🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

A high-throughput and memory-efficient inference and serving engine for LLMs

Transforms complex documents like PDFs into LLM-ready markdown/JSON for your Agentic workflows.

Natural Language Processing for the next decade. Tokenization, Part-of-Speech Tagging, Named Entity Recognition, Syntactic & Semantic Dependency Parsing, Document Classification

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

Open-Sora: Democratizing Efficient Video Production for All

🤗 smolagents: a barebones library for agents that think in code.

Graph Neural Network Library for PyTorch

Fast and memory-efficient exact attention

verl: Volcano Engine Reinforcement Learning for LLMs

Ongoing research training transformer models at scale

Transfer learning / domain adaptation / domain generalization / multi-task learning etc. Papers, codes, datasets, applications, tutorials.-迁移学习

Open Machine Learning Compiler Framework

Easy-to-use and powerful LLM and SLM library with awesome model zoo.

TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. Tensor…

Automated Machine Learning with scikit-learn

UFO³: Weaving the Digital Agent Galaxy

Large World Model -- Modeling Text and Video with Millions Context

An implementation of model parallel autoregressive transformers on GPUs, based on the Megatron and DeepSpeed libraries

Open-source observability for your GenAI or LLM application, based on OpenTelemetry

Supercharge Your LLM with the Fastest KV Cache Layer

[CVPR 2024] Real-Time Open-Vocabulary Object Detection

Nightly release of ControlNet 1.1

An Open-Source Framework for Prompt-Learning.

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

搜索所有中文NLP数据集，附常用英文NLP数据集

[ECCV 2024] Champ: Controllable and Consistent Human Image Animation with 3D Parametric Guidance

Fengshenbang-LM(封神榜大模型)是IDEA研究院认知计算与自然语言研究中心主导的大模型开源体系，成为中文AIGC和认知智能的基础设施。

A Tensorflow implementation of CapsNet(Capsules Net) in paper Dynamic Routing Between Capsules

PyTorch code and models for V-JEPA self-supervised learning from video.