CooLLaMACEO commited on
Commit
0f79223
·
verified ·
1 Parent(s): caa3198

Update Dockerfile

Browse files
Files changed (1) hide show
  1. Dockerfile +11 -10
Dockerfile CHANGED
@@ -1,19 +1,20 @@
1
  FROM python:3.10-slim
2
 
3
- # Install system dependencies
4
- RUN apt-get update && apt-get install -y \
5
- build-essential \
6
- python3-dev \
7
- wget \
8
- && rm -rf /var/lib/apt/lists/*
9
 
10
- # The -q flag makes it quiet so the logs don't hit the 2MB limit
11
  RUN wget -q -O model.gguf https://huggingface.co/maddes8cht/mosaicml-mpt-7b-chat-gguf/resolve/main/mosaicml-mpt-7b-chat-Q2_K.gguf
12
 
13
- # Install modern llama-cpp and web server
14
- RUN pip install --no-cache-dir llama-cpp-python fastapi uvicorn
 
 
 
 
 
15
 
16
  COPY app.py .
17
 
18
- # Port 7860 is required for Hugging Face Spaces
19
  CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]
 
1
  FROM python:3.10-slim
2
 
3
+ # Install wget to get the model
4
+ RUN apt-get update && apt-get install -y wget && rm -rf /var/lib/apt/lists/*
 
 
 
 
5
 
6
+ # 1. Download the GGUF model (Quiet mode)
7
  RUN wget -q -O model.gguf https://huggingface.co/maddes8cht/mosaicml-mpt-7b-chat-gguf/resolve/main/mosaicml-mpt-7b-chat-Q2_K.gguf
8
 
9
+ # 2. Install llama-cpp-python using a PRE-BUILT wheel
10
+ # This skips the "Building wheel" phase and finishes in seconds
11
+ RUN pip install --no-cache-dir \
12
+ https://huggingface.co/Luigi/llama-cpp-python-wheels-hf-spaces-free-cpu/resolve/main/llama_cpp_python-0.3.22-cp310-cp310-linux_x86_64.whl
13
+
14
+ # 3. Install web server basics
15
+ RUN pip install --no-cache-dir fastapi uvicorn
16
 
17
  COPY app.py .
18
 
19
+ # Port 7860 for Hugging Face
20
  CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]