Upgrade get_dataset.tokenize() to multiprocessing #24
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
get_dataset.tokenize() is to slow on a single CPU. Therefore it is upgraded to multiprocessing by implementing the multiprocessing target function worker_tokenize(args_list). Additionally a multiprocessing debug logger mp_logger was added together with logger.debug() and mp_logger.debug() message to track progress in the python console.
thomwolf
reviewed
Aug 20, 2019
thomwolf
reviewed
Aug 20, 2019
|
Looks nice, thanks! |
DrStoop
commented
Aug 20, 2019
| dataset = tokenize(dataset) | ||
| # dataset = tokenize(dataset) |
There was a problem hiding this comment.
absolutely!
Suggested change
| # dataset = tokenize(dataset) |
| personachat = tokenize(personachat) | ||
| torch.save(personachat, dataset_cache) | ||
| # torch.save(personachat, dataset_cache) |
There was a problem hiding this comment.
of course!
Suggested change
| # torch.save(personachat, dataset_cache) | |
| torch.save(personachat, dataset_cache) |
|
The question would be, if |
|
@thomwolf , please could we get this merged? Thank you. |
|
@thomwolf, before merging: i did some work on parallelizing the complete preprocessing chain affecting quite some code in ‚train.py‘ and ‚utils.py‘. i could clean the code & create a new pull request with e.g. 2 new files ‚utils_multiprocessing.py‘ and ‚train_multiprocessing.py‘. This way merging would become very easy & backward compatibility for everybody is guaranteed. Just let me know if you have interest in merging such a speedup ⏩ 💨 |
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Outdated suggestions cannot be applied.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.
get_dataset.tokenize() on a single CPU is very slow. Therefore in this pull request it is upgraded to multiprocessing by implementing the multiprocessing target function worker_tokenize(args_list). Additionally a multiprocessing debug logger mp_logger was added together with
logger.debug() and mp_logger.debug() message to track progress in the python console.