Ameeeee/Cultural_diversity
Viewer
ā¢
Updated
ā¢
7
ā¢
10
Ame Vi
Ameeeee
AI & ML interests
None yet
Recent Activity
updated
a dataset
5 days ago
Ameeeee/Cultural_diversity
published
a dataset
5 days ago
Ameeeee/Cultural_diversity
Organizations
Ameeeee's activity
updated
a
dataset
5 days ago
published
a
dataset
5 days ago
reacted to
tomaarsen's
post with š„
14 days ago
Post
6497
An assembly of 18 European companies, labs, and universities have banded together to launch šŖšŗ EuroBERT! It's a state-of-the-art multilingual encoder for 15 European languages, designed to be finetuned for retrieval, classification, etc.
šŖšŗ 15 Languages: English, French, German, Spanish, Chinese, Italian, Russian, Polish, Portuguese, Japanese, Vietnamese, Dutch, Arabic, Turkish, Hindi
3ļøā£ 3 model sizes: 210M, 610M, and 2.1B parameters - very very useful sizes in my opinion
ā”ļø Sequence length of 8192 tokens! Nice to see these higher sequence lengths for encoders becoming more common.
āļø Architecture based on Llama, but with bi-directional (non-causal) attention to turn it into an encoder. Flash Attention 2 is supported.
š„ A new Pareto frontier (stronger *and* smaller) for multilingual encoder models
š Evaluated against mDeBERTa, mGTE, XLM-RoBERTa for Retrieval, Classification, and Regression (after finetuning for each task separately): EuroBERT punches way above its weight.
š Detailed paper with all details, incl. data: FineWeb for English and CulturaX for multilingual data, The Stack v2 and Proof-Pile-2 for code.
Check out the release blogpost here: https://huggingface.co/blog/EuroBERT/release
* EuroBERT/EuroBERT-210m
* EuroBERT/EuroBERT-610m
* EuroBERT/EuroBERT-2.1B
The next step is for researchers to build upon the 3 EuroBERT base models and publish strong retrieval, zero-shot classification, etc. models for all to use. I'm very much looking forward to it!
šŖšŗ 15 Languages: English, French, German, Spanish, Chinese, Italian, Russian, Polish, Portuguese, Japanese, Vietnamese, Dutch, Arabic, Turkish, Hindi
3ļøā£ 3 model sizes: 210M, 610M, and 2.1B parameters - very very useful sizes in my opinion
ā”ļø Sequence length of 8192 tokens! Nice to see these higher sequence lengths for encoders becoming more common.
āļø Architecture based on Llama, but with bi-directional (non-causal) attention to turn it into an encoder. Flash Attention 2 is supported.
š„ A new Pareto frontier (stronger *and* smaller) for multilingual encoder models
š Evaluated against mDeBERTa, mGTE, XLM-RoBERTa for Retrieval, Classification, and Regression (after finetuning for each task separately): EuroBERT punches way above its weight.
š Detailed paper with all details, incl. data: FineWeb for English and CulturaX for multilingual data, The Stack v2 and Proof-Pile-2 for code.
Check out the release blogpost here: https://huggingface.co/blog/EuroBERT/release
* EuroBERT/EuroBERT-210m
* EuroBERT/EuroBERT-610m
* EuroBERT/EuroBERT-2.1B
The next step is for researchers to build upon the 3 EuroBERT base models and publish strong retrieval, zero-shot classification, etc. models for all to use. I'm very much looking forward to it!
- 1 reply
reacted to
fdaudens's
post with š„
26 days ago
Post
3094
Is this the best tool to extract clean info from PDFs, handwriting and complex documents yet?
Open source olmOCR just dropped and the results are impressive.
Tested the free demo with various documents, including a handwritten Claes Oldenburg letter. The speed is impressive: 3000 tokens/second on your own GPU - that's 1/32 the cost of GPT-4o ($190/million pages). Game-changer for content extraction and digital archives.
To achieve this, Ai2 trained a 7B vision language model on 260K pages from 100K PDFs using "document anchoring" - combining PDF metadata with page images.
Best part: it actually understands document structure (columns, tables, equations) instead of just jumbling everything together like most OCR tools. Their human eval results back this up.
š Try the demo: https://olmocr.allenai.org
Going right into the AI toolkit: JournalistsonHF/ai-toolkit
Open source olmOCR just dropped and the results are impressive.
Tested the free demo with various documents, including a handwritten Claes Oldenburg letter. The speed is impressive: 3000 tokens/second on your own GPU - that's 1/32 the cost of GPT-4o ($190/million pages). Game-changer for content extraction and digital archives.
To achieve this, Ai2 trained a 7B vision language model on 260K pages from 100K PDFs using "document anchoring" - combining PDF metadata with page images.
Best part: it actually understands document structure (columns, tables, equations) instead of just jumbling everything together like most OCR tools. Their human eval results back this up.
š Try the demo: https://olmocr.allenai.org
Going right into the AI toolkit: JournalistsonHF/ai-toolkit
- 3 replies
reacted to
burtenshaw's
post with š
26 days ago
Post
5500
I made a real time voice agent with FastRTC, smolagents, and hugging face inference providers. Check it out in this space:
š burtenshaw/coworking_agent
š burtenshaw/coworking_agent
- 9 replies
upvoted
an
article
28 days ago
Article
Synthetic data: save money, time and carbon with open source
ā¢
63
liked
a
Space
29 days ago
liked
a
Space
about 1 month ago
liked
a
Space
3 months ago
updated
2
datasets
3 months ago
liked
2
datasets
3 months ago
updated
a
dataset
3 months ago
liked
a
Space
3 months ago
published
an
article
3 months ago
Article
Introducing the Synthetic Data Generator - Build Datasets with Natural Language
By
and 5 others
ā¢
ā¢
118
updated
2
datasets
3 months ago