CNN algorithms like DenseNet DenseNet stress parameter efficiency, which usually results in less FLOPs. However, what I am struggling to understand is why this is important. For DenseNet, in particular, it has low inference speed. Isn't the purpose of decreased parameter size/FLOPs to decrease the time for inference? Is there another real world reason, such as perhaps less energy used, for these optimizations?
Asked
Active
Viewed 775 times
1
-
Iām voting to close this question because it is not about programming as defined in the [help] but about ML theory and/or methodology - please see the intro and NOTE in the `machine-learning` [tag info](https://stackoverflow.com/tags/machine-learning/info). ā desertnaut Sep 20 '21 at 10:33
1 Answers
1
There is a difference between overall inference time vs. per parameter/FLOPs training efficiency. Having lower parameter/FLOPs in training does not guarantee higher speed in inference. Because overall inference depends on the architecture and how predictions are computed.

patagonicus
- 203
- 2
- 10