Sign inSign up

irfanuruchi/local-llm-service

By irfanuruchi

Updated 7 months ago

Self-hosted local LLM service with persistent multi-chat history

Image
Machine learning & AI
Web servers
Databases & storage
0

1.6K

irfanuruchi/local-llm-service repository overview

Local LLM Service

Self-hosted local LLM service with:

  • Persistent multi-chat history
  • Conversation rename support
  • Default baked-in GGUF model
  • Model override via environment variables
  • Multi-architecture support (linux/amd64, linux/arm64)
  • Clean minimal dark UI

🚀 Quick Start

docker run -p 8000:8000 irfanuruchi/local-llm-service:latest

Open:

http://localhost:8000

With persistence

docker run \
  -v $(pwd)/data:/data \
  -p 8000:8000 \
  irfanuruchi/local-llm-service:latest

Override model

docker run \
  -v $(pwd)/models:/models \
  -e MODEL_PATH=/models/custom.gguf \
  -p 8000:8000 \
  irfanuruchi/local-llm-service:latest

Environment variables

VariableDefaultDescription
MODEL_PATH/models/model.ggufPath to GGUF model
MODEL_NAMElocal-llmModel display name
N_GPU_LAYERS35GPU layers for inference
N_CTX4096Context window size
CHAT_DB_PATH/data/chat_history.dbSQLite DB location

Source code

Github: https://github.com/irfanuruchi/local-llm-service

Tag summary

Content type

Image

Digest

sha256:15563f622

Size

2.1 GB

Last updated

7 months ago

docker pull irfanuruchi/local-llm-service