Spaces:
Running on Zero
Running on Zero
| title: Qwen2.5 Coder 7B Tools | |
| emoji: 🛠️ | |
| colorFrom: green | |
| colorTo: indigo | |
| sdk: gradio | |
| sdk_version: 5.50.0 | |
| app_file: app.py | |
| pinned: false | |
| short_description: Qwen2.5-Coder-7B (GGUF) with tool calling on ZeroGPU | |
| python_version: "3.12" | |
| startup_duration_timeout: 1h | |
| # Qwen2.5-Coder-7B-Instruct (GGUF) — tool calling | |
| Qwen2.5-Coder-7B-Instruct (GGUF, `q4_k_m`) on Hugging Face **ZeroGPU**. | |
| Exposes two Gradio API endpoints: | |
| - `respond` — streaming chat (the UI). | |
| - `tool_chat` — non-streaming generation with OpenAI-style `tools`; the model's | |
| `<tool_call>...</tool_call>` blocks are returned as raw text so a client can | |
| parse them into OpenAI `tool_calls`. | |
| Used by the local `hf-qwen-bridge` to provide an OpenAI-compatible | |
| `/v1/chat/completions` provider (with tool calling) to Nexus. | |