{ "effective_tokens_per_sec": 2658.117272588518, "epoch": 4.0, "total_flos": 3.454020596210336e+18, "train_loss": 0.7742519322697508, "train_runtime": 25098.6052, "train_samples_per_second": 2.089, "train_steps_per_second": 0.07 }