search

google-research / medical-ai-research-foundations

Apache License 2.0
189 stars 21 forks source link

Unable to load the weights from `path-152x2-remedis-m` #4

Closed afilt closed 1 year ago

afilt commented 1 year ago

Hello,

Congratulations for this amazing work, which is super useful to the computational pathology community ! I'm trying to load the weights via the instructions you gave, however I encounter the following error which I don't (for now) manage to solve:

FileNotFoundError: Op type not registered 'DisableCopyOnRead' in binary running on <name_of_my_machine>. Make sure the Op and Kernel are registered in the binary running in this process. Note that if you are loading a saved graph which used ops from tf.contrib, accessing (e.g.) `tf.contrib.resampler` should be done before importing the graph, as contrib ops are lazily registered when the module is first accessed.
 You may be trying to load on a different device from the computational device. Consider setting the `experimental_io_device` option in `tf.saved_model.LoadOptions` to the io_device such as '/job:localhost'.

Do you have any idea how to solve it ? I tried with different tf.saved_model.LoadOptions but nothing's working.

Thank you very much for your answer !

afilt commented 1 year ago

Solved by installing tensorflow 2.12, not further due to this current install bug.