Dependency Installation

Installing the Rec SDK TensorFlow requires the following environment dependencies and operations. Refer to Table 1 to prepare the installation environment.

**Table 1** Rec SDK required environment dependencies
Dependency/Operation	Recommended Version	How to Obtain
CANN and Ascend-adapted TensorFlow plugin	CANN 8.5.0	Click here. In the Select Resource area on the left, filter the required software packages, confirm the version, and download the software packages. Obtain the software and perform installation by referring to the CANN Software Installation Guide: Ascend-cann-toolkit_{version}_linux-{arch}.run and Ascend-cann-{chip_type}-ops_{version}_linux-{arch}.run. click here to obtain the Ascend-adapted TensorFlow plugin. npu_device-2.6.5 for TensorFlow 2.6.5 and npu_bridge-1.15.0 for TensorFlow 1.15.0.
Driver and firmware of Ascend hardware products	Ascend HDK 25.5.0	Click here. In the Select Resource area on the left, filter the required software packages, confirm the version, and download the software packages. For driver and firmware installation details, see the Driver and Firmware Installation and Upgrade Guide that matches your hardware product.
Ascend Docker Runtime	MindCluster 7.3.0	For details, see "Installation" > "Installation and Deployment" in MindCluster Cluster Scheduling User Guide.
Device NIC configuration	-	Refer to the "Configuring a Training Node" in the Ascend Training Solution Networking Guide. Use HCCN Tool to configure the device IP addresses of NPU network ports.
TensorFlow	TensorFlow 1.15.0 and TensorFlow 2.6.5	Obtain the source code from the TensorFlow repository. The TensorFlow official website does not provide the corresponding .whl package for the Arm environment. If you need to use the .whl package in the Arm environment, obtain it from here. NOTE: If the .whl package fails to be downloaded, you can copy the link and open it in a new tab page to complete the download.
Python 3.7.5	Python 3.7.5	Obtain the dependency packages from the Python official website.

For custom open-source and third-party software, track the vulnerabilities and issues in the corresponding community and fix them in a timely manner. You can confirm the known vulnerabilities of the corresponding open-source software version on the CVE official website, and fix the vulnerabilities through version upgrade or patch package update.

Parent topic: Installation and Deployment