Introduce padding op to mimic pytorch semantics in ONNX export (#2069)

In pytorch, after pad_packed_sequence, the "extra" elements (after the
ends of the sequences) are reset. In the equivalent Caffe2 graph
exported via ONNX, they contained some leftover values, which caused
tests to fail. Probably no one depends on these values, but just in
case, set them to zero to mimic pytorch semantics.
3 files changed
tree: 5c30d75fcc5560542a029ff30f22d138e848d0db
  1. .github/
  2. .jenkins/
  3. .travis/
  4. caffe/
  5. caffe2/
  6. cmake/
  7. conda/
  8. docker/
  9. docs/
  10. modules/
  11. scripts/
  12. third_party/
  13. .gitattributes
  14. .gitignore
  15. .gitmodules
  16. .travis.yml
  17. appveyor.yml
  18. CMakeLists.txt
  19. LICENSE
  20. Makefile
  21. NOTICE
  22. README.md
  23. release-notes.md
  24. setup.py
  25. VERSION_NUMBER
README.md

Caffe2

License Jenkins Build Status Appveyor Build Status

Caffe2 is a lightweight, modular, and scalable deep learning framework. Building on the original Caffe, Caffe2 is designed with expression, speed, and modularity in mind.

Questions and Feedback

Please use Github issues (https://github.com/caffe2/caffe2/issues) to ask questions, report bugs, and request new features.

Please participate in our survey (https://www.surveymonkey.com/r/caffe2). We will send you information about new releases and special developer events/webinars.

License

Caffe2 is released under the Apache 2.0 license. See the NOTICE file for details.

Further Resources on Caffe2.ai