Package com.codedstream.otterstream.inference.config

Class ModelConfig

java.lang.Object

com.codedstream.otterstream.inference.config.ModelConfig

public class ModelConfig
extends Object

Configuration for ML models in Otter Stream inference framework.

Defines how to locate, load, and identify a model. Supports both local models (file system, HDFS, S3) and remote models (REST endpoints, cloud services).

Model Sources:

• Local: file://, hdfs://, s3://
• Remote: HTTP/HTTPS endpoints
• Cloud: AWS SageMaker, Google Vertex AI, Azure ML

Usage Examples:

Local TensorFlow Model:

 ModelConfig config = ModelConfig.builder()
     .modelId("fraud-detector")
     .modelPath("file:///models/fraud_detection")
     .format(ModelFormat.TENSORFLOW_SAVEDMODEL)
     .modelVersion("2.0")
     .build();
 

Remote SageMaker Endpoint:

 AuthConfig auth = AuthConfig.builder()
     .apiKey("AWS_ACCESS_KEY")
     .build();

 ModelConfig config = ModelConfig.builder()
     .modelId("recommendation-engine")
     .format(ModelFormat.SAGEMAKER)
     .endpointUrl("https://runtime.sagemaker.us-east-1.amazonaws.com/...")
     .authConfig(auth)
     .build();
 

ONNX Model with Options:

 ModelConfig config = ModelConfig.builder()
     .modelId("image-classifier")
     .modelPath("s3://my-bucket/models/resnet50.onnx")
     .format(ModelFormat.ONNX)
     .modelOptions(Map.of(
         "providers", List.of("CUDAExecutionProvider"),  // Use GPU
         "intra_op_num_threads", 4
     ))
     .build();
 

Since:

1.0.0

Author:

Nestor Martourez, Sr Software and Data Streaming Engineer @ CodedStreams

See Also:

ModelFormat, AuthConfig

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	ModelConfig.Builder	Builder for creating ModelConfig instances.

Constructor Summary

Constructors

Constructor	Description
ModelConfig(String modelId, String modelPath, ModelFormat format, String modelName, String modelVersion, Map<String,Object> modelOptions, String endpointUrl, AuthConfig authConfig)	Constructs model configuration.

Method Summary

Modifier and Type	Method	Description
static ModelConfig.Builder	builder()	Creates a new builder for ModelConfig.
AuthConfig	getAuthConfig()
String	getEndpointUrl()
ModelFormat	getFormat()
String	getModelId()
String	getModelName()
Map<String,Object>	getModelOptions()
String	getModelPath()
String	getModelVersion()
boolean	isRemote()	Checks if this is a remote model configuration.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ModelConfig

public ModelConfig(String modelId,
                   String modelPath,
                   ModelFormat format,
                   String modelName,
                   String modelVersion,
                   Map<String,Object> modelOptions,
                   String endpointUrl,
                   AuthConfig authConfig)

Constructs model configuration.

Parameters:

modelId - unique identifier for this model

modelPath - file system or URL path to model

format - model format (TensorFlow, ONNX, PyTorch, etc.)

modelName - human-readable model name

modelVersion - version string

modelOptions - engine-specific model options

endpointUrl - remote endpoint URL (for remote models)

authConfig - authentication configuration (for remote models)

Method Detail

builder

public static ModelConfig.Builder builder()

Creates a new builder for ModelConfig.

Returns:

a new builder instance

getModelId

public String getModelId()

getModelPath

public String getModelPath()

getFormat

public ModelFormat getFormat()

getModelName

public String getModelName()

getModelVersion

public String getModelVersion()

getModelOptions

public Map<String,Object> getModelOptions()

getEndpointUrl

public String getEndpointUrl()

getAuthConfig

public AuthConfig getAuthConfig()

isRemote

public boolean isRemote()

Checks if this is a remote model configuration.

Returns:

true if model is accessed via remote endpoint