🔄 In a Training Loop

Stefan Schweter PRO

stefan-it

·

https://schweter.bayern

AI & ML interests

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

Recent Activity

updated a Space 2 days ago

published a dataset 2 days ago

histde/dta-documents

updated a dataset 2 days ago

histde/dta-documents

View all activity

Organizations

upvoted a paper 3 days ago

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

Paper • 2608.19026 • Published 5 days ago • 1

upvoted a paper 9 days ago

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

Paper • 2608.12894 • Published 11 days ago • 1

upvoted 3 articles 13 days ago

Article

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

+2

pcuenq, merve, burtenshaw, ariG23498

•

14 days ago

• 106

Article

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

finebooks

•

13 days ago

• 25

Article

Making Knowledge Distillation Cheap Enough to Run at Scale

MultiverseComputingCAI

•

13 days ago

• 35

upvoted a paper 16 days ago

Beyond Sequence Order: Syntax-Informed Positional Embeddings for Transformers

Paper • 2608.06111 • Published 18 days ago • 6

upvoted a paper 18 days ago

When Attention Goes Blind: Numerical Failure in ALiBi Positional Encodings

Paper • 2608.03994 • Published 20 days ago • 8

upvoted a paper 23 days ago

Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory

Paper • 2607.27919 • Published 25 days ago • 59

upvoted a collection 23 days ago

MemoryDecoder-at-Scale

10 items • Updated 24 days ago • 11

upvoted a paper 25 days ago

Every Time I Hire a Linguist, Inference Costs Go Down: On Linguistic Rules as Effective Prompt Compressors

Paper • 2607.25335 • Published 27 days ago • 1

upvoted a paper 26 days ago

From Data to Device: ELMOD An Efficient German-First 2.7B Language Model for Mobile Inference

Paper • 2607.24585 • Published 28 days ago • 1

upvoted a collection 26 days ago

ELMOD-2.7b

Efficient Language Model for On-Device Deployment (ELMOD) a compact (2.7B) German LM designed for efficient inference on constrained hardware. • 2 items • Updated 12 days ago • 2

upvoted a paper 26 days ago

Kimi K3: Open Frontier Intelligence

Paper • 2607.24653 • Published 28 days ago • 498

upvoted a paper 27 days ago

Building a European Multilingual Evaluation Dataset: The MMLU Localisation Project within the EMT Network

Paper • 2607.18432 • Published Jul 20 • 2

upvoted a paper 30 days ago

Separating Representation from Reconstruction Enables Scalable Text Encoders

Paper • 2607.04011 • Published Jul 4 • 1

upvoted a collection about 1 month ago

Apertus v1.5

8B and 70B open-data open-weights models, globally multilingual with multimodal input and reasoning support • 2 items • Updated 2 days ago • 31

upvoted a paper about 1 month ago

news-crawler-LM: A Small Long-Context Model For High-Quality News Crawling

Paper • 2607.21284 • Published Jul 23 • 2

upvoted 2 articles about 1 month ago

Article

Be Ready Before the Attack: A Practical Guide to Self-Hosting an Open Model for Cyber Defense

jeffboudier

•

Jul 20

• 20

Article

Security incident disclosure — July 2026

system

•

Jul 16

• 776

upvoted a paper about 1 month ago

A Sovereign, Open-Source Foundation Model for German and English

Paper • 2607.09424 • Published Jul 10 • 15