DeepSeek Coder comprises a series of code language models trained from scratch on both 87% code and 13% natural language in English and Chinese, with each model pre-trained on 2T tokens. We provide various sizes of the code model, ranging from 1B to 33B versions. Each model is pre-trained on repo-level code corpus by employing a window size of 16K... See more

Our new Omni-Embed-Nemotron cross-modal retrieval model can search text, images, audio, or video using a single bi-encoder architecture that process all modalities in a shared embedding space for simpler, faster retrieval-- all in one open-source architecture 🛠️🌐 Try out the code yourself ➡️ https://t.co/X1JkRIc4Xp

TorchMultimodal (Beta Release)

Introduction

TorchMultimodal is a PyTorch library for training state-of-the-art multimodal multi-task models at scale. It provides:

• A repository of modular and composable building blocks (models, fusion layers, loss functions, datasets and utilities).


• A repository of examples that show how to combine these building

... See more