Eugenio Lomurno

icon indicating an email [email protected]

icon location PhD student in Human Centered Artificial Intelligence @ Politecnico di Milano

Results 1 comments of Eugenio Lomurno

ValueError: Dimension size must be evenly divisible by 10 but is 1 for '{{node training/Reshape}} = Reshape[T=DT_FLOAT, Tshape=DT_INT32](loss/mul, training/Reshape/shape)' with input shapes: [], [2] and with input tensors computed as partial shapes: input[1] = [10,?]

I am still having the same issue, moreover, my training accuracy is tucked around 10%. This issue seems to be solved by reducing both batch_size and num_microbatches from 128 to...