aws-samples / pace-genai-demos

This repository features three demos that can be effortlessly integrated into your AWS environment. They serve as a practical guide to leveraging AWS services for crafting a sophisticated Large Language Model (LLM) Generative AI, geared towards creating a responsive Question and Answer Bot and localizing content generation.

Harnessing the power of enterprise data with generative AI: Insights from Amazon Kendra, LangChain, and large language models

This repository provides code samples for three Generative AI demos, licensed under MIT-0 license.

1. Amazon Kendra with foundational LLM: Utilizes the deep search capabilities of Amazon Kendra combined with the expansive knowledge of Large Language Models. This integration provides precise and context-aware answers to complex queries by drawing from a diverse range of sources.

2. Embeddings model with foundational LLM: Merges the power of embeddings—a technique to capture semantic meanings of words and phrases—with the vast knowledge base of LLMs. This synergy enables more accurate topic modeling, content recommendation, and semantic search capabilities.

3. Foundation Models Pharma Ad Generator: A specialized application tailored for the pharmaceutical industry. Harnessing the generative capabilities of foundational models, this tool creates convincing and compliant pharmaceutical advertisements, ensuring content adheres to industry standards and regulations

These demos can be seamlessly deployed in your AWS account, offering foundational insights and guidance on utilizing AWS services to create a state-of-the-art Large Language Model (LLM) Generative AI Question and Answer Bot and content generation.

You can deploy these demo's independent of each other. Please refer to the Readme files in each of the folders for deployment instructions.

Refer to the blog post for details on how these solutions work.

Authors

• Jin Tan Ruan - AWSI PACE Prototyping Developer
• Aravind Kodandaramaiah - AWSI PACE Senior Prototyping Developer

Troubleshoot

Unzipped size must be smaller than 262144000 bytes (Service: AWSLambdaInternal; Status Code: 400)

1. Delete the Existing Lambda Layer Folder: Begin by removing the lambda_langchain_layer folder from your project. This action ensures that any corrupted or oversized files are cleared.
2. Recreate the Layer: After deletion, recreate the lambda layer using the deploy.sh command. This process should generate a fresh, size-compliant layer with the necessary components.
3. Clean Docker Resources: It's also crucial to ensure that no residual Docker images or containers are occupying unnecessary space. Clean all running Docker images and containers to free up space and avoid any potential conflicts.

Reporting Bugs/Feature Requests

When filing an issue, please check existing open, or recently closed, issues to make sure somebody else hasn't already reported the issue. Please try to include as much information as you can. Details like these are incredibly useful:

• A reproducible test case or series of steps
• The version of our code being used
• Any modifications you've made relevant to the bug
• Anything unusual about your environment or deployment

License

This library is licensed under the MIT-0 License. See the LICENSE file.

• Changelog of the project.
• License of the project.
• Code of Conduct of the project.
• CONTRIBUTING for more information.

pace-genai-demos's Issues

embeddingsbedrockLangchainLayer54390634 Lambda too big

I cloned the repo on 14/11/23, and from the tip of the main branch (commit hash: a01bb1b) tried to deploy the Embeddings-Foundational-LLM-ChatBot example.

But on the cloudformation deployment step in the guru-chatbot stack I ran into this error for the logical resource: embeddingsbedrockLangchainLayer54390634

Resource handler returned message: "Unzipped size must be smaller than 262144000 bytes (Service: AWSLambdaInternal; Status Code: 400; Error Code: InvalidParameterValueException; Request ID: 5c2f3a4b-c0c9-48ba-ab7e-af70401b9f31; Proxy: null)" (RequestToken: 579e2a4b-d216-c234-1769-ff474b9ab1d3, HandlerErrorCode: GeneralServiceException)

Which I think means this Lambda layer is >250MB in size unzipped.

Checked and its 302MB unzipped. Lots of large python libs:

du -h . | sort -hr | head -n 10
302M .
81M ./botocore
78M ./botocore/data
70M ./faiss
36M ./numpy.libs
35M ./numpy
19M ./sqlalchemy
18M ./numpy/core
16M ./langchain
9.2M ./pydantic

How might this be resolved/worked-around?

Error: "Unzipped size must be smaller than 262144000 bytes (Service: AWSLambdaInternal; Status Code: 400"

Hi,

Deploying this solution fails with error:

"Resource handler returned message: "Unzipped size must be smaller than 262144000 bytes (Service: AWSLambdaInternal; Status Code: 400; "

This is on the lambda layer resource.

Please can you check this?

[Feature Request] Add support for deployment via Cloud9 instances

Opening this issue as a feature request to add support for deploying this solution via AWS Cloud9.

Currently, when using Cloud9 to deploy the "Embeddings-Foundational-LLM-ChatBot" solution, we run into an error on the "lambda_langchain_layer" unzipped size being larger than lambda's 250Mb size constraint. As noted in the ReadMe for "Embeddings-Foundational-LLM-ChatBot", this solution may not function correctly when deployed via Cloud9 or similar cloud-based development environments.