Appearance

API Reference ​

The PaiTIENT Secure Model Service provides comprehensive SDKs for both Python and JavaScript (Node.js), as well as a REST API for direct integration.

SDK Overview ​

Our SDKs are designed to provide a simple, consistent interface for interacting with the PaiTIENT Secure Model Service. They handle authentication, error handling, and provide a convenient abstraction over the underlying REST API.

Available SDKs ​

  • Python SDK: For data scientists, ML engineers, and backend developers
  • JavaScript SDK: For web developers and Node.js applications

Common Concepts ​

All SDKs share common concepts and structures:

Authentication ​

All requests to the PaiTIENT Secure Model Service require authentication using an API key and client ID. These credentials should be kept secure and never exposed publicly.

python
# Python example
from paitient_secure_model import SecureModelClient

client = SecureModelClient(
    api_key="your-api-key",
    client_id="your-client-id"
)
javascript
// JavaScript example
const { SecureModelClient } = require('paitient-secure-model');

const client = new SecureModelClient({
  apiKey: "your-api-key",
  clientId: "your-client-id"
});

Model Deployments ​

A deployment represents a specific model that has been deployed to your secure environment. Deployments can be created, updated, and deleted as needed.

python
# Python example - Create a deployment
deployment = client.deploy_model(
    model_name="ZimaBlueAI/HuatuoGPT-o1-8B",
    deployment_name="my-secure-model",
    use_gpu=True
)

# Get deployment status
status = client.get_deployment_status(deployment_id=deployment.id)
javascript
// JavaScript example - Create a deployment
const deployment = await client.deploy({
  modelName: "ZimaBlueAI/HuatuoGPT-o1-8B",
  deploymentName: "my-secure-model",
  useGpu: true
});

// Get deployment status
const status = await client.getDeploymentStatus(deployment.id);

Text Generation ​

Once a model is deployed, you can use it to generate text using a prompt.

python
# Python example
result = client.generate(
    deployment_id="deployment-123",
    prompt="Explain the importance of HIPAA compliance in healthcare AI",
    max_tokens=500,
    temperature=0.7
)

print(result.text)
javascript
// JavaScript example
const result = await client.generate({
  deploymentId: "deployment-123",
  prompt: "Explain the importance of HIPAA compliance in healthcare AI",
  maxTokens: 500,
  temperature: 0.7
});

console.log(result.text);

Subscription Management ​

You can check your subscription status, including usage limits and current usage.

python
# Python example
subscription = client.get_subscription()
print(f"Tier: {subscription.tier}")
print(f"Usage: {subscription.current_usage}/{subscription.usage_limit}")
javascript
// JavaScript example
const subscription = await client.getSubscription();
console.log(`Tier: ${subscription.tier}`);
console.log(`Usage: ${subscription.currentUsage}/${subscription.usageLimit}`);

Error Handling ​

All SDKs use a consistent error handling approach. Errors are categorized as:

  • Authentication Errors: Invalid API key or client ID
  • Validation Errors: Invalid input parameters
  • Resource Errors: Problems with the requested resource
  • Rate Limit Errors: Too many requests
  • Service Errors: Internal server errors

Example error handling:

python
# Python example
from paitient_secure_model import SecureModelClient, AuthenticationError, ValidationError

try:
    client = SecureModelClient(api_key="invalid-key", client_id="invalid-id")
    client.deploy_model(model_name="ZimaBlueAI/HuatuoGPT-o1-8B")
except AuthenticationError as e:
    print(f"Authentication error: {e}")
except ValidationError as e:
    print(f"Validation error: {e}")
except Exception as e:
    print(f"Other error: {e}")
javascript
// JavaScript example
const { SecureModelClient, AuthenticationError, ValidationError } = require('paitient-secure-model');

try {
  const client = new SecureModelClient({
    apiKey: "invalid-key", 
    clientId: "invalid-id"
  });
  await client.deploy({ modelName: "ZimaBlueAI/HuatuoGPT-o1-8B" });
} catch (error) {
  if (error instanceof AuthenticationError) {
    console.error("Authentication error:", error.message);
  } else if (error instanceof ValidationError) {
    console.error("Validation error:", error.message);
  } else {
    console.error("Other error:", error.message);
  }
}

REST API ​

For direct integration without using our SDKs, you can use our REST API. All endpoints require authentication using an API key and client ID passed in the request headers.

Authorization: Bearer your-api-key
X-Client-ID: your-client-id

The base URL for all API requests is:

https://api.paitient.ai/v1

See the REST API documentation for detailed information on all available endpoints.

Next Steps ​

Released under the MIT License.

Copyright © 2023 PaiTIENT.ai