chore: deprecate eval task (Sync updates from stainless branch: main) (#150)

# What does this PR do?
- See llamastack/llama-stack#1186

## Test Plan
- Test in llamastack/llama-stack#1186

[//]: # (## Documentation)
[//]: # (- [ ] Added a Changelog entry if the change is significant)

Author: yanxi0830

src/llama_stack_client/_client.py

@@ -40,7 +40,6 @@
     telemetry,
     vector_io,
     benchmarks,
-    eval_tasks,
     toolgroups,
     vector_dbs,
     batch_inference,
@@ -94,7 +93,6 @@ class LlamaStackClient(SyncAPIClient):
     datasetio: datasetio.DatasetioResource
     scoring: scoring.ScoringResource
     scoring_functions: scoring_functions.ScoringFunctionsResource
-    eval_tasks: eval_tasks.EvalTasksResource
     benchmarks: benchmarks.BenchmarksResource
     with_raw_response: LlamaStackClientWithRawResponse
     with_streaming_response: LlamaStackClientWithStreamedResponse
@@ -177,7 +175,6 @@ def __init__(
         self.datasetio = datasetio.DatasetioResource(self)
         self.scoring = scoring.ScoringResource(self)
         self.scoring_functions = scoring_functions.ScoringFunctionsResource(self)
-        self.eval_tasks = eval_tasks.EvalTasksResource(self)
         self.benchmarks = benchmarks.BenchmarksResource(self)
         self.with_raw_response = LlamaStackClientWithRawResponse(self)
         self.with_streaming_response = LlamaStackClientWithStreamedResponse(self)
@@ -312,7 +309,6 @@ class AsyncLlamaStackClient(AsyncAPIClient):
     datasetio: datasetio.AsyncDatasetioResource
     scoring: scoring.AsyncScoringResource
     scoring_functions: scoring_functions.AsyncScoringFunctionsResource
-    eval_tasks: eval_tasks.AsyncEvalTasksResource
     benchmarks: benchmarks.AsyncBenchmarksResource
     with_raw_response: AsyncLlamaStackClientWithRawResponse
     with_streaming_response: AsyncLlamaStackClientWithStreamedResponse
@@ -395,7 +391,6 @@ def __init__(
         self.datasetio = datasetio.AsyncDatasetioResource(self)
         self.scoring = scoring.AsyncScoringResource(self)
         self.scoring_functions = scoring_functions.AsyncScoringFunctionsResource(self)
-        self.eval_tasks = eval_tasks.AsyncEvalTasksResource(self)
         self.benchmarks = benchmarks.AsyncBenchmarksResource(self)
         self.with_raw_response = AsyncLlamaStackClientWithRawResponse(self)
         self.with_streaming_response = AsyncLlamaStackClientWithStreamedResponse(self)
@@ -533,7 +528,6 @@ def __init__(self, client: LlamaStackClient) -> None:
         self.datasetio = datasetio.DatasetioResourceWithRawResponse(client.datasetio)
         self.scoring = scoring.ScoringResourceWithRawResponse(client.scoring)
         self.scoring_functions = scoring_functions.ScoringFunctionsResourceWithRawResponse(client.scoring_functions)
-        self.eval_tasks = eval_tasks.EvalTasksResourceWithRawResponse(client.eval_tasks)
         self.benchmarks = benchmarks.BenchmarksResourceWithRawResponse(client.benchmarks)
 
 
@@ -565,7 +559,6 @@ def __init__(self, client: AsyncLlamaStackClient) -> None:
         self.scoring_functions = scoring_functions.AsyncScoringFunctionsResourceWithRawResponse(
             client.scoring_functions
         )
-        self.eval_tasks = eval_tasks.AsyncEvalTasksResourceWithRawResponse(client.eval_tasks)
         self.benchmarks = benchmarks.AsyncBenchmarksResourceWithRawResponse(client.benchmarks)
 
 
@@ -597,7 +590,6 @@ def __init__(self, client: LlamaStackClient) -> None:
         self.scoring_functions = scoring_functions.ScoringFunctionsResourceWithStreamingResponse(
             client.scoring_functions
         )
-        self.eval_tasks = eval_tasks.EvalTasksResourceWithStreamingResponse(client.eval_tasks)
         self.benchmarks = benchmarks.BenchmarksResourceWithStreamingResponse(client.benchmarks)
 
 
@@ -631,7 +623,6 @@ def __init__(self, client: AsyncLlamaStackClient) -> None:
         self.scoring_functions = scoring_functions.AsyncScoringFunctionsResourceWithStreamingResponse(
             client.scoring_functions
         )
-        self.eval_tasks = eval_tasks.AsyncEvalTasksResourceWithStreamingResponse(client.eval_tasks)
         self.benchmarks = benchmarks.AsyncBenchmarksResourceWithStreamingResponse(client.benchmarks)
 
 

src/llama_stack_client/resources/__init__.py

@@ -128,14 +128,6 @@
     BenchmarksResourceWithStreamingResponse,
     AsyncBenchmarksResourceWithStreamingResponse,
 )
-from .eval_tasks import (
-    EvalTasksResource,
-    AsyncEvalTasksResource,
-    EvalTasksResourceWithRawResponse,
-    AsyncEvalTasksResourceWithRawResponse,
-    EvalTasksResourceWithStreamingResponse,
-    AsyncEvalTasksResourceWithStreamingResponse,
-)
 from .toolgroups import (
     ToolgroupsResource,
     AsyncToolgroupsResource,
@@ -326,12 +318,6 @@
     "AsyncScoringFunctionsResourceWithRawResponse",
     "ScoringFunctionsResourceWithStreamingResponse",
     "AsyncScoringFunctionsResourceWithStreamingResponse",
-    "EvalTasksResource",
-    "AsyncEvalTasksResource",
-    "EvalTasksResourceWithRawResponse",
-    "AsyncEvalTasksResourceWithRawResponse",
-    "EvalTasksResourceWithStreamingResponse",
-    "AsyncEvalTasksResourceWithStreamingResponse",
     "BenchmarksResource",
     "AsyncBenchmarksResource",
     "BenchmarksResourceWithRawResponse",

src/llama_stack_client/resources/eval/eval.py

@@ -67,7 +67,7 @@ def with_streaming_response(self) -> EvalResourceWithStreamingResponse:
 
     def evaluate_rows(
         self,
-        task_id: str,
+        benchmark_id: str,
         *,
         input_rows: Iterable[Dict[str, Union[bool, float, str, Iterable[object], object, None]]],
         scoring_functions: List[str],
@@ -89,10 +89,10 @@ def evaluate_rows(
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
-        if not task_id:
-            raise ValueError(f"Expected a non-empty value for `task_id` but received {task_id!r}")
+        if not benchmark_id:
+            raise ValueError(f"Expected a non-empty value for `benchmark_id` but received {benchmark_id!r}")
         return self._post(
-            f"/v1/eval/tasks/{task_id}/evaluations",
+            f"/v1/eval/benchmarks/{benchmark_id}/evaluations",
             body=maybe_transform(
                 {
                     "input_rows": input_rows,
@@ -151,7 +151,7 @@ def evaluate_rows_alpha(
 
     def run_eval(
         self,
-        task_id: str,
+        benchmark_id: str,
         *,
         task_config: BenchmarkConfigParam,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -171,10 +171,10 @@ def run_eval(
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
-        if not task_id:
-            raise ValueError(f"Expected a non-empty value for `task_id` but received {task_id!r}")
+        if not benchmark_id:
+            raise ValueError(f"Expected a non-empty value for `benchmark_id` but received {benchmark_id!r}")
         return self._post(
-            f"/v1/eval/tasks/{task_id}/jobs",
+            f"/v1/eval/benchmarks/{benchmark_id}/jobs",
             body=maybe_transform({"task_config": task_config}, eval_run_eval_params.EvalRunEvalParams),
             options=make_request_options(
                 extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout
@@ -242,7 +242,7 @@ def with_streaming_response(self) -> AsyncEvalResourceWithStreamingResponse:
 
     async def evaluate_rows(
         self,
-        task_id: str,
+        benchmark_id: str,
         *,
         input_rows: Iterable[Dict[str, Union[bool, float, str, Iterable[object], object, None]]],
         scoring_functions: List[str],
@@ -264,10 +264,10 @@ async def evaluate_rows(
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
-        if not task_id:
-            raise ValueError(f"Expected a non-empty value for `task_id` but received {task_id!r}")
+        if not benchmark_id:
+            raise ValueError(f"Expected a non-empty value for `benchmark_id` but received {benchmark_id!r}")
         return await self._post(
-            f"/v1/eval/tasks/{task_id}/evaluations",
+            f"/v1/eval/benchmarks/{benchmark_id}/evaluations",
             body=await async_maybe_transform(
                 {
                     "input_rows": input_rows,
@@ -326,7 +326,7 @@ async def evaluate_rows_alpha(
 
     async def run_eval(
         self,
-        task_id: str,
+        benchmark_id: str,
         *,
         task_config: BenchmarkConfigParam,
         # Use the following arguments if you need to pass additional parameters to the API that aren't available via kwargs.
@@ -346,10 +346,10 @@ async def run_eval(
 
           timeout: Override the client-level default timeout for this request, in seconds
         """
-        if not task_id:
-            raise ValueError(f"Expected a non-empty value for `task_id` but received {task_id!r}")
+        if not benchmark_id:
+            raise ValueError(f"Expected a non-empty value for `benchmark_id` but received {benchmark_id!r}")
         return await self._post(
-            f"/v1/eval/tasks/{task_id}/jobs",
+            f"/v1/eval/benchmarks/{benchmark_id}/jobs",
             body=await async_maybe_transform({"task_config": task_config}, eval_run_eval_params.EvalRunEvalParams),
             options=make_request_options(
                 extra_headers=extra_headers, extra_query=extra_query, extra_body=extra_body, timeout=timeout

src/llama_stack_client/types/__init__.py

@@ -86,7 +86,6 @@
 from .vector_io_query_params import VectorIoQueryParams as VectorIoQueryParams
 from .benchmark_list_response import BenchmarkListResponse as BenchmarkListResponse
 from .dataset_register_params import DatasetRegisterParams as DatasetRegisterParams
-from .eval_task_list_response import EvalTaskListResponse as EvalTaskListResponse
 from .list_providers_response import ListProvidersResponse as ListProvidersResponse
 from .scoring_fn_params_param import ScoringFnParamsParam as ScoringFnParamsParam
 from .toolgroup_list_response import ToolgroupListResponse as ToolgroupListResponse
@@ -98,7 +97,6 @@
 from .benchmark_register_params import BenchmarkRegisterParams as BenchmarkRegisterParams
 from .dataset_retrieve_response import DatasetRetrieveResponse as DatasetRetrieveResponse
 from .eval_evaluate_rows_params import EvalEvaluateRowsParams as EvalEvaluateRowsParams
-from .eval_task_register_params import EvalTaskRegisterParams as EvalTaskRegisterParams
 from .list_tool_groups_response import ListToolGroupsResponse as ListToolGroupsResponse
 from .toolgroup_register_params import ToolgroupRegisterParams as ToolgroupRegisterParams
 from .vector_db_register_params import VectorDBRegisterParams as VectorDBRegisterParams

src/llama_stack_client/types/benchmark_config_param.py

@@ -3,7 +3,7 @@
 from __future__ import annotations
 
 from typing import Dict
-from typing_extensions import Literal, Required, TypedDict
+from typing_extensions import Required, TypedDict
 
 from .eval_candidate_param import EvalCandidateParam
 from .scoring_fn_params_param import ScoringFnParamsParam
@@ -16,6 +16,4 @@ class BenchmarkConfigParam(TypedDict, total=False):
 
     scoring_params: Required[Dict[str, ScoringFnParamsParam]]
 
-    type: Required[Literal["benchmark"]]
-
     num_examples: int