Description
Hello Aaron, first of all thank you for your effort!
I am trying to build the latest tensorflow for my old Intel i7 without Avx.
I followed your instruction and was able to get, at the end of the 'docker build... ' a:
Successfully built 0ac7f2497a59
Successfully tagged tf_build:latest
The only error messages (out of 1,320 output lines were):
warning: /var/cache/yum/x86_64/7/cuda-rhel7-x86_64/packages/cuda-12-2-12.2.2-1.x86_64.rpm: Header V4 RSA/SHA512 Signature, key ID d42d0685: NOKEY
Non-fatal POSTIN scriptlet failure in rpm package 3:nvidia-driver-latest-545.23.08-1.el7.x86_64
but when I run 'docker run...' I get:
docker: Error response from daemon: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: Auto-detected mode as 'legacy'
nvidia-container-cli: initialization error: load library failed: libnvidia-ml.so.1: cannot open shared object file: no such file or directory: unknown.
Any help on how to get unstuck would be great.
Thanks
Luca
Software:
OS: Ubuntu 22.04.3 LTS
Kernel: Linux 6.2.0-37-generic (x86_64)
Python Version: 3.10.12
NVIDIA Driver 545.29.02
CUDA Version 12.3
cudnn Version 8.9.7.29
TensorRT Version 8.6.1 (imports ok in python but not sure it was installed correctly)
Current TensorFlow Version, from community wheels): 2.12.0
Clang Version 16.0.0
Bazelisk / Bazel 6.1.0
Hardware:
Processor: Intel(R) Core(TM) i7 CPU 950 @ 3.07GHz
Instruction Set: 64-bit
Instruction Set Extensions: Intel® SSE4.2 , (NB: NO AVX)
GPU: NVIDIA GeForce RTX 3060 Ti
GPU Engine Specs: NVIDIA CUDA® Cores 4864
RAM: 24 GB
Attachment:
output_of_docker_build.zip