Skip to content

Changes in RNNBase and Embedding for compatibility with DataParallel #660

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.

Sign up for GitHub

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

Jump to bottom
Merged
soumith merged 6 commits into from
Feb 9, 2017
Merged

Changes in RNNBase and Embedding for compatibility with DataParallel #660

soumith merged 6 commits into from
Feb 9, 2017

Conversation

@bmccann
Copy link
Contributor

@bmccann bmccann commented Jan 31, 2017 •
edited
Loading

@bmccann bmccann mentioned this pull request Jan 31, 2017
Closed
@bmccann bmccann changed the title Replacing stale references in RNNBase Changes in RNNBase and Embedding for compatibility with DataParallel Feb 1, 2017
@apaszke
Copy link
Contributor

apaszke commented Feb 3, 2017

The PR looks good, but it will break all checkpoints. I think you'd also need to add __setstate__ that will modify the all_weights dict to match the new types.

@apaszke apaszke mentioned this pull request Feb 6, 2017
Closed
@apaszke
Copy link
Contributor

apaszke commented Feb 7, 2017

@pytorchbot test this please

@soumith
Copy link
Contributor

soumith commented Feb 7, 2017

ok, last checks the tests are failing. Here's the log:

======================================================================
ERROR: test_RNN_cpu_vs_cudnn_no_dropout (__main__.TestNN)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "test_nn.py", line 30, in wrapper
    return fn(*args, **kwargs)
  File "test_nn.py", line 1165, in test_RNN_cpu_vs_cudnn_no_dropout
    self._test_RNN_cpu_vs_cudnn(0)
  File "test_nn.py", line 1135, in _test_RNN_cpu_vs_cudnn
    False, rnn, input_val, hx_val, rnn.all_weights)
  File "test_nn.py", line 1062, in forward_backward
    x.data.copy_(y.data)
AttributeError: 'str' object has no attribute 'data'

======================================================================
ERROR: test_RNN_cpu_vs_cudnn_with_dropout (__main__.TestNN)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "test_nn.py", line 30, in wrapper
    return fn(*args, **kwargs)
  File "test_nn.py", line 1171, in test_RNN_cpu_vs_cu* Closing connection 0
dnn_with_dropout
    self._test_RNN_cpu_vs_cudnn(1)
  File "test_nn.py", line 1135, in _test_RNN_cpu_vs_cudnn
    False, rnn, input_val, hx_val, rnn.all_weights)
  File "test_nn.py", line 1062, in forward_backward
    x.data.copy_(y.data)
AttributeError: 'str' object has no attribute 'data'

@ngimel ngimel mentioned this pull request Feb 9, 2017
Closed
@soumith soumith merged commit 6c77fa9 into pytorch:master Feb 9, 2017
@soumith
Copy link
Contributor

soumith commented Feb 9, 2017

thanks for fixing this Bryan!

@bmccann bmccann mentioned this pull request Feb 9, 2017
Closed
ashishfarmer added a commit to ashishfarmer/pytorch that referenced this pull request May 22, 2020
@ashishfarmer
Merge pull request pytorch#660 from jithunnair-amd/ifu_20200521_2
89c64f2 
Ifu 20200521 2
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

1 more reviewer

@apaszke apaszke apaszke approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

open source

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@bmccann @apaszke @soumith @ezyang