-
In the final lab of week 1, it says “the size of the input data specified in the model.fit statement below”. But below where it has model.fit, I don’t see the size of input specified. So a bit confused of that sentence.
-
When we use function model.layers[2].weights or layer3.get_weights(), we all get W3 and b3. I thought weights only refer to W and b is called parameter. So b is considered weight too?
Hello, @flyunicorn
The upper screenshot tells us two ways for Tensorflow to know the input shape. The lower screenshot shows us the first way: we give model.fit the input data, then tensorflow determines the input shape from the input data.
Yes - both w and b are considered weights in Tensorflow’s terminology. Any trainable parameters are weights in Tensorflow.
Cheers,
Raymond
1 Like