Preprocess dataset size fix #1131
Merged
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
|
also, when running training, make sure |
NanoCode012
reviewed
Jan 17, 2024
Do you mean when loading from cache or we need it all the time now? |
the opposite, pre-process always skips the cache and overwrites it |
it's more of a reminder to make sure you use that when you've pre-processed the data. |
NanoCode012
approved these changes
Jan 17, 2024
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Outdated suggestions cannot be applied.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.
Description
This PR fixes the obscene disk usage when preprocessing datasets. this is mostly during the tokenization map step. By keeping this step completely in memory, it solves ALL the additional disk requirements (which is as much as 10x). Additionally, when running the
-m axolotl.cli.preprocesssometimes it would rely on the cache, but since pre-processing is an explicit step, the expectation is that it should ignore the cache in most cases.