xtuner

xtuner is a fine-tuning toolkit - A Next-Generation Training Engine Built for Ultra-Large MoE Models.

Overview

• [2025/09] XTuner V1 Released! A Next-Generation Training Engine Built for Ultra-Large MoE Models

XTuner V1 is a next-generation LLM training engine specifically designed for ultra-large-scale MoE models. Unlike traditional 3D parallel training architectures, XTuner V1 is optimized for the mainstream MoE training scenarios prevalent in today's academic research.

XTuner V1 is committed to continuously improving training efficiency for pre-training, instruction fine-tuning, and reinforcement learning of ultra-large MoE models, with special focus on Ascend NPU optimization.

Our vision is to establish XTuner V1 as a versatile training backend that seamlessly integrates with the broader open-source ecosystem.

The algorithm component is actively evolving. We welcome community contributions - with XTuner V1, scale your algorithms to unprecedented sizes!

• You can use GraphGen to create synthetic data for fine-tuning.

We appreciate all contributions to XTuner. Please refer to CONTRIBUTING.md for the contributing guideline.

The development of XTuner V1's training engine has been greatly inspired by and built upon the excellent work of the open-source community. We extend our sincere gratitude to the following pioneering projects:

We are deeply grateful to all contributors and maintainers of these projects for advancing the field of large-scale model training.

This project is released under the Apache License 2.0. Please also adhere to the Licenses of models and datasets being used.

xtuner is open-source, written primarily in Python, with 5,152 GitHub stars under the Apache 2.0 license. The latest release is v0.2.0 (2025-07-11).

Key capabilities

From the project's documentation:

• [2025/09] XTuner V1 Released! A Next-Generation Training Engine Built for Ultra-Large MoE Models
• Robust performance: Maintains stability despite expert load imbalance during long sequence training
• Massive scale: Supports MoE training up to 1T parameters
• Hardware optimization: Achieves training efficiency on Ascend A3 Supernode that exceeds NVIDIA H800
• ✅ Multimodal Pre-training - Full support for vision-language model training
• ✅ Multimodal Supervised Fine-tuning - Optimized for instruction following

How it fits a local-AI stack

xtuner runs on your own hardware, so pair it with a model and a GPU sized to your needs. Use the VRAM calculator to pick a model that fits your card, and see what you can run for hardware guidance. Related fine-tuning toolkits in the directory:

• Axolotl
• LitGPT
• LLaMA-Factory
• Torchtune

Sources

• Source code & docs: InternLM/xtuner
• Official website: https://xtuner.readthedocs.io/zh-cn/latest/

Stats from GitHub, 2026-06-08.