Installation

Installation

1.1. Prerequisites

1.2. Install from Binary

1.3. Install from Source
System Requirements

2.1. Validated Hardware Environment

2.2. Validated Software Environment

Installation

Prerequisites

The following prerequisites and requirements must be satisfied for a successful installation:

Python version: 3.10 or 3.11 or 3.12 or 3.13

Notes:

If you get some build issues, please check frequently asked questions at first.

Install Framework

Install torch for CPU

pip install torch --index-url https://download.pytorch.org/whl/cpu

Use Docker Image with torch installed for HPU

https://docs.habana.ai/en/latest/Installation_Guide/Bare_Metal_Fresh_OS.html#bare-metal-fresh-os-single-click

Install torch/intel_extension_for_pytorch for Intel GPU

https://intel.github.io/intel-extension-for-pytorch/index.html#installation

Install torch for other platform

https://pytorch.org/get-started/locally

Install tensorflow

pip install tensorflow

Install from Binary

Install from Pypi

# Framework extension API for PyTorch/Tensorflow
pip install neural-compressor
# Framework extension API + specific dependency
pip install neural-compressor[pt]
pip install neural-compressor[tf]

# Framework extension API + PyTorch dependency
pip install neural-compressor-pt

# Framework extension API + TensorFlow dependency
pip install neural-compressor-tf

Install from Source

The latest code on master branch may not be stable. Feel free to open an issue if you encounter an error.

git clone https://github.com/intel/neural-compressor.git
cd neural-compressor

# PyTorch framework extension API + PyTorch dependency
INC_PT_ONLY=1 pip install .

# TensorFlow framework extension API + TensorFlow dependency
INC_TF_ONLY=1 pip install .

# JAX framework extension API + JAX dependency
INC_JAX_ONLY=1 pip install .

System Requirements

Validated Hardware Environment

Intel® Neural Compressor supports HPUs based on heterogeneous architecture with two compute engines (MME and TPC):

Intel Gaudi Al Accelerators (Gaudi2, Gaudi3)

Intel® Neural Compressor supports CPUs based on Intel 64 architecture or compatible processors:

Intel Xeon Scalable processor (Sapphire Rapids, Emerald Rapids, Granite Rapids)
Intel Xeon CPU Max Series (Sapphire Rapids HBM)
Intel Core Ultra Processors (Meteor Lake, Lunar Lake)

Intel® Neural Compressor supports GPUs built on Intel’s Xe architecture:

Intel Data Center GPU Flex Series (Arctic Sound-M)
Intel Data Center GPU Max Series (Ponte Vecchio)
Intel® Arc™ B-Series Graphics (Battlemage)

Validated Software Environment

OS version: CentOS 8.4, Ubuntu 24.04, MacOS Ventura 13.5, Windows 11
Python version: 3.10, 3.11, 3.12, 3.13

Framework	TensorFlow	PyTorch	Intel® Extension for PyTorch*
Version	2.16.1 2.15.0 2.14.1	2.8.0 2.7.1 2.6.0	2.8.0 2.7.0 2.6.0