Plato_based_on_rugpt3large_based_on_gpt2/all_results.json

{
    "epoch": 10.0,
    "total_flos": 4.144250115588096e+16,
    "train_loss": 2.163431647516066,
    "train_runtime": 6094.6449,
    "train_samples": 496,
    "train_samples_per_second": 0.814,
    "train_steps_per_second": 0.102
}