Exception: ("We can't determine the cudnn version as it is not available",

[lawlite19]

I have installed cuda and cudnn, and I run THEANO_FLAGS='device=cpu, floatX=float32, nvcc.fastmath=True' /usr/bin/python iGAN_main.py --model_name outdoor_64, the error message is:

Traceback (most recent call last):
  File "iGAN_main.py", line 41, in <module>
    model = model_class.Model(model_name=args.model_name, model_file=args.model_file)
  File "/root/iGAN/model_def/dcgan_theano.py", line 27, in __init__
    self._gen = self.def_gen(self.gen_params, self.gen_batchnorm, n_layers=self.n_layers, n_f=self.n_f, nc=self.nc)
  File "/root/iGAN/model_def/dcgan_theano.py", line 40, in def_gen
    gx = gen_test(z, gen_params, gen_pl, n_layers=n_layers, n_f=n_f, nc=nc, use_tanh=False)
  File "/root/iGAN/model_def/dcgan_theano.py", line 227, in gen_test
    hout = relu(batchnorm(deconv(hin, w, subsample=(2, 2), border_mode=(2, 2)), u=u, s=s, g=g, b=b))
  File "/root/iGAN/lib/ops.py", line 95, in deconv
    d_img = GpuDnnConvGradI()(kerns, img, out, desc)
  File "/usr/local/lib/python2.7/dist-packages/theano/sandbox/cuda/dnn.py", line 838, in __init__
    if version() < (4000, 4000):
  File "/usr/local/lib/python2.7/dist-packages/theano/sandbox/cuda/__init__.py", line 407, in dnn_version
    dnn_available.msg)
Exception: ("We can't determine the cudnn version as it is not available", "Can not compile with cuDNN. We got this error:\nnvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).\n/usr/bin/ld: cannot find -lcudnn\ncollect2: error: ld returned 1 exit status\n")

Theano (0.8.2) cuda (8.0) cudnn(5.1)

envionment variable:

Thank you!

[lawlite19]

When I run: THEANO_FLAGS='device=gpu0, floatX=float32, nvcc.fastmath=True' /usr/bin/python iGAN_main.py --model_name outdoor_64

Using gpu device 0: Tesla P100-PCIE-16GB (CNMeM is enabled with initial size: 50.0% of memory, cuDNN not available)
[n_iters] = 40
[top_k] = 16
[average] = False
[morph_steps] = 16
[model_file] = ./models/outdoor_64.dcgan_theano
[d_weight] = 0.0
[interp] = linear
[batch_size] = 64
[framework] = theano
[model_type] = dcgan_theano
[shadow] = False
[model_name] = outdoor_64
[win_size] = 384
LOADING...
('n_layers=', 3)
load model from ./models/outdoor_64.dcgan_theano
3.81 seconds to load theano models
Traceback (most recent call last):
  File "iGAN_main.py", line 41, in <module>
    model = model_class.Model(model_name=args.model_name, model_file=args.model_file)
  File "/root/iGAN/model_def/dcgan_theano.py", line 27, in __init__
    self._gen = self.def_gen(self.gen_params, self.gen_batchnorm, n_layers=self.n_layers, n_f=self.n_f, nc=self.nc)
  File "/root/iGAN/model_def/dcgan_theano.py", line 40, in def_gen
    gx = gen_test(z, gen_params, gen_pl, n_layers=n_layers, n_f=n_f, nc=nc, use_tanh=False)
  File "/root/iGAN/model_def/dcgan_theano.py", line 227, in gen_test
    hout = relu(batchnorm(deconv(hin, w, subsample=(2, 2), border_mode=(2, 2)), u=u, s=s, g=g, b=b))
  File "/root/iGAN/lib/ops.py", line 95, in deconv
    d_img = GpuDnnConvGradI()(kerns, img, out, desc)
  File "/usr/local/lib/python2.7/dist-packages/theano/sandbox/cuda/dnn.py", line 838, in __init__
    if version() < (4000, 4000):
  File "/usr/local/lib/python2.7/dist-packages/theano/sandbox/cuda/__init__.py", line 407, in dnn_version
    dnn_available.msg)
Exception: ("We can't determine the cudnn version as it is not available", "Can not compile with cuDNN. We got this error:\nnvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).\n/usr/bin/ld: cannot find -lcudnn\ncollect2: error: ld returned 1 exit status\n")

[junyanz]

The code is only tested with an older Theano version as well as old GPUs and CUDNN. Could you run other Theano demo code such as this one?

[alanhzh]

mine is similar, I am on windows and i have this problem:

[model_file] = ./models/outdoor_64.dcgan_theano
[shadow] = False
[top_k] = 16
[n_iters] = 40
[framework] = theano
[interp] = linear
[win_size] = 384
[model_type] = dcgan_theano
[morph_steps] = 16
[batch_size] = 64
[model_name] = outdoor_64
[d_weight] = 0.0
[average] = False
LOADING...
n_layers= 3
load model from ./models/outdoor_64.dcgan_theano
3.09 seconds to load theano models
Traceback (most recent call last):
  File "iGAN_main.py", line 41, in <module>
    model = model_class.Model(model_name=args.model_name, model_file=args.model_file)
  File "C:\Users\Alan\Desktop\gits\iGAN\model_def\dcgan_theano.py", line 27, in __init__
    self._gen = self.def_gen(self.gen_params, self.gen_batchnorm, n_layers=self.n_layers, n_f=self.n_f, nc=self.nc)
  File "C:\Users\Alan\Desktop\gits\iGAN\model_def\dcgan_theano.py", line 40, in def_gen
    gx = gen_test(z, gen_params, gen_pl, n_layers=n_layers, n_f=n_f, nc=nc, use_tanh=False)
  File "C:\Users\Alan\Desktop\gits\iGAN\model_def\dcgan_theano.py", line 227, in gen_test
    hout = relu(batchnorm(deconv(hin, w, subsample=(2, 2), border_mode=(2, 2)), u=u, s=s, g=g, b=b))
  File "C:\Users\Alan\Desktop\gits\iGAN\lib\ops.py", line 95, in deconv
    d_img = GpuDnnConvGradI()(kerns, img, out, desc)
  File "C:\ProgramData\Anaconda3\lib\site-packages\theano\sandbox\cuda\dnn.py", line 859, in __init__
    if version() < (5000, 5000):
  File "C:\ProgramData\Anaconda3\lib\site-packages\theano\sandbox\cuda\__init__.py", line 424, in dnn_version
    dnn_available.msg)
Exception: ("We can't determine the cudnn version as it is not available", 'CUDA not available')

[junyanz]

The code was developed around 2015. I am pretty sure it doesn't work with cudnn 5.1. See here for the Theano version.