Environment Requirements

Before porting the model, ensure that the environment meets the hardware and software requirements. The hardware configuration includes CPUs, and the software configuration includes the OS and applications.

Hardware Requirements

Table 1 lists the hardware requirements.

**Table 1** Hardware requirements
Item	Description
CPU	Kunpeng 920 processor

OS and Software Requirements

Table 2 and Table 3 describe the OSs compatible with the DLRM and the software and dependencies required for porting the DLRM.

**Table 2** OS requirements
Item	Version	Description	Download URL
OS	openEuler 22.03 LTS SP3	OS compatible with the software to be ported. When installing an OS, choose Minimal Install and select Development Tools to minimize manual operations.	https://repo.openeuler.org/openEuler-22.03-LTS-SP3/ISO/aarch64

**Table 3** Software requirements
Item	Version	Description	Download URL
Python	3.9.19	Python is a high-level, general-purpose, interpreted, and object-oriented programming language.	It can be installed using Miniconda.
TensorFlow	2.13.0	TensorFlow is a deep learning framework developed by Google for research on machine learning and deep neural networks.	It can be installed using pip.
Miniconda	24.3.0	Miniconda is a compact Python environment management tool.	https://repo.anaconda.com/miniconda/Miniconda3-py39_24.3.0-0-Linux-aarch64.sh
tqdm	4.66.2	tqdm is a Python library used to display task progress bars on the command-line interface (CLI) or graphical user interface (GUI).	It can be installed by running pip commands.
h5py	3.11.0	h5py is a Python API on which TensorFlow depends. It is used to interact with Hierarchical Data Format version 5 (HDF5) files, helping to read and write HDF5 datasets in Pythonic mode.	It can be installed by running conda commands.