prabeakarve/microgpt
在 GitHub 打开 ↗A from-scratch, dependency-free GPT implementation in pure NumPy: byte-level tokenizer, multi-head self-attention, Adam trainer, and top-k sampler.
Star50
24h 新增—
7 天新增—
Fork0
Issue0
语言Python
项目简介
A from-scratch, dependency-free GPT implementation in pure NumPy: byte-level tokenizer, multi-head self-attention, Adam trainer, and top-k sampler.
Star 趋势
话题标签
attentiondeep-learningfrom-scratchgptlanguage-modelllmmachine-learningnumpypythontransformer
仓库信息
| License | MIT |
| 创建时间 | 2026-11-05 |
| 最后推送 | 23 分钟前 |
| 热度分 | 5.0 |