jikechao commented on PR #15334: URL: https://github.com/apache/tvm/pull/15334#issuecomment-1638609630
@echuraev Thanks for your comments! If `unit =0`, the output shape will contain a zero dimension. Such models will be meaningless. It looks like a bug in the old version of Keras, and this bug was fixed in the later version. Thus, I would like to reject these invalid models.  In the new version of Keras, the called statement `layer = keras.layers.LSTM(units=0)` will crash immediately by checking in Keras. Thus, the newly added assertion statement only works for that old version of Keras . -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
