Towards Best Practices for Open Datasets for LLM Training
Paper
β’
2501.08365
β’
Published
β’
40
HuggingFaceFW
Enterprise
community
AI & ML interests
None defined yet.
Recent Activity
View all activity
HuggingFaceFW's activity
hynkyΒ
authored
a
paper
2 days ago
lvwerraΒ
authored
a
paper
2 days ago
guipenedoΒ
authored
a
paper
2 days ago
thomwolfΒ
authored
a
paper
2 days ago
Post
2829
π«...And we're live!π« Seasonal newsletter from ethicsy folks at Hugging Face, exploring the ethics of "AI Agents"
https://huggingface.co/blog/ethics-soc-7
Our analyses found:
- There's a spectrum of "agent"-ness
- *Safety* is a key issue, leading to many other value-based concerns
Read for details & what to do next!
With @evijit , @giadap , and @sasha
https://huggingface.co/blog/ethics-soc-7
Our analyses found:
- There's a spectrum of "agent"-ness
- *Safety* is a key issue, leading to many other value-based concerns
Read for details & what to do next!
With @evijit , @giadap , and @sasha
davanstrienΒ
posted
an
update
5 days ago
Post
2901
Introducing scandi-fine-web-cleaner
davanstrien/scandi-fine-web-cleaner, the first model trained on FineWeb-C community annotations!
FineWeb2 is a massive multilingual dataset for pre-training language models. Like any web-scale dataset, it contains low-quality content. How can we improve it?
Over the past months, an amazing community of 400+ annotators has been labelling content quality (using Argilla) across 23 languages through the FineWeb-C initiative.
Today, I'm happy to share the first classifier trained on this data.
π What we've built:
- A lightweight classifier that efficiently removes low-quality content
- 90%+ precision demonstrated on Danish & Swedish
- Can process the 43M+ documents in Danish FineWeb2 with minimal compute
π Why this matters: The approach can be reproduced for any of the 23 languages in FineWeb-C ( data-is-better-together/fineweb-c). We can improve training data quality at scale without massive compute resources by starting with community annotations and training small, efficient classifiers.
Want to build a classifier for your language? Check out the full blog post with code examples and implementation details: https://danielvanstrien.xyz/posts/2025/FineWeb-c/scandinavian-content-filtering-fineweb.html
FineWeb2 is a massive multilingual dataset for pre-training language models. Like any web-scale dataset, it contains low-quality content. How can we improve it?
Over the past months, an amazing community of 400+ annotators has been labelling content quality (using Argilla) across 23 languages through the FineWeb-C initiative.
Today, I'm happy to share the first classifier trained on this data.
π What we've built:
- A lightweight classifier that efficiently removes low-quality content
- 90%+ precision demonstrated on Danish & Swedish
- Can process the 43M+ documents in Danish FineWeb2 with minimal compute
π Why this matters: The approach can be reproduced for any of the 23 languages in FineWeb-C ( data-is-better-together/fineweb-c). We can improve training data quality at scale without massive compute resources by starting with community annotations and training small, efficient classifiers.
Want to build a classifier for your language? Check out the full blog post with code examples and implementation details: https://danielvanstrien.xyz/posts/2025/FineWeb-c/scandinavian-content-filtering-fineweb.html
-
- 1 reply
davanstrienΒ
posted
an
update
8 days ago
Post
2074
The
data-is-better-together/fineweb-c dataset is growing!
This week a few more languages have got 1,000 annotations for the educational quality of data from HuggingFaceFW/fineweb-2.
Why should you care?
The quality of pre-training data can have a big impact on the performance of downstream language models trained on that data ( HuggingFaceFW/blogpost-fineweb-v1).
Being able to filter by educational quality is on way of improving the quality of the data you use for training an LLM. Very importantly this approach can also reduce the amount of data needed for pertaining.
Why not use an LLM?
LLMs can be used to annotate educational quality for a subset of data. This data can then be used to train a smaller encoder only model to label the full dataset. However, this may not work well for languages outside of english. This is where fineweb-c (community) comes in.
The community is annotating the educational quality of fineweb2 data. Currently 114 languages have some annotations. These annotations will enable a number of things:
- Evaluate whether an LLM can label the educational quality for texts in that language well
- Directly be used for training quality classifiers
- Help discover other rules and huerisitcs for refining fineweb2 further for different languages.
This week the following languages where done:
Swedish thanks to: @Lauler @AntonVic @ohallstrom @bjarlestam @menbom @Ekgren @apsod
Ukrainian thanks to: @hannayukhymenko @robinhad @realPivo @RabotiahovDmytro @reciprocate
Assamese thanks to: @moyoor97 @Arpanjyoti @nawaf-helmi123 @pahigogoi1 @aelhence @kishorekashyap
Want to learn more: https://huggingface.co/blog/davanstrien/fineweb2-community
Contribute yourself here: data-is-better-together/fineweb-c
This week a few more languages have got 1,000 annotations for the educational quality of data from HuggingFaceFW/fineweb-2.
Why should you care?
The quality of pre-training data can have a big impact on the performance of downstream language models trained on that data ( HuggingFaceFW/blogpost-fineweb-v1).
Being able to filter by educational quality is on way of improving the quality of the data you use for training an LLM. Very importantly this approach can also reduce the amount of data needed for pertaining.
Why not use an LLM?
LLMs can be used to annotate educational quality for a subset of data. This data can then be used to train a smaller encoder only model to label the full dataset. However, this may not work well for languages outside of english. This is where fineweb-c (community) comes in.
The community is annotating the educational quality of fineweb2 data. Currently 114 languages have some annotations. These annotations will enable a number of things:
- Evaluate whether an LLM can label the educational quality for texts in that language well
- Directly be used for training quality classifiers
- Help discover other rules and huerisitcs for refining fineweb2 further for different languages.
This week the following languages where done:
Swedish thanks to: @Lauler @AntonVic @ohallstrom @bjarlestam @menbom @Ekgren @apsod
Ukrainian thanks to: @hannayukhymenko @robinhad @realPivo @RabotiahovDmytro @reciprocate
Assamese thanks to: @moyoor97 @Arpanjyoti @nawaf-helmi123 @pahigogoi1 @aelhence @kishorekashyap
Want to learn more: https://huggingface.co/blog/davanstrien/fineweb2-community
Contribute yourself here: data-is-better-together/fineweb-c
- 1 reply
Cannot load with datasets
3
#4 opened 10 days ago
by
mbanon
guipenedoΒ
updated
a
dataset
10 days ago
davanstrienΒ
posted
an
update
22 days ago
Post
3166
πΈπ° Hovorte po slovensky? Help build better AI for Slovak!
We only need 90 more annotations to include Slovak in the next Hugging Face FineWeb2-C dataset ( data-is-better-together/fineweb-c) release!
Your contribution will help create better language models for 5+ million Slovak speakers.
Annotate here: data-is-better-together/fineweb-c.
Read more about why we're doing it: https://huggingface.co/blog/davanstrien/fineweb2-community
We only need 90 more annotations to include Slovak in the next Hugging Face FineWeb2-C dataset ( data-is-better-together/fineweb-c) release!
Your contribution will help create better language models for 5+ million Slovak speakers.
Annotate here: data-is-better-together/fineweb-c.
Read more about why we're doing it: https://huggingface.co/blog/davanstrien/fineweb2-community
-
- 3 replies
davanstrienΒ
posted
an
update
29 days ago
Post
1769
Introducing FineWeb-C ππ, a community-built dataset for improving language models in ALL languages.
Inspired by FineWeb-Edu the community is labelling the educational quality of texts for many languages.
318 annotators, 32K+ annotations, 12 languages - and growing! π
data-is-better-together/fineweb-c
Inspired by FineWeb-Edu the community is labelling the educational quality of texts for many languages.
318 annotators, 32K+ annotations, 12 languages - and growing! π
data-is-better-together/fineweb-c
Post
2234
Introducing ππ
π’π§πππππ‘: the best public math pre-training dataset with 50B+ tokens!
HuggingFaceTB/finemath
Math remains challenging for LLMs and by training on FineMath we see considerable gains over other math datasets, especially on GSM8K and MATH.
We build the dataset by:
π οΈ carefully extracting math data from Common Crawl;
π iteratively filtering and recalling high quality math pages using a classifier trained on synthetic annotations to identify math reasoning and deduction.
We conducted a series of ablations comparing the performance of Llama-3.2-3B-Base after continued pre-training on FineMath and observe notable gains compared to the baseline model and other public math datasets.
We hope this helps advance the performance of LLMs on math and reasoning! π
Weβre also releasing all the ablation models as well as the evaluation code.
HuggingFaceTB/finemath-6763fb8f71b6439b653482c2
HuggingFaceTB/finemath
Math remains challenging for LLMs and by training on FineMath we see considerable gains over other math datasets, especially on GSM8K and MATH.
We build the dataset by:
π οΈ carefully extracting math data from Common Crawl;
π iteratively filtering and recalling high quality math pages using a classifier trained on synthetic annotations to identify math reasoning and deduction.
We conducted a series of ablations comparing the performance of Llama-3.2-3B-Base after continued pre-training on FineMath and observe notable gains compared to the baseline model and other public math datasets.
We hope this helps advance the performance of LLMs on math and reasoning! π
Weβre also releasing all the ablation models as well as the evaluation code.
HuggingFaceTB/finemath-6763fb8f71b6439b653482c2
Post
4868
We are proud to announce
HuggingFaceFW/fineweb-2: A sparkling update to
HuggingFaceFW/fineweb with 1000s of π£οΈlanguages.
We applied the same data-driven approach that led to SOTA English performance inπ· FineWeb to thousands of languages.
π₯ FineWeb2 has 8TB of compressed text data and outperforms other multilingual datasets in our experiments.
The dataset is released under the permissive π ODC-By 1.0 license, and the π» code to reproduce it and our evaluations is public.
We will very soon announce a big community project, and are working on a π blogpost walking you through the entire dataset creation process. Stay tuned!
In the mean time come ask us question on our chat place: HuggingFaceFW/discussion
H/t @guipenedo @hynky @lvwerra as well as @vsabolcec Bettina Messmer @negar-foroutan and @mjaggi
We applied the same data-driven approach that led to SOTA English performance inπ· FineWeb to thousands of languages.
π₯ FineWeb2 has 8TB of compressed text data and outperforms other multilingual datasets in our experiments.
The dataset is released under the permissive π ODC-By 1.0 license, and the π» code to reproduce it and our evaluations is public.
We will very soon announce a big community project, and are working on a π blogpost walking you through the entire dataset creation process. Stay tuned!
In the mean time come ask us question on our chat place: HuggingFaceFW/discussion
H/t @guipenedo @hynky @lvwerra as well as @vsabolcec Bettina Messmer @negar-foroutan and @mjaggi
-
- 2 replies
clefourrierΒ
authored
a
paper
about 1 month ago
Post
1262
Exponentially growing number of open-source AI models over the course of the past 30 months β from a few thousands to over 1 million and more
Interactive data viz: huggingface/open-source-ai-year-in-review-2024
Interactive data viz: huggingface/open-source-ai-year-in-review-2024
Post
1440
Most liked and most downloaded open-source AI models from 2022 to 2024
Interactive viz: https://aiworld.eu/embed/model/model/treemap
Discussion: huggingface/open-source-ai-year-in-review-2024
Interactive viz: https://aiworld.eu/embed/model/model/treemap
Discussion: huggingface/open-source-ai-year-in-review-2024
davanstrienΒ
posted
an
update
about 2 months ago
Post
513
Increasingly, LLMs are becoming very useful for helping scale annotation tasks, i.e. labelling and filtering. When combined with the structured generation, this can be a very scalable way of doing some pre-annotation without requiring a large team of human annotators.
However, there are quite a few cases where it still doesn't work well. This is a nice paper looking at the limitations of LLM as an annotator for Low Resource Languages: On Limitations of LLM as Annotator for Low Resource Languages (2411.17637).
Humans will still have an important role in the loop to help improve models for all languages (and domains).
However, there are quite a few cases where it still doesn't work well. This is a nice paper looking at the limitations of LLM as an annotator for Low Resource Languages: On Limitations of LLM as Annotator for Low Resource Languages (2411.17637).
Humans will still have an important role in the loop to help improve models for all languages (and domains).
davanstrienΒ
posted
an
update
about 2 months ago
Post
2488
First dataset for the new Hugging Face Bluesky community organisation:
bluesky-community/one-million-bluesky-posts π¦
π 1M public posts from Bluesky's firehose API
π Includes text, metadata, and language predictions
π¬ Perfect to experiment with using ML for Bluesky π€
Excited to see people build more open tools for a more open social media platform!
π 1M public posts from Bluesky's firehose API
π Includes text, metadata, and language predictions
π¬ Perfect to experiment with using ML for Bluesky π€
Excited to see people build more open tools for a more open social media platform!
davanstrienΒ
posted
an
update
about 2 months ago
Post
1357
The Bluesky AT Protocol unlocks exciting possibilities:
- Building custom feeds using ML
- Creating dashboards for data exploration
- Developing custom models for Bluesky
To gather Bluesky resources on the Hub, I've created a community org: https://huggingface.co/bluesky-community
My first rather modest contribution is a dashboard that shows the number of posts every second. Drinking straight from the firehose API π°
bluesky-community/bluesky-posts-over-time
- Building custom feeds using ML
- Creating dashboards for data exploration
- Developing custom models for Bluesky
To gather Bluesky resources on the Hub, I've created a community org: https://huggingface.co/bluesky-community
My first rather modest contribution is a dashboard that shows the number of posts every second. Drinking straight from the firehose API π°
bluesky-community/bluesky-posts-over-time
- 1 reply
Post
1905
Making SmolLM2 reproducible: open-sourcing our training & evaluation toolkit π οΈ https://github.com/huggingface/smollm/
- Pre-training code with nanotron
- Evaluation suite with lighteval
- Synthetic data generation using distilabel (powers our new SFT dataset HuggingFaceTB/smoltalk)
- Post-training scripts with TRL & the alignment handbook
- On-device tools with llama.cpp for summarization, rewriting & agents
Apache 2.0 licensed. V2 pre-training data mix coming soon!
Which other tools should we add next?
- Pre-training code with nanotron
- Evaluation suite with lighteval
- Synthetic data generation using distilabel (powers our new SFT dataset HuggingFaceTB/smoltalk)
- Post-training scripts with TRL & the alignment handbook
- On-device tools with llama.cpp for summarization, rewriting & agents
Apache 2.0 licensed. V2 pre-training data mix coming soon!
Which other tools should we add next?