-
Notifications
You must be signed in to change notification settings - Fork 486
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
core dump when copying tensor from cuda to xla gpu device #3466
Comments
I think eventually it will call xla/torch_xla/csrc/tensor_util.cpp Line 520 in 94d7165
not exactly sure what this will output for a |
Hmmm I think the Easy fix would probably to move the source to cpu first :) but then you're twice as slow. I'm not 100% familiar with the API's you'd need to do a cuda device-to-host copy to an xla memory buffer, but you'd probably need to include some cuda headers? |
Although I wonder if we can do something inside of that function like:
|
🐛 Bug
core dump when copying a tensor from cuda to xla gpu device
To Reproduce
Expected behavior
exit successfully
Environment
Additional context
core dump call stack
The text was updated successfully, but these errors were encountered: