We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
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
accelerate==1.2.0
Bug is from this commit: d7b1b36
max_memory initialization was moved into _init_infer_auto_device_map, which does not return the max_memory value.
max_memory
_init_infer_auto_device_map
So if max_memory=None is passed to infer_auto_device_map (the default value), then it will still be None at line 1415:
infer_auto_device_map
accelerate/src/accelerate/utils/modeling.py
Line 1415 in cb8b7c6
Leading to error: TypeError: 'NoneType' object is not subscriptable
max_memory=None when passed to infer_auto_device_map does not throw an error.
The text was updated successfully, but these errors were encountered:
@Nech-C
Sorry, something went wrong.
Sorry for the oversight. I will fix it ASAP. Thanks for pointing it out!
Successfully merging a pull request may close this issue.
System Info
Reproduction
Bug is from this commit:
d7b1b36
max_memory
initialization was moved into_init_infer_auto_device_map
, which does not return themax_memory
value.So if max_memory=None is passed to
infer_auto_device_map
(the default value), then it will still be None at line 1415:accelerate/src/accelerate/utils/modeling.py
Line 1415 in cb8b7c6
Leading to error: TypeError: 'NoneType' object is not subscriptable
Expected behavior
max_memory=None when passed to
infer_auto_device_map
does not throw an error.The text was updated successfully, but these errors were encountered: