Hugging Face API Agent Evaluation Runner

Set HF_TOKEN as a Space secret.
Optionally set HF_MODEL_ID, HF_PROVIDER, HF_ASR_MODEL_ID, HF_VQA_MODEL_ID.
Log in and run the evaluation. The table includes submitted answers, local answer-key comparison, and trace.

Use only for study/debug when HF Inference Provider credits are depleted.

Use public validation fallback

Run Status / Submission Result

Questions, Answers, Local Score, and Trace

Questions, Answers, Local Score, and Trace