feat: Switch synthetic data generation API to jobs pattern

The API should behave in a way similar to existing training and eval
flows where a long running task is sent to background; the client
receives a job ID to follow status and extract artifacts.

Note: there are no providers for this API implemented yet, so no
implementation changes seem to be needed.

Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>

llama_stack/apis/synthetic_data_generation/synthetic_data_generation.py

@@ -4,11 +4,13 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
+from datetime import datetime
 from enum import Enum
-from typing import Any, Dict, List, Optional, Protocol, Union
+from typing import Any, Dict, List, Optional, Protocol
 
 from pydantic import BaseModel
 
+from llama_stack.apis.common.job_types import JobStatus
 from llama_stack.apis.inference import Message
 from llama_stack.schema_utils import json_schema_type, webmethod
 
@@ -34,18 +36,58 @@ class SyntheticDataGenerationRequest(BaseModel):
 
 
 @json_schema_type
-class SyntheticDataGenerationResponse(BaseModel):
-    """Response from the synthetic data generation. Batch of (prompt, response, score) tuples that pass the threshold."""
+class SyntheticDataGenerationJob(BaseModel):
+    job_uuid: str
+
+
+@json_schema_type
+class SyntheticDataGenerationJobStatusResponse(BaseModel):
+    """Status of a synthetic data generation job."""
+
+    job_uuid: str
+    status: JobStatus
+
+    scheduled_at: Optional[datetime] = None
+    started_at: Optional[datetime] = None
+    completed_at: Optional[datetime] = None
+
+    synthetic_data: List[Dict[str, Any]]
+    statistics: Optional[Dict[str, Any]] = None
+
+
+class ListSyntheticDataGenerationJobsResponse(BaseModel):
+    data: List[SyntheticDataGenerationJob]
+
+
+@json_schema_type
+class SyntheticDataGenerationJobArtifactsResponse(BaseModel):
+    job_uuid: str
 
     synthetic_data: List[Dict[str, Any]]
     statistics: Optional[Dict[str, Any]] = None
 
 
 class SyntheticDataGeneration(Protocol):
-    @webmethod(route="/synthetic-data-generation/generate")
+    @webmethod(route="/synthetic-data-generation/generate", method="POST")
     def synthetic_data_generate(
         self,
         dialogs: List[Message],
         filtering_function: FilteringFunction = FilteringFunction.none,
         model: Optional[str] = None,
-    ) -> Union[SyntheticDataGenerationResponse]: ...
+    ) -> SyntheticDataGenerationJob: ...
+
+    @webmethod(route="/synthetic-data-generation/jobs", method="GET")
+    async def get_synthetic_data_generation_jobs(self) -> ListSyntheticDataGenerationJobsResponse: ...
+
+    @webmethod(route="/synthetic-data-generation/job/status", method="GET")
+    async def get_synthetic_data_generation_job_status(
+        self, job_uuid: str
+    ) -> Optional[SyntheticDataGenerationJobStatusResponse]: ...
+
+    @webmethod(route="/synthetic-data-generation/job/cancel", method="POST")
+    async def cancel_synthetic_data_generation_job(self, job_uuid: str) -> None: ...
+
+    @webmethod(route="/synthetic-data-generation/job/artifacts", method="GET")
+    async def get_synthetic_data_generation_job_artifacts(
+        self, job_uuid: str
+    ) -> Optional[SyntheticDataGenerationJobArtifactsResponse]: ...