Dataset Viewer
Auto-converted to Parquet Duplicate
id
int64
0
31.8k
image
imagewidth (px)
164
500
image_id
stringlengths
1
5
filename
stringlengths
9
14
captions
listlengths
5
5
caption
stringlengths
32
403
image_emb
list
text_emb
list
0
0
1000092795.jpg
[ "Two young guys with shaggy hair look at their hands while hanging out in the yard .", "Two young White males are outside near many bushes .", "Two men in green shirts are standing in a yard .", "A man in a blue shirt standing in a garden .", "Two friends enjoy time spent together ." ]
Two young guys with shaggy hair look at their hands while hanging out in the yard .
[ 0.0282440185546875, 0.10577392578125, -0.0289459228515625, 0.01776123046875, -0.039276123046875, -0.004016876220703125, 0.027099609375, -0.03131103515625, 0.014556884765625, 0.100830078125, 0.0054931640625, -0.057098388671875, -0.01183319091796875, 0.0255279541015625, -0.0161590576171875...
[ 0.05828857421875, -0.01180267333984375, 0.00717926025390625, 0.0809326171875, -0.056610107421875, 0.031890869140625, 0.031951904296875, 0.00904083251953125, 0.06951904296875, 0.06414794921875, 0.0772705078125, -0.01549530029296875, 0.01482391357421875, -0.037017822265625, 0.0050048828125...
1
1
10002456.jpg
[ "Several men in hard hats are operating a giant pulley system .", "Workers look down from up above on a piece of equipment .", "Two men working on a machine wearing hard hats .", "Four men on top of a tall structure .", "Three men on a large rig ." ]
Several men in hard hats are operating a giant pulley system .
[ -0.0159454345703125, -0.06890869140625, -0.1019287109375, -0.005550384521484375, -0.040740966796875, -0.00974273681640625, -0.0231475830078125, 0.00026798248291015625, 0.00897216796875, 0.02783203125, -0.0256195068359375, -0.002010345458984375, 0.00925445556640625, 0.0233154296875, 0.029...
[ -0.02410888671875, -0.020355224609375, -0.08514404296875, 0.01220703125, -0.006160736083984375, 0.0254974365234375, -0.046630859375, 0.025726318359375, 0.0226898193359375, 0.00437164306640625, -0.0087127685546875, 0.05767822265625, 0.0193023681640625, 0.00978851318359375, 0.0263366699218...
2
2
1000268201.jpg
[ "A child in a pink dress is climbing up a set of stairs in an entry way .", "A little girl in a pink dress going into a wooden cabin .", "A little girl climbing the stairs to her playhouse .", "A little girl climbing into a wooden playhouse ", "A girl going into a wooden building ." ]
A child in a pink dress is climbing up a set of stairs in an entry way .
[ 0.0148468017578125, 0.252197265625, -0.0090179443359375, -0.0014019012451171875, -0.0300445556640625, 0.014892578125, 0.036895751953125, 0.004268646240234375, -0.014068603515625, 0.100341796875, -0.01296234130859375, -0.06256103515625, 0.003631591796875, 0.032501220703125, -0.01495361328...
[ -0.0162811279296875, 0.01187896728515625, 0.05780029296875, -0.04705810546875, -0.05609130859375, -0.026458740234375, -0.032958984375, 0.0309600830078125, 0.03302001953125, 0.0537109375, -0.035797119140625, -0.02655029296875, 0.032440185546875, 0.038177490234375, -0.0023136138916015625, ...
3
3
1000344755.jpg
[ "Someone in a blue shirt and hat is standing on stair and leaning against a window .", "A man in a blue shirt is standing on a ladder cleaning a window .", "A man on a ladder cleans the window of a tall building .", "man in blue shirt and jeans on ladder cleaning windows", "a man on a ladder cleans a window...
Someone in a blue shirt and hat is standing on stair and leaning against a window .
[ -0.0033054351806640625, 0.11492919921875, -0.083740234375, -0.01617431640625, 0.0013685226440429688, 0.038055419921875, 0.006072998046875, 0.029510498046875, 0.00421905517578125, 0.08538818359375, -0.028106689453125, -0.00930023193359375, -0.0230255126953125, -0.01041412353515625, -0.016...
[ -0.01251220703125, -0.029754638671875, -0.003772735595703125, -0.0809326171875, -0.01244354248046875, -0.0318603515625, -0.026458740234375, 0.03369140625, 0.07354736328125, 0.0650634765625, -0.007526397705078125, 0.016021728515625, 0.00327301025390625, 0.0655517578125, -0.061065673828125...
4
4
1000366164.jpg
[ "Two men one in a gray shirt one in a black shirt standing near a stove .", "Two guy cooking and joking around with the camera .", "Two men in a kitchen cooking food on a stove .", "Two men are at the stove preparing food .", "Two men are cooking a meal ." ]
Two men one in a gray shirt one in a black shirt standing near a stove .
[ -0.0614013671875, 0.055938720703125, 0.0540771484375, 0.0672607421875, -0.0285186767578125, 0.01035308837890625, -0.009033203125, -0.00687408447265625, -0.022613525390625, 0.00843048095703125, -0.00785064697265625, -0.0121002197265625, -0.0094146728515625, -0.039154052734375, -0.01518249...
[ -0.00257110595703125, -0.0738525390625, 0.08563232421875, 0.082763671875, -0.048126220703125, -0.01155853271484375, -0.061431884765625, -0.058624267578125, 0.0206756591796875, 0.01088714599609375, 0.0281219482421875, 0.0021114349365234375, -0.0208282470703125, 0.0002161264419555664, -0.0...
5
5
1000523639.jpg
[ "Two people in the photo are playing the guitar and the other is poking at him .", "A man in green holds a guitar while the other man observes his shirt .", "A man is fixing the guitar players costume .", "a guy stitching up another man 's coat .", "the two boys playing guitar" ]
Two people in the photo are playing the guitar and the other is poking at him .
[ 0.05902099609375, 0.0143890380859375, -0.0517578125, 0.0158843994140625, -0.00046515464782714844, -0.01354217529296875, -0.0011262893676757812, 0.005306243896484375, 0.01296234130859375, -0.02386474609375, 0.00817108154296875, -0.0007147789001464844, 0.0294189453125, -0.003826141357421875,...
[ 0.026611328125, 0.0012369155883789062, 0.0257568359375, 0.08392333984375, -0.01195526123046875, -0.043487548828125, -0.018096923828125, -0.01317596435546875, 0.0096893310546875, -0.01690673828125, -0.003925323486328125, -0.06976318359375, 0.01053619384765625, -0.008148193359375, 0.004730...
6
6
1000919630.jpg
[ "A man sits in a chair while holding a large stuffed animal of a lion .", "A man is sitting on a chair holding a large stuffed animal .", "A man completes the finishing touches on a stuffed lion .", "A man holds a large stuffed lion toy .", "A man is smiling at a stuffed lion" ]
A man sits in a chair while holding a large stuffed animal of a lion .
[ 0.0265960693359375, -0.043243408203125, -0.1435546875, 0.00717926025390625, -0.0184478759765625, 0.01056671142578125, 0.049713134765625, -0.0112762451171875, 0.0177764892578125, 0.005290985107421875, -0.0390625, 0.027587890625, -0.0006451606750488281, -0.0149078369140625, -0.038238525390...
[ 0.037353515625, -0.03216552734375, -0.01336669921875, -0.028594970703125, 0.0166778564453125, 0.04180908203125, -0.0672607421875, -0.0750732421875, -0.001415252685546875, -0.01386260986328125, -0.016510009765625, 0.042572021484375, -0.0241241455078125, -0.02972412109375, -0.029541015625,...
7
7
10010052.jpg
[ "A girl is on rollerskates talking on her cellphone standing in a parking lot .", "A trendy girl talking on her cellphone while gliding slowly down the street .", "A young adult wearing rollerblades holding a cellular phone to her ear .", "there is a young girl on her cellphone while skating .", "Woman tal...
A girl is on rollerskates talking on her cellphone standing in a parking lot .
[ -0.0029964447021484375, -0.0210113525390625, -0.043548583984375, 0.0143890380859375, 0.0061187744140625, 0.00846099853515625, -0.0164337158203125, 0.00893402099609375, -0.0161590576171875, -0.01073455810546875, 0.019134521484375, -0.0238800048828125, 0.01187896728515625, 0.043914794921875,...
[ -0.0816650390625, -0.033233642578125, 0.0286712646484375, -0.03521728515625, 0.04248046875, -0.020172119140625, 0.01053619384765625, 0.0236663818359375, -0.053741455078125, 0.031463623046875, -0.07025146484375, 0.006404876708984375, 0.035186767578125, 0.05731201171875, 0.0108261108398437...
8
8
1001465944.jpg
[ "An asian man wearing a black suit stands near a dark-haired woman and a brown-haired woman .", "Three people are standing outside near large pipes and a metal railing .", "A young woman walks past two young people dressed in hip black outfits .", "A woman with a large purse is walking by a gate .", "Severa...
An asian man wearing a black suit stands near a dark-haired woman and a brown-haired woman .
[ 0.03094482421875, -0.0030231475830078125, -0.12261962890625, 0.1253662109375, -0.0103607177734375, -0.0426025390625, -0.0491943359375, -0.032318115234375, -0.001384735107421875, 0.0357666015625, 0.0229034423828125, -0.00308990478515625, 0.0173187255859375, 0.0159912109375, -0.04663085937...
[ -0.005039215087890625, -0.0594482421875, 0.045654296875, 0.00817108154296875, -0.00815582275390625, -0.024322509765625, -0.022003173828125, 0.009765625, 0.0203094482421875, -0.01202392578125, -0.0283050537109375, 0.035888671875, 0.01404571533203125, -0.026641845703125, -0.04193115234375,...
9
9
1001545525.jpg
[ "Two men in Germany jumping over a rail at the same time without shirts .", "Two youths are jumping over a roadside railing at night .", "Boys dancing on poles in the middle of the night .", "Two men with no shirts jumping over a rail .", "two guys jumping over a gate together" ]
Two men in Germany jumping over a rail at the same time without shirts .
[ 0.00862884521484375, 0.06280517578125, -0.183349609375, 0.06402587890625, -0.005939483642578125, 0.007659912109375, 0.047607421875, 0.01027679443359375, -0.03387451171875, -0.0146331787109375, -0.0014801025390625, -0.0021820068359375, -0.03448486328125, -0.03704833984375, -0.010246276855...
[ -0.019287109375, -0.03961181640625, -0.058624267578125, 0.08447265625, -0.03302001953125, 0.055511474609375, -0.040985107421875, 0.041656494140625, -0.0180511474609375, 0.00519561767578125, -0.0186767578125, -0.06195068359375, 0.0458984375, -0.0986328125, -0.01611328125, -0.01242065429...
End of preview. Expand in Data Studio

Flickr30k (Lance Format)

Lance-formatted version of Flickr30k (re-distributed via lmms-lab/flickr30k) — 31,783 images, each paired with 5 human-written captions, with CLIP image and text embeddings stored inline and pre-built ANN indices on both.

Key features

  • Inline images — full JPEG bytes per row.
  • Pre-computed CLIP embeddings for both image and caption text — IVF_PQ indices on both columns let you do cross-modal retrieval (image→caption or caption→image) without any model at query time.
  • Full-text inverted index on the canonical caption.
  • Self-contained: no sidecar files or external image downloads.

Schema

Column Type Notes
id int64 Row index
image large_binary Inline JPEG bytes
image_id string Original Flickr image id
filename string Original filename (e.g. 1000092795.jpg)
captions list<string> All 5 captions for the image
caption string First caption — used as canonical text for FTS / quick browsing
image_emb fixed_size_list<float32, 512> CLIP image embedding (cosine-normalized)
text_emb fixed_size_list<float32, 512> CLIP text embedding of the canonical caption

Pre-built indices

  • IVF_PQ on image_embmetric=cosine
  • IVF_PQ on text_embmetric=cosine (cross-modal retrieval works out of the box)
  • INVERTED on caption
  • BTREE on image_id

Splits

A single train.lance table containing all 31,783 rows (the lmms-lab/flickr30k redistribution exposes them as a single split). The original train/val/test labels are not preserved in the source parquet.

Load with Lance

import lance

ds = lance.dataset("hf://datasets/lance-format/flickr30k-lance/data/train.lance")
print(ds.count_rows(), ds.schema.names, ds.list_indices())

Cross-modal text→image search

import lance
import pyarrow as pa
import open_clip
import torch

# 1. Encode the query text once with the same CLIP model used at conversion.
model, _, _ = open_clip.create_model_and_transforms("ViT-B-32", pretrained="laion2b_s34b_b79k")
tokenizer = open_clip.get_tokenizer("ViT-B-32")
model = model.eval().cuda().half()
with torch.no_grad():
    q = model.encode_text(tokenizer(["a man surfing at sunset"]).cuda())
    q = (q / q.norm(dim=-1, keepdim=True)).float().cpu().numpy()[0]

ds = lance.dataset("hf://datasets/lance-format/flickr30k-lance/data/train.lance")
emb_field = ds.schema.field("image_emb")
query = pa.array([q.tolist()], type=emb_field.type)

# 2. Nearest-neighbour search against the image embedding index.
hits = ds.scanner(
    nearest={"column": "image_emb", "q": query[0], "k": 10, "nprobes": 16, "refine_factor": 30},
    columns=["image_id", "caption"],
).to_table().to_pylist()
for h in hits:
    print(h)

Image→caption (image-to-text retrieval)

ds = lance.dataset("hf://datasets/lance-format/flickr30k-lance/data/train.lance")
ref = ds.take([0], columns=["image_emb", "caption"]).to_pylist()[0]
emb_field = ds.schema.field("text_emb")
query = pa.array([ref["image_emb"]], type=emb_field.type)
neighbors = ds.scanner(
    nearest={"column": "text_emb", "q": query[0], "k": 10},
    columns=["caption"],
).to_table().to_pylist()

Full-text search on captions

import lance
ds = lance.dataset("hf://datasets/lance-format/flickr30k-lance/data/train.lance")
hits = ds.scanner(
    full_text_query="dog playing in the snow",
    columns=["image_id", "caption"],
    limit=10,
).to_table().to_pylist()

Working with images

from pathlib import Path
import lance
ds = lance.dataset("hf://datasets/lance-format/flickr30k-lance/data/train.lance")
row = ds.take([0], columns=["image", "filename"]).to_pylist()[0]
Path(row["filename"]).write_bytes(row["image"])

Why Lance?

  • One dataset carries images + image embeddings + text embeddings + indices — no sidecar files.
  • On-disk vector and full-text indices live next to the data, so search works on local copies and on the Hub.
  • Schema evolution: add columns (new captions, alternate embeddings, moderation labels) without rewriting the data.

Source & license

Converted from lmms-lab/flickr30k, which is itself a parquet redistribution of the original Flickr30k corpus. Original images come from Flickr; review the Flickr30k licensing terms before redistribution.

Citation

@article{young2014image,
  title={From image descriptions to visual denotations: New similarity metrics for semantic inference over event descriptions},
  author={Young, Peter and Lai, Alice and Hodosh, Micah and Hockenmaier, Julia},
  journal={Transactions of the Association for Computational Linguistics},
  volume={2},
  pages={67--78},
  year={2014}
}
Downloads last month
27