googlecloudplatform / terraform-genai-doc-summarization

Summarizes document using OCR and Vertex Generative AI LLM

Generative AI Document Summarization

Description

Tagline

Create summaries of a large corpus of documents using Generative AI.

Detailed

This solution showcases how to summarize a large corpus of documents using Generative AI. It provides an end-to-end demonstration of document summarization going all the way from raw documents, detecting text in the documents and summarizing the documents on-demand using Vertex AI LLM APIs, Document AI Optical Character Recognition (OCR), and BigQuery.

PreDeploy

To deploy this blueprint you must have an active billing account and billing permissions.

Architecture

• User uploads a new document triggering the webhook Cloud Function.
• Document AI extracts the text from the document file.
• A Vertex AI Large Language Model summarizes the document text.
• The document summaries are stored in BigQuery.

Documentation

• Generative AI Document Summary

Deployment Duration

Configuration: 1 mins Deployment: 5 mins

Cost

Cost Details

Inputs

Name	Description	Type	Default	Required
disable_services_on_destroy	Whether project services will be disabled when the resources are destroyed.	bool	false	no
documentai_location	Document AI location, see https://cloud.google.com/document-ai/docs/regions	string	"us"	no
labels	A set of key/value label pairs to assign to the resources deployed by this blueprint.	map(string)	{}	no
project_id	The Google Cloud project ID to deploy to	string	n/a	yes
region	The Google Cloud region to deploy to	string	"us-central1"	no
unique_names	Whether to use unique names for resources	bool	false	no

Outputs

Name	Description
bigquery_dataset_id	The name of the BigQuery dataset created
bucket_docs_name	The name of the docs bucket created
bucket_main_name	The name of the main bucket created
documentai_processor_id	The full Document AI processor path ID
neos_walkthrough_url	The URL to launch the in-console tutorial for the Generative AI Document Summarization solution
unique_id	The unique ID for this deployment

Requirements

These sections describe requirements for using this module.

Software

The following dependencies must be available:

• Terraform v0.13
• Terraform Provider for GCP plugin v3.0

Service Account

A service account with the following roles must be used to provision the resources of this module:

• Storage Admin: roles/storage.admin

APIs

A project with the following APIs enabled must be used to host the resources of this module:

• Google Cloud Storage JSON API: storage-api.googleapis.com

Contributing

Refer to the contribution guidelines for information on contributing to this module.

Security Disclosures

Please see our security disclosure process.

terraform-genai-doc-summarization's Issues

Failing periodic

Passing periodic at 083d7da. Closing this issue.

Failing periodic

Passing periodic at 083d7da. Closing this issue.

document_extract doesn't use my own credentials

I am running from Jump Start Soltion center this example which uses gen_ai_jss notebook

Cell 29

bucket = "arxiv-dataset"
pdf_name = "arxiv/cmp-lg/pdf/9404/9404002v1.pdf"
output_bucket = f"{PROJECT_ID}_output"

complete_text = document_extract(bucket=bucket,
                                 name=pdf_name,
                                 output_bucket=output_bucket,
                                 project_id=PROJECT_ID)

# Entire text is long; print just first 1000 characters
print(complete_text[:1000])

fails for me with the below error

_InactiveRpcError                         Traceback (most recent call last)
[/usr/local/lib/python3.10/dist-packages/google/api_core/grpc_helpers.py](https://localhost:8080/#) in error_remapped_callable(*args, **kwargs)
     71         try:
---> 72             return callable_(*args, **kwargs)
     73         except grpc.RpcError as exc:

9 frames
_InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
	status = StatusCode.PERMISSION_DENIED
	details = "Cloud Vision API has not been used in project 522309567947 before or it is disabled. Enable it by visiting https://console.developers.google.com/apis/api/vision.googleapis.com/overview?project=522309567947 then retry. If you enabled this API recently, wait a few minutes for the action to propagate to our systems and retry."
	debug_error_string = "UNKNOWN:Error received from peer ipv4:172.253.63.95:443 {created_time:"2023-10-22T18:48:21.432087561+00:00", grpc_status:7, grpc_message:"Cloud Vision API has not been used in project 522309567947 before or it is disabled. Enable it by visiting https://console.developers.google.com/apis/api/vision.googleapis.com/overview?project=522309567947 then retry. If you enabled this API recently, wait a few minutes for the action to propagate to our systems and retry."}"
>

The above exception was the direct cause of the following exception:

PermissionDenied                          Traceback (most recent call last)
[/usr/local/lib/python3.10/dist-packages/google/api_core/grpc_helpers.py](https://localhost:8080/#) in error_remapped_callable(*args, **kwargs)
     72             return callable_(*args, **kwargs)
     73         except grpc.RpcError as exc:
---> 74             raise exceptions.from_grpc_error(exc) from exc
     75 
     76     return error_remapped_callable

PermissionDenied: 403 Cloud Vision API has not been used in project 522309567947 before or it is disabled. Enable it by visiting https://console.developers.google.com/apis/api/vision.googleapis.com/overview?project=522309567947 then retry. If you enabled this API recently, wait a few minutes for the action to propagate to our systems and retry. [links {
  description: "Google developers console API activation"
  url: "https://console.developers.google.com/apis/api/vision.googleapis.com/overview?project=522309567947"
}
, reason: "SERVICE_DISABLED"
domain: "googleapis.com"
metadata {
  key: "consumer"
  value: "projects/522309567947"
}
metadata {
  key: "service"
  value: "vision.googleapis.com"
}
]

although you might wonder if I forget to enable something, its not the case.
Actually I don't recognize the project number, so it's not mine.

Somehow the script even if the notebook is authentication for my user, it tries to use a different project to run the Vision API call, perhaps it's from the "bucket" owner. This should be fixed.

Alianzas productivas “mejoramiento del manejo, producción y Comercialización del grano de cacao con 84 productores del Municipio de pueblo rico - Risaralda

Matriz ODS_grupo 93 (1).pdf

Evaluate whether to replace or refactor text truncation function

The utils.truncate_complete_text function uses a heuristic to extract the abstract and conclusion from a OCR result. This approach has multiple issues (cannot handle corner cases; doesn't capture all or only abstracts & conclusions).

I recommend replacing or refactoring this module in one of the following ways:

1. Replace the heuristic string manipulation code with a call to an in-memory NLP or LLM model.
2. Investigate improvements to Document AI templating to get better results from OCR.
3. Use regex (ugh) to better isolate the abstract and conclusion
4. Other?

Failing periodic

Passing periodic at 6b6b0dc. Closing this issue.

Failing periodic

Passing periodic at 6406a24. Closing this issue.

404 Not Found The page you requested could not be found.

Can't able to access :- https://registry.terraform.io/modules/GoogleCloudPlatform/terraform-genai-doc-summarization/google

• Please correct URL

Failing periodic

Passing periodic at d00bc10. Closing this issue.

Failing periodic

Passing periodic at e730587. Closing this issue.

Dependency Dashboard

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Pending Status Checks

These updates await pending status checks. To force their creation now, click the checkbox below.

• chore(deps): update dependency functions-framework to v3.7.0

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• chore(deps): update dependency google-cloud-documentai to v2.27.1
• chore(deps): update dependency pytest to v8.2.1
• chore(deps): Update module github.com/GoogleCloudPlatform/cloud-foundation-toolkit/infra/blueprint-test to v0.15.0
• chore(deps): update dependency google-cloud-aiplatform to v1.52.0
• chore(deps): update dependency google-cloud-bigquery to v3.23.1
• chore(deps): Update Terraform terraform-google-modules/project-factory/google to v15
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

github-actions

.github/workflows/periodic-reporter.yaml

• actions/github-script v7

.github/workflows/webhook.yml

• actions/checkout v4
• actions/setup-python v5
• actions/checkout v4
• hashicorp/setup-terraform v3
• actions/setup-python v5
• google-github-actions/auth v2
• actions/setup-python v5
• actions/cache v4

gomod

test/integration/go.mod

• go 1.21
• go 1.22.1
• github.com/GoogleCloudPlatform/cloud-foundation-toolkit/infra/blueprint-test v0.14.0
• github.com/stretchr/testify v1.9.0

pip_requirements

webhook/requirements-test.txt

• mypy ==1.10.0
• pytest ==8.2.0

webhook/requirements.txt

• flask ==3.0.3
• functions-framework ==3.5.0
• google-cloud-aiplatform ==1.51.0
• google-cloud-documentai ==2.27.0
• google-cloud-bigquery ==3.22.0
• google-cloud-storage ==2.16.0

regex

Makefile

• cft/developer-tools 1

build/int.cloudbuild.yaml

• cft/developer-tools 1

terraform

examples/simple_example/main.tf

main.tf

• terraform-google-modules/project-factory/google ~> 14.5

test/setup/main.tf

• terraform-google-modules/project-factory/google ~> 14.5

test/setup/versions.tf

• google >= 5.24, < 6
• google-beta >= 5.24, < 6
• hashicorp/terraform >= 0.13

versions.tf

• archive ~> 2.4
• google >= 5.24, < 6
• google-beta >= 5.24, < 6
• random ~> 3.5
• hashicorp/terraform >= 0.13

---

• Check this box to trigger a request for Renovate to run again on this repository

pubsub service account missing roles/iam.serviceAccountTokenCreator

I was able to reconstruct an issue when hiting Deploy from this url
https://console.cloud.google.com/products/solutions/deployments/details/us-central1/generative-ai-document-summarization

I used an existing project

all deployments succeed, notebook is working, file is uploaded, EventArc picks up the event, Pub/Sub receives the message, but
Pub/Sub is not calling the Cloud Function, although everything looks setup. the GCF exists

I did this many times by undeploying and deploying.

After that I was able to observ in the Edit subscription page of Pub/Sub that it complaints the roles/iam.serviceAccountTokenCreator is not applied. Once I manually granted this, the subscription started to fire the push job.

Please fix the scripts. and apply this role.

Failing periodic

Passing periodic at ee8ed5b. Closing this issue.

Failing periodic

Passing periodic at 6b6b0dc. Closing this issue.

Location is defined explicitly to us-central1

def predict_large_language_model(
project_id: str,
model_name: str,
temperature: float,
max_decode_steps: int,
top_p: float,
top_k: int,
content: str,
location: str = "us-central1",

Remove test skip from document_extract_test

The integration test for document_extract sometimes encounters a race condition where it attempts to access a OCR page that doesn't exist in the Storage bucket. We need to adjust the test so that this race condition is ameliorated.