Congrats, and good job!

During your request, make sure you are using the us-east-1 region. Here are the details to submit your request. docs.aws.amazon.com/bedrock/latest/userguide/model-access.html

I think it requires subscription but you definitely need access granted for it to work.

@ I gave up and switched to cloudflare

Thank you!

Please also delete the vector store, otherwise you will be charged.

@@techwithzoum Can you please guide me how to delete vector store?

@@SalwaSayeedulHasan the steps are mentioned here from the 28th minute kzbin.info/www/bejne/e4HUYatopq-NjZo

@@techwithzoum Thank you so much 😊

@@SalwaSayeedulHasan you're very welcome!

Thank you!

I am glad it was clear enough to help you understand!

on AWS EC2 or use Serverless

Thank you!

You first need to be connected to the console as an IAM user with an Admin account to create the database. Have you followed all the steps before? Step 1 explains how to create an IAM user. Please let me know if you have any questions.

​@@techwithzoum Thanks for the response. Yes, I assigned all needed privilegdes, as in video, and after I see notification that DB is created, I have issue with the following: USER_XY is not authorized to perform: bedrock:CreateKnowledgeBase I don't know could it be maybe my region, since I am in Europe(Ireland)?

Just to update: The issue was region.. thank you!

@@milicacvjeticanin1439 perfect!

@@milicacvjeticanin1439 What did you need to switch the region to?

Thank you!

is the image app shared publically available?

I have shared the source code in the description section to help you reproduce the same thing.

I am glad it helps. How may I help?

Hey, I am also from India, What kind of project are you working on?

Thanks!

thank you!

Thank you for bringing this to my attention! Here is the link to the notebook! I have also updated in the description section! github.com/keitazoumana/LLMs/blob/main/Multi_Agents_For_Content_Creation.ipynb

I am really glad it did

Can you please share the error you are getting?

You're welcome! Stay tuned for the next video dropping this Thursday explaining how to create a graphical interface outside AWS using Streamlit, API Gateway, and Lambda function!

Second part is available: kzbin.info/www/bejne/e4HUYatopq-NjZo

Merci beaucoup, Noellie! De rien!

Noellie, the second part of the video is available! kzbin.info/www/bejne/e4HUYatopq-NjZo

Thank you, PapolAVetorial-oe5nj!

This tutorial aims to give a sweet of AWS Bedrock via Bedrock interface. Next week, I will post a video showing how to build one with UI, away from AWS! Stay tuned!

Here is the video you have been waiting for! It allows you to build chatbot with graphical interface. Let me know what you think! kzbin.info/www/bejne/e4HUYatopq-NjZo

Yes, they do

The easiest I can think of is to generate a requirements.txt file and install all the libraries from your current environment into the new environment. Main steps: - Generate requirements.txt from the current Env: pip freeze > requirements.txt - Export it into your new computer - From that new computer, create a new virtual environment - Finally, from where you have created the new environment install all the libraries as follows: pip install -r requirements.txt After these steps, your new environment in your other laptop should have everything installed.

@@techwithzoum but the point is the other computer will not have internet at all then also i can do it?

@@Gokujr77 In that case, creating a wheel directory might help. Generate the requirements.txt the same way as above. From the same first laptop: - Create a wheel directory: mkdir wheel_dir - Download all those packages as a wheel: pip download -r requirements.txt -d wheel_dir Transfer the requirements.txt file and the wheel_dir directory to the new laptop using a USB drive or other means. On the new laptop - Create a new virtual environment: python -m venv new_env Activate the new virtual environment. - Install the packages from the wheel directory: pip install -r requirements.txt --no-index --find-links wheel_dir I hope this helps

@@techwithzoum thanks i will try this

@@techwithzoum i am from India,can you help me get a remote job as a fresher, i have done my BS in AI and Data Science i have good skill set of Numpy Pandas , data visualisation libraries like matplotlib and seaborn and also Machine Learning scikit learn , Hadoop, SQL, spark , kafka , good communication skills and very hardworking also

You are very welcome Mr. Traore Bamba!

You are welcome, Anna! This is happening because the transcription model could not find the file you want to transcribe. Are you making sure to provide the exact file location for transcription?

@@techwithzoum i found out that the problem was in ffmpeg.exe file that was not found by the code (the file to transcribe was correct saved). I was not the only person with similar issue and i i found the solution on stackoverflow :) Thank you any way. Again, great tutorial. I am happy to build the same project for my own based on your videos and codebase on github!!!

​@@techwithzoum is it possible to communicate with you in a private message? i have solved the issue described above (got help from stackoverflow). But i have got the new issue and i don't find a solution. may be it is something that you had experience with. The issue appears when i run partition_pdf and error is again: PermissionError: [Errno 13] Permission denied: 'C:\\\\Users\\\\Konsumer\\\\AppData\\\\Local\\\\Temp\\\\tmpt4iifmwt'" }. But this is nothing really with permission. ". from chatGPT: "You're right that the error message can be misleading. In your case, the problem isn't actually related to file permissions, but rather to how the unstructured library is trying to handle temporary files and NLTK data downloads."

Thank you, and I am glad it helps!

Hi, My Multimodal RAG video covers that

​@@techwithzoum No I have checked that one. Could u test it with another pdf. It is not properly printing text and image

@@Know_Ur_World the actual implementation does not print images, but the reference to the images is provided and you can retrieve the base64 encoding of those images to print them. For text, the current implementation is already doing it.

Hi @Know_Ur_World, My latest video about multimodal RAG might provide the answer you are looking for. For a given question, the model answers with the source (like document page, paragraph number, image) containing the response to the question. The link is below: kzbin.info/www/bejne/embMfKasgJyWjsk

@@techwithzoum like when im asking a question : "give the steps to install neo4j desktop". The answer should print all steps included in the pdf and associated images(screenshots of steps in this case)

@@Know_Ur_World yes, it is possible and it all starts from getting the image containing the response, which can be performed and steps are given in the above video. The next will be to display that image along with the result using your favorite graphical library (matplotlib, seaborn, etc.)

Thanks, @thibauteka4046, and that sounds like a great idea!

Thank you! I am glad they are helpful. Please feel free to share with others! Don't forget to subscribe, and like as well! :)

Nice video with an great explanation. What about creating a JSON schema based on an user prompt and using the schema like a tool to get the data extracted from the PDF?

@@torstenschindler1965 Thank you!

@@torstenschindler1965 the current approach is generating the final response as a JSON, as specified in the prompt. However, I am not quite sure that I understand your inquiry. Can you please elaborate on what you would like to achieve? This way I will be able to help you.