Spaces:
Build error
Build error
File size: 1,535 Bytes
266954d 1efb585 1dcc055 266954d 1efb585 df729c3 266954d 1efb585 df729c3 1efb585 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 | ---
title: Qwen3.5 0.8B OpenAI API
emoji: "🧠"
colorFrom: yellow
colorTo: gray
sdk: docker
app_port: 7860
startup_duration_timeout: 2h
pinned: false
---
# Hugging Face Space Template (OpenAI-Compatible Qwen 0.8B)
This folder is ready to be used as a Docker Space that serves `Qwen3.5-0.8B` behind OpenAI-style endpoints:
- `GET /v1/models`
- `POST /v1/chat/completions` (with streaming)
## 1) Create the Space
1. Go to Hugging Face -> **New Space**.
2. Select **Docker** SDK.
3. Choose hardware:
- For free testing: **CPU Basic**.
4. Create the Space.
## 2) Upload these files
Upload all files from this folder to the root of that Space repository:
- `Dockerfile`
- `requirements.txt`
- `start_server.py`
- `.dockerignore`
- `README.md` (this file)
## 3) Set Space Variables (Settings -> Variables and secrets)
Recommended defaults:
- `MODEL_REPO=unsloth/Qwen3.5-0.8B-GGUF`
- `MODEL_FILE=Qwen3.5-0.8B-Q4_K_M.gguf`
- `N_CTX=2048`
- `N_THREADS=2`
- `CHAT_FORMAT=chatml`
Optional:
- `API_KEY=<your-secret>` to require bearer auth.
- `HF_TOKEN=<token>` if your model repo is private.
## 4) Connect frontend
In this app's Settings:
- Preset: `Hugging Face Space`
- Base URL: `https://<your-space-name>.hf.space/v1`
- Model Name: `Qwen3.5-0.8B-Q4_K_M.gguf`
- API Key: only if you set `API_KEY` in the Space
## Notes
- Free CPU Spaces can sleep when idle and cold-start slowly.
- First boot includes model download, so startup may take a few minutes.
- If you hit memory pressure, use a smaller GGUF quantization file.
|