Porting nanochat to a TPU: what carries over from PyTorch, and what breaks
28 points
3 days ago
| 2 comments
| github.com
| HN
sciences44
1 hour ago
[-]
Nice work. Regarding your first insight, that's a sneaky one ^^. I experiment a similar issue with fine-tuning a small model with MLX. It was very hard to find out.
reply
adityashankar
14 minutes ago
[-]
For people that are like me : This entire text is AI generated, i feel weird reading it personally, i guess others may not
reply
embedding-shape
12 minutes ago
[-]
Don't tell me you don't enjoy and learn a ton from barely coherent nuggets like this?

> The most important thing the port taught me: if you can avoid it, don't. I decided to do it anyway, so here are the problems I ran into and some insights of my own, trimmed down to five. Each item is tagged with which side it hurt: quality (CORE) or performance (MFU). The first is the bug that fooled us the longest.

reply