Sign inSign up

mrhappynice/bplus-xngllamasearch

By mrhappynice

•Updated 10 months ago

Image
0

1.7K

mrhappynice/bplus-xngllamasearch repository overview

bplus🤷🏻‍♂️ XNGllamasearch

Private local search with SearXNG, Custom bplus-searchrs + other apps⁠ app(search without MCP), and llama.cpp to run any models in the single container. Vulkan backend, this script ready for AMD/Intel with Vulkan drivers installed.

  • run.sh:
docker run --name xngllamasearch-cont -d \
  --network host \
  --device=/dev/dri \
  --group-add video \
  -v "$PWD/config:/etc/searxng" \
  -v "$PWD/data:/var/cache/searxng" \
  -e INSTANCE_NAME="SearXNG" \
  -e BASE_URL="http://localhost:8080/" \
  -e SEARXNG_URL=http://127.0.0.1:8080 \
  -e AUTH_USERNAME= \
  -e AUTH_PASSWORD= \
  -e OPENAI_API_BASE=http://localhost:1234/v1 \
  -e LMSTUDIO_API_BASE="http://localhost:1234/v1" \
  -e OPENAI_API_KEY="sk-proj-xxxxx-x-x-x-hwA" \
  -e OPENROUTER_API_KEY="sk-oxxxxr-xx-xxxx" \
  -e GOOGLE_API_KEY="xxxxx" \
  -e MODEL="bartowski/Llama-3.2-1B-Instruct-GGUF:Q4_K_M" \
  -e USE_NATIVE=1 \
  mrhappynice/bplus-xngllamasearch
  • add your keys and llama.cpp compatible gguf in the MODEL e-var, from huggingface with the quant as shown

Tag summary

Content type

Image

Digest

sha256:5d9621442…

Size

511.7 MB

Last updated

10 months ago

docker pull mrhappynice/bplus-xngllamasearch