ludwig
LLM FrameworksDeclarative deep learning framework for training, fine‑tuning, and deploying LLMs, multimodal models, and tabular AI using YAML configs with zero boilerplate Python
Features
- Train or fine‑tune LLMs (e.g., LoRA, PiSSA) with a single YAML config
- Support for multimodal vision‑language models via `is_multimodal: true`
- Advanced time‑series forecasting encoders (PatchTST, N‑BEATS) and dedicated forecast API
- Game‑theoretic multi‑task loss balancing (Nash‑MTL, Pareto‑MTL)
- Automatic YAML config generation from natural‑language prompts
Recent releases
View all 22 releases →
Review required
v0.17.3
Breaking risk
Dependencies
Breaking upgrade
Dependency bump + AutoML change
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
About
Stars
11,739
Forks
1,217
Languages
Python
Jupyter Notebook
Dockerfile
Install & Platforms
Install via
pip
