volcengine/veScale

PyTorch-native post-training at scale

An Efficient and User-Friendly Scaling Library for Reinforcement Learning with Large Language Models

🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support

Modalities, a PyTorch-native framework for distributed and reproducible foundation model training.

An industrial extension library of pytorch to accelerate large scale model training

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

Distributed Training Over-The-Internet

An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Async RL)

Supercharge Your Model Training