TypeError: Cannot convert Type TensorType(float32, matrix) (of Variable Subtensor{int64:int64:}.0) into Type TensorType(float64, matrix). You can try to manually convert Subtensor{int64:int64:}.0 into a TensorType(float64, matrix). about made

[mgermain]

3 things from the top of my head.

1. Delete experiments/ef0f220f13f115d34798a5f7e16e8c539e4eaee42564d07ed8f893b7fadaa8a0
2. Make sure your theano is in float32.
3. You might take a look at the head which has been updated to be compatible with he new version and backend of Theano.

Should not be necessary be here's my ~/.theanorc:

[global]
device=cuda
#device=gpu
floatX=float32
exception_verbosity=high
#lib.cnmem=0.85
gpuarray.preallocate=0.85

[nvcc]
fastmath = True

The commented parts are for the old backend.

[hatimuradin]

I appreciate your help, my foat32 problem solved by configuring theanorc file.

[mgermain]

Great.