Hello Nvidia Developers, I have tried to use Transfer Learning Toolkit. And there are things make me confused. After I train my network, the output shape=(None,3,640,480). More specifically, the input dims have 3 channels and 640 x 480 resolution. But the batch size is None. Could anyone have me, what is the meaning of batch size = None? And also the Params# show some values. What actually was it?
I am training on DetectNet_V2 model
“None” means the batch dimension is variable. Any batch size will be accepted.
“Params#” means the total trainable and non-trainable params for this layer.
does it mean that the number of samples that use on training in 1 batch is varied?
and Params mean the weight of the layer?
“None” tells that any batch size will be accepted. Set to None, then the bs is not bounded by a specific number.
Params means each layer’s trainable and non-trainable parameters.
See more from python - Keras model.summary() result - Understanding the # of Parameters - Stack Overflow