H Company Releases NeoMME, an Open-Source Multimodal Encoder Family

Unite.AI
Read full post
H Company launched NeoMME, an open-source multimodal and multilingual encoder family with 260M and 800M parameters, trained from scratch for efficient visual document retrieval tasks. NeoMME uses a shared bidirectional Transformer for text and images, achieving top performance on the ViDoRe v3 benchmark without relying on pretrained vision-language models.

More in Machine Learning

Machine Learning3 min read

Nvidia and Palantir fine-tune a 30B Nemotron model for Nvidia’s supply chain. It beats a model 18 times its size.

Covered by 3 sources
Machine Learning6 min read

CoreWeave Puts Field Engineers Inside Customer Teams for Physical AI

Covered by 2 sources
Machine Learning2 min read

Weatherwatch: AI model beats standard methods at predicting cyclones

The Guardian