Help Center/ ModelArts/ Troubleshooting/ Inference Deployment/ AI Application Management/ Failed to Download a pip Package When an AI Application Is Created Using OBS
Updated on 2022-12-08 GMT+08:00

Failed to Download a pip Package When an AI Application Is Created Using OBS

Symptom

Creating an AI application using OBS failed. Logs showed that downloading the pip package failed, for example, downloading the NumPy 1.16 package failed.

Possible Causes

Possible causes are as follows:

  1. The package is not available in the pip source. The default pip source is pypi.org. Check whether the package of the target version is available in pypi.org and check the package installation restrictions.
  2. The downloaded package does not match the architecture in the base image. For example, an x86 package is downloaded for Arm, or a Python 3 package is downloaded for Python 2. For details about the runtime environment of a base image, see Available Inference Base Images.
  3. The sequence of configuring package dependencies is incorrect.

Solution

  1. Log in to pypi.org and check whether the required installation package is available. If the package is unavailable, use the WHL package and place it into the OBS directory where the model is stored.
  2. Check whether the installation restrictions and dependencies of the package are met.
  3. If there are package dependencies, configure the dependencies in a correct sequence. For details, see How Do I Edit the Installation Package Dependency Parameters in a Model Configuration File When Importing a Model?