Dataset Viewer
Auto-converted to Parquet Duplicate
kind
stringclasses
3 values
candidate_idx
int64
0
19
score
float64
0.42
0.78
embedding
listlengths
46
779
query_text
null
null
[ [ 0.0238037109375, -0.0849609375, -0.076171875, 0.1318359375, -0.07568359375, -0.062255859375, 0.03173828125, 0.1962890625, -0.005889892578125, 0.0223388671875, -0.123046875, 0.060302734375, -0.00543212890625, 0.025390625, 0.041748046875, -0.1044921875...
query_image
null
null
[[0.0203857421875,-0.08544921875,-0.07177734375,0.1220703125,-0.07421875,-0.041015625,0.026123046875(...TRUNCATED)
candidate
0
0.466475
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
1
0.459239
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
2
0.456392
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
3
0.447216
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
4
0.441792
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
5
0.439896
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
6
0.437075
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
candidate
7
0.433155
[[0.0206298828125,-0.08642578125,-0.0712890625,0.1220703125,-0.076171875,-0.041259765625,0.025268554(...TRUNCATED)
End of preview. Expand in Data Studio

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Pathology Retrieval Benchmark — Rerank Cache

Privacy-safe embedding cache for rerank-only reproduction (no document IDs or file paths).

Files

One Parquet file per query: entry_{id}.parquet (100 entries).

Schema

Column Description
kind query_text, query_image, or candidate
candidate_idx 0–19 for candidates; null for queries
score Text-retrieval score (candidates only)
embedding ColQwen multi-vector embedding (N, 128)

Each file contains 2 query rows + 20 candidate rows.

Usage

import pandas as pd
df = pd.read_parquet("entry_1.parquet")
text = df[df.kind == "query_text"].iloc[0].embedding
candidates = df[df.kind == "candidate"].sort_values("candidate_idx")

Generated from local retrieval cache via scripts/export_cache_parquet.py.

Downloads last month
24