Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

tiktoken is a fast BPE tokeniser for use with OpenAI's models.

Official repository of paper "IRG-MotionLLM: Interleaving Motion Generation, Assessment and Refinement for Text-to-Motion Generation"

Simple DirectMedia Layer

Model Context Protocol Servers

Agent framework and applications built upon Qwen>=3.0, featuring Function Calling, MCP, Code Interpreter, RAG, Chrome extension, etc.

A Transparent Generalist Model towards Holistic Medical Vision-Language Understanding

We present the first systematic study on the scaling property of raw agents instantiated by LLMs. We find that performance scales with the increase in the number of agents, using the simple(st) way…

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

Making large AI models cheaper, faster and more accessible

Ongoing research training transformer models at scale

[IEEE Medcial Imaging 2025] FairFedMed: Benchmarking Group Fairness in Federated Medical Imaging with FairLoRA

text and image to video generation: CogVideoX (2024) and CogVideo (ICLR 2023)

LLM inference in C/C++

Unsloth Studio is a web UI for training and running open models like Qwen, DeepSeek, gpt-oss and Gemma locally.

ModelScope: bring the notion of Model-as-a-Service to life.

Minimal reproduction of DeepSeek R1-Zero

A curated list of resources for using LLMs to develop more competitive grant applications.

Downloads videos and playlists from YouTube

DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

Effortless Real-Time Sign Language Translation

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model

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

Let us control diffusion models!

Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.

DIAMOND (DIffusion As a Model Of eNvironment Dreams) is a reinforcement learning agent trained in a diffusion world model. NeurIPS 2024 Spotlight.

Ultralytics YOLO 🚀

NeurIPS 2024 | 🤸‍♂️💥🚗Pedestrian-Centric 3D Pre-collision Pose and Shape Estimation from Dashcam Perspective

[ECCV-2024] DailyDVS-200: A Comprehensive Benchmark Dataset for Event-Based Action Recognition

Heterogeneous Pre-trained Transformer (HPT) as Scalable Policy Learner.