{ "effective_tokens_per_sec": 6380.59331163488, "epoch": 3.0, "total_flos": 3.0241141054567875e+18, "train_loss": 0.8476084547685393, "train_runtime": 9225.3576, "train_samples_per_second": 4.891, "train_steps_per_second": 0.076 }