v41_v29_baseline_general_replay_lr4e7_ep10/train_results.json

{
    "epoch": 1.0,
    "total_flos": 3.68251671698516e+17,
    "train_loss": 1.7103283592543677,
    "train_runtime": 6160.8721,
    "train_samples_per_second": 0.496,
    "train_steps_per_second": 0.124
}