Dan Saattrup Nielsen's picture

Dan Saattrup Nielsen

saattrupdan

·

https://saattrupdan.com

AI & ML interests

NLP for low-resource languages.

Recent Activity

reacted to davanstrien's post with 🔥 about 4 hours ago

Introducing scandi-fine-web-cleaner https://huggingface.co/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. 🔍 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 🌍 Why this matters: The approach can be reproduced for any of the 23 languages in FineWeb-C (https://huggingface.co/datasets/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

liked a model about 4 hours ago

davanstrien/scandi-fine-web-cleaner

reacted to davanstrien's post with 🤗 about 4 hours ago

Introducing scandi-fine-web-cleaner https://huggingface.co/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. 🔍 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 🌍 Why this matters: The approach can be reproduced for any of the 23 languages in FineWeb-C (https://huggingface.co/datasets/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

View all activity

Organizations

saattrupdan's activity

New activity in answerdotai/ModernBERT-base 1 day ago

How to use ModernBERT with the AutoModelForQuestionAnswering class?

#15 opened 21 days ago by

New activity in allenai/OLMo-2-1124-7B 6 days ago

Add text generation pipeline tag

#8 opened 6 days ago by

New activity in allenai/OLMo-2-1124-13B 6 days ago

Add text generation pipeline tag

#5 opened 6 days ago by

New activity in BSC-LT/salamandraTA-2B 6 days ago

Set pipeline tag to text generation

#1 opened 6 days ago by

New activity in NbAiLab/nb-llama-3.1-70B 18 days ago

Change dtype to bf16

#1 opened 18 days ago by

New activity in ThatsGroes/gemma-2-27b-it-FP8-Dynamic 24 days ago

Add pipeline tag to model

#1 opened 24 days ago by

New activity in yhavinga/squad_v2_dutch 25 days ago

Translation model used?

#2 opened 28 days ago by

New activity in corona-tweet/dutch_social 28 days ago

Sentiment label source?

#5 opened 28 days ago by

New activity in AI-Sweden-Models/Llama-3-8B-instruct about 1 month ago

Finetuning datasets used?

#6 opened about 1 month ago by

New activity in saattrupdan/nbailab-base-ner-scandi about 2 months ago

Is there a for the model to show more than 4 results?

#6 opened about 2 months ago by

New activity in NLPnorth/snakmodel-7b-base about 2 months ago

Vocab size does not match tokenizer config

#1 opened about 2 months ago by

New activity in alexandrainst/scandi-nli-base about 2 months ago

Adding `safetensors` variant of this model

#2 opened about 2 months ago by

New activity in NLPnorth/snakmodel-7b-base about 2 months ago

Update tokenizer_config.json

#3 opened about 2 months ago by

Update tokenizer_config.json

#2 opened about 2 months ago by

New activity in DDSC/da-wikipedia-queries 2 months ago

Update README.md

#3 opened 2 months ago by

Librarian Bot: Add language metadata for dataset

#2 opened 2 months ago by

New activity in alexandrainst/foqa 3 months ago

Update README.md

#2 opened 3 months ago by

New activity in alexandrainst/scandi-reddit 4 months ago

[bot] Conversion to Parquet

#1 opened 4 months ago by

parquet-converter

New activity in alexandrainst/ftspeech 4 months ago

Update README.md

#1 opened 4 months ago by

New activity in alexandrainst/radial-plot-generator 6 months ago

Update README.md

#1 opened 6 months ago by