Hugging Face's logo

Datasets:
gist-sparse-attention
/
GSA-PT-Llama-3.2-1B-chunk4-chunk4-data

Tags:
gsa
gist-sparse-attention
long-context
pretraining
License:
Dataset card Files
xet
 Community
Dataset Viewer
The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.
Couldn't infer the same data file format for all splits. Got {NamedSplit('train'): ('arrow', {}), NamedSplit('test'): ('json', {})}
Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

GSA-PT-Llama-3.2-1B-chunk4-chunk4-data

This is the continue pretraining dataset used for training GSA (Gist Sparse Attention) models with chunk size chunk4-chunk4.

Each sample is tokenized and formatted with GSA gist tokens for continued pretraining.

Paper

GSA: Gist Sparse Attention via Learnable Compression and Selective Unfolding

Related Models

Downloads last month
59
Source
FineWeb (sample-10BT)
Filtering
tokens
Chunk size
chunk4-chunk4
Splits
eval
Total file size:
29.7 GB

Collection including gist-sparse-attention/GSA-PT-Llama-3.2-1B-chunk4-chunk4-data