Skip to content

feat(edge-agentic): add BFCL v4 edge-agentic workload to edge benchma…#331

Draft
Palanivelg wants to merge 6 commits into
mlcommons:masterfrom
Palanivelg:feat/edge-agentic-rules
Draft

feat(edge-agentic): add BFCL v4 edge-agentic workload to edge benchma…#331
Palanivelg wants to merge 6 commits into
mlcommons:masterfrom
Palanivelg:feat/edge-agentic-rules

Conversation

@Palanivelg

Copy link
Copy Markdown

Summary

Adds the Edge Agentic workload (BFCL v4 function-calling accuracy) to the
MLPerf Inference edge benchmarks and required-scenarios tables.

Changes

  • Edge benchmarks table: new row — Language | Agentic Function Calling | Qwen3.6-27B | gorilla-llm/gorilla-eval-set (BFCL v4)
  • Edge required-scenarios table: new row — Language | Agentic Function Calling | Offline (accuracy-only)

TBD (blocking merge)

  • QSL size — pending WG agreement
  • Multi-turn overall accuracy threshold — pending full-dataset run
  • Accuracy thresholds (99% of reference) — pending WG review

Related

@github-actions

github-actions Bot commented Jun 8, 2026

Copy link
Copy Markdown

MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant