torchstat
torchstat copied to clipboard
Swall0w
Model analyzer in PyTorch
i got this error in the running time: File "col_cub_gan_gn.py", line 513, in main train(args) File "col_cub_gan_gn.py", line 232, in train stat(clss,temp) File "/home/xujia/anaconda3/envs/xujia-py37/lib/python3.7/site-packages/torchstat/statistics.py", line 70, in stat ms =...
When computing the GFLOPs of ResNet50, the answer is around 100GFLOPs. However, the result in other paper is around 500. Is there something wrong with your computing method?
After installing torchstat, the following code won't work! ``` model.cuda() tensor = tensor.cuda() model(tensor) ``` and the error traces back to 👍 ``` ~/miniconda3/lib/python3.7/site-packages/torchstat/model_hook.py in wrap_call(module, *input, **kwargs) 45 46...
`torchstat.stat` does not work with arbitrarily shaped input, it only accepts inputs with shape `(channels, height, width) I tried running it with an AutoEncoder for time series and it failed....
I think this tool may be quite useful, but there are still something not convenient for me... I want to get the data of my model, but since my model's...
I want to use it to compute parameters and memory for my model, but my model needs other things to input, how can I input the arguments?
Thank you for your work! It is a very useful tool. But if the inputs of the network are more than one image, like stereo matching task which needs input...
What's the different between MAdd and the GFlops which are showed in the output? I look at the ResNet paper and found that the MAdd is equal to the GFLOPs....
