Datasets:

tuskanny
/

nfcorpus_modern_colbert

PLAIN-2 string	MED-2427 string	2 int64
PLAIN-2	MED-10	2
PLAIN-2	MED-2429	2
PLAIN-2	MED-2430	2
PLAIN-2	MED-2431	2
PLAIN-2	MED-14	2
PLAIN-2	MED-2432	2
PLAIN-2	MED-2428	1
PLAIN-2	MED-2440	1
PLAIN-2	MED-2434	1
PLAIN-2	MED-2435	1
PLAIN-2	MED-2436	1
PLAIN-2	MED-2437	1
PLAIN-2	MED-2438	1
PLAIN-2	MED-2439	1
PLAIN-2	MED-3597	1
PLAIN-2	MED-3598	1
PLAIN-2	MED-3599	1
PLAIN-2	MED-4556	1
PLAIN-2	MED-4559	1
PLAIN-2	MED-4560	1
PLAIN-2	MED-4828	1
PLAIN-2	MED-4829	1
PLAIN-2	MED-4830	1
PLAIN-12	MED-2513	2
PLAIN-12	MED-5237	2
PLAIN-12	MED-2517	2
PLAIN-12	MED-2518	2
PLAIN-12	MED-2519	2
PLAIN-12	MED-2520	2
PLAIN-12	MED-2521	2
PLAIN-12	MED-2514	1
PLAIN-12	MED-2943	1
PLAIN-12	MED-5322	1
PLAIN-12	MED-5323	1
PLAIN-12	MED-5324	1
PLAIN-12	MED-5325	1
PLAIN-12	MED-5326	1
PLAIN-12	MED-5327	1
PLAIN-12	MED-5328	1
PLAIN-12	MED-5329	1
PLAIN-12	MED-5330	1
PLAIN-12	MED-5331	1
PLAIN-12	MED-5332	1
PLAIN-12	MED-5333	1
PLAIN-12	MED-5334	1
PLAIN-12	MED-5335	1
PLAIN-12	MED-5363	1
PLAIN-12	MED-5337	1
PLAIN-12	MED-5338	1
PLAIN-12	MED-5339	1
PLAIN-12	MED-5340	1
PLAIN-12	MED-5341	1
PLAIN-12	MED-5342	1
PLAIN-23	MED-2644	2
PLAIN-23	MED-2646	2
PLAIN-23	MED-2651	2
PLAIN-23	MED-118	2
PLAIN-23	MED-2652	2
PLAIN-23	MED-2653	2
PLAIN-23	MED-2655	2
PLAIN-23	MED-2659	2
PLAIN-23	MED-2661	2
PLAIN-23	MED-2662	2
PLAIN-23	MED-4551	1
PLAIN-23	MED-2643	1
PLAIN-23	MED-2645	1
PLAIN-23	MED-2647	1
PLAIN-23	MED-2648	1
PLAIN-23	MED-2627	1
PLAIN-23	MED-2649	1
PLAIN-23	MED-2650	1
PLAIN-23	MED-2654	1
PLAIN-23	MED-2656	1
PLAIN-23	MED-2657	1
PLAIN-23	MED-2658	1
PLAIN-23	MED-2660	1
PLAIN-23	MED-3021	1
PLAIN-23	MED-3023	1
PLAIN-23	MED-2904	1
PLAIN-23	MED-2905	1
PLAIN-23	MED-2906	1
PLAIN-23	MED-2907	1
PLAIN-23	MED-3024	1
PLAIN-23	MED-3025	1
PLAIN-23	MED-2910	1
PLAIN-23	MED-3027	1
PLAIN-23	MED-3028	1
PLAIN-23	MED-2913	1
PLAIN-23	MED-3012	1
PLAIN-23	MED-3013	1
PLAIN-23	MED-3030	1
PLAIN-23	MED-2917	1
PLAIN-23	MED-3033	1
PLAIN-23	MED-3034	1
PLAIN-23	MED-3035	1
PLAIN-23	MED-2921	1
PLAIN-23	MED-3585	1
PLAIN-23	MED-3586	1
PLAIN-23	MED-3587	1
PLAIN-23	MED-3588	1

End of preview. Expand in Data Studio

NFCorpus, GTE-ModernColBERT

Token-level (late-interaction) embeddings of the BEIR NFCorpus corpus and queries, encoded with GTE-ModernColBERT, in the TACHIOM multivector format.

Source

BEIR NFCorpus, test split. Corpus, queries and qrels were read from the official BEIR files via ir_datasets (beir/nfcorpus/test); PyLate only did the encoding
3,633 documents, 323 queries, 12,334 qrels
Text given to the encoder for each document: title + " " + text (BEIR title and body joined by a space). The text itself is not included, only its vectors
Row order follows the BEIR corpus and query files; row i of doc_ids.npy / queries_ids.npy identifies row i of doclens.npy / queries.npy

Encoding

Model: lightonai/GTE-ModernColBERT-v1 @ 25f6f7bb8237b7ae25ae1d9b805ce17c0d1cc639
Library: PyLate 1.6.0, CPU
Document length cap: 300 tokens (model default)
Query length: 48 tokens (model default)
Query expansion: no (model default)
The model defaults come from config_sentence_transformers.json in the model repository. We did not override any of them
Documents: punctuation tokens are dropped (PyLate skiplist), as are padding tokens
Vectors: 128-d, L2-normalized

Statistics


Token vectors (N)	862,599
Avg vectors per document	237.4 (max 286)
Vectors per query	variable, 4 to 25 (no query expansion), zero-padded to 25
Avg vectors per query	8.6

Files

File	dtype	Shape	Content
`documents.npy`	float16 (`<f2`)	`[862599, 128]`	All document vectors, concatenated document by document
`doclens.npy`	int32	`[3633]`	Vectors per document; `sum == N`
`token_ids.npy`	uint32	`[862599]`	Input token id of each row of `documents.npy`
`doc_ids.npy`	string	`[3633]`	BEIR doc id of each document
`queries.npy`	float32	`[323, 25, 128]`	Query vectors, zero-padded at the end
`query_lens.npy`	int32	`[323]`	True number of vectors per query
`queries_ids.npy`	string	`[323]`	BEIR query id of each query
`qrels.test.tsv`	TREC	12334 lines	`qid \t 0 \t docid \t relevance`
`groundtruth/gt_top100.tsv`	TSV	32300 lines	Exhaustive top-100: `query_idx \t doc_idx \t rank \t score` (0-based positions)
`groundtruth/gt_ids.npy`	int32	`[323, 100]`	Same, as doc positions
`groundtruth/gt_scores.npy`	float32	`[323, 100]`	Same, MaxSim scores

Zero padding does not change any score: a zero query vector adds 0 to the MaxSim of every document.

Exhaustive-search effectiveness

Exact MaxSim over the full collection (vectorium compute_groundtruth_multivec). These are the reference numbers for approximate search on this data.

nDCG@10	R@100
0.3792	0.3222

Downloads last month: 18

Collection including tuskanny/nfcorpus_modern_colbert

multivector_datasets

Collection

11 items • Updated 2 days ago