Added support for multidimensional tensors in PReLU#932
Added support for multidimensional tensors in PReLU#932goelhardik wants to merge 2 commits intopytorch:masterfrom goelhardik:issue-724
Conversation
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
|
Made some comments inline. Logically, PR looks good. |
|
Also, add a test case for PReLU 3d in test_nn.py |
|
I have added a basic PReLU test that tests the forward and backward pass. If you had any other ideas, please let me know. |
|
@pytorchbot add to whitelist |
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
|
Should be ready, I suppose |
This comment was marked as off-topic.
This comment was marked as off-topic.
Sorry, something went wrong.
|
I've rebased this branch to only have tests. The other changes are coming through subtree updates from respective upstreams in THNN / THCUNN |
|
merged into master via 6b7aef6 Thanks a lot Hardik! |
Fixes #724
Also, for tensors with more than 1 dimension, the number of PReLU channels is now to be input in the second tensor dimension.