has anybody else faced the error type:"ValueError: Input 0 of layer "model" is incompatible with the layer: expected shape=(None, 100, 217), found shape=(None, 36, 6)"???
Posted by: Enricomagliulo @ Dec. 10, 2022, 3:32 p.m.If I'm not mistaken your input layer of the network has an input size of (None, 100, 217), which means that somehow you trained your network on some data points that instead of being time series of 36x6, were 100x217. Maybe you did some preprocessing that resulted in a change of shape? I suggest you to check any preprocessing that you are doing and the shape of the input layer that you defined.
I actually made a similar mistake in the first homework since I assumed that the images were 48x48 and in reality, the images were 96x96.
My input layer was trained on some smaller preprocessed images and thus expected as input images that were 48x48 and when presented with 96x96 images, an error was raised.
indeed, I tried once again, having my input and output elements that had as shape (36,6), to male sure to fix this error; I tried to submit the model, but it gave me the following error:
ValueError: operands could not be broadcast together with shapes (827,36) (827,)
and I actually don't know why