Show HN: I Built an Image Captioning Tool Using Llama.cpp
a day ago
- #Phoenix LiveView
- #EXIF Metadata
- #AI Image Processing
- Phoenix LiveView application for batch processing JPEG images with AI-generated descriptions and keywords.
- Workflow includes directory scanning, EXIF metadata extraction, AI caption generation, and metadata writing.
- Prioritizes existing EXIF data over LLM generation; uses local LLM if no EXIF data is present.
- Requires Elixir 1.15+, Erlang/OTP 26+, exiftool, fd, and a vision-capable AI model.
- Recommends using a local llama.cpp server to avoid high API costs.
- Setup involves cloning the project, checking dependencies, and starting the Phoenix server.
- Configuration for LLM parameters is done via environment variables.
- Application is licensed under MIT License, copyright Jeff Sandberg.