Amazon SageMaker Notebooks - Intro to Jupyter and hands on!

Amazon SageMaker Notebooks are managed Jupyter Notebook servers and a great place to center your machine learning and data science projects.
Soon we'll look at SageMaker Studio, but before we do lets dig into SageMaker Notebooks and how they work.
Follow me:
- LinkedIn: / mikegchambers
- Twitter: / mikegchambers
And if you're interested in my AWS Machine Learning course, take a look here: link.mls-c01.com

[@nexus888]

you have a very good presentation style - keep going with this approach. I hope you get lots of followers!

[@mikegchambers]

Thanks! Me too! :)

[@CloudBart]

Like that live notes/paper setup Mike!

[@mikegchambers]

Thanks Bart! It’s kinda fun! Gonna have to break out the LEGO soon! :)

[@Venkatesh_UV]

Hi, May I request you to please create a separate Playlist for all your AWS SageMaker videos? Your videos are crystal clear and your presentation style helps someone in the industry who's new to the tool.

[@mikegchambers]

Hello. Thanks for the comment. I’m working on the best way to structure content. Your idea seems the logical solution!

[@tibinlukose]

is it available@@mikegchambers

[@AnilKumar-np9mh]

Thank you Mike! Very helpful. You have explained it very well. Loved it.

[@saranrajv3]

I like ur way of explaining - Simple and Precise :) Thanks Mike

[@mikegchambers]

Thanks Saran!

[@jamiewalker9571]

Awesome video. Cheers, Mike.

[@sjsocially]

Very Clear and Clean.....

[@patawee313]

Great channel, wish met this channel sooner. Always stuck with Colab 😅

[@marceli-wac]

Really great content! Thanks for sharing this!

[@brunochagas2500]

Great video. Thanks.

[@earlbovell8965]

Bit late since this was posted a year ago, but can you describe your setup? Your youtube video is unusually crisp and clear. Thanks

[@mikegchambers]

Thanks! Let me know what parts your particularly referring to, I use a Canon80d for video, whatever lights I have around (I have many :)) and I record the screen output with QuickTime. Hope that helps. :)

[@vishalmahesh]

Very nice video. It got me unstuck of the issue I was getting. Can you please create a video of loading data from S3 using Notebook.

[@mikegchambers]

I will try to make sure I include that in a future video. In the meantime the easy way is to use the AWS CLI. Create a code cell, add and exclamation mark ‘!’ to let the cell know you’re using CLI, then issue a command like: ! aws s3 cp -recursive s3://bucket/folder /path/in/notebook/server/ Hope that helps.

[@esitagirlikcimuhendis4593]

Your video is really helpful. Actually I m stuck on using my dataset from s3 bucket. I uploaded images to s3 bucket but I don’t know how I can use tensorflow image data generator to read those images automatically 😶 is there anyone to help me pleaassseee

[@gauravsharmahsetsrhu4139]

Unable to understand why so low count in views...

[@mikegchambers]

Kinda niche I guess. Feel free to share! :)

[@Minifigpriceguide]

Do you have a video on how to connect to MySQL rds instance?

[@mikegchambers]

I don’t. What do you want to know?

[@optionsunleashed6486]

Great - but example fails when I copy paste same code with error during model.predict with a sample image. Input 0 of layer "dense_10" is incompatible with the layer: expected axis -1 of input shape to have value 784, but received input with shape (None, 28) Any ideas? The sample code on TensorFlow does not work with version in sagemaker. conda tensorflow2_p310 - if that helps >> model summary Model: "sequential_5" _________________________________________________________________ Layer (type) Output Shape Param # ================================================================= flatten_5 (Flatten) (None, 784) 0 dense_10 (Dense) (None, 128) 100480 dropout_5 (Dropout) (None, 128) 0 dense_11 (Dense) (None, 10) 1290 ================================================================= Total params: 101770 (397.54 KB) Trainable params: 101770 (397.54 KB) Non-trainable params: 0 (0.00 Byte)

[@optionsunleashed6486]

I think model needs an array of list... image_array= np.array([image]) model.predict(image_array) this works.