Home
Amazon Neptune
Ml Create Model Training Job Request Structure
Ml Create Model Training Job Request Structure
CreateModelTrainingJobRequest schema from Neptune
Type: object
Properties: 20
Required: 2
Database Graph Database Gremlin Neptune Property Graph RDF SPARQL
CreateModelTrainingJobRequest is a JSON Structure definition published by Amazon Neptune, describing 20 properties, of which 2 are required. It conforms to the https://json-structure.org/meta/core/v0/# meta-schema.
Properties
id
dataProcessingJobId
trainModelS3Location
previousModelTrainingJobId
sagemakerIamRoleArn
neptuneIamRoleArn
modelName
baseProcessingInstanceType
trainingInstanceType
trainingInstanceVolumeSizeInGB
trainingTimeOutInSeconds
maxHPONumberOfTrainingJobs
maxHPOParallelTrainingJobs
subnets
securityGroupIds
volumeEncryptionKMSKey
s3OutputEncryptionKMSKey
enableInterContainerTrafficEncryption
enableManagedSpotTraining
customModelTrainingParameters
Meta-schema: https://json-structure.org/meta/core/v0/#
JSON Structure
{
"$schema": "https://json-structure.org/meta/core/v0/#",
"$id": "https://raw.githubusercontent.com/api-evangelist/amazon-neptune/refs/heads/main/json-structure/ml-create-model-training-job-request-structure.json",
"name": "CreateModelTrainingJobRequest",
"description": "CreateModelTrainingJobRequest schema from Neptune",
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Unique identifier for the job."
},
"dataProcessingJobId": {
"type": "string",
"description": "Job ID of the completed data processing job."
},
"trainModelS3Location": {
"type": "string",
"description": "S3 location for model artifacts output."
},
"previousModelTrainingJobId": {
"type": "string",
"description": "Job ID of a previous training job for incremental training."
},
"sagemakerIamRoleArn": {
"type": "string"
},
"neptuneIamRoleArn": {
"type": "string"
},
"modelName": {
"type": "string",
"description": "The model type to train: rgcn (relational graph convolutional network), transe, distmult, rotate, or custom.",
"enum": [
"rgcn",
"transe",
"distmult",
"rotate",
"custom"
]
},
"baseProcessingInstanceType": {
"type": "string",
"description": "ML instance type for data preparation step."
},
"trainingInstanceType": {
"type": "string",
"description": "ML instance type for the training step.",
"default": "ml.p3.2xlarge"
},
"trainingInstanceVolumeSizeInGB": {
"type": "int32",
"description": "Disk volume size for training instance in GB."
},
"trainingTimeOutInSeconds": {
"type": "int32",
"description": "Training job timeout in seconds.",
"default": 86400
},
"maxHPONumberOfTrainingJobs": {
"type": "int32",
"description": "Maximum total training jobs for hyperparameter tuning. Minimum 10 recommended for meaningful results.",
"default": 2
},
"maxHPOParallelTrainingJobs": {
"type": "int32",
"description": "Maximum parallel training jobs.",
"default": 2
},
"subnets": {
"type": "array",
"items": {
"type": "string"
}
},
"securityGroupIds": {
"type": "array",
"items": {
"type": "string"
}
},
"volumeEncryptionKMSKey": {
"type": "string"
},
"s3OutputEncryptionKMSKey": {
"type": "string"
},
"enableInterContainerTrafficEncryption": {
"type": "boolean",
"default": true
},
"enableManagedSpotTraining": {
"type": "boolean",
"description": "Whether to use EC2 spot instances for training.",
"default": false
},
"customModelTrainingParameters": {
"type": "object",
"description": "Custom model training configuration.",
"properties": {
"sourceS3DirectoryPath": {
"type": "string",
"description": "S3 path to the custom training script directory."
},
"trainingEntryPointScript": {
"type": "string",
"description": "Name of the training entry point script."
},
"transformEntryPointScript": {
"type": "string",
"description": "Name of the transform entry point script."
}
}
}
},
"required": [
"dataProcessingJobId",
"trainModelS3Location"
]
}