multivector_datasets
Collection
11 items • Updated
Datasets:
Search is not available for this dataset
300674 int64 2 1.1M | 0 int64 0 0 | 7067032 int64 3.86k 8.01M | 1 int64 1 1 |
|---|---|---|---|
125,705 | 0 | 7,067,056 | 1 |
94,798 | 0 | 7,067,181 | 1 |
9,083 | 0 | 7,067,274 | 1 |
174,249 | 0 | 7,067,348 | 1 |
320,792 | 0 | 7,067,677 | 1 |
1,090,270 | 0 | 7,067,796 | 1 |
1,101,279 | 0 | 7,067,891 | 1 |
201,376 | 0 | 7,068,066 | 1 |
54,544 | 0 | 7,068,203 | 1 |
118,457 | 0 | 7,068,493 | 1 |
178,627 | 0 | 7,068,519 | 1 |
178,627 | 0 | 7,068,520 | 1 |
1,101,278 | 0 | 7,068,907 | 1 |
68,095 | 0 | 7,069,266 | 1 |
87,892 | 0 | 7,069,601 | 1 |
257,309 | 0 | 4,959,637 | 1 |
1,090,242 | 0 | 7,070,556 | 1 |
211,691 | 0 | 7,070,643 | 1 |
165,002 | 0 | 7,070,877 | 1 |
1,101,276 | 0 | 7,070,950 | 1 |
264,827 | 0 | 7,071,066 | 1 |
342,285 | 0 | 7,071,436 | 1 |
372,586 | 0 | 7,071,494 | 1 |
89,786 | 0 | 7,071,501 | 1 |
118,448 | 0 | 7,071,642 | 1 |
92,542 | 0 | 7,072,003 | 1 |
206,117 | 0 | 7,072,155 | 1 |
206,117 | 0 | 7,072,156 | 1 |
206,117 | 0 | 7,072,160 | 1 |
141,472 | 0 | 7,072,290 | 1 |
293,992 | 0 | 2,790,193 | 1 |
196,232 | 0 | 7,072,326 | 1 |
352,818 | 0 | 7,072,358 | 1 |
45,924 | 0 | 5,167,800 | 1 |
45,924 | 0 | 7,072,691 | 1 |
208,145 | 0 | 7,072,838 | 1 |
208,145 | 0 | 7,072,843 | 1 |
79,891 | 0 | 7,073,211 | 1 |
208,494 | 0 | 7,073,272 | 1 |
319,564 | 0 | 7,073,381 | 1 |
155,234 | 0 | 502,713 | 1 |
14,151 | 0 | 7,073,772 | 1 |
67,802 | 0 | 7,074,071 | 1 |
1,090,184 | 0 | 7,074,235 | 1 |
323,382 | 0 | 4,778,293 | 1 |
323,998 | 0 | 7,074,377 | 1 |
91,711 | 0 | 1,956,185 | 1 |
125,898 | 0 | 7,074,710 | 1 |
289,812 | 0 | 262,205 | 1 |
333,486 | 0 | 7,075,218 | 1 |
1,090,171 | 0 | 7,075,317 | 1 |
73,257 | 0 | 7,075,398 | 1 |
1,090,170 | 0 | 7,075,411 | 1 |
237,373 | 0 | 7,075,449 | 1 |
127,876 | 0 | 7,075,540 | 1 |
85,095 | 0 | 430,598 | 1 |
1,090,165 | 0 | 7,075,801 | 1 |
259,417 | 0 | 7,075,870 | 1 |
1,101,271 | 0 | 7,075,989 | 1 |
281,930 | 0 | 7,076,058 | 1 |
205,107 | 0 | 3,489,272 | 1 |
205,107 | 0 | 7,076,253 | 1 |
307,118 | 0 | 7,076,536 | 1 |
87,019 | 0 | 7,076,754 | 1 |
335,710 | 0 | 7,076,765 | 1 |
127,984 | 0 | 7,076,927 | 1 |
1,090,151 | 0 | 3,624,309 | 1 |
46,711 | 0 | 7,077,382 | 1 |
1,090,146 | 0 | 7,077,624 | 1 |
1,090,132 | 0 | 7,078,051 | 1 |
1,090,115 | 0 | 7,078,204 | 1 |
1,090,110 | 0 | 7,078,250 | 1 |
1,090,107 | 0 | 7,078,279 | 1 |
1,090,102 | 0 | 7,078,336 | 1 |
1,090,100 | 0 | 7,078,350 | 1 |
1,090,086 | 0 | 7,078,525 | 1 |
1,090,077 | 0 | 7,078,615 | 1 |
1,090,072 | 0 | 7,078,672 | 1 |
1,090,063 | 0 | 7,078,754 | 1 |
1,090,054 | 0 | 7,078,842 | 1 |
1,090,043 | 0 | 7,078,956 | 1 |
1,101,259 | 0 | 7,078,991 | 1 |
1,090,029 | 0 | 7,079,065 | 1 |
1,090,029 | 0 | 7,079,067 | 1 |
1,089,983 | 0 | 7,079,501 | 1 |
1,089,966 | 0 | 7,079,658 | 1 |
1,089,964 | 0 | 7,079,676 | 1 |
1,089,945 | 0 | 7,079,883 | 1 |
1,089,940 | 0 | 7,079,948 | 1 |
1,089,925 | 0 | 7,080,091 | 1 |
1,089,906 | 0 | 7,080,202 | 1 |
1,089,896 | 0 | 7,080,300 | 1 |
1,101,236 | 0 | 7,080,466 | 1 |
1,089,868 | 0 | 7,080,589 | 1 |
1,089,846 | 0 | 7,080,803 | 1 |
1,089,832 | 0 | 7,080,937 | 1 |
1,089,810 | 0 | 7,081,127 | 1 |
1,101,228 | 0 | 7,081,141 | 1 |
1,089,805 | 0 | 7,081,193 | 1 |
1,089,804 | 0 | 7,081,198 | 1 |
End of preview. Expand in Data Studio
MS MARCO v1 Passage, ColBERTv2
Token-level (late-interaction) ColBERTv2 embeddings of the MS MARCO v1 passage collection and the dev/small queries.
Source
- Collection: MS MARCO v1 passage (
ir_datasetsmsmarco-passage), 8,841,823 passages - Queries: dev/small, 6,980 queries and 7,437 qrels (
msmarco-passage/dev/small) - Document order: passage id order (row i is pid i)
Encoding
- Model: ColBERTv2 (
colbert-ir/colbertv2.0, BERT-base-uncased tokenizer) - Documents start with
[CLS]and the ColBERT[D]marker (token id 2) - Queries: always 32 vectors (
[MASK]expansion, no zero padding) - Vectors: 128-d, L2-normalized
- Not recorded: exact checkpoint revision, encoding library, document length cap (the longest document has 300 vectors), and whether punctuation was dropped
Statistics
| Token vectors (N) | 597,909,919 |
| Avg vectors per document | 67.6 (min 4, max 300) |
| Vectors per query | 32 |
Files
| File | dtype | Shape | Content |
|---|---|---|---|
documents.npy |
uint16 (<u2) |
[597909919, 128] |
Raw float16 bit patterns stored as uint16. Read with .view(np.float16) |
doclens.npy |
int32 | [8841823] |
Vectors per document; sum == N |
token_ids_per_token.npy |
int64 | [597909919] |
Input token id of each row of documents.npy |
doc_ids.npy |
int64 | [8841823] |
MS MARCO pid (equals the row index) |
queries.npy |
float32 | [6980, 32, 128] |
Query vectors |
queries_ids.npy |
int64 | [6980] |
MS MARCO qid of each query |
Document ids are MS MARCO passage ids and equal the row index: row i of doclens.npy is passage i.
documents.npy stores float16 values as their raw 16-bit patterns, with dtype uint16.
In numpy, read it with np.load("documents.npy", mmap_mode="r").view(np.float16).
Relevance judgments
The official MS MARCO passage dev/small qrels (qrels.dev.small.tsv): 7,437 relevant
query-passage pairs over the 6,980 queries, all with relevance 1 (ir_datasets msmarco-passage/dev/small).
Standard metric: MRR@10 (RR@10 in ir_measures).
- Downloads last month
- 117