-
Notifications
You must be signed in to change notification settings - Fork 21.4k
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
[quantization] fix conv2d #23690
[quantization] fix conv2d #23690
Conversation
torch/nn/quantized/functional.py
Outdated
prepacked_weight, bias, | ||
stride, padding, dilation, | ||
groups, scale, zero_point).permute([0, 3, 1, 2]) | ||
return _ops.quantized.fbgemm_conv2d(input.permute([0, 2, 3, 1]), |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
maybe remove as _ops
in the import?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The reason it was there was so that the user wouldn't have ops
in their top level after from torch.quantized.functional import *
, but I guess we don't need that restriction
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
maybe use torch.ops directly then?
[quantization] fix conv2d gh-metadata: pytorch pytorch 23690 gh/jamesr66a/28/head
@jamesr66a merged this pull request in 3314d60. |
Stack from ghstack:
Differential Revision: D16610734