i made env for gpu so i installed ubuntu 18.04, docker , NVIDIA Docker and download docker image tensorflow/tensorflow: latest-gpu-jupyter then i make container follow command on google chrome and firefox
sudo docker run --gpus all -it -p 8888:8888 -p 0.0.0.0:6006:6006 -v "$(pwd)"/Desktop/docker:/tf/docker tensorflow/tensorflow:latest-gpu-jupyter
everything goes well except tensorboard
then i checked tensor board log file and re installed tensorbord ,sadly it does`n work well i userd this command for tensorboard
%reload_ext tensorboard
%tensorboard --logdir {logs_base_dir}
%tensorboard --logdir=./logs/fit/20210308-085859
%tensorboard --logdir=runs
%tensorboard --logdir tf_logs/
but return like this
Reusing TensorBoard on port 6007 (pid 58), started 0:03:41 ago. (Use '!kill 58' to kill it.)
127.0.0.1 refused to connect
i used command "!kill 58" and restart tensorboard but same log returned
then used this command
sudo docker exec -it 9dff3d5802df bash
#in the container
tensorboard --logdir=./logs/fit/20210308-085859
but connection was refused
so i bilud up to local env (not virtual env)
tensorboard goes well
?what wrong ? i hope to use jupyter notebook, docker env for tensorboard