[object Object]
For scenarios where compilation and execution only need to occur in a single environment without cross-OS version compatibility concerns, Scenario 1 is recommended for maximum development efficiency. Conversely, if the compiled software package needs to be deployed to a legacy operating system, Scenario 2 should be selected. (It is recommended to first use the image from Scenario 1 to complete software debugging, ensure its stability, and then switch to the Scenario 2 image for final compilation, thereby achieving a balance between development efficiency and runtime compatibility.)
Use the CANN official container image as the compilation environment. For image details, refer to .
Select an [object Object] image similar to the following version: [object Object] (where xxx should be filled in based on your Ascend AI processor model).
Taking Atlas A2 training products/Atlas A2 inference products as the example, the pull command is as follows:
Obtain the corresponding operator development and compilation Docker image from the Huawei CLOUD official container image repository based on your specific environment.
- x86 architecture:
- Arm architecture:
Run the following command to download:
[object Object]
Parameter Description:
Command Format:
Execution Example:
Expected Output: After successful startup, you will enter the container. The terminal displays the command prompt inside the container, waiting for command input:
[object Object]
After entering the container, run the following command:
Run the following command to write the CANN environment variable configuration to the [object Object] file to ensure it takes effect permanently:
You can configure and save Git credentials using the following command: