metadata
language:
- en
license: apache-2.0
tags:
- agentfile
- moe
datasets:
- Idavidrein/gpqa
- TIGER-Lab/MMLU-Pro
- ScaleAI/SWE-bench_Pro
metrics:
- accuracy
AgentFile Model Merger
Official Reasoning Suite
- suite:
official_reasoning_suite - generated_at:
None
This repo includes official-style eval results for GPQA and MMLU-Pro.