metadata
license: apache-2.0
Subset of cerebras/SlimPajama-627B,
consisting of 10% of the train split and 100% of the test and
validation splits.
The train split consists of chunk2 from the original
[cerebras/SlimPajama-627B] dataset, split into five zstd-compressed jsonl
files for efficient loading. The dataset is 70 GB compressed, 249 GB
uncompressed.
@misc{cerebras2023slimpajama,
author = {Soboleva, Daria and Al-Khateeb, Faisal and Myers, Robert and Steeves, Jacob R and Hestness, Joel and Dey, Nolan},
title = {{SlimPajama: A 627B token cleaned and deduplicated version of RedPajama}},
month = June,
year = 2023,
howpublished = {\url{https://www.cerebras.net/blog/slimpajama-a-627b-token-cleaned-and-deduplicated-version-of-redpajama}},
url = {https://huggingface.co/datasets/cerebras/SlimPajama-627B},
}