TensorFlow <2.3>=2.0 states NumPy<2.0.0 as a dependency, but is compatible with NumPy<1.19.0
NumPy in versions >=1.19.0<2.0.0 has a breaking ABI change making it incompatible with TensorFlow in versions <2.3>=2.0.
Issue description
TensorFlow releases matching version range specification <2.3>=2.0 stated NumPy<2.0.0 as a dependency. A new release 1.19.0 of NumPy introduced a breaking change which made the version range specification of TensorFlow not valid for a more recent NumPy releases.
Affected packages
tensorflow>=2.0<2.3
tensorflow-cpu>=2.0<2.3
tensorflow-gpu>=2.0<2.3
intel-tensorflow>=2.0<2.3
numpy>=1.19.0<2.0.0
Severity
- WARNING
Issue fix
The recommendation engine will try to resolve a software stack with a compatible NumPy release without ABI breaking change.
Pipeline units
Recommendation types
All the recommendation types can produce this warning message, except for latest:
- performance
- security
- stable
- testing
See this document that describes recommendation types listed.