Closed flugenheimer closed 5 years ago
Hmm solved it by going back to ubuntu 16.04 with CUDA 9.0 and tensorflow 1.12. I knew that this worked at some point. Issue could be my installation of CUDA, an issue with CUDA 10 and pyopengl or a lot of other things. If I have the time I will try to look a bit more into it.
If anyone else runs into this, it appears to be a video driver issue. I personally ran into this issue while accessing my Linux machine through NoMachine (remote desktop software). It was loading an video driver which I had noticed only supported an older version of OpenGL.
Hey @dfraska-ftl , I faced this issue and somehow manage to resolve this. Following are my configurations: python: 3.6.9 TF-GPU: 1.15.0 PyOpenGL: 3.1.5 PyOpenGL Accelerate: 3.1.5 CUDA: 10.0 I did not had to downgrade PyOpenGL. I was running this inside an ubuntu18.04-based docker container. Therefore in the end, I had to go for headless rendering using EGL. Refer to this for anyone using docker container: https://github.com/DLR-RM/AugmentedAutoencoder/issues/19#issuecomment-522597053
Hey @dfraska-ftl , I faced this issue and somehow manage to resolve this. Following are my configurations: python: 3.6.9 TF-GPU: 1.15.0 PyOpenGL: 3.1.5 PyOpenGL Accelerate: 3.1.5 CUDA: 10.0 I did not had to downgrade PyOpenGL. I was running this inside an ubuntu18.04-based docker container. Therefore in the end, I had to go for headless rendering using EGL. Refer to this for anyone using docker container: #19 (comment)
@saqib1707 I export PYOPENGL_PLATFORM='egl' but Not allowed to force software rendering when API explicitly selects a hardware device.
Hey @dfraska-ftl , I faced this issue and somehow manage to resolve this. Following are my configurations: python: 3.6.9 TF-GPU: 1.15.0 PyOpenGL: 3.1.5 PyOpenGL Accelerate: 3.1.5 CUDA: 10.0 I did not had to downgrade PyOpenGL. I was running this inside an ubuntu18.04-based docker container. Therefore in the end, I had to go for headless rendering using EGL. Refer to this for anyone using docker container: #19 (comment)
Hi, I face the same problem now and think this is because of the driver, so what's the version of your Nvidia driver?
I just tried to install and run the scripts on a new pc but I am running into this error when running the ae_train exp_group/my_autoencoder -d:
File "src/errorchecker.pyx", line 53, in OpenGL_accelerate.errorchecker._ErrorChecker.glCheckError OpenGL.error.GLError: GLError( err = 1282, description = 'invalid operation', baseOperation = glNamedFramebufferTexture, cArguments = ( array([1], dtype=uint32), GL_COLOR_ATTACHMENT0, 1, 0, ) )
I tried using pip to install pyopengl, and also tried it by using the branch directly, which fixed another opegl bug: https://github.com/mcfletch/PyOpenGL
I am using CUDA 10.1 and tensorflow 1.14, and the nvidia 390.87 driver on a Quadro P2000 GPU.