ValueError: You are tring to use the old GPU back-end.

[speed8928]

Hi I use windows 10 to implement the project and it turns out the error below: ValueError: You are tring to use the old GPU back-end. It was removed from Theano. Use device=cuda*

I tried to follow your link that from error https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28gpuarray%29 and install theano pygpu, but nothing good come out.

I also tried change the Theano variable device = gpu to device = cuda0 and it gives me the error below:

(py3) C:\Users\speed\source\3D-R2N2>python demo.py prediction.obj WARNING (theano.tensor.blas): Using NumPy C-API based implementation for BLAS functions. ERROR (theano.gpuarray): pygpu was configured but could not be imported or is too old (version 0.7 or higher required) NoneType Downloading a pretrained model Traceback (most recent call last): File "demo.py", line 84, in main() File "demo.py", line 55, in main download_model(DEFAULT_WEIGHTS) File "demo.py", line 34, in download_model '--create-dirs', '-o', fn]) File "C:\Users\speed\AppData\Local\Programs\Python\Python35\Lib\subprocess.py", line 557, in call with Popen(*popenargs, **kwargs) as p: File "C:\Users\speed\AppData\Local\Programs\Python\Python35\Lib\subprocess.py", line 947, in init restore_signals, start_new_session) File "C:\Users\speed\AppData\Local\Programs\Python\Python35\Lib\subprocess.py", line 1224, in _execute_child startupinfo) FileNotFoundError: [WinError 2] The system cannot find the file specified

Could you please advise me the way to compile it, thanks.

[SJTUPanda]

Hi, I have the same error with you. Have you solved this problem?

[speed8928]

I lowered the Theano version 7.x then it worked.

[SJTUPanda]

Thanks! I use the Theano version 0.8.0 then it works.

[CloudChoSony]

I found another way to solve at ValueError «You are trying to use the old GPU back-end» when importing keras According to cgl, please type in Linux Terminal "export THEANO_FLAGS=mode=FAST_RUN,device=cuda,floatX=float32"

[ShubhayanS]

Thanks! I use the Theano version 0.8.0 then it works.

I am getting - AttributeError: ('This name is already taken', 'floatX')