ubix commited on
Commit
8244fa9
·
verified ·
1 Parent(s): a6d8710

Update Dockerfile

Browse files
Files changed (1) hide show
  1. Dockerfile +5 -14
Dockerfile CHANGED
@@ -1,24 +1,15 @@
1
  FROM ghcr.io/ggml-org/llama.cpp:full
2
 
3
- # Install dependencies
4
- RUN apt-get update && apt-get install -y wget && rm -rf /var/lib/apt/lists/*
5
 
6
- # Set a stable working directory
7
- WORKDIR /models
8
 
9
- # Download the model into a known location
10
- RUN wget -O deepseek-coder-6.7b-instruct.Q5_K_M.gguf \
11
- "https://huggingface.co/TheBloke/deepseek-coder-6.7B-instruct-GGUF/resolve/main/deepseek-coder-6.7b-instruct.Q5_K_M.gguf"
12
-
13
- # Expose port (optional but good practice)
14
- EXPOSE 7860
15
-
16
- # Run server with correct model path
17
  CMD ["--server", \
18
- "-m", "/models/deepseek-coder-6.7b-instruct.Q5_K_M.gguf", \
19
  "--port", "7860", \
20
  "--host", "0.0.0.0", \
21
- "--ctx-size", "8192", \
22
  "--n-predict", "-1", \
23
  "--threads", "2", \
24
  "--jinja"]
 
1
  FROM ghcr.io/ggml-org/llama.cpp:full
2
 
3
+ RUN apt update && apt install wget -y
 
4
 
5
+ # Download the coding-specialized model
6
+ RUN wget "https://huggingface.co/unsloth/Qwen3.5-9B-GGUF/resolve/main/Qwen3.5-9B-Q4_K_M.gguf" -O /Qwen3.5-9B-Q4_K_M.gguf
7
 
 
 
 
 
 
 
 
 
8
  CMD ["--server", \
9
+ "-m", "/Qwen3.5-9B-Q4_K_M.gguf", \
10
  "--port", "7860", \
11
  "--host", "0.0.0.0", \
12
+ "--ctx-size", "32768", \
13
  "--n-predict", "-1", \
14
  "--threads", "2", \
15
  "--jinja"]