CoolFace
Datasetpublic

JackHsieh/statML-arxiv-42M-11M-L-1024-llama32-pause7-injected

A pause-injected variant of JackHsieh/statML-arxiv-42M-11M-L-1024-llama32: every document token is preceded by a sentinel-wrapped stretch of pause tokens, so each row is exactly 8x longer (1_024 -> 8_192 tokens). Nothing else changes — same papers, same windows, same train/test split, same schema and column order as the parent. How it was derived For each row of the parent, the token sequence [t0, t1, ...] becomes: <|reserved_special_token_1|> <|reserved_special_token_0|> x5… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/statML-arxiv-42M-11M-L-1024-llama32-pause7-injected.

sourceHugging Faceupdated 2mo agoView on Hugging Face
1likes113downloads
4 commits on main
6e7a2e52mo ago

Upload README.md with huggingface_hub

JackHsieh
a20a0282mo ago

Upload README.md with huggingface_hub

JackHsieh
328df252mo ago

Upload dataset

JackHsieh
d390db22mo ago

initial commit

JackHsieh