 27.723134994506836, | |
| "learning_rate": 1e-05, | |
| "loss": 7.6666, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 18.65027618408203, | |
| "learning_rate": 1.2e-05, | |
| "loss": 5.3314, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 6.373291015625, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 3.1671, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 2.784041404724121, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.0792, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2142857142857144, | |
| "grad_norm": 1.8078209161758423, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.4497, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 1.2121152877807617, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9399, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.928571428571429, | |
| "grad_norm": 0.5274812579154968, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 0.6974, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.44586724042892456, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.5473, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.642857142857143, | |
| "grad_norm": 0.3544086813926697, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.5048, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.2498295158147812, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 0.4708, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "grad_norm": 0.31819191575050354, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4269, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 0.19436120986938477, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.4093, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.071428571428571, | |
| "grad_norm": 0.19919425249099731, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 0.4568, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 0.19824030995368958, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.3915, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 6.785714285714286, | |
| "grad_norm": 0.19597011804580688, | |
| "learning_rate": 3.8e-05, | |
| "loss": 0.4281, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 0.19348232448101044, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3779, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "eval_gen_len": 234.1551, | |
| "eval_loss": 0.38291501998901367, | |
| "eval_rouge1": 0.9021, | |
| "eval_rouge2": 0.8305, | |
| "eval_rougeL": 0.8829, | |
| "eval_runtime": 46.7129, | |
| "eval_samples_per_second": 9.526, | |
| "eval_steps_per_second": 0.15, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.14055049419403076, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.362, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 0.15631766617298126, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.4151, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.214285714285714, | |
| "grad_norm": 0.15075279772281647, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.3821, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 0.1391923427581787, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.3422, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 8.928571428571429, | |
| "grad_norm": 0.15504033863544464, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3775, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 0.14940936863422394, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 0.3694, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 9.642857142857142, | |
| "grad_norm": 0.12456735968589783, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 0.3817, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.16763611137866974, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.3245, | |
| "step": 280 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 280, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.08333608730624e+16, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |