{"slug":"sediman-agent-huggingface-community-evals","source_name":"sediman-agent/huggingface-community-evals","name":"Sediman Agent/Huggingface Community Evals","description":"Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware. Use for backend selection, local GPU evals, and choosing between vLLM / Transformers / accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publication, or community-evals automation.","version":1,"lift":{"pass_rate_delta_pts":68.18,"pass_rate_pct":100,"total_cases":22,"passed_cases":22,"tokens_delta_pct":22.4,"turns_delta_pct":0,"verdict":"pass","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-08-20T01:49:33.341558+00:00"},"skill_score":0.9953271028037383,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":68.18,"with_pass_pct":100,"without_pass_pct":31.8,"tokens_delta_pct":22.4,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":21,"verdict":"pass","never_hurt":true,"completed_at":"2026-08-20T01:49:33.341558+00:00","run_id":"0d805ae8-ca0a-4610-9445-a38d73bc2010","latest_attempt":{"id":"0d805ae8-ca0a-4610-9445-a38d73bc2010","runner_model":"gemini-3.6-flash","overall_verdict":"pass","completed_at":"2026-08-20T01:49:33.341558+00:00","gate":null},"version_number":1,"is_latest_version":true,"gate":null}],"trust":{"skill_safety":"passed","safety_status":"clean","intent_verdict":"safe","content_status":"clean","indexable":true},"license":"NOASSERTION","install_count":3,"total_uses":175,"total_users":8,"manifest_hash":"cd231337617e311d9771974ed63e44fc79e61e54dc1a5c33efd510d4122a08ae","install_name":"sediman-agent-huggingface-community-evals","version_id":"c5bc348b-8c05-4890-9340-30a1576f7e3f","body_sha256":"cd231337617e311d9771974ed63e44fc79e61e54dc1a5c33efd510d4122a08ae","render_hash":"d9516655cfd508aa190b970a63bde106b0c594a97ed0019fdfe7bab95e3e0af6","raw_url":"https://app.decimal.ai/s/sediman-agent-huggingface-community-evals/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/sediman-agent-huggingface-community-evals"}