vision_QA_medgemma / app /inference_client.py

Commit History

  1. Detect and flag truncated responses, raise token budget to 2048

    fd88749

  1. Initial commit: MedGemma medical VQA chat UI

    cbee686