simba

simba copied to clipboard

Thank you for your repository, at this time, I have the problem as follows: " File "main_debug.py", line 495, in main train_stats = train_one_epoch( File "/home/ubuntu/workspace/mamba/simba/classification/engine.py", line 48, in train_one_epoch...

duc-anh-2002

Always got nan loss while token_label is False.

5

[03/29 11:44:29] train INFO: Test: [520/521] eta: 0:00:00 loss: 6.7130 (6.8320) acc1: 0.0000 (0.5440) acc5: 0.0000 (2.0020) time: 0.4453 data: 0.0002 max mem: 15878 [03/29 11:44:29] train INFO: Test: Total...

lqniunjunlper

Hi there, I read the research paper, however one thing I was wondering is what would happen if one were to mate EinFFT to a different architecture, for example standard...

Xynonners

Hello, Great work on the model. I was wondering if the timeseries forecasting benchmark code would be added sometime soon? I was not able to find it in the repository,...

chris7neves

Like the title suggests, can we use simba and ignore the labels (just pass all image label as zero) and then train via self-distillation (SSL)? Would you expect the results...

chokevin8

Dear Authors, Thank you for you great work in propose a new channel mixing technique, but I have a question about why don't you expand the dimension of the input...

Kiet0712

Hello, Is there any way in which you can make the pretrained weights for the base model available? I have lots of computational power limitations, so it would be really...

Ale0311

For deployment on the jetson NX

woshituobaye