KShivendu/nq-500k-iso-moderncolbert-token-vectors
NQ 500k Iso-ModernColBERT token vectors Per-token ColBERT vectors from topk-io/Iso-ModernColBERT (revision b95d9608ad424fdd7bfd045001576f59cbb89a98, bf16) for a 500,000-passage sample of the BEIR Natural Questions corpus and 500 of its test queries. 47.5M kept document tokens, 128 dims each. They were made to test how first stages for ColBERT (BM25, token-type centroids, SMVE) hold up as the corpus grows. The corpus is built so that its first 5,000, 50,000 and 500,000 rows are… See the full description on the dataset page: https://huggingface.co/datasets/KShivendu/nq-500k-iso-moderncolbert-tok
| Author / organization | KShivendu |
|---|---|
| License (as declared) | cc-by-sa-3.0 |
| Downloads | 156 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 7 October 2026 |
| Last updated at source | 8 October 2026 |
More from KShivendu
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 21 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.