-
-
Notifications
You must be signed in to change notification settings - Fork 39
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BUG] Why is the backpropagation calculation so slow when I use the mamba network? #220
Labels
Comments
Hello there, thank you for opening an Issue ! 🙏🏻 The team was notified and they will get back to you asap. |
@1325116124 its using the mamba scan, or SSM, it should be updated soon! |
I had the same problem. I don't suppose you managed to fix it right ? |
@Alex-Naxitus yes, it should be updated now! |
@Alex-Naxitus @1325116124 it was the backscan that was really slow, let me know so I can close this issuee! |
Stale issue message |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When I used the mamba network, I defined a loss to test backpropagation and found that the calculation was very slow. Setting the len length to 1024 requires a long waiting time. code show as below:
`import torch
import torch.nn as nn
from zeta.nn import MambaBlock
block = MambaBlock(dim=512, depth=1)
x = torch.randn(1, 1024, 512)
target = torch.randn(1, 1024, 512)
loss_fn = nn.MSELoss()
y = block(x)
loss = loss_fn(y, target)
loss.backward()
print("Output shape:", y.shape)
print("Loss value:", loss.item())
`
Upvote & Fund
The text was updated successfully, but these errors were encountered: