Fix TF version mismatch error in the Colab training tutorial (#10423)

[yumemio]

Description

This is a fix for issue #10423, which raised UnknownError at detection_model.predict(image, shapes).
I modified the code so that the latest TensorFlow installation (which is pre-installed in Colab VMs) is used, since the cause of the issue is TF/CUDA-cuDNN version mismatch (Colab switched to use CUDA11.2 / cuDNN 7.6.5).

Another option I thought about was to install tensorflow==2.7.0 instead of 2.2.0, but I discarded it because that would not prevent similar versioning issues in the future.

Type of change

• Bug fix (non-breaking change which fixes an issue)

Tests

Test Configuration:

1. Open this notebook in Colab.
2. Switch the runtime to GPU.
3. (optional) Uncomment the ready-made gt_boxes and comment out colab_utils.annotate().
4. Execute all cells, and verify that UnknownError is not raised.

Checklist

• I have signed the Contributor License Agreement.
• I have read guidelines for pull request.
• My code follows the coding guidelines.
• I have performed a self code review of my own code.
• I have commented my code, particularly in hard-to-understand areas.
• I have made corresponding changes to the documentation.
• My changes generate no new warnings.
• I have added tests that prove my fix is effective or that my feature works.
  (comment: see above)

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB