Yes. I am investigating that route now.
Hi @<1576381444509405184:profile|ManiacalLizard2> , can you please elaborate more on your specific use case? And yes, ClearML supports working only with a specific user currently. What do you have in mind to expand this?
@<1523701070390366208:profile|CostlyOstrich36>
So we have 3 python package, store in github.com
On the dev machine, the datascientist (DS) will add the local ssh key to his github account as authorized ssh keys, account level.
With the DS can run git clone git@github.com:org/repo1
then install that python package via pip install -e .
Do that for all 3 python packages, each in its own repo1
, repo2
and repo3
. All 3 can be clone using the same key that the DS added to his account.
The DS run a training, clearml log the "Installed package" as:
[...]
-e
...
-e
...
-e
...
Now we want to reproduce the run inside another dedicated PC, running clearml-agent
This dedicated PC is not own by a single person. We don't want to put the DS personal ssh key in there. Especially, with that key, you can mess-up all repos that the DS have access to.
Ideally, we want to use a SSH Deploy Key that have readonly permission and accessibilty to restricted number of repo. Problem: ssh deploy only allow 1 single repo. We have 3 repos ... which mean that we would need 3 ssh keys ... but so far, I can only see clearml-agent "supporting" single ssh key, which is the default ssh key ...
Or is there a better way to handle this use case ?
@<1576381444509405184:profile|ManiacalLizard2> can you perhaps use user/PAT in that agent's setup? All users can still use their SSH keys, but only the agent will use user/PAT to check out these?