What is the problem?

When I run in local mode I see lots of the following message being printed many times such that the console isn't really useable:

INFO worker.py:439 -- ray.get_gpu_ids() currently does not work in LOCAL MODE.

I'm not sure why this message is being printed as I'm not using a GPU (num_gpus=0, although my machine does have a GPU enabled). As I recall some time ago I was running in local mode and found that the GPU was being used when it wasn't supposed to be however. I first see this message printed with code version 9bfc2c4, and so assume it was caused by #7670. I'm not sure why this message is being printed when I shouldn't be using a GPU, and also I think the message should only be printed once and then suppressed in future instances when it would otherwise be printed

Ray version and other system information (Python version, TensorFlow version, OS):
Running on Azure NC6s_v2 machine
ray: 0.9.0.dev0 (latest wheels installed on 4/27/20)
python: 3.7.6
tensorflow: 2.1
torch: 1.4.0
OS: Ubuntu 16.04.6 LTS

Reproduction (REQUIRED)

Please provide a script that can be run to reproduce the issue. The script should have no external library dependencies (i.e., use fake or mock data / environments):

I'm just running the following example, but changing ray.init() to ray.init(local_mode=True)

python examples/custom_env.py

If we cannot run your script, we cannot fix your issue.

• I have verified my script runs in a clean environment and reproduces the issue.
• I have verified the issue also occurs with the latest wheels.