idle timeout for codegen

This commit is contained in:
Denton Social
2026-09-24 04:53:06 -05:00
parent 0a9b602ec7
commit e92b3e7228
5 changed files with 176 additions and 22 deletions
+5 -2
View File
@@ -13,10 +13,13 @@
},
"llm": {"base_url": "http://localhost:11434/v1", "model": "qwen3.5:4b"},
"codegen": {
"_comment": "OpenAI-compatible model that writes runnable skill bodies. Larger/slower than the decision or self-assessment model. No max_tokens cap: qwen38-iq3s reasons extensively (~7 min) before emitting the body; reasoning is filtered automatically. timeout is seconds.",
"_comment": "OpenAI-compatible model that writes runnable skill bodies. Larger/slower than the decision or self-assessment model. No max_tokens cap: qwen38-iq3s reasons extensively (~7 min) before emitting the body; reasoning is filtered automatically. timeout is total seconds; idle_warn/idle_timeout surface a silent (wedged) stream in minutes instead of blocking on the total timeout.",
"base_url": "http://localhost:11434/v1",
"model": "qwen38-iq3s",
"timeout": 1200
"timeout": 1200,
"stream": false,
"idle_warn": 60,
"idle_timeout": 180
},
"skill_bodies": "data/skills",
"skills": {