Datasets:

Formats:
parquet
Languages:
English
ArXiv:
Tags:
image
Libraries:
Datasets
Dask
License:
Nick Padgett commited on
Commit
757982f
1 Parent(s): bd2d12c

Updated README with parquet file structure. Adding parquet files for dataset.

Browse files
README.md CHANGED
@@ -13,10 +13,10 @@ tags:
13
  ![PD12M](header.jpg)
14
 
15
  # Summary
16
- **PD12M** is a collection of 12 million CC0/PD image-caption pairs for the purpose of training generative image models.
17
 
18
  # About
19
- PD12M was built and curated with [Source.Plus] (htts://source.plus) with the aim of resolving many of the data quality issues that arise in web-scraped training data: the presence of copyrighted material, low quality images and captions, violent or nsfw content, PII, decaying dataset quality via broken links, etc.
20
 
21
  PD12M consists of entirely public domain and CC0 licensed images, with automated recaptioning of image data, and quality and safety filtering. Images in PD12M are also hosted on dedicated cloud storage, separate from the original image hosts, to avoid placing an undue burden on those hosts or impacting services for regular users. This also ensures the dataset remains wholly intact over its lifetime.
22
 
@@ -27,13 +27,12 @@ This dataset has two components. The first is the `metadata`, which contains the
27
  The metadata is made available through a series of parquet files with the following schema:
28
  - `id`: A unique identifier for the image.
29
  - `url`: The URL of the image.
30
- - `s3_key`: The S3 file key of the image.
31
  - `caption`: A caption for the image.
32
- - `md5_hash`: The MD5 hash of the image file.
33
- - `mime_type`: The MIME type of the image file.
34
  - `width`: The width of the image in pixels.
35
  - `height`: The height of the image in pixels.
36
- - `license_type`: The URL of the license.
 
 
37
 
38
  ## Images
39
  The image files are all hosted in the AWS S3 bucket `pd12m`. The URLs to the images files are all maintained in the metadata files.
 
13
  ![PD12M](header.jpg)
14
 
15
  # Summary
16
+ **PD12M** is a collection of 12.4 million CC0/PD image-caption pairs for the purpose of training generative image models.
17
 
18
  # About
19
+ PD12M was built and curated with [Source.Plus](https://source.plus) with the aim of resolving many of the data quality issues that arise in web-scraped training data: the presence of copyrighted material, low quality images and captions, violent or nsfw content, PII, decaying dataset quality via broken links, etc.
20
 
21
  PD12M consists of entirely public domain and CC0 licensed images, with automated recaptioning of image data, and quality and safety filtering. Images in PD12M are also hosted on dedicated cloud storage, separate from the original image hosts, to avoid placing an undue burden on those hosts or impacting services for regular users. This also ensures the dataset remains wholly intact over its lifetime.
22
 
 
27
  The metadata is made available through a series of parquet files with the following schema:
28
  - `id`: A unique identifier for the image.
29
  - `url`: The URL of the image.
 
30
  - `caption`: A caption for the image.
 
 
31
  - `width`: The width of the image in pixels.
32
  - `height`: The height of the image in pixels.
33
+ - `mime_type`: The MIME type of the image file.
34
+ - `hash`: The MD5 hash of the image file.
35
+ - `license`: The URL of the license.
36
 
37
  ## Images
38
  The image files are all hosted in the AWS S3 bucket `pd12m`. The URLs to the images files are all maintained in the metadata files.
pd12m.00.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ee70206eb4111680e9a353eee8bf9632d55f99b9616cee77f5a2f1c81bbd235
3
+ size 189808004
pd12m.01.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2eb76f98dfaee31c0618ce82f3bb2f294d9833fa810ee031f0a84e37345aacc8
3
+ size 189736792
pd12m.02.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58d619ce906118235d1bceeaa5e34fcdeb50aa0e08958fd108856b6f6dfd5667
3
+ size 189751474
pd12m.03.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c125a1ee063eb395381cb8f3f0323c9609d38ea790c471d354a606f7bdd40472
3
+ size 189750808
pd12m.04.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04fc801d56c0a722d7c9e229030ce8243f4b3b7f81aca56785c7af41a7adae7e
3
+ size 189725500
pd12m.05.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fca8ac4a562a8c7a05bd01e7d5850a47fbc08cf084e4f867157936df72167a90
3
+ size 189845940
pd12m.06.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d656dc24e8eadfe9ada5f6034a910d49c504489237cfc96ac6b1c55a851bb3d
3
+ size 189617861
pd12m.07.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6441185caa16513763385318ed87d3c0b8f24450552674dab3443887f5a614d8
3
+ size 189749633
pd12m.08.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e49a2033080ca692c2885cc63dd55e9af51b3ab09122019622929756c39d843
3
+ size 189740174
pd12m.09.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:171668e5136c93c15411e59b0e40539faa6b70cb7d838b5910bf3285a673b74c
3
+ size 189609084
pd12m.10.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56ee394b8ab84b7d82072d554967f73dc53aad0af8ae716b8c1f62ae063145d7
3
+ size 189706562
pd12m.11.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1dca3c6905cde393e16230aee88ff6c86121652c482e4b0588dcd2d62f48aa83
3
+ size 189705721
pd12m.12.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c60a3baecd2bb91c976de2bce9ee409dbea16d7de35ba27435944b7eb479a496
3
+ size 77356752