I installed TensorFlow 2.8.0 in my Windows 10 PC with CPU only. No Nvidia/CUDA hardware/software installed at all. I run a simple Python image processing project with CNN and I got bunch of no sense messages:
2022-02-23 07:31:55.511878: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'cudart64_110.dll'; dlerror: cudart64_110.dll not found 2022-02-23 07:31:55.512119: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine. 2022-02-23 07:32:16.820716: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'nvcuda.dll'; dlerror: nvcuda.dll not found 2022-02-23 07:32:16.820934: W tensorflow/stream_executor/cuda/cuda_driver.cc:269] failed call to cuInit: UNKNOWN ERROR (303) 2022-02-23 07:32:16.823066: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:169] retrieving CUDA diagnostic information for host: DESKTOP-7OCT2M4 2022-02-23 07:32:16.823285: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:176] hostname: DESKTOP-7OCT2M4 2022-02-23 07:32:16.823612: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX AVX2 To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
Why I'm getting these CUDA messages? I never installed anything for CUDA. I did try the following like everyone recommended in Google:
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3' And still getting these messages. WHY? I really don't want to see any CUDA message because I did not install nothing for it. This is very bad for the TensorFlow deployment team. Why no one test this before in Windows?
I'm waiting for any suggestion to fix this TensorFlow installation issues.