README.md

metadata

title: Qwen3.5 Reasoning Chat (Claude Distilled)
emoji: 🧠
colorFrom: blue
colorTo: indigo
sdk: gradio
app_file: app.py
pinned: false
license: apache-2.0

Qwen3.5 Reasoning Chat (Claude 4.6 Distilled)

This project runs the Jackrong/Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2-GGUF model using llama-cpp-python.

It is specifically tuned to follow a Chain-of-Thought (CoT) reasoning process using <think> tags, distilled from Claude 4.6 Opus trajectories.

Features

• GGUF Optimized: Runs efficiently on CPU.
• Reasoning First: Designed to "think" before answering.
• No API Key Required: Runs the model directly within the Space environment.

Deployment

1. Log in: huggingface-cli login
2. Create Space: huggingface-cli repo create qwen3.5-reasoning --repo-type space --sdk gradio
3. Upload: huggingface-cli upload <username>/qwen3.5-reasoning . --repo-type space