import os
import time

from scale_gp import SGPClient

client = SGPClient(api_key=api_key)

evaluation = client.evaluations.create(
    type="builder",
    account_id=account_id,
    application_spec_id=spec_id,
    application_variant_id=variant_id,
    description="description",
    evaluation_dataset_id=manual_evaluation_dataset.id,
    name="Report card evaluation",
    evaluation_config_id=evaluation_config.id,
)
create_response = client.application_variant_reports.create(
    application_variant_id=builder.application.variant_id,
    evaluation_dataset_ids=[manual_evaluation_dataset.id],
    account_id=account_id,
)
print(create_response)
# retrieve the application variant report, will still be PENDING if immediately retrieved
application_variant_report = client.application_variant_reports.retrieve(
    application_variant_report_id=create_response.id,
    view=["AsyncJobs"],
)
print(application_variant_report)


ApplicationVariantWithScores(
    id='cba0f08d-cb98-4221-a96d-d3eacb1c7a38',
    account_id='66049ada2fc77c99ef015be7',
    application_spec_id='ef0f2d50-1304-4b40-a0cc-1faf449b9555',
    application_variant_id='37d7d6ed-b7e8-4c3b-a488-03c96bb23d5d',
    created_at=datetime.datetime(2024, 9, 26, 20, 34, 6, 45000),
    updated_at=datetime.datetime(2024, 9, 26, 20, 34, 6, 45000),
    category_scores=None,
    evaluation_datasets=None,
    score=None
)
Follow the instructions in the Quickstart Guide to setup the SGP Client
from scale_gp import SGPClient

client = SGPClient(api_key=api_key)
Create an evaluation by following any previous evaluation recipe such as Multistage Evaluation
evaluation = client.evaluations.create(
    type="builder",
    account_id=account_id,
    application_spec_id=spec_id,
    application_variant_id=variant_id,
    description="description",
    evaluation_dataset_id=evaluation_dataset.id,
    name="Report card evaluation",
    evaluation_config_id=evaluation_config.id,
)
A report card will provide a summary of the performance of the application variant against an evaluation dataset Three main scores will be produced: Accuracy, Quality and Trust & Safety. The performance in each category will contribute to an overall Scale Confidence Score for the variant.You can view generated report cards in the UI at: https://egp.dashboard.scale.com/applications/{application_spec_id}/{application_variant_id}/report-card/overview
create_response = client.application_variant_reports.create(
    application_variant_id=variant_id,
    evaluation_dataset_ids=[manual_evaluation_dataset.id],
    account_id=account_id,
)
# retrieve the application variant report, will still be PENDING if immediately retrieved
application_variant_report = client.application_variant_reports.retrieve(
    application_variant_report_id=create_response.id,
    view=["AsyncJobs"],
)

import os
import time

from scale_gp import SGPClient

client = SGPClient(api_key=api_key)

evaluation = client.evaluations.create(
    type="builder",
    account_id=account_id,
    application_spec_id=spec_id,
    application_variant_id=variant_id,
    description="description",
    evaluation_dataset_id=manual_evaluation_dataset.id,
    name="Report card evaluation",
    evaluation_config_id=evaluation_config.id,
)
create_response = client.application_variant_reports.create(
    application_variant_id=builder.application.variant_id,
    evaluation_dataset_ids=[manual_evaluation_dataset.id],
    account_id=account_id,
)
print(create_response)
# retrieve the application variant report, will still be PENDING if immediately retrieved
application_variant_report = client.application_variant_reports.retrieve(
    application_variant_report_id=create_response.id,
    view=["AsyncJobs"],
)
print(application_variant_report)


ApplicationVariantWithScores(
    id='cba0f08d-cb98-4221-a96d-d3eacb1c7a38',
    account_id='66049ada2fc77c99ef015be7',
    application_spec_id='ef0f2d50-1304-4b40-a0cc-1faf449b9555',
    application_variant_id='37d7d6ed-b7e8-4c3b-a488-03c96bb23d5d',
    created_at=datetime.datetime(2024, 9, 26, 20, 34, 6, 45000),
    updated_at=datetime.datetime(2024, 9, 26, 20, 34, 6, 45000),
    category_scores=None,
    evaluation_datasets=None,
    score=None
)