| { | |
| "best_metric": 0.34489724040031433, | |
| "best_model_checkpoint": "./checkpoint-my/checkpoint-1500", | |
| "epoch": 7.56789812129618, | |
| "eval_steps": 500, | |
| "global_step": 7500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.7375328083989503e-05, | |
| "loss": 4.7871, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "eval_bleu": 26.4661, | |
| "eval_gen_len": 31.9891, | |
| "eval_loss": 1.3954318761825562, | |
| "eval_runtime": 189.1048, | |
| "eval_samples_per_second": 5.352, | |
| "eval_steps_per_second": 1.338, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 1.4750656167979002e-05, | |
| "loss": 0.3945, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "eval_bleu": 27.1091, | |
| "eval_gen_len": 32.2134, | |
| "eval_loss": 0.34945473074913025, | |
| "eval_runtime": 187.8434, | |
| "eval_samples_per_second": 5.387, | |
| "eval_steps_per_second": 1.347, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 1.2125984251968505e-05, | |
| "loss": 0.145, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_bleu": 27.3119, | |
| "eval_gen_len": 32.2885, | |
| "eval_loss": 0.34489724040031433, | |
| "eval_runtime": 188.4135, | |
| "eval_samples_per_second": 5.371, | |
| "eval_steps_per_second": 1.343, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 1.650043744531934e-05, | |
| "loss": 0.3623, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "eval_bleu": 3.0191, | |
| "eval_gen_len": 46.4427, | |
| "eval_loss": 0.7403104901313782, | |
| "eval_runtime": 245.2236, | |
| "eval_samples_per_second": 4.127, | |
| "eval_steps_per_second": 1.032, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 1.562554680664917e-05, | |
| "loss": 0.3224, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "eval_bleu": 3.0664, | |
| "eval_gen_len": 46.9269, | |
| "eval_loss": 0.7314952611923218, | |
| "eval_runtime": 247.2109, | |
| "eval_samples_per_second": 4.094, | |
| "eval_steps_per_second": 1.023, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 1.4750656167979002e-05, | |
| "loss": 0.3066, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "eval_bleu": 2.9075, | |
| "eval_gen_len": 47.915, | |
| "eval_loss": 0.726075291633606, | |
| "eval_runtime": 280.6593, | |
| "eval_samples_per_second": 3.606, | |
| "eval_steps_per_second": 0.901, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 1.3875765529308838e-05, | |
| "loss": 0.2952, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "eval_bleu": 2.9214, | |
| "eval_gen_len": 47.5, | |
| "eval_loss": 0.7223652005195618, | |
| "eval_runtime": 268.7931, | |
| "eval_samples_per_second": 3.765, | |
| "eval_steps_per_second": 0.941, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "learning_rate": 1.3000874890638671e-05, | |
| "loss": 0.2859, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "eval_bleu": 3.0026, | |
| "eval_gen_len": 47.8577, | |
| "eval_loss": 0.7183992266654968, | |
| "eval_runtime": 271.7722, | |
| "eval_samples_per_second": 3.724, | |
| "eval_steps_per_second": 0.931, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.1257, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "eval_bleu": 27.2567, | |
| "eval_gen_len": 32.3172, | |
| "eval_loss": 0.3456554710865021, | |
| "eval_runtime": 190.3961, | |
| "eval_samples_per_second": 5.315, | |
| "eval_steps_per_second": 1.329, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 1.3162393162393164e-05, | |
| "loss": 0.1218, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "eval_bleu": 27.4171, | |
| "eval_gen_len": 32.2915, | |
| "eval_loss": 0.3452778160572052, | |
| "eval_runtime": 193.9304, | |
| "eval_samples_per_second": 5.218, | |
| "eval_steps_per_second": 1.305, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "learning_rate": 1.247863247863248e-05, | |
| "loss": 0.1189, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 5.64, | |
| "eval_bleu": 27.4615, | |
| "eval_gen_len": 32.2204, | |
| "eval_loss": 0.34533679485321045, | |
| "eval_runtime": 186.2569, | |
| "eval_samples_per_second": 5.433, | |
| "eval_steps_per_second": 1.358, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 1.1794871794871796e-05, | |
| "loss": 0.1166, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "eval_bleu": 27.6321, | |
| "eval_gen_len": 32.2816, | |
| "eval_loss": 0.34552034735679626, | |
| "eval_runtime": 185.1476, | |
| "eval_samples_per_second": 5.466, | |
| "eval_steps_per_second": 1.366, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "learning_rate": 1.125462495795493e-05, | |
| "loss": 0.1171, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "eval_bleu": 27.3986, | |
| "eval_gen_len": 32.4397, | |
| "eval_loss": 0.34709280729293823, | |
| "eval_runtime": 204.1804, | |
| "eval_samples_per_second": 4.956, | |
| "eval_steps_per_second": 1.239, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "learning_rate": 1.0581903800874538e-05, | |
| "loss": 0.1157, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "eval_bleu": 27.5083, | |
| "eval_gen_len": 32.2816, | |
| "eval_loss": 0.34769660234451294, | |
| "eval_runtime": 209.4596, | |
| "eval_samples_per_second": 4.831, | |
| "eval_steps_per_second": 1.208, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 7.57, | |
| "learning_rate": 9.909182643794148e-06, | |
| "loss": 0.1121, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 7.57, | |
| "eval_bleu": 27.4437, | |
| "eval_gen_len": 32.2194, | |
| "eval_loss": 0.34789395332336426, | |
| "eval_runtime": 209.9641, | |
| "eval_samples_per_second": 4.82, | |
| "eval_steps_per_second": 1.205, | |
| "step": 7500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 14865, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "total_flos": 5.2005742820563354e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |