Add chainer_torch_function and TorchCainerFunction
#27
+229
−0
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
emcastillo
force-pushed
the
chainer-torch-function
branch
from
b7c0d80 to
fd80272
Compare
|
Thanks for the PR. As long as I tested, this function works pretty well. It would be helpful for other users also. |
emcastillo
force-pushed
the
chainer-torch-function
branch
from
ea3635c to
dd4d223
Compare
emcastillo
changed the title
chainer_torch_functionchainer_torch_function and TorchCainerFunction
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This allows to call a PyTorch function and combine it with chainer functions and the opposite.
The backprop of both Chainer and PyTorch is mixed, being the backprop of torch performed inside the chainer graph.
The main difference with
TorchModuleis that the latter only holds one graph in PyTorch and the changes are reflected in the parameters of a chainer link. This PR allows us to construct mixed graphs. The same is done for the opposite user case.This is useful when doing migrations since it allows us to replace functions in the model part by part.