Hi ExuberantParrot61 , if it's on the same machine you can use different configuration files. for example, when running the agent you can use --config-file <PATH_TO_USER_CONFIG_FILE>

You have a self hosted server, several agents that are running and two users developing on their own machines but the training is done elsewhere?

What is the setup that they do the training on?

One self hosted server with GPU (training server, with one agent running in docker mode), then me and another user developing on own machines, intending to deploy training jobs to that hosted server.

the self hosted server has an agent running listening to the “default” queue

Hi CostlyOstrich36 it’s on a different machine. Its a central training server. We’d like to somehow configure different git creds for two different users who will access clearml hosted, but develop on their own laptops, but train on that central server. LMK if that makes sense.

ExuberantParrot61 , I'm not sure I understand the entire setup. can you please elaborate?

Closing thread, I just passed the Git credentials via the docker arguments within the experiment (not scalable).