The input training data files (multiple files in glob format). #7717
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Outdated suggestions cannot be applied.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.
Very often Corpus comes in split files (book-large-p1.txt, book-large-p2.txt); Also splitting large files to smaller files can prevent language_modeling's tokenizer going out of memory in environment like Colab that does not have swap memory and limited to Standard (12Gb) or High RAM (25Gb) instances.
To avoid making any assumption and prematurely truncate the file to avoid such error, we add support to concatenate training data on Dataset level. User can split files to multiple 512Mb, in that case language_modeling's tokenizer would be less to go out of memory.
In addition, it would even further enhance the memory limitation if we keep pytorch tensor in memory instead of python list. We leave this in future work.
@LysandreJik @sgugger