Get Alignment Data
Retrieve detailed information about a specific alignment project.
This endpoint fetches comprehensive details about an alignment project including metadata, associated documents, training progress, status, performance metrics, and evaluation information.
Path Parameters:
id: Unique identifier of the alignment project
Returns:
id: Alignment project identifiername: Project namebase_model: Base model identifier used for alignmentstatus: Project status (pending,completed,failed,deploying,evaluating,undeployed,evaluated)created_date: ISO timestamp when project was createdcompleted_date(optional): ISO timestamp when project completeddocuments: List of document IDs used for trainingdocument_count: Number of documents usedcreator: Username or ID of the project creatorperformance_metrics(optional): Training performance metrics (loss, accuracy, etc.)progress(optional): Training progress percentage (0-100)estimated_completion(optional): Estimated completion timeerror(optional): Error message if project failedevaluation_id(optional): Auto-evaluation ID if triggeredmodel_id(optional): Deployed model ID for evaluation
Raises:
404: If alignment project not found or doesn’t belong to user
Example Request:
GET /api/v3/alignment/alignment-xyz789
Headers: {"Authorization": "Bearer <api_key>"}
Example Response (Completed):
{
"id": "alignment-xyz789",
"name": "Customer Support Domain Alignment",
"base_model": "nugen-flash-instruct",
"status": "completed",
"created_date": "2024-01-15T10:30:00Z",
"completed_date": "2024-01-15T14:45:00Z",
"documents": ["doc-abc123", "doc-xyz456", "doc-def789"],
"document_count": 3,
"creator": "user-123",
"performance_metrics": {
"final_loss": 0.15,
"accuracy": 0.92,
"perplexity": 2.3
},
"evaluation_id": "eval-def456",
"model_id": "alignment-xyz789-deployed"
}
Example Response (In Progress):
{
"id": "alignment-abc456",
"name": "Technical Documentation Alignment",
"base_model": "nugen-flash-instruct",
"status": "pending",
"created_date": "2024-01-16T09:00:00Z",
"documents": ["doc-111", "doc-222"],
"document_count": 2,
"creator": "user-123",
"progress": 45,
"estimated_completion": "2024-01-16T11:30:00Z"
}
Notes:
- Returns full project details including training progress and performance
model_idis only available for projects that have been deployed for evaluationevaluation_idis only present if auto-evaluation was triggeredperformance_metricsare populated after training completesprogressandestimated_completionare available during trainingerrorfield contains failure details if status isfailed
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Path Parameters
Response
Returns detailed information about a specific alignment project, including metadata, associated documents, training progress, status, performance metrics, and evaluation information
Unique identifier for the alignment project
"alignment-abc123"
"alignment-xyz789"
Human-readable name for the alignment project
"Customer Support Model"
"Legal Document Assistant"
"Medical Q&A Bot"
ID of the base model used for alignment training
"nugen-flash-instruct"
"model-20240229"
Current status of the alignment project (PENDING, COMPLETED, FAILED)
pending, completed, deploying, evaluating, undeployed, evaluated, failed ISO 8601 timestamp when the project was created
"2024-01-15T10:30:00Z"
"2024-02-01T09:15:00Z"
List of document IDs used for alignment training
["doc-abc123", "doc-def456"]["doc-xyz789"]Total number of documents used for training
x >= 05
12
3
Email/username of the user who created this project
"user@example.com"
"admin@company.com"
ISO 8601 timestamp when the project completed (null if still in progress or failed)
"2024-01-15T12:45:00Z"
Performance comparison metrics showing before/after alignment improvements (available after evaluation)
Training progress percentage (0-100). Null if not started or completed.
0 <= x <= 1000
Estimated completion time (ISO 8601 timestamp). Null if not available or completed.
"2024-01-15T14:00:00Z"
Error message if the alignment project failed. Null if no error.
"Training failed: insufficient data"
Auto-evaluation ID if automatic evaluation was triggered after alignment completion
"eval-abc123"
Deployed model ID if the aligned model was automatically deployed for evaluation
"alignment-abc123-deployed"