Can't use google-colab

[fisakhan]

While running the following code, I get an error:

import datetime
import os
import tensorflow as tf
import tensorflow_cloud as tfc
import tensorflow_datasets as tfds
from tensorflow import keras
from tensorflow.keras import layers
from tensorflow.keras.models import Model

if not tfc.remote():
  from google.colab import files

  key_upload = files.upload()
  key_path = list(key_upload.keys())[0]
  os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = key_path
  os.system(f"gcloud auth activate-service-account --key-file {key_path}")

Details of the error:

$ python3 dogs_classification.py 
2021-06-28 12:17:54.852448: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcudart.so.11.0
/home/t/.virtualenvs/tf_cloud/lib/python3.6/site-packages/IPython/utils/traitlets.py:5: UserWarning: IPython.utils.traitlets has moved to a top-level traitlets package.
  warn("IPython.utils.traitlets has moved to a top-level traitlets package.")
<IPython.core.display.HTML object>
Traceback (most recent call last):
  File "dogs_classification.py", line 24, in <module>
    key_upload = files.upload()
  File "/home/t/.virtualenvs/tf_cloud/lib/python3.6/site-packages/google/colab/files.py", line 64, in upload
    input_id=input_id, output_id=output_id))
  File "/home/t/.virtualenvs/tf_cloud/lib/python3.6/site-packages/google/colab/output/_js.py", line 35, in eval_js
    kernel = _ipython.get_kernel()
  File "/home/t/.virtualenvs/tf_cloud/lib/python3.6/site-packages/google/colab/_ipython.py", line 28, in get_kernel
    return get_ipython().kernel
AttributeError: 'InteractiveShell' object has no attribute 'kernel'

I copied the above code from cloud/src/python/tensorflow_cloud/core/tests/examples/dogs_classification.ipynb, and want to run it from dogs_classification.py on my computer.

google-auth-1.32.0
google-colab==1.0.0
tensorflow==2.5.0
tensorflow-cloud==0.1.16

[eddydecena]

The library google-colab is not designed to run in a normal Python REPL, which means that you have to run this in an ipython environment like Jupiter notebooks or colab rather than the python interpreter directly.