astrocyted

Hi, Thanks for the reply. Good to know that the source code for this metric is actively maintained! I could not find the repo on github, and the link you...

I just created the issue on the psds_eval repo, where i mentioned the logic for slicing [:-1] too. While we're on the subject of intersection F1 calculation , another issue...

> ```python > lr_monitor: > _target_: pytorch_lightning.callbacks.LearningRateMonitor > logging_interval: null > log_momentum: False > ``` did you even read the title of the issue?

> Isn't `Gammatonegram` already like this? The output is already [Batch, N_filters X Signal length]. Or am I understanding your question wrongly? Just saw your reply now. No. clearly thats...

https://github.com/detly/gammatone/blob/master/gammatone/filters.py This is one example of its implementation. the output of erb_filterbank() function is what im asking for. its quite slow though. I tried to do it in torch too...

Hi Yuan, I would like to double down on this issue. because I don't think it is about whether you use ```mean(dim=0)``` or ```sum(dim=0)``` in where you aggregate the output...

On a different note, I see you normalize the attention values across temporal axis : https://github.com/YuanGongND/psla/blob/7f8fafa23ef707ad63c4f3965ea1a3f0a4bb1bff/src/models/HigherModels.py#L162 this would seemingly encourage the model to attend to one single temporal unit (in...