An error ocurred while starting the kernel

Description of your problem

My code was running fine in the beginning, but after maybe 20 seconds, it shows following errors:

2017󈚰󈚲 12:55:35.692732: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations. 2017󈚰󈚲 12:55:35.692779: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations. 2017󈚰󈚲 12:55:35.692785: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations. 2017󈚰󈚲 12:55:35.692789: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations. 2017󈚰󈚲 12:55:35.692793: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations. 2017󈚰󈚲 12:55:35.777783: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (𔂫), but there must be at least one NUMA node, so returning NUMA node zero 2017󈚰󈚲 12:55:35.778333: I tensorflow/core/common_runtime/gpu/gpu_device.cc:955] Found device 0 with properties: name: GeForce GTX 950M major: 5 minor: 0 memoryClockRate (GHz) 1.124 pciBusID 0000:0a:00.0 Total memory: 3.95GiB Free memory: 3.67GiB 2017󈚰󈚲 12:55:35.778357: I tensorflow/core/common_runtime/gpu/gpu_device.cc:976] DMA: 0 2017󈚰󈚲 12:55:35.778371: I tensorflow/core/common_runtime/gpu/gpu_device.cc:986] 0: Y 2017󈚰󈚲 12:55:35.778382: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1045] Creating TensorFlow device (/gpu:0) ‑> (device: 0, name: GeForce GTX 950M, pci bus id: 0000:0a:00.0) 2017󈚰󈚲 12:55:40.437879: I tensorflow/stream_executor/dso_loader.cc:129] Couldn’t open CUDA library libcupti.so.8.0. LD_LIBRARY_PATH: 2017󈚰󈚲 12:55:40.437937: F ./tensorflow/stream_executor/lib/statusor.h:212] Non‑OK‑status: status_ status: Failed precondition: could not dlopen DSO: libcupti.so.8.0; dlerror: libcupti.so.8.0: cannot open shared object file: No such file or directory What steps will reproduce the problem?

1. Open spyder from Anaconda Navigator
2. Run my code, in the beginning, it was running fine, but after maybe 20 seconds, it shows these errors

What is the expected output? What do you see instead?

Please provide any additional information below

Versions and main components

• Spyder Version:3.2.4
• Python Version:3.5
• Qt Version:5.6.2
• PyQt Version:5.6.0
• Operating system: Ubuntu 16.04

Dependencies

Please go to the menu entry Help > Optional Dependencies (or Help > Dependencies), press the button Copy to clipboard and paste the contents below: IPython >=4.0 : 6.2.1 (OK) cython >=0.21 : None (NOK) jedi >=0.9.0 : 0.11.0 (OK) nbconvert >=4.0 : 5.3.1 (OK) numpy >=1.7 : 1.13.3 (OK) pandas >=0.13.1 : 0.21.0 (OK) psutil >=0.3 : 5.4.1 (OK) pycodestyle >=2.3: 2.3.1 (OK) pyflakes >=0.6.0 : 1.6.0 (OK) pygments >=2.0 : 2.2.0 (OK) pylint >=0.25 : 1.7.4 (OK) qtconsole >=4.2.0: 4.3.1 (OK) rope >=0.9.4 : 0.10.7 (OK) sphinx >=0.6.6 : 1.6.3 (OK) sympy >=0.7.3 : None (NOK)

[estathop]

the case is I loaded a keras model which occupied my whole GPU Memory, then used

cuda.select_device(0)
cuda.close()

in order to try to release the GPU memory and give it to a pytorch model I overcame this problem with a naive way

[CAM-Gerlach]

Okay, thanks for clarifying further and glad you found a solution!

Just to be clear, Spyder has no control over your GPU memory or (to our knowledge) how Keras and Pytorch use it, as you seem to have figured out.