cxlssd-raw-medium / README.md
shadowcollecter's picture
Upload README.md with huggingface_hub
a87b3b5 verified
metadata
license: other
task_categories:
  - other
language:
  - en
tags:
  - recommendation-system
  - dlrm
  - merci
  - cxl-ssd
  - paper-archive
size_categories:
  - 10G<n<100G
pretty_name: CXL-SSD Archetype Routing  Raw Datasets (medium tier)

CXL-SSD Archetype Routing — Raw Datasets (medium tier)

Stage-00 snapshot (2026-05-02). Mirrors the research_data/raw/ tree excluding criteo_terabyte/ (the 381 GB Terabyte dataset is hosted in a separate repo: cxlssd-archetype-raw-criteo-tb).

Contains 27 source datasets used by the CXL-SSD page-oriented embedding lookup paper:

Dataset Size Source
amazon 46 GB UCSD Amazon Reviews
criteo_kaggle 33 GB Kaggle Display Ad Challenge
alibaba_ifashion 31 GB Alibaba iFashion
yelp 8.7 GB Yelp Open
avazu 7.1 GB Kaggle Avazu CTR
taobao 4.4 GB Alibaba Taobao
lastfm_1k 2.4 GB last.fm
jd 2.2 GB jd.com
steam 2.1 GB Steam Reviews
netflix 2.0 GB Netflix Prize
yoochoose 1.8 GB RecSys 2015
tmall 1.7 GB tmall.com
lastfm_360k 1.6 GB last.fm
movielens_25m 1.1 GB GroupLens
retailrocket 942 MB Kaggle
movielens_32m 912 MB GroupLens
amazon_m2 832 MB Amazon M2
mind 610 MB MS News
gowalla 398 MB Stanford SNAP
movielens_10m 257 MB GroupLens
million_song 197 MB million-song-dataset
nyc_taxi 145 MB NYC TLC
anime 108 MB MyAnimeList
bookcrossing 103 MB Book-Crossing
foursquare 98 MB Foursquare
movielens_1m 24 MB GroupLens
(+ DATASETS.md) metadata

License

license: other — each subdirectory inherits its upstream dataset's licence. Do not redistribute fields commercially. Use academic reproducibility only.

See DATASETS.md (root) for upstream URLs and citation entries.