This is very nit-picky, but I happened to look at the RNN constructor and see that a ValueError is raised when, by definition, the exception should be a TypeError, right?
https://github.com/keras-team/keras/blob/764ed95651c1c7dfa71ca523287f6eeb514ccf77/keras/src/layers/rnn/rnn.py#L221-L225
From TypeError: python org docs
exception TypeError Raised when an operation or function is applied to an object of inappropriate type. The associated value is a string giving details about the type mismatch.
and ValueError: python org docs
exception ValueError Raised when an operation or function receives an argument that has the right type but an inappropriate value, and the situation is not described by a more precise exception such as IndexError.
The state_size argument in __init__(...) must be one of the specified types. The currently raised ValueError implies that state_size is the correct type but incorrect value, but that implication is directly contradicted by "state_size must be an integer, or a list/tuple of integers".
This is an extremely minor fix, and am happy to open a PR to do so, but wanted to open an issue first! I appreciate any feedback!
Comment From: dhantule
Hi @jfdev001, thanks for the report.
As per this comment Keras will raise ValueError for an invalid user-provided argument, can we close this issue if you don't have anymore questions?
Comment From: jfdev001
@dhantule I see the tag still awaiting response is present, apologies for not explicitly responding! I have no further questions.
Comment From: dhantule
@jfdev001, Thank you! Feel free to open another issue.