Latest posts tagged with #modernbert on Bluesky
ModernBERT & ColBERT Enhance Biomedical Retrieval-Augmented Generation
A two-stage retriever using ModernBERT and ColBERTv2 lifted Recall@3 by up to 4.2 pts and reached 0.4448 accuracy on the MIRAGE benchmark, slightly beating MedCPT’s 0.4436. getnews.me/modernbert-colbert-enhan... #modernbert #colbertv2 #biomedicalai
ModernVBERT is a vision model
Suite of compact 250M-parameter vision-language encoders, achieving state-of-the-art performance in this size class, matching the performance of models up to 10x larger.
MIT licensed
#AI #LLM #ModernBERT
huggingface.co/collections/...
ModernBERT Pretrained for Patent Text Outperforms Baselines
Researchers released three ModernBERT models—base‑PT, base‑VX, and Mosaic‑BERT‑large—trained on 60 million patent records, achieving up to three‑times faster inference than PatentBERT. Read more: getnews.me/modernbert-pretrained-fo... #modernbert #patentnlp
L'été à Paris a un air tropical. Je teste des modèles de langage pour des tâches de NLI en zero shot classification. Premier résultat : #ModernBert va très vite. Là, c'est sur ma machine, la suite demain sur un H100. #NLP #productivité
Read more about the journey behind ModernBERT here:
www.lighton.ai/lighton-blog...
#ACL2025 #LightOnResearch #NLP #ModernBERT #LanguageModels #MachineLearning #openscience
Built on #ModernBERT and #PyLate library, this 150M parameter model outperforms models 45x larger on #BRIGHT, the gold standard for reasoning-intensive retrieval.
Adam Kovacs presents LettuceDetect, a lightweight hallucination detector for Retrieval-Augmented Generation (#RAG) pipelines. It is an encoder-based model built on #ModernBERT, released under the MIT license with ready-to-use #Python packages and pretrained models.
🚀 Fine-tuning ModernBERT for Radiology 🚀. Part 2 of my 3-part series shows how to use the ROCOv2 radiology dataset to fine-tune a transformer 🤗.
www.johnpaulett.com/2025/modernb...
#bert #modernbert #nlp #ml
Discover ModernBERT: the faster, smarter, and longer-context successor to BERT, delivering state-of-the-art performance for encoding tasks without GenAI hype.
#AI #ModernBERT #BERT #LLM
aidisruptionpub.com/p/google-ber...
Are there any tutorials or starter scripts on training a new #ModernBERT models (or fine-tuning the Answer.ai's model for a language from a different family)? In both cases, a new tokenizer will be trained.
Happy to share my latest guide on #ModernBERT, including a fine-tuning notebook!
Learn what's new in ModernBERT's architecture, from extended context windows to advanced attention patterns, and explore hands-on model training with our Colab tutorial.
Most exciting update to encoder-only models in a long time! Love to use them for classification tasks where LLMs are an overkill #ModernBERT
Finally, a Replacement for BERT: Introducing ModernBERT #bert #modernbert #nlp #machinelearning
