microndla / mobilevitv3 Goto Github PK

Could you please help me with mobilevit_block.py?
There is this import that I can't fine in original ml_cvnets repository: from ..misc.profiler import module_profile

https://github.com/apple/ml-cvnets/tree/main/cvnets/misc

Hello, your work is wonderful! And I wonder if I could get your pre-trained modules as you have shown on ImageNet-1K, like MobileViTv3-XXS, MobileViTv3-XS, and so on.

Thanks for your great work!

These days I try to compress MobileVit, and reduce the FLOPs and parameters, but the inference time almost doesn't change.

Only when I remove some MV2 blocks, the inference time reduce a little.

My method I used for compressing:
1: Try LinFormer.(This method reduces the FLOPs, but increases the length of sequence. It seems longer sequence means slower speed)
2: Decrease the channels of some layers.
3: Other methods, such as replacing conv33 with conv11(learning from you)

Could you please tell me the reason?

Hi, it's a wonderful work.
Could you please show the code of drawing the Fig, 1&3 in the paper or give some code demos?
I don't know how to draw the analogous scatterplot as follows:

I really appreciate it.

In code:
self.conv_1x1_exp = Identity()

In paper:

File "C:\Users\72344.conda\envs\MobileViTv2\Scripts\cvnets-train.exe_main_.py", line 4, in
ModuleNotFoundError: No module named 'main_train'

And I tried to download this module, but show
"ERROR: Could not find a version that satisfies the requirement main_train (from versions: none)
ERROR: No matching distribution found for main_train"

What can I do?

I obey the README-training-and-evaluation.md, and I stuck by this step: cvnets-train --common.config-file config/classification/mobilevit_small.yaml --common.results-loc results_mobilevit_small.
it always said: cvnets-train command not found. what is 'cvnets-train' for?
I'm not familiar with linux, so I'm confused for a long time. Hoping to find the answer. Thanks!