Marcos Reinan de Assis Conceição
Marcos Reinan de Assis Conceição
> I'm using the latest version of tensorflow (2.3.0) with python 3.6.10 and cuda 10.1 and facing the same issue as well in Ubuntu 18.04. export CUDA_VISIBLE_DEVICES=0 or export CUDA_VISIBLE_DEVICES=''...
> Limiting the GPU memory solves this issue for me: > > ```python > import tensorflow as tf > > MEMORY_LIMIT = 1024 > gpus = tf.config.experimental.list_physical_devices('GPU') > if gpus:...
Okey, downgrading tensorflow to version 2.2 did it. ``` pip install --force-reinstall tensorflow-gpu==2.2 ``` Also, if you have ever used `pip install` with `--ignore-installed` to install tensorflow versions or dependencies,...
For some reason, returning the old task instead of the modified one (as in the wiki) is a current workaround: ```sh #!/usr/bin/env sh read old_task read new_task # workaround for...
@smemsh Exactly.
> nothing in the python interface changed that would have caused this - must be due to the hdf5 version that is linked. Can you run the 'checkversion.py' script the...
I installed it with pip (`pip install netcdf4`).
I just tested this PR and it worked fine on my cluster (training on 12 GPUs).