8 ms·
> Quantization in ML refers to a technique used to make it possible to run certain AI models on less powerful computers and chips by replacing specific long num
by janalsncm 3y ago
> Quantization in ML refers to a technique used to make it possible to run certain AI models on less powerful computers and chips by replacing specific long numeric sequences in a model’s architecture with shorter ones.
It’s always amusing when the press tries to explain technical concepts. Quantization just means substituting high precision numeric types with lower precision ones. And not specific “numeric sequences”, all numbers.
- SilverBirch 3y agoIt's funny how the original phrase sounds like it was generated by chatgpt. I'd guess the reason they didn't phrase it how you suggest is they probably thought that people who don't know what quantization means aren't going to be happy if you throw precision numerics at them. I don't knwo why they didn't just say "some of the numbers are rounded" though.