To run unittests it would be great if all our benchmarks in principle can run on CPU.
Right now the logistic regression benchmark does not run on a CPU and crashes with the following error message
TypeError: ('An update must have the same type as the original shared variable (shared_var=W, shared_var.type=TensorType(float32, matrix), update_val=Elemwise{sub,no_inplace}.0, update_val.type=TensorType(float64, matrix)).', 'If the difference is related to the broadcast pattern, you can call the tensor.unbroadcast(var, axis_to_unbroadcast[, ...]) function to remove broadcastable dimensions.') F.
To run on CPU, Logreg requires
theano.floatX = "float64"
whereas the convolutional network and the fully connected network need
theano.floatX = "float32"
To run unittests it would be great if all our benchmarks in principle can run on CPU.
Right now the logistic regression benchmark does not run on a CPU and crashes with the following error message
TypeError: ('An update must have the same type as the original shared variable (shared_var=W, shared_var.type=TensorType(float32, matrix), update_val=Elemwise{sub,no_inplace}.0, update_val.type=TensorType(float64, matrix)).', 'If the difference is related to the broadcast pattern, you can call the tensor.unbroadcast(var, axis_to_unbroadcast[, ...]) function to remove broadcastable dimensions.') F.
@aaronkl: Can this be fixed?