🔄 In a Training Loop

Stefan Schweter PRO

stefan-it

P(doom) <0.001%

·

https://schweter.bayern

AI & ML interests

Flair + Flukes Library 💕, NER & PoS Tagging, LM Pretraining (mostly encoder-only & encoder-decoder), Historical Language Models, German Language Models, Bavarian NLP 🥨, xLSTM

Recent Activity

commentedon a paper 3 days ago

Alice: A Large-Scale German Benchmark for Rubric-Based Multi-Dimensional Automatic Short Answer Scoring

upvoted an article 4 days ago

One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO

liked a model 5 days ago

StrandsAgents/strands-decider-2B-hobson-v21

View all activity

Organizations

upvoted an article 4 days ago

Article

One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO

nvidia

•

4 days ago

• 16

upvoted an article 5 days ago

Article

Can LLMs Correct Historical OCR Without Rewriting History?

emanuelaboros

•

5 days ago

• 2

upvoted a paper 6 days ago

Beyond Single-Negative Preference: Multi-Negative DPO for LLM-Centric Historical Entity Linking

Paper • 2609.07379 • Published Sep 10 • 1

upvoted a changelog 6 days ago

Hugging Face Changelog

P(doom) on user profiles

6 days ago

• 339

upvoted an article 10 days ago

Article

Introducing Olmo-core 3: Open, scalable training infrastructure for large MoEs

allenai

•

10 days ago

• 25

upvoted 2 collections 18 days ago

Flair 2.0 - Named Entity Recognition Models

New state-of-the-art NER models for Flair! Commercial use possible. • 6 items • Updated 10 days ago • 1

Flair 1.0 - All Classic Models

All classic NLP models that ship with Flair. Noncommercial use only. • 23 items • Updated 18 days ago • 1

upvoted an article 20 days ago

Article

tokenizers v1: encode, decode and scaling, measured

+2

ArthurZ, sbrandeis, mcpotato, lysandre

•

21 days ago

• 89

upvoted a paper 25 days ago

Augmented Lagrangian Predictive Coding

Paper • 2605.31022 • Published May 29 • 2

upvoted 2 collections about 1 month ago

OCR on the Hub

Curated OCR models for documents, languages, handwriting and text in images. Browse five collections with short practical notes. • 5 items • Updated 25 days ago • 13

NeoMME

Meet NeoMME: a family of 260M and 800M Multimodal-Native Multilingual Encoders • 12 items • Updated 14 days ago • 33

upvoted a paper about 1 month ago

Jina-OCR-v1: Efficient Document Parsing with Speculative Decoding and Dense Verifiable Rewards

Paper • 2609.03181 • Published Sep 2 • 11

upvoted a collection about 1 month ago

Kraken PP-OCRv6 text recognition models

Hub mirrors of Benjamin Kiessling's multilingual PP-OCRv6 line-recognition family for Kraken: tiny, small, and medium. • 3 items • Updated Sep 3 • 9

upvoted a paper about 1 month ago

Institutional Newspapers Pipeline: Deriving billions of high quality tokens from historical newspapers

Paper • 2608.18972 • Published Aug 19 • 13

upvoted 2 collections about 1 month ago

Institutional Newspapers

A growing corpus of newspapers, parsed and optimized for computational access. • 6 items • Updated 27 days ago • 10

Nemotron-Personas

A collection of multilingual, region-specific synthetic persona datasets that support sovereign AI development across many countries and regions. • 10 items • Updated Aug 11 • 78

upvoted 2 papers about 2 months ago

Institutional Books - Enriched Text: A customizable multilingual open-source pipeline for denoising, deduplicating, and annotating OCR text at scale

Paper • 2608.19026 • Published Aug 19 • 3

BavGround: A Benchmark for Regional Cultural Grounding and Dialect Competence in Bavarian

Paper • 2608.12894 • Published Aug 13 • 1

upvoted 2 articles 2 months ago

Article

Meta is back with Muse Glimmer: local, agentic, multimodal, and open source

+2

pcuenq, merve, burtenshaw, ariG23498

•

Aug 10

• 117

Article

FineBooks: are open OCR models good enough to unlock historical knowledge?

finebooks

•

Aug 10

• 26