Sachin Goyal
Sachin Goyal
@oindrilasaha would be good if you can have a look at this PR
@adityakusupati , incorporated the changes from PR#173.
> I'm waiting for a fresh machine to replicate these steps. Once I get my hands on one, I'll try to simplify things. > > The VSCode/IoT Devkit version hasn't...
Hey, To all those who are unable to resolve this issue in linux Shift to the windows version, I can confirm that it works perfectly there without this issue. This...
As I mentioned in my previous comment, this is a known issue with the python version of the Montreal Forced Aligner. Try using the windows version of the same and...
For ImageNet, the validation set is used as the test set. The place where you downloaded ImageNet should have a val_dir folder as well.
Hi, Can you check if you are able to get correct results with B/16? I personally have tried B/16, L/14, L/14-336px and even many other people in the community have...
Hi, Thanks for pointing this out. I just checked my dataset, and it has 23.7k images only, probably some issue would have happened when I downloaded. I will update the...
The results should be much better even without weight ensembling. I am not really sure about what the baseline for say standard cross entropy finetuning would be, but still FLYP...
Hi, Were you able to check this on other models?