feat: include candidates and response in benchmark JSON output

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-07-28 14:47:22 -07:00 · 2026-02-14 09:25:48 -05:00
parent bebf008a40
commit a8982fcd9e
1 changed files with 2 additions and 0 deletions
@@ -159,6 +159,8 @@ def benchmark_model(url, model, prompt, num_ctx):
    candidates = filter_candidates(response_text)
    result["candidate_count"] = len(candidates)
    result["unique_candidates"] = len(set(candidates))
+    result["candidates"] = candidates
+    result["response"] = response_text

    return result