An Open Source Machine Learning Framework for Everyone

Tensors and Dynamic neural networks in Python with strong GPU acceleration

LLM inference in C/C++

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

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance…

One-for-All Multimodal Evaluation Toolkit Across Text, Image, Video, and Audio Tasks

Train transformer language models with reinforcement learning.

Models and examples built with TensorFlow

Supercharge Your LLM with the Fastest KV Cache Layer

Scalable data pre processing and curation toolkit for LLMs

🤗 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.

Reference implementations of MLPerf® inference benchmarks

NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.

scikit-learn: machine learning in Python

Awesome LLM compression research papers and tools.

Fine-tuning & Reinforcement Learning for LLMs. 🦥 Train OpenAI gpt-oss, DeepSeek-R1, Qwen3, Gemma 3, TTS 2x faster with 70% less VRAM.

DeepEP: an efficient expert-parallel communication library

Making large AI models cheaper, faster and more accessible

Minimal examples of data structures and algorithms in Python

✨✨Latest Advances on Multimodal Large Language Models

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

[TMLR] A curated list of language modeling researches for code (and other software engineering activities), plus related datasets.

Democratizing Reinforcement Learning for LLMs

Accessible large language models via k-bit quantization for PyTorch.

MNBVC(Massive Never-ending BT Vast Chinese corpus)超大规模中文语料集。对标chatGPT训练的40T数据。MNBVC数据集不但包括主流文化，也包括各个小众文化甚至火星文的数据。MNBVC数据集包括新闻、作文、小说、书籍、杂志、论文、台词、帖子、wiki、古诗、歌词、商品介绍、笑话、糗事、聊天记录等一切形式的纯文本中文数据。

An Easy-to-use, Scalable and High-performance RLHF Framework based on Ray (PPO & GRPO & REINFORCE++ & vLLM & Ray & Dynamic Sampling & Async Agentic RL)

A framework for few-shot evaluation of language models.

Algorithms, 4th edition textbook code and libraries

[ICLR 2025] PEARL: Parallel Speculative Decoding with Adaptive Draft Length