Skip to content

Latest commit

 

History

History
46 lines (31 loc) · 2.05 KB

new_joiner.md

File metadata and controls

46 lines (31 loc) · 2.05 KB

Git access

Model and model server code are hosted on github. Please join https://github.com/AIforGoodSimulator organization by first signing up to github and then contacting Kostya Titorenko or Peter Hung with your user name.

Note if you are admin to add new user go to https://github.com/settings/organizations.

https://github.com/AIforGoodSimulator/model-server repository contains model server and migrated models. See readme.md there for details.

Task board

Not started / in progress tasks can be seen here https://github.com/orgs/AIforGoodSimulator/projects/1

You would need to be member of DB team to see tasks there, if you can not see tasks contact person who setup github access for you.

Communication

We use Slack workspace https://aiforgood-simulator.slack.com/x-p1063252470320-1167393800146-1187864846674/ to communicate. Check out #db and #general channels to start with once you join.

Use join link https://join.slack.com/t/aiforgood-simulator/shared_invite/zt-e2zp3l12-yBmAEHKgj35Ld2cbZf7cTQ to join.

Our version of Slack does not have screensharing, you can use private webex to have screensharing session.

Azure

Model server production deployment is on Azure. Contact Stefan Cronjaeger and Andrei Varanovich on slack or email to get access. There is a known bug with Azure sandbox we use where you can not use existing Microsoft account to connect to Azure. Setup virgin email address and send that to Stefan/Andrei and be careful to never create account, instead always use invite link.

To use az console you would need use service principal and associated password, contact Kostya Titorenko on slack to request one.

Model server is setup to deploy on commit, see https://github.com/AIforGoodSimulator/model-server/blob/master/README.md for details how to deploy.

High level plan

Our minimal aim should be to migrate all existing useful models to model server and ensure these can be used by target audience on the ground. We already have compartmental model migrated, next up are network and agent models as well as some infrastructure work.