space.yaml

title: Chat with DeepSeek Coder 1.3B
emoji: 💬
colorFrom: blue
colorTo: indigo
sdk: gradio
sdk_version: 4.25.0 # Or your target Gradio version
python_version: "3.10" # Or "3.9", "3.11", etc.
app_file: app.py
pinned: false
# If you were using secrets (like HF_TOKEN for an API):
# secrets:
#   - HF_TOKEN
# If you needed persistent storage:
# persistent_storage:
#   - path: /data
#     mount_to: /home/user/app/data # Example mount path
---

# Chat with Quantized DeepSeek Coder 1.3B

This is a Gradio application that allows you to chat with the `deepseek-ai/deepseek-coder-1.3b-instruct` model.
The model is loaded directly within the Space. If a GPU is available, 8-bit quantization is attempted; otherwise, the model is loaded on CPU in its native precision.

**Note:** Model loading can take a few minutes, especially on the first run or after a rebuild.