Documentation

AMCT (TensorFlow)

This section uses Ubuntu 20.04 as an example to describe the pre-installation actions. The version mapping is as follows:

Table 1 Version mapping

Type

Version Requirement

How to Obtain

Remarks

OS and version

EulerOS release 2.0 (SP10) aarch64

cat /etc/*release && uname -m

Only quantization on CPU is supported.

OS and version

Ubuntu 20.04 x86_64

Download the required software release from the Ubuntu official website and install it. Run the following command to query the installation:

cat /etc/*release && uname -m

Quantization on CPU or GPU is supported.

OS and version

Ubuntu 20.04 aarch64

Download the required software release from the Ubuntu official website and install it. Run the following command to query the installation:

cat /etc/*release && uname -m

Quantization on CPU or GPU is supported.

OS and version

KylinV10SP1 x86_64

-

Only quantization on CPU is supported.

OS and version

openEuler 20.03 x86_64

-

Only quantization on CPU is supported.

OS and version

UOS20 aarch64

-

Only quantization on CPU is supported.

CUDA toolkit/CUDA driver

11.2 or 10.0

Obtain required packages for installation. For example, you can obtain the Toolkit package from the following URL, which contains the Driver package.

https://developer.nvidia.com/cuda-toolkit-archive

You need to install the CUDA software to perform quantization on the GPU.

  • TensorFlow 2.6.5 works with CUDA 11.2.
  • TensorFlow 1.15.0 works with CUDA 10.0.

Python

Python3.7.x, Python3.8.x, Python3.9.x

Python 3.9.2 is recommended.

For Ubuntu, Kylin V10 SP1, openEuler, and UOS, see Python 3.9.2 Installation on Ubuntu.

For EulerOS, see Python Python3.9.2 Installation on EulerOS.
  • Ensure that the server has Internet access.
  • This document uses Python 3.9.2 as an example. The environment variables and installation commands are subject to the actual Python version.

TensorFlow

2.6.5 or 1.15.0

For details, see Dependency Installation.

Mapping between TensorFlow and Python versions:

  • Version 2.6.5 matches the following Python versions: Python3.7.x (3.7.5-3.7.11), Python3.8.x and Python 3.9.x.
  • Version 1.15.0 matches Python 3.7.x (3.7.5-3.7.11).

numpy

1.16.0–1.23.5 or 1.20.0–1.23.5

For details, see Dependency Installation.
  • If your TensorFlow version is 2.6.5, the NumPy version must be 1.20.0–1.23.5.
  • If your TensorFlow version is 1.15.0, the NumPy version must be 1.16.0–1.23.5.

protobuf

3.13.0+

For details, see Dependency Installation.

-