Hi Data G! It's taken me a while, but if you're still looking for things on SageMaker Feature Store, this video just went live: ruclips.net/video/BP0_QYRN8zU/видео.html. Enjoy! :)
Thanks for doing this ! Prevented me from setting up and configuring an EC2 instance for my ML project. I don't have time to "RTFM" so your videos have allowed me to kick-start my effort. Keep up the good work!
The Studio console may be recently changed, but changed a lot ! Can't find the Inference, Endpoints sections in the page. Do you have latest video of latest Studio UI ?
Yeah, the UI changes a lot...it's hard to keep up! 😊 I don't have any more recent videos, but they're on my list for future videos! In the meantime, the best bet is probably the AWS documentation: docs.aws.amazon.com/sagemaker/latest/dg/studio-ui.html
Thank you very much! I tried another tutorial where I launched the Domain using the "Standard Setup" approach, but when I was ready to launch Studio I got hit with an error alert indicating the subnet I selected was not supported by my availability zone... I'll look into that later but went ahead and I deleted that Domain (I learned a lot about VPCs, subnets, CIDRs, etc. which was nice but I needed to get into Studio). I found your video, and went with the "Quick Setup" approach and that worked. I am working on a POC to better understand feature engineering capabilities, data wrangler, and the feature store in SageMaker - I haven't finished this video but will browse your videos to see if there's anything on those specific topics. Thanks again!
Hi sampathkumar! To learn AWS SageMaker, it helps if you're already familiar with the core AWS services (EC2, S3, IAM, etc.). And knowing some basic Python/Jupyter Notebooks will also help too. But to actually get set up and start using it, you can just follow the instructions here: docs.aws.amazon.com/sagemaker/latest/dg/gs.html. There are several "JumpStarts" that you can walk through to try things out. Hope that helps! :)
@@TinyTechnicalTutorials thanks for the response ☺️ I’m familiar with core AWS services. So, I need to learn about Jupyter notebooks. For that, do I need to learn python ?
I have setup EC2 on free tier Account earlier Trying to setup sagemaker studio It’s asking for domain name Not sure if they charge for this And what name shoul we give as it is not accepting any name I give
Hi Sunitha! 😊 The SageMaker domain is mostly for configuration (of things like users, security, network, etc.), but it does also contain an Elastic File System (EFS) volume. If you're on the Free Tier still, there should be no charge for that, but just FYI. As far as naming your domain, I can't find the exact naming conventions allowed, but you can't use spaces or special characters (although you can use hyphens and underscores). But as long as you follow those rules, you should be able to choose whatever name you want. Then after you create the domain, you can set up SageMaker Studio. SageMaker Studio also has some usage available on the Free Tier so you're probably good there too: aws.amazon.com/sagemaker/pricing/. But to clarify, you don't need to set up a separate EC2 instance in order to use SageMaker Studio. SageMaker Studio will create EC2 instances automatically when you create new notebooks. Hope that helps! 😊
One more question 🙋♀️ I was trying to xgboost example given in tutorial. While we get only ml.t3.medium, on free trial version, we require atleast ml. m4.xlarge. Even that got exceeded and I couldn’t run while excitement.🙃 Do I need to go for request or we have a limit per day. Thanks in advance 🙏
Hmmm...which tutorial are you using? By default, most accounts won't have the ml.m4.xlarge instance type enabled (meaning the service limit or quota is set to 0). That's because it can be quite expensive, especially if you're just using it to learn with. 😊 But for future reference, you can request to increase the quota by navigating to "Service quotas" (from the top menu). Then on the left-hand navigation, go to "AWS services." Type in "SageMaker" and select that. Then type in "ml.m4.xlarge" and select "ml.m4.xlarge for notebook instance usage." Then you'll need to "Request quota increase" and say that you want 1. That sends a ticket to the AWS Support team. They'll review it and make a decision on whether to increase the quota. Depending on the type of support plan you have, it can take a few days to get approval. But FYI, I've found that if you have a new account, sometimes they won't approve increases like this. So you might want to find a tutorial that uses a smaller instance type! 😊
Hey DAMBY! Not at all a stupid question...it can be a little hard to keep track of all the different SageMaker things! :) Studio and Canvas are two separate products and are meant for two different audiences: -Canvas is for business analysts/data analysts (i.e., not developers/data scientists/ML experts). The Canvas tool is visual and doesn't require any code. -Studio IS for the developers/data scientists/ML experts, who want to dig into code and have a lot of control over the whole ML lifecycle. It's the "big guns" part of the SageMaker family. Behind the scenes, they both leverage the same technology to prepare data, train the models and make predictions. It's just that the UIs/audiences are different. But to answer your question directly...no, if you use Studio, you're not also starting a Canvas. Hope that helps?? :)
![Noob To Max With DRAGON REWORK In Blox Fruits [FULL MOVIE]](http://i.ytimg.com/vi/LnBMOinoOvA/mqdefault.jpg)
What else do you want to learn about AWS? Let me know below in the comments!
Hi there! Anything on feature engineering capabilities, data wrangler, and the feature store in SageMaker?! Great channel. A lot to learn! Subscribed.
Welcome to the channel, Data G, and thanks for the comments! :) I'll add these to my list...I definitely plan to do more videos on SageMaker.
Hey Data G! Just released a video about SageMaker Data Wrangler: ruclips.net/video/tB0WrVlYhc4/видео.html. Enjoy! :)
Hi Data G! It's taken me a while, but if you're still looking for things on SageMaker Feature Store, this video just went live: ruclips.net/video/BP0_QYRN8zU/видео.html. Enjoy! :)
Thanks for doing this ! Prevented me from setting up and configuring an EC2 instance for my ML project. I don't have time to "RTFM" so your videos have allowed me to kick-start my effort. Keep up the good work!
Thanks for such a nice comment! Glad it helped, and good luck! :)
Amazing! We'll need an updated tutorial on SageMaker Studio 😁 the new version and Ui seem pretty awesome. Thank you for putting these videos up!
Thanks, Miguel!! Yes, the SageMaker videos are getting out of date...all the UI updates make it hard to keep up! It's on my list! 🤓🙏
The Studio console may be recently changed, but changed a lot ! Can't find the Inference, Endpoints sections in the page. Do you have latest video of latest Studio UI ?
Yeah, the UI changes a lot...it's hard to keep up! 😊 I don't have any more recent videos, but they're on my list for future videos! In the meantime, the best bet is probably the AWS documentation: docs.aws.amazon.com/sagemaker/latest/dg/studio-ui.html
Any update on the latest video ?
Hey @abhis3kh! 👋 I'll definitely be doing more SageMaker videos, but it may be a few months! Have to find the time. 🤓
Thank you very much! I tried another tutorial where I launched the Domain using the "Standard Setup" approach, but when I was ready to launch Studio I got hit with an error alert indicating the subnet I selected was not supported by my availability zone... I'll look into that later but went ahead and I deleted that Domain (I learned a lot about VPCs, subnets, CIDRs, etc. which was nice but I needed to get into Studio). I found your video, and went with the "Quick Setup" approach and that worked. I am working on a POC to better understand feature engineering capabilities, data wrangler, and the feature store in SageMaker - I haven't finished this video but will browse your videos to see if there's anything on those specific topics. Thanks again!
Awesome! Glad you were able to get it working with Quick Setup!
May I know what are the prerequisites to learn Aws sagemaker and work as an infrastructure engineer?
Hi sampathkumar! To learn AWS SageMaker, it helps if you're already familiar with the core AWS services (EC2, S3, IAM, etc.). And knowing some basic Python/Jupyter Notebooks will also help too. But to actually get set up and start using it, you can just follow the instructions here: docs.aws.amazon.com/sagemaker/latest/dg/gs.html. There are several "JumpStarts" that you can walk through to try things out. Hope that helps! :)
@@TinyTechnicalTutorials thanks for the response ☺️
I’m familiar with core AWS services. So, I need to learn about Jupyter notebooks.
For that, do I need to learn python ?
@@sampathkumarbasa4774 yes , python will help.
For new SageMaker Studio > we will need to go to Studio Classic > start a Studio Classic space (create if there is none) > run and launch the studio.
Thanks for posting this! 👍🤓
Right. But how to start it?
Hi Radoslaw! 😊 Can you clarify? Start SageMaker Studio?
Great Job! Thanks
Thanks so much!! 🙏😊🌟
Did you find Hadwriting solution from begining of video? I couldn't find it. Thanks.
I have setup EC2 on free tier Account earlier
Trying to setup sagemaker studio
It’s asking for domain name
Not sure if they charge for this
And what name shoul we give as it is not accepting any name I give
Hi Sunitha! 😊 The SageMaker domain is mostly for configuration (of things like users, security, network, etc.), but it does also contain an Elastic File System (EFS) volume. If you're on the Free Tier still, there should be no charge for that, but just FYI. As far as naming your domain, I can't find the exact naming conventions allowed, but you can't use spaces or special characters (although you can use hyphens and underscores). But as long as you follow those rules, you should be able to choose whatever name you want.
Then after you create the domain, you can set up SageMaker Studio. SageMaker Studio also has some usage available on the Free Tier so you're probably good there too: aws.amazon.com/sagemaker/pricing/.
But to clarify, you don't need to set up a separate EC2 instance in order to use SageMaker Studio. SageMaker Studio will create EC2 instances automatically when you create new notebooks.
Hope that helps! 😊
@@TinyTechnicalTutorials Thanks a lot for your response😊
It’s working now 👍
Yay! 😎
One more question 🙋♀️
I was trying to xgboost example given in tutorial.
While we get only ml.t3.medium, on free trial version, we require atleast ml. m4.xlarge.
Even that got exceeded and I couldn’t run while excitement.🙃
Do I need to go for request or we have a limit per day.
Thanks in advance 🙏
Hmmm...which tutorial are you using? By default, most accounts won't have the ml.m4.xlarge instance type enabled (meaning the service limit or quota is set to 0). That's because it can be quite expensive, especially if you're just using it to learn with. 😊
But for future reference, you can request to increase the quota by navigating to "Service quotas" (from the top menu). Then on the left-hand navigation, go to "AWS services." Type in "SageMaker" and select that. Then type in "ml.m4.xlarge" and select "ml.m4.xlarge for notebook instance usage." Then you'll need to "Request quota increase" and say that you want 1. That sends a ticket to the AWS Support team. They'll review it and make a decision on whether to increase the quota. Depending on the type of support plan you have, it can take a few days to get approval. But FYI, I've found that if you have a new account, sometimes they won't approve increases like this. So you might want to find a tutorial that uses a smaller instance type! 😊
Hello,
Can I ask you a (stupid) question?
Whats the difference between Studio and Canvas in Sagemaker? If I use Studio, am I also starting a Canvas?
Hey DAMBY! Not at all a stupid question...it can be a little hard to keep track of all the different SageMaker things! :)
Studio and Canvas are two separate products and are meant for two different audiences:
-Canvas is for business analysts/data analysts (i.e., not developers/data scientists/ML experts). The Canvas tool is visual and doesn't require any code.
-Studio IS for the developers/data scientists/ML experts, who want to dig into code and have a lot of control over the whole ML lifecycle. It's the "big guns" part of the SageMaker family.
Behind the scenes, they both leverage the same technology to prepare data, train the models and make predictions. It's just that the UIs/audiences are different. But to answer your question directly...no, if you use Studio, you're not also starting a Canvas.
Hope that helps?? :)
@@TinyTechnicalTutorials thank you. Really
Great🎉
Thank you! Cheers! 🙌