llama-factory/saves/Qwen2-7B-Instruct/train_results.json

{
    "epoch": 5.9786476868327405,
    "total_flos": 2.4604999894985933e+17,
    "train_loss": 1.4173777671087355,
    "train_runtime": 2918.7063,
    "train_samples_per_second": 9.214,
    "train_steps_per_second": 0.072
}