From 3d551df9f5ccec0bcc7ac71ccf0c6bc2c5347bc0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Agust=C3=ADn=20Castro?= Date: Thu, 27 Jun 2024 16:18:08 -0300 Subject: [PATCH] Add comment that this is the develop branch --- docs/challenge.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/challenge.md b/docs/challenge.md index 07dcc7e..11ccf5a 100644 --- a/docs/challenge.md +++ b/docs/challenge.md @@ -124,5 +124,5 @@ The training script `./challenge/train.py` used all the data we have for trainin **REMARKS**: - There are better ways of dealing with this last problem. Instead of training inside the `cd.yml`, we could have a model registry (using for example, MLflow), and then deploying our API with a model we have available in our model registry. - Even though some variables like the project name and the key for my Service Account are treated as secret variables in the repository, I am aware that some variables are still hardcoded (and even the project name which is 'secret', is explicitly mentioned a few times in this markdown documentation just to simplify the exposition). One hardcoded variable is for example `STRESS_URL` in the Makefile, which shouldn't be harcoded. However, I still added a few steps in the `cd.yml` to get the URL to my API when it deploys with gcloud, and uses that URL to run the stress-test. -- After everything worked with the main branch, I modified my `cd.yml` file to handle deployments based on different branches (main, develop, and release/*), by including conditional logic within the deployment steps or create separate jobs for each branch scenario. +- After everything worked with the main branch, I modified my `cd.yml` file to handle deployments based on different branches (main, develop, and release/*), by including conditional logic within the deployment steps or create separate jobs for each branch scenario. The current branch you are seeing corresponds to the develop branch.