3 ms·
In fairness, for the last major scientific/academic ML-based project I was working on, we had to make a lot of compromises due to speed and performance.
by Fomite 5y ago
In fairness, for the last major scientific/academic ML-based project I was working on, we had to make a lot of compromises due to speed and performance.
- toxik 5y agoMy experience is that you use as large a model as your target platform and application allows. Bigger is better.