Vincent Delbar

Hi @viniciuspg , You can use either PowerShell or cmd.exe to execute docker pull, create or run commands. Then I believe the easiest way would be to create a persistent...

Hi @remicres , may be this repo could be useful : https://github.com/RustingSword/tensorboard_logger

Hi @FerdinandKlingenberg, may be you should try with TF 2.4.3 and CUDA 11.0.3 (with TF_CUDA_COMPUTE_CAPABILITIES=7.0) ? Sometimes, the latest version is buggy... And you only need CUDA 11.2 / TF...

May be insufficient memory space (https://github.com/tensorflow/models/issues/3647 )... You can decrease number of concurrent process (for example `bazel build -j 8`). Could you provide more information regarding your build ? System...

Looking at the bazel output, you're building with 8 cores. So you probably need more RAM, or less cores. But to be sure, you should read this file "'/home/yla18/.cache/bazel/_bazel_yla18/d4ef3948567f7f7d65fbd5757d6da7fb/server/jvm.out". Also...

You should avoid building from the TF master branch. Just checkout a release `git checkout v2.4.2` for CUDA==11.0 or `git checkout v2.5.0` for CUDA==11.2. Then just omit the `--noincompatible_do_not_split_linking_cmdline` option....

Yes you can use TF without GPU / CUDA. Bazel is kind of messy, it could be some dependency version mismatch or something else related to your env / VM....

Which bazel version are you running ? If you're following this file https://github.com/remicres/otbtf/blob/develop/doc/HOWTOBUILD.md you should `git checkout v2.1.4` because TF installation steps have changed a bit. Make sure you're running...

> Thank you for your quick reply. I used the bazel 3.7.2. In that case you should checkout v2.5.0. For TF 2.4.2 the required bazel version is 3.1.0.

``` docker pull mdl4eo/otbtf2.4:cpu docker run -it mdl4eo/otbtf2.4:cpu ``` Check this file : https://github.com/remicres/otbtf/blob/develop/doc/DOCKERUSE.md