Porting nanochat to a TPU: what carries over from PyTorch, and what breaks(github.com)
github.com
Porting nanochat to a TPU: what carries over from PyTorch, and what breaks
https://github.com/tucan9389/nanochat-jax/discussions/1
12 comments
Every engineer who's worked with "agentic" workflows can tell that this is entirely generated with very little tweaks. It's embarrassing.
Even the title is a Claudeism, it makes me sad
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.
For people that are like me : This entire text is AI generated, i feel weird reading it personally, i guess others may not
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.
> 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.
TBH llm generated text is usually better than this..
TBF it kind of depends on what weights you end up using, the quality gap can be pretty wide.
he's just dogfooding us all...
nice