Building, training and deploying machine learning models with Amazon SageMaker (July 2020)

[@sasha786123]

Amazon SageMaker 1.Prepare -- 6:56 2. Build -- 11:59 3. Train & Tune -- 25:44 4. Deloy -- 33:51 5. Demo -- 38:14 6. Links & Conclusion -- 58:41

[@KokkeOP]

Clear and concise. I've always appreciated the similarity between SageMaker and Jupyter notebooks in their UI distribution. Perhaps I'm a nostalgic Data Scientist who prefers a simple canvas to work on. Hehe.

[@vidhyalakshmiparthasarathy8573]

Hi @Julien Simon - Thank you sir for your great videos. I've learnt AWS SageMaker by following your youtube channel. Your explanation and contents are amazing. 👍 😀

[@juliensimonfr]

Great to hear!

[@caiyu538]

Great lectures.

[@imnischaygowda]

In detail. Helpful. Thank you

[@juliensimonfr]

You're welcome!

[@omarpasha2968]

I learned a lot!

[@juliensimonfr]

great, thanks for watching.

[@denisbrinzarei6411]

Julien, could you please share the GIT repo for the project?

[@juliensimonfr]

This should be the one: gitlab.com/juliensimon/awsdevdays2020/-/blob/master/mls1/XGBoost.ipynb

[@NV3009]

Julien, Good one really, I have a quick question. Is Appstream 2.0 can be deployed via SCCM? Can i have the steps if this can be done?

[@juliensimonfr]

I know nothing about Appstream and SCCM.

[@soumyajitganguly2593]

how can we invoke the endpoint from outside the sagemaker studio session? any other video where this is explained?

[@juliensimonfr]

Sure :) The endpoint is a regular HTTPS endpoint. You can HTTP POST to it, or use a Predictor in the SageMaker SDK docs.aws.amazon.com/sagemaker/latest/dg/neo-requests-sdk.html

[@MirrorNeuron]

Hi @Julien, thanks for the video. I am looking for a tutorial where I can call a Sagemaker endpoint with a restAPI and continuously run inference. Where can I find that? Please guide.

[@juliensimonfr]

This should help aws.amazon.com/blogs/machine-learning/creating-a-machine-learning-powered-rest-api-with-amazon-api-gateway-mapping-templates-and-amazon-sagemaker/

[@MirrorNeuron]

@@juliensimonfr Thank you Sir :)