ML Workbench
ML Workbench
  • 9
  • 28 985
Enriching Prompts with Assumptions using Amazon Bedrock
In this video, I will walk through a prompt template that I use to make user input more explicit when passing data to an LLM for Generative AI use-cases.
To see the example prompt used in this video:
1. github.com/chaeAclark/literate-eureka/tree/main/demos/contextual-prompts
For information on my Prompt Development application:
1. github.com/chaeAclark/literate-eureka/tree/main/demos/prompt-library
2. th-cam.com/video/ARV2JaDhmVY/w-d-xo.html
#Bedrock #GenerativeAI #GenAI #AWS #PromptEngineering
มุมมอง: 145

วีดีโอ

Prompt Engineering for Generative AI Applications with Amazon Bedrock
มุมมอง 1815 หลายเดือนก่อน
This is a walkthrough of a tool I commonly use when building prompts for generative AI applications. Code Repository: github.com/chaeAclark/literate-eureka/tree/main/demos/prompt-library If you are interested in more projects like this, please let us know in the comments.
Using Streamlit in Amazon SageMaker (Notebooks, Studio, JupyterLab)
มุมมอง 3.2K9 หลายเดือนก่อน
In this video, I will quickly take you through using streamlit to stand up applications in Amazon SageMaker. We will do this for: 1. SageMaker Studio Classic 2. SageMaker Studio JupyterLab 3. and SageMaker Notebooks Chapters: 00:00 Intro 00:36 SageMaker Studio Classic 05:24 SageMaker Studio JupyterLab 07:47 SageMaker Notebooks
Understanding Research Papers with Generative AI and Amazon Bedrock
มุมมอง 19710 หลายเดือนก่อน
In this video I will show you how to use AWS's Generative AI service, Amazon Bedrock, as well as Streamlit to create an AI Research Assistant that allows you to upload PDFs and produce detailed and informative summaries, code examples, and podcast episodes. Specifically, we will: 1. Demonstrate how the application works and the services used, 2. Walkthrough the implementation for generating the...
Using Amazon Kendra to Build your Enterprise Product Search
มุมมอง 15Kปีที่แล้ว
In this video, we'll use Amazon Kendra build a product search capability. This includes: 1. Downloading and processing an e-commerce dataset 2. Creating a Kendra index 3. Connecting a data source 4. Tuning our search results 5. and finally, using the python SDK to programmatically query our index. Links: The notebooks I used in this tutorial is available here: github.com/chaeAclark/literate-eur...
SageMaker GroundTruth Hands-On Tutorial: Labeling Images and Text
มุมมอง 8Kปีที่แล้ว
In this video, we give a quick tutorial using SageMaker GroundTruth to label both images and sensor logs (CSV). In this tutorial we will: 1. Upload data 2. Create a private workgroup 3. Label a single-class dataset 4. Label a multi-class dataset If there are any other tutorials that would be helpful, please comment below (We read them all). Links: Automated Labeling Information: aws.amazon.com/...
Using Amazon Rekognition to Baseline Classification Performance
มุมมอง 219ปีที่แล้ว
In this video we're going to take you through a quick tutorial using Amazon Rekognition to train and evaluate a baseline image classification model. In this tutorial, I will run through: 1. Creating a project and loading your data into Rekognition 2. And training a custom model and evaluating on an evaluation data set The dataset used for this is a collection of white blood cell images covering...
Advanced Import and Export Options with SageMaker Data Wrangler
มุมมอง 748ปีที่แล้ว
In this video, I’m going to take you through a few of the more advanced import and export options available with SageMaker Data Wrangler. In this tutorial, I will quickly run through: 1. Importing and concatenating multiple datasets 2. Setting multiple data destinations 3. And exporting data to Sagemaker FeatureStore If there are any other tutorials that would be helpful, please comment below (...
SageMaker Data Wrangler Hands-On Tutorial
มุมมอง 2.2Kปีที่แล้ว
This hands-on tutorial uses a stroke prediction dataset from Kaggle and processes it using Amazon SageMaker Data Wrangler. This is Amazon’s low-code/no-code solution for pre-processing datasets for machine Learning tasks. In this video, we will: 1. Create a Data Wrangler Flow 2. Load a dataset into Data Wrangler 3. Perform data exploration 4. Use Data Wrangler’s built-in feature transformations...

ความคิดเห็น

  • @esjuan8953
    @esjuan8953 หลายเดือนก่อน

    How do you know what is malicious and what's not

  • @mattd2185j
    @mattd2185j หลายเดือนก่อน

    Awesome!

  • @adarshreddy0099
    @adarshreddy0099 หลายเดือนก่อน

    Brrooo. Crazy. Even though the UI is different now this video is very helpful. Thank you

  • @toppu10animes
    @toppu10animes 3 หลายเดือนก่อน

    Ok, so what do you do after this?

  • @gerardorosiles8918
    @gerardorosiles8918 3 หลายเดือนก่อน

    How does Kendra know what to index without providing any information about the contents of the JSON files?

  • @cohenbenr
    @cohenbenr 5 หลายเดือนก่อน

    Great video. Have you tried this with CodeEditor in the new Studio? I'm trying to get this to work now.

  • @GENAI-pf3ok
    @GENAI-pf3ok 5 หลายเดือนก่อน

    Nice, But what is the tool you are using for drafting prompt and choosing models ?

  • @GENAI-pf3ok
    @GENAI-pf3ok 5 หลายเดือนก่อน

    Wonderfully explained ! Very Helpful ! - TIP - Please raise your voice and decrease the speed, Rest everything is great !!

  • @sinjini3189
    @sinjini3189 6 หลายเดือนก่อน

    Thank you so much for this!!

  • @jameslee7712
    @jameslee7712 6 หลายเดือนก่อน

    Method 3 works perfectly!

  • @berliansafriprakoso3
    @berliansafriprakoso3 7 หลายเดือนก่อน

    nice video 👏

  • @JiyuKim-sr1mi
    @JiyuKim-sr1mi 8 หลายเดือนก่อน

    I keep having this issues whenever I try to install it on SageMaker notebook: ERROR: Could not install packages due to an OSError: [Errno 2] No such file or directory: '/opt/conda/lib/python3.10/site-packages/pytz-2023.3.dist-info/METADATA'

  • @Abhisekgev
    @Abhisekgev 8 หลายเดือนก่อน

    I tried accessing the streamlit and followed your steps. The webpage hosted is reachable but appears to be completely blank.

    • @ken262849
      @ken262849 8 หลายเดือนก่อน

      I'm also having this problem. Did you manage to solve it? I will test in another browser other than Chrome.

    • @Abhisekgev
      @Abhisekgev 8 หลายเดือนก่อน

      One needs to upgrade the jupyter version to >= 2 and I tried a lot in doing so but the version of jupyter remained the same@@ken262849

    • @maks6770
      @maks6770 8 หลายเดือนก่อน

      The same for me but 3rd approach with --server.baseUrlPath works in my case for SageMaker Studio JupyterLab

    • @m.caneda4205
      @m.caneda4205 7 หลายเดือนก่อน

      I have the same problem. The third approach also does not work for me. Anyone have an idea?

    • @Abhisekgev
      @Abhisekgev 7 หลายเดือนก่อน

      @@m.caneda4205 One has to upgrade the jupyter version to greater than 2. I tried a lot with all conda environments but after upgrading the jupyter lab reboots and it shows version 1

  • @sailochanar3546
    @sailochanar3546 9 หลายเดือนก่อน

    can we upload documents from ui ?? instead of pre-loading documents in data sources.

  • @sailochanar3546
    @sailochanar3546 9 หลายเดือนก่อน

    can I know what is benefit of using kendra over using bedrock fms to build such product search application in cost base,I want to know charges info .

  • @sanjaybhatikar
    @sanjaybhatikar 9 หลายเดือนก่อน

    Thanks for this video. I did not realize this wouldn't work with Code Editor and wasted a lot of time going down the rabbit-hole of IAM roles and policies that the error messages had me thinking were the root cause.

  • @sanjaybhatikar
    @sanjaybhatikar 9 หลายเดือนก่อน

    AWS has surprisingly many rough edges. Amazon should so more to earn the reputation of customer obsession that they claim to have.

  • @benharlander8831
    @benharlander8831 10 หลายเดือนก่อน

    Thanks! Trying to get this working with Code Editor. Let me know if you have solved that one too!

    • @sabyr15
      @sabyr15 8 หลายเดือนก่อน

      Did you have a chance to make it work in Code editor

  • @musicworld-tm5uz
    @musicworld-tm5uz 10 หลายเดือนก่อน

    Please make a video from initail (from new account to create ah job) like 3d lidar jobs

  • @sirojiddinkomolov2820
    @sirojiddinkomolov2820 11 หลายเดือนก่อน

    Thanks for the tutorial

  • @ameyshahane
    @ameyshahane 11 หลายเดือนก่อน

    Awsome video! Super easy to understand! Thanks for this🙌🙌

  • @RadioMcRadioface
    @RadioMcRadioface ปีที่แล้ว

    "Add a pop of paddling fun to your bed"... I'm not making that up, its in the video. Kendra sure sounds fun so far doesnt it? I'm only 5 minutes in to this thing.

  • @hskhawaja
    @hskhawaja ปีที่แล้ว

    I'm facing this error and unable to figure out the issue with S3 bucket. S3 Connector Error code: S3-5110 Error message: Unable to connect with provided S3 bucket, Please try again with valid bucket. Can anyone help?

    • @karthikKumar-ew7wr
      @karthikKumar-ew7wr ปีที่แล้ว

      This happened to me as well. Next day I was able to crawl it with same bucket with new index. Hope it helps. Try command - s3 list from terminal if bucket shows. may be due to replication takes time

    • @HanXolo
      @HanXolo ปีที่แล้ว

      I've got the same error as well just now. You should go to the Sync history tab of the data source to see the run history of failed sync. Under the Details column there will be a link to CloudWatch. Click on that and run the query in CloudWatch Insights. For me it shows 'No Permissions to Access bucket.' This is a simple issue to fix with some edits to the IAM role used by Kendra.

  • @bibhutibaibhavbora8770
    @bibhutibaibhavbora8770 ปีที่แล้ว

    Great tutorial❤ Please make videos on S3 and EC2.

  • @Mairah21
    @Mairah21 ปีที่แล้ว

    Thanks for the tutorial. Can this be done on iPad or is this just laptop desktop thing like most 😢

  • @amandla1020
    @amandla1020 ปีที่แล้ว

    amazing tutorial, thanks !

  • @apoorrv0555
    @apoorrv0555 ปีที่แล้ว

    also object of type 'int' has no len() on colors[:10]

  • @apoorrv0555
    @apoorrv0555 ปีที่แล้ว

    i am getting this error [Errno 2] No such file or directory: 'jsons/doc_0_prod_297896.json'

    • @leandrosoares8350
      @leandrosoares8350 ปีที่แล้ว

      Did you fix this? im facing the same

    • @leandrosoares8350
      @leandrosoares8350 ปีที่แล้ว

      I fix. Just create a folder with name "jsons" in the lab

  • @solihinsolay4645
    @solihinsolay4645 ปีที่แล้ว

    Very good

  • @Reborn4games
    @Reborn4games ปีที่แล้ว

    Well, i don't really know anything related to this video, I never really searched anything related to this and I almost click off the video before 5 seconds... But idk why, after it started I couldn't leave, it was quite interesting and pleasurable to experience. Good work there, keep it up, o/ (btw, it was so good that it made me want to try programing again)