adding batch gemini predict sample #12606

CadillacBurgess1 · 2024-09-18T15:22:38Z

Description

Fixes #

Note: Before submitting a pull request, please open an issue for discussion if you are not associated with Google.

Checklist

I have followed Sample Guidelines from AUTHORING_GUIDE.MD
README is updated to include all relevant information
Tests pass: nox -s py-3.9 (see Test Environment Setup)
Lint pass: nox -s lint (see Test Environment Setup)
These samples need a new API enabled in testing projects to pass (let us know which ones)
These samples need a new/updated env vars in testing projects set to pass (let us know which ones)
This sample adds a new sample directory, and I updated the CODEOWNERS file with the codeowners for this sample
This sample adds a new Product API, and I updated the Blunderbuss issue/PR auto-assigner with the codeowners for this sample
Please merge this PR for me once it is approved

snippet-bot · 2024-09-18T15:22:46Z

Here is the summary of changes.

You are about to add 1 region tag.

generative_ai/batch_predict/batch_gemini_predict.py:41, tag generativeaionvertexai_batch_predict_gemini_createjob

This comment is generated by snippet-bot.
If you find problems with this result, please file an issue at:
https://github.com/googleapis/repo-automation-bots/issues.
To update this comment, add snippet-bot:force-run label or use the checkbox below:

Refresh this comment

generative_ai/batch_predict/batch_gemini_predict.py

holtskinner · 2024-09-18T15:51:13Z

generative_ai/batch_predict/batch_gemini_predict.py

+
+
+def batch_prediction_gemini_example(
+    input_uri: str = None, output_uri: str = None


Suggested change

input_uri: str = None, output_uri: str = None

input_uri: Optional[str] = None, output_uri: Optional[str] = None

Also add:

from typing import Optional

I think this is the same as it was before. Also, should these parameters be optional/default None? Aren't those required parameters for the method?

This is an old version

updated params

generative_ai/batch_predict/batch_gemini_predict.py

… uri structure

holtskinner · 2024-09-18T17:56:44Z

generative_ai/batch_predict/batch-predict_gemini.py

Looks like a duplicate file. Change the name to use all underscores and replace the previous file.

CadillacBurgess1 · 2024-09-19T13:44:08Z

The issue was importing the preview resource. When I first created this sample it was working using the preview namespace (which as fine when I first started working on this sample), but the samples and testing in the batch predict folder uses the aiplatform namespace. Both aiplatform and vertexai.preview have a BatchPredictionJob class and they work a little different. I updated my sample to use the BatchPredictionJob class in the aiplatform namespace.

holtskinner · 2024-09-19T15:48:56Z

The issue was importing the preview resource. When I first created this sample it was working using the preview namespace (which as fine when I first started working on this sample), but the samples and testing in the batch predict folder uses the aiplatform namespace. Both aiplatform and vertexai.preview have a BatchPredictionJob class and they work a little different. I updated my sample to use the BatchPredictionJob class in the aiplatform namespace.

I think this sample should be using the vertexai.preview BatchPredictionJob class as it is intended for use with Gemini. (Different than other batch prediction samples for non-Gemini models)

I'll defer to the Vertex AI SDK product team @Ark-kun (or other)

This notebook from @gericdong also uses the vertexai.preview space

https://github.com/GoogleCloudPlatform/generative-ai/pull/1138/files

holtskinner

Btw, something I missed in our sync. The older PaLM samples use aiplatform instead of vertexai because the Batch Prediction methods weren't implemented yet in vertexai when PaLM was released.

holtskinner · 2024-09-19T18:04:56Z

generative_ai/batch_predict/batch_gemini_predict.py

+import os
+from typing import Optional
+
+from google.cloud.aiplatform import BatchPredictionJob


Suggested change

from google.cloud.aiplatform import BatchPredictionJob

from vertexai.preview.batch_prediction import BatchPredictionJob

holtskinner · 2024-09-19T18:05:23Z

generative_ai/batch_predict/batch_gemini_predict.py

+    """
+
+    # [START generativeaionvertexai_batch_predict_gemini_createjob]
+    from google.cloud import aiplatform


Suggested change

from google.cloud import aiplatform

import vertexai

from vertexai.preview.batch_prediction import BatchPredictionJob

holtskinner · 2024-09-19T18:05:35Z

generative_ai/batch_predict/batch_gemini_predict.py

+    # output_uri = bq://example_project.example_dataset.example_table or gs://mybucket/sampledataset.json
+
+    # Initialize
+    aiplatform.init(project=PROJECT_ID, location=LOCATION)


Suggested change

aiplatform.init(project=PROJECT_ID, location=LOCATION)

vertexai.init(project=PROJECT_ID, location=LOCATION)

holtskinner · 2024-09-19T18:05:43Z

generative_ai/batch_predict/batch_gemini_predict.py

+    aiplatform.init(project=PROJECT_ID, location=LOCATION)
+
+    # Create the batch prediction job using BatchPredictionJob
+    batch_prediction_job = aiplatform.BatchPredictionJob.create(


Suggested change

batch_prediction_job = aiplatform.BatchPredictionJob.create(

batch_prediction_job = BatchPredictionJob.create(

holtskinner · 2024-09-20T19:49:11Z

Looks like you're getting an error in the tests:

ValueError: Please either specify output_uri_prefix or set staging_bucket in vertexai.init().

holtskinner · 2024-09-20T19:49:43Z

generative_ai/batch_predict/batch_gemini_predict.py

+    print(f"Job output location: {batch_prediction_job.output_location}")
+
+    # Example response:
+    #  Job output location: gs://yourbucket/gen-ai-batch-prediction/prediction-model-year-month-dayThour:minute:second.12345


Suggested change

# Job output location: gs://yourbucket/gen-ai-batch-prediction/prediction-model-year-month-dayThour:minute:second.12345

# Job output location: gs://yourbucket/gen-ai-batch-prediction/prediction-model-year-month-day-hour:minute:second.12345

holtskinner · 2024-09-20T19:50:52Z

generative_ai/batch_predict/batch_gemini_predict.py

+
+
+def batch_prediction_gemini_example(
+    input_uri: str = None, output_uri: str = None


I think this is the same as it was before. Also, should these parameters be optional/default None? Aren't those required parameters for the method?

adding batch gemini predict sample

a83f725

CadillacBurgess1 requested review from a team as code owners September 18, 2024 15:22

product-auto-label bot added the samples Issues that are directly related to samples. label Sep 18, 2024

blunderbuss-gcf bot assigned iennae Sep 18, 2024

update spacing

9a7fb45

holtskinner requested changes Sep 18, 2024

View reviewed changes

CadillacBurgess1 added 4 commits September 18, 2024 12:02

removing additonal copyright information and spacing / changing input…

3232433

… uri structure

updating spacing

09af586

updated formatting

fb42928

remvoing whitespaces

70e94bb

iennae assigned holtskinner Sep 18, 2024

holtskinner reviewed Sep 18, 2024

View reviewed changes

CadillacBurgess1 added 15 commits September 18, 2024 14:03

updating import statement order

b309848

renaming

d5502fd

removing white space

f43f49d

modifying import

31240a2

removing whitespace

b2d325b

removing extra line

9fff375

adding line

edf6aac

fixing attribute on test file

e2c2cf1

modifying testing file to point to target in cloud storage bucket

dcadb92

removing extra spaces

a6144fb

adding new line at the end of testing file

1c63c18

added jsonl file in gcs bucket for testing

3abcf5d

removing extra line

37a2973

changing lambda name

04d5329

changing def name to conform to naming convention

9ea7a6a

CadillacBurgess1 requested a review from iennae September 19, 2024 13:27

holtskinner assigned CadillacBurgess1 and unassigned iennae and holtskinner Sep 19, 2024

holtskinner reviewed Sep 19, 2024

View reviewed changes

CadillacBurgess1 added 10 commits September 20, 2024 10:03

using vertex.preview

4211286

adding spacing

e7635d9

renaming values

34b4d4b

updating function name

fe60580

updating naming convention

3783b1f

updating naming convention

d26a721

updating statement

4bb2bd1

updating resource

a130960

updating resource

fcfad6b

updating output info

6344980

CadillacBurgess1 requested a review from holtskinner September 20, 2024 19:48

holtskinner requested changes Sep 20, 2024

View reviewed changes

CadillacBurgess1 added 10 commits September 20, 2024 16:45

added input_uri

b3dd538

updating import statement

f575d09

updating import order

572f3fa

updating import

f824d0c

update testing file

0916266

update spacing

fb0200e

updating function

35a475c

update to str

362b980

update testing

e4364a5

updating input

a5ba6c3

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

adding batch gemini predict sample #12606

adding batch gemini predict sample #12606

CadillacBurgess1 commented Sep 18, 2024 •

edited

Loading

snippet-bot bot commented Sep 18, 2024 •

edited

Loading

holtskinner Sep 18, 2024

CadillacBurgess1 Sep 18, 2024

holtskinner Sep 20, 2024

CadillacBurgess1 Sep 20, 2024

CadillacBurgess1 Sep 20, 2024

holtskinner Sep 18, 2024 •

edited

Loading

CadillacBurgess1 commented Sep 19, 2024

holtskinner commented Sep 19, 2024 •

edited

Loading

holtskinner left a comment

holtskinner Sep 19, 2024

holtskinner Sep 19, 2024

holtskinner Sep 19, 2024

holtskinner Sep 19, 2024

holtskinner commented Sep 20, 2024

holtskinner Sep 20, 2024

CadillacBurgess1 Sep 20, 2024

holtskinner Sep 20, 2024



		def batch_prediction_gemini_example(
		input_uri: str = None, output_uri: str = None

	input_uri: str = None, output_uri: str = None
	input_uri: Optional[str] = None, output_uri: Optional[str] = None

	from google.cloud.aiplatform import BatchPredictionJob
	from vertexai.preview.batch_prediction import BatchPredictionJob

	from google.cloud import aiplatform
	import vertexai
	from vertexai.preview.batch_prediction import BatchPredictionJob

	aiplatform.init(project=PROJECT_ID, location=LOCATION)
	vertexai.init(project=PROJECT_ID, location=LOCATION)

	batch_prediction_job = aiplatform.BatchPredictionJob.create(
	batch_prediction_job = BatchPredictionJob.create(

	# Job output location: gs://yourbucket/gen-ai-batch-prediction/prediction-model-year-month-dayThour:minute:second.12345
	# Job output location: gs://yourbucket/gen-ai-batch-prediction/prediction-model-year-month-day-hour:minute:second.12345

adding batch gemini predict sample #12606

Are you sure you want to change the base?

adding batch gemini predict sample #12606

Conversation

CadillacBurgess1 commented Sep 18, 2024 • edited Loading

Description

Checklist

snippet-bot bot commented Sep 18, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

holtskinner Sep 18, 2024 • edited Loading

Choose a reason for hiding this comment

CadillacBurgess1 commented Sep 19, 2024

holtskinner commented Sep 19, 2024 • edited Loading

holtskinner left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

holtskinner commented Sep 20, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

CadillacBurgess1 commented Sep 18, 2024 •

edited

Loading

snippet-bot bot commented Sep 18, 2024 •

edited

Loading

holtskinner Sep 18, 2024 •

edited

Loading

holtskinner commented Sep 19, 2024 •

edited

Loading