Daniel van Strien
@danielvanstrien.bsky.social
Machine Learning Librarian at @hf.co
Reposted by Daniel van Strien
Built a 2.5MB image classifier that runs in the browser in an evening with Claude Code.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
December 19, 2025 at 12:08 PM
Built a 2.5MB image classifier that runs in the browser in an evening with Claude Code.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
Built a 2.5MB image classifier that runs in the browser in an evening with Claude Code.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
December 19, 2025 at 12:08 PM
Built a 2.5MB image classifier that runs in the browser in an evening with Claude Code.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
I used a dataset I labelled in 2022 and left on @hf.co for 3 years 😬.
It finds illustrated pages in historical books. No server. No GPU.
Just posted my slides from the AI4LAM #FF2025 workshop on open source AI for GLAMs.
Probably slides on their own aren't that useful, but they do feature one of my growing collection of libraries-and-AI memes, so there's that danielvanstrien.xyz/slides.html
Probably slides on their own aren't that useful, but they do feature one of my growing collection of libraries-and-AI memes, so there's that danielvanstrien.xyz/slides.html
December 9, 2025 at 10:13 AM
Just posted my slides from the AI4LAM #FF2025 workshop on open source AI for GLAMs.
Probably slides on their own aren't that useful, but they do feature one of my growing collection of libraries-and-AI memes, so there's that danielvanstrien.xyz/slides.html
Probably slides on their own aren't that useful, but they do feature one of my growing collection of libraries-and-AI memes, so there's that danielvanstrien.xyz/slides.html
Building datasets to train smaller, task-focused models used to be incredibly time-consuming.
Very excited to see SAM3 massively lower that barrier. Describe the class you want to detect and get annotated datasets automatically!
Try it yourself: huggingface.co/datasets/uv-...!
Very excited to see SAM3 massively lower that barrier. Describe the class you want to detect and get annotated datasets automatically!
Try it yourself: huggingface.co/datasets/uv-...!
November 21, 2025 at 1:30 PM
Building datasets to train smaller, task-focused models used to be incredibly time-consuming.
Very excited to see SAM3 massively lower that barrier. Describe the class you want to detect and get annotated datasets automatically!
Try it yourself: huggingface.co/datasets/uv-...!
Very excited to see SAM3 massively lower that barrier. Describe the class you want to detect and get annotated datasets automatically!
Try it yourself: huggingface.co/datasets/uv-...!
Very much looking forward to presenting at this tomorrow. I will be making my usual pitch that datasets are the foundational infrastructure for cultural heritage to benefit from and create useful AI models and tools.
Be warned, I did fire up the meme generator for my slides...
Be warned, I did fire up the meme generator for my slides...
November 5, 2025 at 5:40 PM
Very much looking forward to presenting at this tomorrow. I will be making my usual pitch that datasets are the foundational infrastructure for cultural heritage to benefit from and create useful AI models and tools.
Be warned, I did fire up the meme generator for my slides...
Be warned, I did fire up the meme generator for my slides...
Reposted by Daniel van Strien
Over the last 24 hours, I have finetuned three Qwen3-VL models (2B, 4B, and 8B) on the CATmuS dataset on @hf.co . The first version of the models are now available on the Small Models for GLAM organization with @danielvanstrien.bsky.social (Links below) Working on improving them further.
October 24, 2025 at 2:59 PM
Over the last 24 hours, I have finetuned three Qwen3-VL models (2B, 4B, and 8B) on the CATmuS dataset on @hf.co . The first version of the models are now available on the Small Models for GLAM organization with @danielvanstrien.bsky.social (Links below) Working on improving them further.
DeepSeek-OCR just got vLLM support 🚀
Currently processing @natlibscot.bsky.social's 27,915-page handbook collection with one command.
Processing at ~350 images/sec on A100
Using @hf.co Jobs + uv - zero setup batch OCR!
Will share final time + cost when done!
Currently processing @natlibscot.bsky.social's 27,915-page handbook collection with one command.
Processing at ~350 images/sec on A100
Using @hf.co Jobs + uv - zero setup batch OCR!
Will share final time + cost when done!
October 22, 2025 at 7:20 PM
DeepSeek-OCR just got vLLM support 🚀
Currently processing @natlibscot.bsky.social's 27,915-page handbook collection with one command.
Processing at ~350 images/sec on A100
Using @hf.co Jobs + uv - zero setup batch OCR!
Will share final time + cost when done!
Currently processing @natlibscot.bsky.social's 27,915-page handbook collection with one command.
Processing at ~350 images/sec on A100
Using @hf.co Jobs + uv - zero setup batch OCR!
Will share final time + cost when done!
Reposted by Daniel van Strien
🤗 Sentence Transformers is joining @hf.co! 🤗
This formalizes the existing maintenance structure, as I've personally led the project for the past two years on behalf of Hugging Face. I'm super excited about the transfer!
Details in 🧵
This formalizes the existing maintenance structure, as I've personally led the project for the past two years on behalf of Hugging Face. I'm super excited about the transfer!
Details in 🧵
October 22, 2025 at 1:04 PM
🤗 Sentence Transformers is joining @hf.co! 🤗
This formalizes the existing maintenance structure, as I've personally led the project for the past two years on behalf of Hugging Face. I'm super excited about the transfer!
Details in 🧵
This formalizes the existing maintenance structure, as I've personally led the project for the past two years on behalf of Hugging Face. I'm super excited about the transfer!
Details in 🧵
OCR is one of AI's oldest challenges (first systems: early 1900s!)
Modern vision-language models have transformed what's possible: handwriting, 100+ languages, math formulas, tables, signature extraction...
New @hf.co guide on OCR
huggingface.co/blog/ocr-ope...
Modern vision-language models have transformed what's possible: handwriting, 100+ languages, math formulas, tables, signature extraction...
New @hf.co guide on OCR
huggingface.co/blog/ocr-ope...
Supercharge your OCR Pipelines with Open Models
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
October 22, 2025 at 8:58 AM
OCR is one of AI's oldest challenges (first systems: early 1900s!)
Modern vision-language models have transformed what's possible: handwriting, 100+ languages, math formulas, tables, signature extraction...
New @hf.co guide on OCR
huggingface.co/blog/ocr-ope...
Modern vision-language models have transformed what's possible: handwriting, 100+ languages, math formulas, tables, signature extraction...
New @hf.co guide on OCR
huggingface.co/blog/ocr-ope...
Small models work great for GLAM but there aren't enough examples!
With @wjbmattingly.bsky.social I'm launching small-models-for-glam on @hf.co to create/curate models that run on modest hardware and address GLAM use cases.
Follow the org to keep up-to-date!
huggingface.co/small-models...
With @wjbmattingly.bsky.social I'm launching small-models-for-glam on @hf.co to create/curate models that run on modest hardware and address GLAM use cases.
Follow the org to keep up-to-date!
huggingface.co/small-models...
October 16, 2025 at 1:22 PM
Small models work great for GLAM but there aren't enough examples!
With @wjbmattingly.bsky.social I'm launching small-models-for-glam on @hf.co to create/curate models that run on modest hardware and address GLAM use cases.
Follow the org to keep up-to-date!
huggingface.co/small-models...
With @wjbmattingly.bsky.social I'm launching small-models-for-glam on @hf.co to create/curate models that run on modest hardware and address GLAM use cases.
Follow the org to keep up-to-date!
huggingface.co/small-models...
Very nice work! IMO, this is the kind of topic that more libraries/GLAM/DH people should be working on. The training of these models is *relatively* simple. As always, the missing ingredient is readily accessible data.
It's been brewing for months: @inriaparisnlp.bsky.social releases CoMMA (Corpus of Multilingual Medieval Archives) !
📚 2.5bn tokens of mostly Latin and French texts
🕰️ 800→1600 CE
📜 23k manuscripts
🖥️ 18k on the reading interface: comma.inria.fr
🔍 Paper: inria.hal.science/hal-05299220v1
(1/🧵)
📚 2.5bn tokens of mostly Latin and French texts
🕰️ 800→1600 CE
📜 23k manuscripts
🖥️ 18k on the reading interface: comma.inria.fr
🔍 Paper: inria.hal.science/hal-05299220v1
(1/🧵)
CoMMA
comma.inria.fr
October 15, 2025 at 3:55 PM
Very nice work! IMO, this is the kind of topic that more libraries/GLAM/DH people should be working on. The training of these models is *relatively* simple. As always, the missing ingredient is readily accessible data.
Reposted by Daniel van Strien
It's been brewing for months: @inriaparisnlp.bsky.social releases CoMMA (Corpus of Multilingual Medieval Archives) !
📚 2.5bn tokens of mostly Latin and French texts
🕰️ 800→1600 CE
📜 23k manuscripts
🖥️ 18k on the reading interface: comma.inria.fr
🔍 Paper: inria.hal.science/hal-05299220v1
(1/🧵)
📚 2.5bn tokens of mostly Latin and French texts
🕰️ 800→1600 CE
📜 23k manuscripts
🖥️ 18k on the reading interface: comma.inria.fr
🔍 Paper: inria.hal.science/hal-05299220v1
(1/🧵)
CoMMA
comma.inria.fr
October 15, 2025 at 2:51 PM
It's been brewing for months: @inriaparisnlp.bsky.social releases CoMMA (Corpus of Multilingual Medieval Archives) !
📚 2.5bn tokens of mostly Latin and French texts
🕰️ 800→1600 CE
📜 23k manuscripts
🖥️ 18k on the reading interface: comma.inria.fr
🔍 Paper: inria.hal.science/hal-05299220v1
(1/🧵)
📚 2.5bn tokens of mostly Latin and French texts
🕰️ 800→1600 CE
📜 23k manuscripts
🖥️ 18k on the reading interface: comma.inria.fr
🔍 Paper: inria.hal.science/hal-05299220v1
(1/🧵)
Reposted by Daniel van Strien
Another week, another VLM-based OCR model!
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
October 13, 2025 at 6:13 PM
Another week, another VLM-based OCR model!
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
Another week, another VLM-based OCR model!
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
October 13, 2025 at 6:13 PM
Another week, another VLM-based OCR model!
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
Nanonets just released OCR2 - a 3B parameter vision-language model for document OCR 📄
You can run it with one command on @hf.co Jobs (no local GPU needed)
Reposted by Daniel van Strien
DoTS.ocr just got native vLLM support!
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
October 7, 2025 at 3:45 PM
DoTS.ocr just got native vLLM support!
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
DoTS.ocr just got native vLLM support!
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
October 7, 2025 at 3:45 PM
DoTS.ocr just got native vLLM support!
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
I built a UV script so you can run SOTA multilingual OCR in seconds with zero setup using @hf.co Jobs
Tested on 1800s library cards - works great ✨
Card catalogues aren't just a relic of the past - many institutions still rely on them because full migration is too expensive. VLMs could help change that.
I uploaded two new @hf.co datasets (~470K cards) for training/evaluating models to extract structured metadata from catalogue cards.
I uploaded two new @hf.co datasets (~470K cards) for training/evaluating models to extract structured metadata from catalogue cards.
October 6, 2025 at 9:30 AM
Card catalogues aren't just a relic of the past - many institutions still rely on them because full migration is too expensive. VLMs could help change that.
I uploaded two new @hf.co datasets (~470K cards) for training/evaluating models to extract structured metadata from catalogue cards.
I uploaded two new @hf.co datasets (~470K cards) for training/evaluating models to extract structured metadata from catalogue cards.
Reposted by Daniel van Strien
We’re hiring for two machine learning roles. A chance to do cutting edge things with ML to make this place a lot more personalized.
jobs.gem.com/bluesky/am9i...
jobs.gem.com/bluesky/am9i...
Bluesky Jobs
Bluesky Jobs
jobs.gem.com
October 1, 2025 at 1:14 AM
We’re hiring for two machine learning roles. A chance to do cutting edge things with ML to make this place a lot more personalized.
jobs.gem.com/bluesky/am9i...
jobs.gem.com/bluesky/am9i...
New @hf.co BigLAM dataset: 9,363 OA books with page images + rich MARC metadata for evaluating (and training) VLMs on metadata extraction.
Libraries are starting to explore AI-assisted cataloguing, but we lack public evaluation data. Hoping this helps fill that gap.
huggingface.co/datasets/big...
Libraries are starting to explore AI-assisted cataloguing, but we lack public evaluation data. Hoping this helps fill that gap.
huggingface.co/datasets/big...
October 2, 2025 at 6:51 PM
New @hf.co BigLAM dataset: 9,363 OA books with page images + rich MARC metadata for evaluating (and training) VLMs on metadata extraction.
Libraries are starting to explore AI-assisted cataloguing, but we lack public evaluation data. Hoping this helps fill that gap.
huggingface.co/datasets/big...
Libraries are starting to explore AI-assisted cataloguing, but we lack public evaluation data. Hoping this helps fill that gap.
huggingface.co/datasets/big...
Blogged: Fine-tuning a VLM for art history in hours, not weeks
iconclass-vlm generates museum catalog codes (fun fact: "71H7131" = "Bathsheba with David's letter"!)
@hf.co TRL + Jobs = magic ✨
Guide here: danielvanstrien.xyz/posts/2025/i...
iconclass-vlm generates museum catalog codes (fun fact: "71H7131" = "Bathsheba with David's letter"!)
@hf.co TRL + Jobs = magic ✨
Guide here: danielvanstrien.xyz/posts/2025/i...
danielvanstrien.xyz
September 4, 2025 at 8:19 PM
Blogged: Fine-tuning a VLM for art history in hours, not weeks
iconclass-vlm generates museum catalog codes (fun fact: "71H7131" = "Bathsheba with David's letter"!)
@hf.co TRL + Jobs = magic ✨
Guide here: danielvanstrien.xyz/posts/2025/i...
iconclass-vlm generates museum catalog codes (fun fact: "71H7131" = "Bathsheba with David's letter"!)
@hf.co TRL + Jobs = magic ✨
Guide here: danielvanstrien.xyz/posts/2025/i...
I fine-tuned a smol VLM to generate specialized art history metadata!
iconclass-vlm: Qwen2.5-VL-3B trained using SFT to generate ICONCLASS codes (think Dewey Decimal for art!)
Trained with @hf.co TRL + Jobs - single UV script, no GPU needed!
Blog soon!
iconclass-vlm: Qwen2.5-VL-3B trained using SFT to generate ICONCLASS codes (think Dewey Decimal for art!)
Trained with @hf.co TRL + Jobs - single UV script, no GPU needed!
Blog soon!
September 3, 2025 at 6:22 PM
I fine-tuned a smol VLM to generate specialized art history metadata!
iconclass-vlm: Qwen2.5-VL-3B trained using SFT to generate ICONCLASS codes (think Dewey Decimal for art!)
Trained with @hf.co TRL + Jobs - single UV script, no GPU needed!
Blog soon!
iconclass-vlm: Qwen2.5-VL-3B trained using SFT to generate ICONCLASS codes (think Dewey Decimal for art!)
Trained with @hf.co TRL + Jobs - single UV script, no GPU needed!
Blog soon!
What if OCR models could show you their thought process?
NuMarkdown-8B-Thinking from NuMind (YC S22) doesn't just extract text - it reasons through documents first.
Could be pretty valuable for weird historical documents?
Example here: davanstrien-ocr-time-capsule.static.hf.space/index.html?d...
NuMarkdown-8B-Thinking from NuMind (YC S22) doesn't just extract text - it reasons through documents first.
Could be pretty valuable for weird historical documents?
Example here: davanstrien-ocr-time-capsule.static.hf.space/index.html?d...
August 7, 2025 at 3:16 PM
What if OCR models could show you their thought process?
NuMarkdown-8B-Thinking from NuMind (YC S22) doesn't just extract text - it reasons through documents first.
Could be pretty valuable for weird historical documents?
Example here: davanstrien-ocr-time-capsule.static.hf.space/index.html?d...
NuMarkdown-8B-Thinking from NuMind (YC S22) doesn't just extract text - it reasons through documents first.
Could be pretty valuable for weird historical documents?
Example here: davanstrien-ocr-time-capsule.static.hf.space/index.html?d...
You can now generate synthetic data using OpenAIs GPT OSS models on @hf.co Jobs!
One command, no setup:
hf jobs uv run --flavor l4x4 [script-url] \
--input-dataset your/dataset \
--output-dataset your/output
Works on L4 GPUs ⚡
huggingface.co/datasets/uv-...
One command, no setup:
hf jobs uv run --flavor l4x4 [script-url] \
--input-dataset your/dataset \
--output-dataset your/output
Works on L4 GPUs ⚡
huggingface.co/datasets/uv-...
uv-scripts/openai-oss · Datasets at Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
August 6, 2025 at 7:38 AM
You can now generate synthetic data using OpenAIs GPT OSS models on @hf.co Jobs!
One command, no setup:
hf jobs uv run --flavor l4x4 [script-url] \
--input-dataset your/dataset \
--output-dataset your/output
Works on L4 GPUs ⚡
huggingface.co/datasets/uv-...
One command, no setup:
hf jobs uv run --flavor l4x4 [script-url] \
--input-dataset your/dataset \
--output-dataset your/output
Works on L4 GPUs ⚡
huggingface.co/datasets/uv-...