-
I am trying to train autoslim on cifar10 with mobilenetV2.
But I found there are some errors when training
![image](https://user-images.githubusercontent.com/1567200/152935581-6016246c-61b2-4cf0…
twmht updated
2 years ago
-
In the paper, AutoSlim-MobileNet-v2 at 305M FLOPs achieves 74.2% top-1 accuracy, 2.4% better than default MobileNet-v2 (301M FLOPs),but your realization is only 72.73% at 320M FLOPs。
What do you me…
-
Hi, I check the test_alphanet.py and find it use the training dataset when testing. From your comment 'bn running stats calibration following Slimmable' , does it mean to caculate the training data's …
-
I didn't find its usage.
-
Thanks for your excellent work!
I seem to only see the block search, no channel search found.
-
* AI News
* Conference
* ICML 2021 rebuttal
![image](https://user-images.githubusercontent.com/11782739/112752311-71041680-900d-11eb-9e4a-afd7d4f27d5f.png)
* ACL 2021 rebuttal
* …
-
![slimmable_networks](https://user-images.githubusercontent.com/22609465/50390872-1b3fb600-0702-11e9-8034-d0f41825d775.png)
-
How to compute CPU Latency and FLOPs in pytorch ?
-
Hi Jiahui! This is Bruce.
I'm new to this field and my mentor asked me to train and test slimmable_networks using Cifar-100 dataset. I just want to know is it possible to change imagenet dataset to C…
-
Hi, thank you for your great works. :)
I have a question about network slimming.
I found that the forward function of SlimmableConv2d is defined as follows:
> def forward(self, input):
…