Marin is an open-source codebase and collaborative lab for training foundation models from scratch. It captures experiments through GitHub issues and pull requests, tracks provenance and results, and publishes analysis dashboards for researchers and developers.
Marin is a Fine-tuning & training project. It focuses on reproducing and collaborating on foundation-model training experiments with shared code, data, provenance, and results. It is built as an open-source project for machine-learning researchers and developers. The project is open source (Apache-2.0). It ships for the web and the command line, and it can be self-hosted.
Behind Marin is Marin community, based in the United States, and it first shipped in 2022. The project is developed in the open on GitHub with 1.9k stars and 1.3k commits in the last 90 days. Key capabilities include experiment preregistration, reproducible experiments, and provenance graphs. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match