phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-06-28 02:53:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

migrate evals to resource (#421)

Browse source 
* migrate evals to resource

* remove listing of providers's evals

* change the order of params in register

* fix after rebase

* linter fix

---------

Co-authored-by: Dinesh Yeduguru <dineshyv@fb.com>
This commit is contained in:
Dinesh Yeduguru 2024-11-11 17:24:03 -08:00 • committed by GitHub
parent b95cb5308f
commit 3802edfc50
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 63 additions and 56 deletions
Download patch file Download diff file Expand all files Collapse all files

28
llama_stack/apis/eval_tasks/eval_tasks.py
View file

@ -7,12 +7,14 @@ from typing import Any, Dict, List, Literal, Optional, Protocol, runtime_checkab
from llama_models.schema_utils import json_schema_type, webmethod
from pydantic import BaseModel, Field
from pydantic import Field
from llama_stack.apis.resource import Resource
@json_schema_type
class EvalTaskDef(BaseModel):
identifier: str
class EvalTask(Resource):
type: Literal["eval_task"] = "eval_task"
dataset_id: str
scoring_functions: List[str]
metadata: Dict[str, Any] = Field(
@ -21,23 +23,21 @@ class EvalTaskDef(BaseModel):
)
@json_schema_type
class EvalTaskDefWithProvider(EvalTaskDef):
type: Literal["eval_task"] = "eval_task"
provider_id: str = Field(
description="ID of the provider which serves this dataset",
)
@runtime_checkable
class EvalTasks(Protocol):
@webmethod(route="/eval_tasks/list", method="GET")
async def list_eval_tasks(self) -> List[EvalTaskDefWithProvider]: ...
async def list_eval_tasks(self) -> List[EvalTask]: ...
@webmethod(route="/eval_tasks/get", method="GET")
async def get_eval_task(self, name: str) -> Optional[EvalTaskDefWithProvider]: ...
async def get_eval_task(self, name: str) -> Optional[EvalTask]: ...
@webmethod(route="/eval_tasks/register", method="POST")
async def register_eval_task(
self, eval_task_def: EvalTaskDefWithProvider
self,
eval_task_id: str,
dataset_id: str,
scoring_functions: List[str],
provider_eval_task_id: Optional[str] = None,
provider_id: Optional[str] = None,
metadata: Optional[Dict[str, Any]] = None,
) -> None: ...