Spaces:

Felladrin
/

MiniSearch-AI-Server

Sleeping

Felladrin commited on Mar 12

Commit

d9391a5

1 Parent(s): 782c406

Use an specific image from llama-server and adjust the CMD arguments

Files changed (1) hide show

Dockerfile CHANGED Viewed

@@ -1,12 +1,11 @@
-FROM ghcr.io/ggml-org/llama.cpp:server
 RUN useradd -m -u 1000 user
 USER user
-ENV PORT=7860
 ENV HOME=/home/user
 ENV PATH=/home/user/.local/bin:$PATH
-EXPOSE $PORT
-CMD ["llama-server", "--port", "$PORT", "--no-mmap", "--flash-attn", "on", "--fit", "on", "--ctx-size", "8192", "--samplers", "min_p", "--min-p", "0.005", "--backend-sampling", "--webui-mcp-proxy", "-hf", "unsloth/Qwen3.5-4B-GGUF:UD-Q4_K_XL", "--chat-template-kwargs", "{ \"enable_thinking\": false }"]

+FROM ghcr.io/ggml-org/llama.cpp:server@sha256:a9b230ae078a6dd947ede5825f227af9fc5fc9e126c4717bebd4762c02601285
 RUN useradd -m -u 1000 user
 USER user
 ENV HOME=/home/user
 ENV PATH=/home/user/.local/bin:$PATH
+EXPOSE 7860
+CMD ["--port", "7860", "--no-mmap", "--flash-attn", "on", "--fit", "on", "--ctx-size", "8192", "--samplers", "min_p", "--min-p", "0.005", "--backend-sampling", "--webui-mcp-proxy", "-hf", "unsloth/Qwen3.5-4B-GGUF:UD-Q4_K_XL", "--chat-template-kwargs", "{ \"enable_thinking\": false }"]