{ "effective_tokens_per_sec": 2517.4633871293017, "epoch": 4.0, "total_flos": 6.460565555370787e+18, "train_loss": 1.1815453414993533, "train_runtime": 50857.5595, "train_samples_per_second": 1.727, "train_steps_per_second": 0.029 }