HOWILLMAKEIT/Sparrow

麻雀虽小。五脏俱全。基于 Decoder-only + MoE 架构的轻量级大语言模型实验项目(196M/A106M),完整覆盖 Pretrain → SFT → DPO 三阶段训练流程。支持 DeepSpeed ZeRO-1/2/3 分布式训练,提供基于 Reward Model 的自动化评测与 DPO 数据合成方案。

16
/ 100
Experimental
No License No Package No Dependents
Maintenance 13 / 25
Adoption 2 / 25
Maturity 1 / 25
Community 0 / 25

How are scores calculated?

Stars

2

Forks

Language

Python

License

Last pushed

Apr 09, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/HOWILLMAKEIT/Sparrow"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.