English | 简体中文

👋 join us on Twitter, Discord and WeChat

---

Latest News 🎉

• [2023/12] Turbomind supports multimodal input. Gradio Demo
• [2023/11] Turbomind supports loading hf model directly. Click here for details.
• [2023/11] TurboMind major upgrades, including: Paged Attention, faster attention kernels without sequence length limitation, 2x faster KV8 kernels, Split-K decoding (Flash Decoding), and W4A16 inference for sm_75
• [2023/09] TurboMind supports Qwen-14B
• [2023/09] TurboMind supports InternLM-20B
• [2023/09] TurboMind supports all features of Code Llama: code completion, infilling, chat / instruct, and python specialist. Click here for deployment guide
• [2023/09] TurboMind supports Baichuan2-7B
• [2023/08] TurboMind supports flash-attention2.
• [2023/08] TurboMind supports Qwen-7B, dynamic NTK-RoPE scaling and dynamic logN scaling
• [2023/08] TurboMind supports Windows (tp=1)
• [2023/08] TurboMind supports 4-bit inference, 2.4x faster than FP16, the fastest open-source implementation🚀. Check this guide for detailed info
• [2023/08] LMDeploy has launched on the HuggingFace Hub, providing ready-to-use 4-bit models.
• [2023/08] LMDeploy supports 4-bit quantization using the AWQ algorithm.
• [2023/07] TurboMind supports Llama-2 70B with GQA.
• [2023/07] TurboMind supports Llama-2 7B/13B.
• [2023/07] TurboMind supports tensor-parallel inference of InternLM.

---

Introduction

LMDeploy is a toolkit for compressing, deploying, and serving LLM, developed by the MMRazor and MMDeploy teams. It has the following core features:

• Efficient Inference Engine (TurboMind): It develops key features like persistent batch(a.k.a. continuous batching), blocked KV cache, dynamic split&fuse, tensor parallelism, high-performance CUDA kernels and so on, ensuring the high throughput and low latency during LLMs inference.

• Interactive Inference Mode: By caching the k/v of attention during multi-round dialogue processes, the engine remembers dialogue history, thus avoiding repetitive processing of historical sessions.

• Quantization: LMDeploy supports various quantization methods and efficient inference of quantized models. The reliability of quantization has been verified on models of different scales.

Performance

The TurboMind engine achieves up to 1.36 ~ 1.85 times higher request throughput compared to vLLM across models of various size. In terms of static inference capabilities, the token throughput (out token/s) of TurboMind's 4bit model inference significantly outperforms FP16/BF16 inference, with an improvement of up to 2.4 times.

For inference benchmarks in more devices and more settings, please refer to the following link:

• A100
• 4090
• 3090
• 2080

Supported Models

LMDeploy has developed two inference engines - Pytorch and TurboMind, each with a different focus. The former strives for ultimate optimization of inference performance, while the latter, developed purely in Python, aims to decrease the barriers for developers.

As shown in the next tables, the inference engines differ in the types of supported models and the inference data type. Users can choose the one that best fits their actual needs.

TurboMind

Model	Size	FP16/BF16	KV INT8	W4A16
Llama	7B - 65B	Yes	Yes	Yes
Llama2	7B - 70B	Yes	Yes	Yes
InternLM	7B - 20B	Yes	Yes	Yes
InternLM-XComposer	7B	Yes	Yes	Yes
QWen	7B - 72B	Yes	Yes	Yes
QWen-VL	7B	Yes	Yes	Yes
Baichuan	7B	Yes	Yes	Yes
Baichuan2	7B	Yes	Yes	Yes
Code Llama	7B - 34B	Yes	No	No

Pytorch

Model	Size	FP16/BF16	KV INT8	W8A8
Llama	7B - 65B	Yes	No	Yes
Llama2	7B - 70B	Yes	No	Yes
InternLM	7B - 20B	Yes	No	Yes
Baichuan2	7B - 13B	Yes	No	Yes
ChatGLM2	6B	Yes	No	No
Falcon	7B - 180B	Yes	No	No

Getting Started

Please overview getting_started section for the basic usage of LMDeploy.

For detailed user guides and advanced guides, please refer to our tutorials:

• User Guide
  • Inference pipeline
  • Inference Engine - TurboMind
  • Inference Engine - PyTorch
  • Serving
  • Quantization
• Advance Guide
  • Add chat template
  • Add a new model
  • gemm tuning
  • Long context inference

Contributing

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

Acknowledgement

• FasterTransformer
• llm-awq
• vLLM
• DeepSpeed-MII

License

This project is released under the Apache 2.0 license.