all_results.json

{"train_loss": 2.553584098815918, "train_perplexity": 12.853088065046574, "eval_loss": 2.5428810119628906, "eval_perplexity": 12.716253927249992}