dongarandClaude Opus 5.5 28092194fe E2E on qwen3.8-27b without thinking ...

The e2e roles move to qwen3.8-27b-nothink, a LiteLLM alias for the same
model with chat_template_kwargs enable_thinking false. With thinking, a
growing share of answers ran to 2-8k tokens (two hit the 8192 cap), which
took one test from 3.5 to 20 minutes.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

2026-10-07 22:49:13 +00:00

..

integration

E2E on qwen3.8-27b without thinking

2026-10-07 22:49:13 +00:00

conftest.py

Pass the capability matrix's models as a pytest option

2026-09-23 23:32:44 +00:00