Is it possible to install earlier version of tensorflow on Kaggle TPU?

Question

Kaggle TPU switched to tf 2.3.1. But I need tf 2.2.0.

I tried:

!pip install tensorflow~=2.2.0

Tensorflow switches to 2.2.1. But I am getting an error:

InvalidArgumentError: NodeDef expected inputs 'string' do not match 0 inputs specified; Op ;attr=T:type; attr=tensor_name:string; attr=send_device:string; attr=send_device_incarnation:int; attr=recv_device:string; attr=client_terminated:bool,default=false; is_stateful=true>; NodeDef: {{node _Send}}

when calling:

tpu = tf.distribute.cluster_resolver.TPUClusterResolver()  # TPU detection
print('Running on TPU ', tpu.cluster_spec().as_dict()['worker'])
tf.config.experimental_connect_to_cluster(tpu)
tf.tpu.experimental.initialize_tpu_system(tpu)

There is no such a problem on Google Colab.

Answer 1

Looks like it is impossible. Refer to:

https://www.kaggle.com/questions-and-answers/195058