QwenLM/qwen-code dsw-eas-full-20260813-r3

DSW EAS Harbor full E2E 2026-08-13 r3

on GitHub

Non-production full SWE-bench Verified E2E validation. The release event prepares the exact Qwen Code cache in ACR before PG submission.

Benchmark-Qwen-Ref: v0.21.11

SWE-bench Verified

• Status: QUARANTINED
• Dataset: swe-bench/swe-bench-verified@2; 500/500 completed
• Results: 13 resolved, 8 unresolved, 7 execution errors, 472 infrastructure failures
• Score: not published (non-scoreable or quarantined run)
• Score denominator: valid grader results (13 + 8 = 21)
• Qwen Code: v0.21.11 (7a48b9278f65); model qwen3.7-plus
• Run: pool-64fd5eaf20114b8a; dispatch workflow: https://github.com/QwenLM/qwen-code/actions/runs/31675289458
• Case trajectories: swe-bench-verified-dsw-eas-full-20260813-r3-trajectories.tar.gz (40 cases)