Recipes
- Evaluations
- Applications
- Datasets
- Inference
Copy
Ask AI
import os
from scale_gp import SGPClient
from scale_gp.types.model_template_create_params import (
VendorConfiguration,
VendorConfigurationBundleConfig,
VendorConfigurationEndpointConfig,
)
account_id = os.environ.get("SGP_ACCOUNT_ID", None)
api_key = os.environ.get("SGP_API_KEY", None)
assert (
account_id is not None
), "You need to set the SGP_ACCOUNT_ID - you can find it at https://egp.dashboard.scale.com/admin/accounts"
assert api_key is not None, "You need to provide your API key - see https://egp.dashboard.scale.com/admin/api-key"
client = SGPClient(api_key=api_key)
bundle_config = VendorConfigurationBundleConfig(image="gemini-pro", registry="aws-registry", tag="latest")
endpoint_config = VendorConfigurationEndpointConfig(
max_workers=3,
)
vendor_configuration = VendorConfiguration(
bundle_config=bundle_config,
endpoint_config=endpoint_config,
)
model_template = client.model_templates.create(
account_id=account_id,
endpoint_type="SYNC",
model_type="COMPLETION",
name="Gemini-Pro Template",
vendor_configuration=vendor_configuration,
)
model_instance = client.models.create(
account_id=account_id,
model_type="COMPLETION",
name="gemini-pro",
model_vendor="GOOGLE",
model_template_id=model_template.id,
)
model_deployment = client.models.deployments.create(
model_instance_id=model_instance.id, name="Gemini-Pro Deployment", account_id=account_id
)
print(model_deployment)
model_completion_response = client.models.deployments.execute(
model_deployment_id=model_deployment.id,
model_instance_id=model_instance.id,
extra_body={"prompts": ["What is the capital of Canada?"]},
)
print(model_completion_response)
Copy
Ask AI
ModelDeployment(
id='d4a457c3-7b56-4b0d-b6f1-45e5809907dd',
account_id='66049ada2fc77c99ef015be7',
created_at=datetime.datetime(2024, 9, 26, 19, 58, 51, 105175),
created_by_user_id='42a5c8af-f698-43d0-923e-ba70102a2887',
name='Gemini-Pro Deployment',
status='READY',
deployment_metadata=None,
model_creation_parameters=None,
model_endpoint_id=None,
model_instance_id='6f6b4a0e-0ae2-43f2-9b46-b783d83a729f',
vendor_configuration=None
)
Inference
Deploy and Execute a Model
Use this recipe to deploy and execute a gemini-pro completion model
Copy
Ask AI
import os
from scale_gp import SGPClient
from scale_gp.types.model_template_create_params import (
VendorConfiguration,
VendorConfigurationBundleConfig,
VendorConfigurationEndpointConfig,
)
account_id = os.environ.get("SGP_ACCOUNT_ID", None)
api_key = os.environ.get("SGP_API_KEY", None)
assert (
account_id is not None
), "You need to set the SGP_ACCOUNT_ID - you can find it at https://egp.dashboard.scale.com/admin/accounts"
assert api_key is not None, "You need to provide your API key - see https://egp.dashboard.scale.com/admin/api-key"
client = SGPClient(api_key=api_key)
bundle_config = VendorConfigurationBundleConfig(image="gemini-pro", registry="aws-registry", tag="latest")
endpoint_config = VendorConfigurationEndpointConfig(
max_workers=3,
)
vendor_configuration = VendorConfiguration(
bundle_config=bundle_config,
endpoint_config=endpoint_config,
)
model_template = client.model_templates.create(
account_id=account_id,
endpoint_type="SYNC",
model_type="COMPLETION",
name="Gemini-Pro Template",
vendor_configuration=vendor_configuration,
)
model_instance = client.models.create(
account_id=account_id,
model_type="COMPLETION",
name="gemini-pro",
model_vendor="GOOGLE",
model_template_id=model_template.id,
)
model_deployment = client.models.deployments.create(
model_instance_id=model_instance.id, name="Gemini-Pro Deployment", account_id=account_id
)
print(model_deployment)
model_completion_response = client.models.deployments.execute(
model_deployment_id=model_deployment.id,
model_instance_id=model_instance.id,
extra_body={"prompts": ["What is the capital of Canada?"]},
)
print(model_completion_response)
Copy
Ask AI
ModelDeployment(
id='d4a457c3-7b56-4b0d-b6f1-45e5809907dd',
account_id='66049ada2fc77c99ef015be7',
created_at=datetime.datetime(2024, 9, 26, 19, 58, 51, 105175),
created_by_user_id='42a5c8af-f698-43d0-923e-ba70102a2887',
name='Gemini-Pro Deployment',
status='READY',
deployment_metadata=None,
model_creation_parameters=None,
model_endpoint_id=None,
model_instance_id='6f6b4a0e-0ae2-43f2-9b46-b783d83a729f',
vendor_configuration=None
)
1. Instantiate Client
1. Instantiate Client
Follow the instructions in the Quickstart Guide to setup the SGP Client
Copy
Ask AI
from scale_gp import SGPClient
client = SGPClient(api_key=api_key)
2. Define model template
2. Define model template
Before creating a model, you must first create a model template. A model
template serves 2 purposes. First, it provides common scaffolding that is static
across multiple models. Second, it exposes several variables that can be
injected at model creation time to customize the model.
Copy
Ask AI
bundle_config = VendorConfigurationBundleConfig(image="gemini-pro", registry="aws-registry", tag="latest")
endpoint_config = VendorConfigurationEndpointConfig(
max_workers=3,
)
vendor_configuration = VendorConfiguration(
bundle_config=bundle_config,
endpoint_config=endpoint_config,
)
model_template = client.model_templates.create(
account_id=account_id,
endpoint_type="SYNC",
model_type="COMPLETION",
name="Gemini-Pro Template",
vendor_configuration=vendor_configuration,
)
3. Create a model instance
3. Create a model instance
Use the created model template to create a model instance
Copy
Ask AI
model_instance = client.models.create(
account_id=account_id,
model_type="COMPLETION",
name="gemini-pro",
model_vendor="GOOGLE",
model_template_id=model_template.id,
)
4. Create a model deployment
4. Create a model deployment
Deploy the created instance
Copy
Ask AI
model_deployment = client.models.deployments.create(
model_instance_id=model_instance.id, name="Gemini-Pro Deployment", account_id=account_id
)
5. Execute the model deployment
5. Execute the model deployment
Execute the model deployment. In this case, we are executing a completion model with a list of prompts.
Copy
Ask AI
execute_result = client.models.deployments.execute(
model_deployment_id=model_deployment.id,
model_instance_id=model_instance.id,
extra_body={"prompts": ["What is the capital of Canada?"]},
)
Copy
Ask AI
import os
from scale_gp import SGPClient
from scale_gp.types.model_template_create_params import (
VendorConfiguration,
VendorConfigurationBundleConfig,
VendorConfigurationEndpointConfig,
)
account_id = os.environ.get("SGP_ACCOUNT_ID", None)
api_key = os.environ.get("SGP_API_KEY", None)
assert (
account_id is not None
), "You need to set the SGP_ACCOUNT_ID - you can find it at https://egp.dashboard.scale.com/admin/accounts"
assert api_key is not None, "You need to provide your API key - see https://egp.dashboard.scale.com/admin/api-key"
client = SGPClient(api_key=api_key)
bundle_config = VendorConfigurationBundleConfig(image="gemini-pro", registry="aws-registry", tag="latest")
endpoint_config = VendorConfigurationEndpointConfig(
max_workers=3,
)
vendor_configuration = VendorConfiguration(
bundle_config=bundle_config,
endpoint_config=endpoint_config,
)
model_template = client.model_templates.create(
account_id=account_id,
endpoint_type="SYNC",
model_type="COMPLETION",
name="Gemini-Pro Template",
vendor_configuration=vendor_configuration,
)
model_instance = client.models.create(
account_id=account_id,
model_type="COMPLETION",
name="gemini-pro",
model_vendor="GOOGLE",
model_template_id=model_template.id,
)
model_deployment = client.models.deployments.create(
model_instance_id=model_instance.id, name="Gemini-Pro Deployment", account_id=account_id
)
print(model_deployment)
model_completion_response = client.models.deployments.execute(
model_deployment_id=model_deployment.id,
model_instance_id=model_instance.id,
extra_body={"prompts": ["What is the capital of Canada?"]},
)
print(model_completion_response)
Copy
Ask AI
ModelDeployment(
id='d4a457c3-7b56-4b0d-b6f1-45e5809907dd',
account_id='66049ada2fc77c99ef015be7',
created_at=datetime.datetime(2024, 9, 26, 19, 58, 51, 105175),
created_by_user_id='42a5c8af-f698-43d0-923e-ba70102a2887',
name='Gemini-Pro Deployment',
status='READY',
deployment_metadata=None,
model_creation_parameters=None,
model_endpoint_id=None,
model_instance_id='6f6b4a0e-0ae2-43f2-9b46-b783d83a729f',
vendor_configuration=None
)
Assistant
Responses are generated using AI and may contain mistakes.