Datasets:
Tasks:
Text2Text Generation
Languages:
English
Multilinguality:
monolingual
Size Categories:
100K<n<1M
Language Creators:
found
Annotations Creators:
no-annotation
Source Datasets:
original
ArXiv:
License:
UnexpectedError: zstd decompress Error: Frame requires too much memory for decoding
#2
by
LLukas22
- opened
I'm getting a 'Frame requires too much memory for decoding' when i try to download any of the dataset splits.
This seams to be a problem with the max_window_size of the zstdDecompressor as mentioned in this thread.
I copied the dataset script to a private repository and changed the max_window_size, which seams to solve the error.
Could we maybe change change the max_window_size in this "official" eli5 repository to?
Thanks for reporting, @LLukas22 .
I am having a look to see how we can fix it.
First we need to fix this: #3
albertvillanova
changed discussion title from
zstdDecompressor Error 'Frame requires too much memory for decoding' when downloading dataset
to UnexpectedError: zstd decompress Error: Frame requires too much memory for decoding