InternVL2-8B-MPO/train_results.json

{
    "epoch": 1.0,
    "train_loss": 0.8237529401029788,
    "train_runtime": 34374.0784,
    "train_samples": 1480532,
    "train_samples_per_second": 43.071,
    "train_steps_per_second": 0.168
}