{ "effective_tokens_per_sec": 7813.8667712883425, "epoch": 4.0, "total_flos": 3.5845266747435254e+18, "train_loss": 0.8291907650812373, "train_runtime": 8889.4262, "train_samples_per_second": 6.385, "train_steps_per_second": 0.213 }