Eval v4: --merged-repo, eos_token_id, deps without unsloth 182872b verified Taimwe commited on 6 days ago
Eval v4: Hermes JSON-string tools, merged-repo loader, eos_token_id 57da0ae verified Taimwe commited on 6 days ago
Eval v2: clean rewrite with PEP 723 deps + inlined helpers 6a38ac4 verified Taimwe commited on 6 days ago
Add eval script (tool-call validity, code sanity, security MCQ) a2ba1ae verified Taimwe commited on 6 days ago