Show HN: GPT-2 inference in pure C#, 0 bytes allocated per token

(github.com) by dev-on-bike | May 17, 2026 | 0 comments on HN