docling-project/webdataset_23_07_2026
20.7 TB
AI & ML interests
None defined yet.
Recent Activity
View all activity
asnassar
updated a
bucket about 6 hours ago
PeterWJStaar
updated a
dataset about 19 hours ago
PeterWJStaar
updated a
model 5 days ago
PeterWJStaar
published a
model 5 days ago
PeterWJStaar
updated a
dataset 22 days ago
PeterWJStaar
published a
dataset 22 days ago
Normalize layout label names
#2 opened 25 days ago
by 
auerchristoph
Normalize layout label names
#4 opened 25 days ago
by auerchristoph
Normalize layout label names
#4 opened 25 days ago
by auerchristoph
Normalize layout label names
#3 opened 25 days ago
by auerchristoph
asnassar
published a
bucket about 2 months ago
Saidgurbuz
authored a
paper 4 months ago
lucas-morin
authored 6
papers 10 months ago
MolGrapher: Graph-based Visual Recognition of Chemical Structures
Paper • 2308.12234 • Published
Docling: An Efficient Open-Source Toolkit for AI-driven Document Conversion
Paper • 2501.17887 • Published • 1
SmolDocling: An ultra-compact vision-language model for end-to-end multi-modal document conversion
Paper • 2503.11576 • Published • 173
Advanced Layout Analysis Models for Docling
Paper • 2509.11720 • Published
MarkushGrapher: Joint Visual and Textual Recognition of Markush Structures
Paper • 2503.16096 • Published • 1
SubGrapher: Visual Fingerprinting of Chemical Structures
Paper • 2504.19695 • Published
andito
authored a
paper 10 months ago
Post
3333
Finally, our new paper is out! "𝗙𝗶𝗻𝗲𝗩𝗶𝘀𝗶𝗼𝗻: 𝗢𝗽𝗲𝗻 𝗗𝗮𝘁𝗮 𝗜𝘀 𝗔𝗹𝗹 𝗬𝗼𝘂 𝗡𝗲𝗲𝗱"! 🥳
FineVision: Open Data Is All You Need (2510.17269)
If you've ever trained a VLM, you know this problem: nobody shares their data mixtures. It's a black box, making replicating SOTA work impossible.
We wanted to change that.
FineVision unifies 200 sources into 24 million samples. With 17.3 million images and 9.5 billion answer tokens, it's the largest open resource of its kind.
In the paper, we share how we built it:
🔍 finding and cleaning data at scale
🧹 removing excessive duplicates across sources
🤗 decontaminating against 66 public benchmarks
My favorite part is Figure 6 (in the video!). It's our visual diversity analysis. It shows that FineVision isn't just bigger; it's more balanced and conceptually richer than other open datasets.
NVIDIA's Eagle 2 paper highlighted just how critical this visual diversity is, and our results confirm it: models trained on FineVision consistently outperform those trained on any other open dataset on 11 benchmarks!
🎉 To celebrate the paper, I’m also releasing a concatenated and shuffled version of the full dataset! 👉
It’s ready to stream, so you can start training your own models right away:
from datasets import load_dataset
d = load_dataset("HuggingFaceM4/FineVision_full_shuffled", split="train", streaming=True)
print(next(iter(d)))
A big shoutout to the first authors: Luis Wiedmann and Orr Zohar. They are rockstars!
FineVision: Open Data Is All You Need (2510.17269)
If you've ever trained a VLM, you know this problem: nobody shares their data mixtures. It's a black box, making replicating SOTA work impossible.
We wanted to change that.
FineVision unifies 200 sources into 24 million samples. With 17.3 million images and 9.5 billion answer tokens, it's the largest open resource of its kind.
In the paper, we share how we built it:
🔍 finding and cleaning data at scale
🧹 removing excessive duplicates across sources
🤗 decontaminating against 66 public benchmarks
My favorite part is Figure 6 (in the video!). It's our visual diversity analysis. It shows that FineVision isn't just bigger; it's more balanced and conceptually richer than other open datasets.
NVIDIA's Eagle 2 paper highlighted just how critical this visual diversity is, and our results confirm it: models trained on FineVision consistently outperform those trained on any other open dataset on 11 benchmarks!
🎉 To celebrate the paper, I’m also releasing a concatenated and shuffled version of the full dataset! 👉
HuggingFaceM4/FineVision_full_shuffled It’s ready to stream, so you can start training your own models right away:
from datasets import load_dataset
d = load_dataset("HuggingFaceM4/FineVision_full_shuffled", split="train", streaming=True)
print(next(iter(d)))
A big shoutout to the first authors: Luis Wiedmann and Orr Zohar. They are rockstars!
- 4 replies