GitHub · Python
logitcai/tinybrain
A Chinese-first small language model built from scratch: open 60B-token dataset, tokenizer, mHC + Engram + MoE model, pretraining through RL, all in PyTorch. 从零做中文小模型,数据和代码全部开源
- Stars
- 252
- Forks
- 9
- Stars gained in 24 hours
- +232
- Forks gained in 24 hours
- +8
Topics: chinese, dataset, from-scratch, language-model, llm, mixture-of-experts, muon, pretraining, pytorch, reinforcement-learning
Latest data point: Nov 20, 2026 · 252 stars · 9 forks