StableQuant/Qwen-Templates-Rebuild-Project

1---2license: apache-2.03base_model:4- Qwen/Qwen3.6-27B5- Qwen/Qwen3.6-35B-A3B6- Qwen/Qwen3.5-27B7- Qwen/Qwen3.5-35B-A3B8pipeline_tag: text-generation9library_name: transformers10tags:11- chat-template12- jinja13- jinja214- qwen315- template-fix16- bugfix17---1809.06.202619### This project is on hold now. For most stable experience currently, please use https://huggingface.co/spiritbuun/buun-Qwen3.6-chat_template20 21 22#### Preserve Thinking:23To enable preserve-thinking (Agent can remember what he thought about in older turns) pass the following parameter to llama.cpp startup:24On Linux&MacOS (Windows need slightly different which cant be correctly displayed here...): --chat-template-kwargs '{"preserve_thinking": true}'25 26#### OpenCode:27If using OpenCode do notice they have an open bug for showing thinking content as plain text. Open Bug since 4 months(https://github.com/anomalyco/opencode/issues/11439). Meanwhile you can use the llama.cpp Server switch to hide thinking completely(but still let the model think) using the additonal llama.cpp parameter:28--reasoning-format deepseek29