Antoine Chaffin's picture

Antoine Chaffin

NohTow

·

https://antoine.chaffin.fr

AI & ML interests

NLP, Computer Vision, Multimodal classification

Recent Activity

upvoted an article 5 days ago

LightOnOCR-2-1B: a lightweight high-performance end-to-end OCR model family

upvoted a collection about 2 months ago

NanoBEIR datasets

updated a dataset about 2 months ago

lightonai/NanoBEIR-en

View all activity

Organizations

upvoted an article 5 days ago

Article

LightOnOCR-2-1B: a lightweight high-performance end-to-end OCR model family

5 days ago

•

58

upvoted a collection about 2 months ago

NanoBEIR datasets

These datasets are compatible with the (Sparse)NanoBEIREvaluator with Sentence Transformers v5.2+. Also CrossEncoderNanoBEIREvaluator if bm25 column • 16 items • Updated Dec 13, 2025 • 12

upvoted an article about 2 months ago

Article

TurkColBERT: A Benchmark of Dense and Late-Interaction Models for Turkish Information Retrieval

Dec 4, 2025

•

19

upvoted an article 3 months ago

Article

LightOnOCR-1B: The Case for End-to-End and Efficient Domain-Specific Vision-Language Models for OCR

Oct 23, 2025

•

70

upvoted 2 papers 3 months ago

Simple Projection Variants Improve ColBERT Performance

Paper • 2510.12327 • Published Oct 14, 2025 • 7

Fantastic (small) Retrievers and How to Train Them: mxbai-edge-colbert-v0 Tech Report

Paper • 2510.14880 • Published Oct 16, 2025 • 19

upvoted an article 5 months ago

Article

Welcome EmbeddingGemma, Google's new efficient embedding model

+4

Sep 4, 2025

•

269

upvoted an article 6 months ago

Article

Introducing ColQwen-Omni: Retrieve in every modality

Jul 17, 2025

•

76

upvoted a paper 6 months ago

Seq vs Seq: An Open Suite of Paired Encoders and Decoders

Paper • 2507.11412 • Published Jul 15, 2025 • 31

upvoted an article 6 months ago

Article

Ettin Suite: SoTA Paired Encoders and Decoders

+4

Jul 16, 2025

•

77

upvoted 2 collections 7 months ago

PyLate 🐕

5 items • Updated 13 days ago • 3

BioClinical ModernBERT

This project was a collaboration between members of the Dana-Farber Cancer Institute, LightOn, MIT, OpenEvidence and Microsoft. • 3 items • Updated Sep 9, 2025 • 11

upvoted an article 11 months ago

Article

Introducing EuroBERT: A High-Performance Multilingual Encoder Model

Mar 10, 2025

•

146

upvoted a paper 11 months ago

Rank1: Test-Time Compute for Reranking in Information Retrieval

Paper • 2502.18418 • Published Feb 25, 2025 • 29

upvoted a collection 12 months ago

Qwen2.5-VL

Vision-language model series based on Qwen2.5 • 11 items • Updated 24 days ago • 553

upvoted an article 12 months ago

Article

Finally, a Replacement for BERT: Introducing ModernBERT

+13

Dec 19, 2024

•

720

upvoted a collection about 1 year ago

ModernGLiNER

GLiNER models based on modern encoder architectures • 2 items • Updated Dec 24, 2024 • 7

upvoted a paper about 1 year ago

Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference

Paper • 2412.13663 • Published Dec 18, 2024 • 159

upvoted a collection about 1 year ago

ModernBERT

Bringing BERT into modernity via both architecture changes and scaling • 3 items • Updated Dec 19, 2024 • 156

upvoted a paper over 1 year ago

Reducing the Footprint of Multi-Vector Retrieval with Minimal Performance Impact via Token Pooling

Paper • 2409.14683 • Published Sep 23, 2024 • 11