| --- |
| license: cc-by-nc-4.0 |
| task_categories: |
| - other |
| language: |
| - en |
| tags: |
| - cxl-ssd |
| - dlrm |
| - merci |
| - preprocessing |
| - paper-archive |
| - cxlssd-archetype-routing |
| size_categories: |
| - 10G<n<100G |
| pretty_name: "CXL-SSD Archetype Routing — Preprocessed Datasets" |
| --- |
| |
| # CXL-SSD Archetype Routing — Preprocessed Datasets |
|
|
| Stage-00 snapshot (2026-05-02). Intermediate MERCI / MaxEmbed / DLRM |
| preprocessing artifacts: filtered CSVs, vocab tables, embedding-table |
| indexes, partition outputs. |
|
|
| These are the **outputs of `MERCI_page_aware/analysis/preprocess_*.py`** |
| applied to each raw dataset; they are the input to trace generation |
| (`research_data/traces/`) and to the Cylon/MQSim/MaxEmbed simulators. |
| |
| ## Tiers |
| |
| - `criteo_terabyte/` — 45 GB |
| - `criteo_kaggle/` — 27 GB |
| - `avazu/` — 11 GB |
| - `alibaba_ifashion/`, `netflix/`, `mind/`, `tmall/` — 1-2 GB each |
| - 24 smaller datasets — < 1 GB each |
| |
| ## Companion repos |
| |
| See https://github.com/shadowcollecter/cxlssd-archetype-routing for the |
| analysis scripts that produced these files. |
| |
| ## License |
| |
| CC-BY-NC-4.0; upstream dataset licences apply for any field-level |
| redistribution. |
| |
