RuntimeError: stack expects a non-empty TensorList

[wenjun90]

trafficstars

Hi @tcapelle Thank you for your great job. Could I ask you a question, please?

I try training with 2 class and have an error like that. How can we fix this error?

** in forward(self, x) 35 def forward(self, x): 36 if self.debug: print(f' input len: {len(x), x[0].shape}') ---> 37 x = torch.stack(x, dim=1) 38 if self.debug: print(f' after stack: {x.shape}') 39 batch_size, seq_length, c, h, w = x.shape

RuntimeError: stack expects a non-empty TensorList**

Thank you very much!

[tcapelle]

Your batch has zero images, check with ''' dls.one_batch() ''' And inspect dimensions.

The problem rises because your list of images is empty, so it cannot stack them together as a image sequence.

On Sat, Jul 31, 2021, 11:02 AM wenjun90 @.***> wrote:

Reopened #26 https://github.com/tcapelle/action_recognition/issues/26.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/tcapelle/action_recognition/issues/26#event-5094181798, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEMWOALU2CGLBW2CRHPRAPLT2O32TANCNFSM5BJZSVVQ .

[wenjun90]

Thank @tcapelle for your response,

One more question. How can you predict a file .avi as well as the own test set and how can get f1-score and AUC.

Thank you very much!

[tcapelle]

That's more than one question 😜, The actual code expects individual frames, so you need to process them. Metrics can be found on sklearn and use the fastai wrapper, so you can pass them to the learner.

On Sat, Jul 31, 2021, 12:46 PM wenjun90 @.***> wrote:

Thank @tcapelle https://github.com/tcapelle for your response,

One more question. How can you predict a file .avi as well as the own test set and how can get f1-score and AUC.

Thank you very much!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/tcapelle/action_recognition/issues/26#issuecomment-890328756, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEMWOAOH5H5HP6GINQ5TEQLT2PIA7ANCNFSM5BJZSVVQ .

[wenjun90]

Thanks @tcapelle So, process with like that: with file .avi --> processing --> sequence image --> input predict --> predict?