https://github.com/artem-zinnatullin/llm-combinator

LLM Combinator exports real-time results of multi-modal Ollama executions with input from Frigate cameras and HA sensors into HomeAssistant, recursively.
https://github.com/artem-zinnatullin/llm-combinator

Last synced: 23 days ago
JSON representation

LLM Combinator exports real-time results of multi-modal Ollama executions with input from Frigate cameras and HA sensors into HomeAssistant, recursively.

• Host: GitHub
• URL: https://github.com/artem-zinnatullin/llm-combinator
• Owner: artem-zinnatullin
• License: mit
• Created: 2024-06-20T04:28:15.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2024-06-30T21:31:15.000Z (4 months ago)
• Last Synced: 2024-07-04T12:17:04.133Z (4 months ago)
• Language: Kotlin
• Homepage:
• Size: 116 KB
• Stars: 7
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# llm-combinator

LLM Combinator exports real-time results of multi-modal Ollama executions with input from Frigate cameras and HA sensors into HomeAssistant, recursively.

# Milestones

- [x] Ollama prompts with multi-modal (images) models

- [x] Frigate latest.jpg over HTTP

- [x] Frigate reactive motion updates over MQTT

- [x] Real-time export of Ollama response fields as HomeAssistant binary sensors

- [x] Extensive YAML configuration file

- [x] Multi-prompting for a single Camera

- [x] Lossy LLM response decoding: "yes" == true, etc

- [x] Docker image (amd64 & arm64) published to [ghcr.io](https://github.com/artem-zinnatullin/llm-combinator/pkgs/container/artem-zinnatullin%2Fllm-combinator) for each commit

- [] OpenAI-compatible APIs support

- [] Authentication with all network components

- [] Recursive inclusion of HomeAssistant sensors into model prompt

- [] Extensive logging configuration

# Local Development

Project is built in Kotlin JVM with Gradle Build System and is fully supported by free version of IntelliJ IDEA.