Show HN: GPT-2 inference in pure C#, 0 bytes allocated per token(github.com)
1 pointby dev-on-bike6 hours ago1 comment