Skip to content
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

When I run the fhe_compatibility, the program is killed #82

Closed
maxwellgodv opened this issue Jun 9, 2023 · 8 comments
Closed

When I run the fhe_compatibility, the program is killed #82

maxwellgodv opened this issue Jun 9, 2023 · 8 comments

Comments

@maxwellgodv
Copy link

qmodel = fhe_compatibility(quant_vgg, data)

print(
f"With {param_c10['dataset_name']}, the circuit has a maximum bit-width of "
f"{qmodel.fhe_circuit.graph.maximum_integer_bit_width()}-bits."
)

1686310285723
@andrei-stoian-zama
Copy link
Collaborator

Most likely your are running out of memory. Compilation may require large amounts of memory and you can check that for example with the htop tool or your System Monitor program. Check if you see a rising amount of memory being used quickly going to 100% and then a crash. How much RAM do you have on your machine ?

@maxwellgodv
Copy link
Author

16

@maxwellgodv
Copy link
Author

How much RAM do I need?

@fd0r
Copy link
Collaborator

fd0r commented Jun 12, 2023

It's hard to say and depends on multiple factors.
Mainly the complexity of your model and the size of the dataset that you use to compile your model.

@andrei-stoian-zama
Copy link
Collaborator

16GB might not be enough for CIFAR, but the test I suggested in my previous reply should help you determine if you're running into memory issues

@maxwellgodv
Copy link
Author

16GB might not be enough for CIFAR, but the test I suggested in my previous reply should help you determine if you're running into memory issues

yes, it is memory issues.If I use the image size of 3 * 224 * 224, how much RAM would I need?

@maxwellgodv
Copy link
Author

It's hard to say and depends on multiple factors. Mainly the complexity of your model and the size of the dataset that you use to compile your model.

If I use the image size of 3 * 224 * 224, how much RAM would I need?

@andrei-stoian-zama
Copy link
Collaborator

That image size might require much more than 16GB, I think you may get away with 32GB for an image oe 3x32x32.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants