athirorg/USS-reward-model-qwen-FT-utterance

Add full raw JSON results

Add metrics summary: ft_utterance_scoring_metrics.md

Add utterance-level scoring predictions

initial commit