5M/final_evaluation/benchmarks.json

{"final_train_loss":1.4081796407699585,"tokens_seen":20004864,"steps":1628,"n_params":5181696,"size_label":"5M","eval_loss":1.370229333639145,"eval_perplexity":3.936253307262444}