Creating assets for model customization in the UI

You can create and manage the dataset and evaluator assets that you can use for model customization in the UI.

Assets

Select Assets in the left hand panel and the Amazon SageMaker Studio UI and then select Datasets.

Choose Upload Dataset to add the dataset that you will use in your model customization jobs. By choosing the Required data input format, you can access a reference of dataset format to use.

Evaluators

You can also add Reward Functions and Reward Prompts for your Reinforcement Learning customization jobs.

The UI also provides guidance on the format required for the reward function or reward prompt.

Assets for model customization using AWS SDK

You can also use the SageMaker AI Python SDK to create assets. See sample code snippet below:

from sagemaker.ai_registry.air_constants import REWARD_FUNCTION, REWARD_PROMPT
from sagemaker.ai_registry.dataset import DataSet, CustomizationTechnique
from sagemaker.ai_registry.evaluator import Evaluator

# Creating a dataset example
dataset = DataSet.create(
            name="sdkv3-gen-ds2",
            source="s3://sample-test-bucket/datasets/training-data/jamjee-sft-ds1.jsonl", # or use local filepath as source.
            customization_technique=CustomizationTechnique.SFT
        )

# Refreshes status from hub
dataset.refresh()
pprint(dataset.__dict__)

# Creating an evaluator. Method : Lambda
evaluator = Evaluator.create(
                name = "sdk-new-rf11",
                source="arn:aws:lambda:us-west-2:<>:function:<function-name>8",
                type=REWARD_FUNCTION
        )

# Creating an evaluator. Method : Bring your own code
evaluator = Evaluator.create(
                name = "eval-lambda-test",
                source="/path_to_local/eval_lambda_1.py",
                type = REWARD_FUNCTION
        )

# Optional wait, by default we have wait = True during create call.
evaluator.wait()

evaluator.refresh()
pprint(evaluator)