Python 2+3 and pip #32
Open
Conversation
This file contains hidden or 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
added 15 commits
July 2, 2018 14:36
The entire script can be done with the Python standard library, which is better than relying on Unix programs.
|
This is great, thanks! I would support this PR. My only concern is that this repo uses outdated Therefore, I would also support adding |
Author
|
@salaniz I have updated to use the version of pycocotools on PyPI, thanks for the feedback! |
Author
|
@tylin @hao-fang @vrama91 @endernewton is there any chance of merging this? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
fixes #27
This PR (1) sets up Python 3 compatibility, and (2) a setup.py so that this repo can be pip installed locally.
I cloned tylin/coco-caption. Then I manually merged with salaniz/pycocoevalcap to get a working version with Python 3. Last, I went back and added support for Python 2. I verified that the ipynb returns the same scores on both Pythons. (Do note the images/captions with low CIDErs are different - Python 2 gives a train image still, Python 3 gives a man on a bicycle. I would expect the difference is due to dictionary key ordering behavior changes in Python 3.6.)
I would love to see coco-caption added to PyPI, but I leave that to the original authors.