| { | |
| "best_metric": 42.5736, | |
| "best_model_checkpoint": "opus_base_simple_tfidf_wce/checkpoint-60000", | |
| "epoch": 2.9574132492113563, | |
| "global_step": 60000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.9753918572555207e-05, | |
| "loss": 0.2098, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_bleu": 35.7996, | |
| "eval_gen_len": 39.4171, | |
| "eval_loss": 0.10897957533597946, | |
| "eval_runtime": 150.8888, | |
| "eval_samples_per_second": 6.912, | |
| "eval_steps_per_second": 0.219, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 1.950746746845426e-05, | |
| "loss": 0.1762, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_bleu": 36.8077, | |
| "eval_gen_len": 39.0182, | |
| "eval_loss": 0.10654434561729431, | |
| "eval_runtime": 116.4461, | |
| "eval_samples_per_second": 8.957, | |
| "eval_steps_per_second": 0.283, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 1.9261016364353314e-05, | |
| "loss": 0.1691, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_bleu": 41.2732, | |
| "eval_gen_len": 34.2013, | |
| "eval_loss": 0.10531464219093323, | |
| "eval_runtime": 133.3831, | |
| "eval_samples_per_second": 7.82, | |
| "eval_steps_per_second": 0.247, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.9014565260252367e-05, | |
| "loss": 0.1635, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_bleu": 41.6402, | |
| "eval_gen_len": 34.6366, | |
| "eval_loss": 0.10416054725646973, | |
| "eval_runtime": 104.9246, | |
| "eval_samples_per_second": 9.94, | |
| "eval_steps_per_second": 0.315, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 1.876811415615142e-05, | |
| "loss": 0.1611, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_bleu": 41.7834, | |
| "eval_gen_len": 35.3193, | |
| "eval_loss": 0.10361454635858536, | |
| "eval_runtime": 153.2125, | |
| "eval_samples_per_second": 6.808, | |
| "eval_steps_per_second": 0.215, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 1.8521786277602524e-05, | |
| "loss": 0.1493, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "eval_bleu": 41.5067, | |
| "eval_gen_len": 35.8773, | |
| "eval_loss": 0.10355237871408463, | |
| "eval_runtime": 92.4584, | |
| "eval_samples_per_second": 11.281, | |
| "eval_steps_per_second": 0.357, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 1.827545839905363e-05, | |
| "loss": 0.1491, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "eval_bleu": 41.1362, | |
| "eval_gen_len": 36.0518, | |
| "eval_loss": 0.10270049422979355, | |
| "eval_runtime": 90.7046, | |
| "eval_samples_per_second": 11.499, | |
| "eval_steps_per_second": 0.364, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 1.8029130520504733e-05, | |
| "loss": 0.1486, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "eval_bleu": 42.0407, | |
| "eval_gen_len": 34.8581, | |
| "eval_loss": 0.10218680649995804, | |
| "eval_runtime": 93.8654, | |
| "eval_samples_per_second": 11.112, | |
| "eval_steps_per_second": 0.352, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 1.778286425473186e-05, | |
| "loss": 0.1472, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_bleu": 40.0238, | |
| "eval_gen_len": 37.8265, | |
| "eval_loss": 0.10189016908407211, | |
| "eval_runtime": 108.5416, | |
| "eval_samples_per_second": 9.609, | |
| "eval_steps_per_second": 0.304, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 1.7536474763406942e-05, | |
| "loss": 0.1463, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_bleu": 42.5557, | |
| "eval_gen_len": 34.1285, | |
| "eval_loss": 0.10188464820384979, | |
| "eval_runtime": 71.1911, | |
| "eval_samples_per_second": 14.651, | |
| "eval_steps_per_second": 0.464, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.7290146884858045e-05, | |
| "loss": 0.1388, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_bleu": 42.1819, | |
| "eval_gen_len": 35.1783, | |
| "eval_loss": 0.10227210074663162, | |
| "eval_runtime": 93.0555, | |
| "eval_samples_per_second": 11.208, | |
| "eval_steps_per_second": 0.355, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 1.704381900630915e-05, | |
| "loss": 0.1378, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "eval_bleu": 42.5524, | |
| "eval_gen_len": 35.2148, | |
| "eval_loss": 0.10202094167470932, | |
| "eval_runtime": 104.8789, | |
| "eval_samples_per_second": 9.945, | |
| "eval_steps_per_second": 0.315, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 1.679755274053628e-05, | |
| "loss": 0.1367, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_bleu": 42.2979, | |
| "eval_gen_len": 34.6817, | |
| "eval_loss": 0.10162690281867981, | |
| "eval_runtime": 90.8884, | |
| "eval_samples_per_second": 11.476, | |
| "eval_steps_per_second": 0.363, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 1.6551163249211358e-05, | |
| "loss": 0.137, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "eval_bleu": 42.4162, | |
| "eval_gen_len": 35.652, | |
| "eval_loss": 0.10157214105129242, | |
| "eval_runtime": 132.7314, | |
| "eval_samples_per_second": 7.858, | |
| "eval_steps_per_second": 0.249, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 1.6304773757886436e-05, | |
| "loss": 0.1373, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_bleu": 42.5736, | |
| "eval_gen_len": 34.418, | |
| "eval_loss": 0.10088738054037094, | |
| "eval_runtime": 113.8031, | |
| "eval_samples_per_second": 9.165, | |
| "eval_steps_per_second": 0.29, | |
| "step": 60000 | |
| } | |
| ], | |
| "max_steps": 324608, | |
| "num_train_epochs": 16, | |
| "total_flos": 1.0348852591617638e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |