jingyaogong/minimind
Lightweight LLM training framework
An open-source project that enables training a small language model from scratch on personal GPUs with minimal cost and time. It provides a complete pipeline for pretraining, fine-tuning, and reinforcement learning using native PyTorch implementations.
- Stars
- 62,500
- Stars gained in 7 days
- +1,000
- Contributors
- 20
- Forks
- 8,100
- Days trending in the last 30
- 17 days
- Mark
- Possible hype — 3,125 stars per contributor. An automatic mark on public metrics: above 500 is watch, above 2,000 possible hype (not a judgement of quality).