forked from phoenix-oss/llama-stack-mirror
983d6ce2df
# What does this PR do?
This PR kills the notion of "ShieldType". The impetus for this is the
realization:
> Why is keyword llama-guard appearing so many times everywhere,
sometimes with hyphens, sometimes with underscores?
Now that we have a notion of "provider specific resource identifiers"
and "user specific aliases" for those and the fact that this works with
models ("Llama3.1-8B-Instruct" <> "fireworks/llama-3pv1-..."), we can
follow the same rules for Shields.
So each Safety provider can make up a notion of identifiers it has
registered. This already happens with Bedrock correctly. We just
generalize it for Llama Guard, Prompt Guard, etc.
For Llama Guard, we further simplify by just adopting the underlying
model name itself as the identifier! No confusion necessary.
While doing this, I noticed a bug in our DistributionRegistry where we
weren't scoping identifiers by type. Fixed.
## Feature/Issue validation/testing/test plan
Ran (inference, safety, memory, agents) tests with ollama and fireworks
providers.
60 lines
1.8 KiB
Python
60 lines
1.8 KiB
Python
# Copyright (c) Meta Platforms, Inc. and affiliates.
|
|
# All rights reserved.
|
|
#
|
|
# This source code is licensed under the terms described in the LICENSE file in
|
|
# the root directory of this source tree.
|
|
from typing import Any, Dict, List, Literal, Optional, Protocol, runtime_checkable
|
|
|
|
from llama_models.schema_utils import json_schema_type, webmethod
|
|
|
|
from pydantic import BaseModel, Field
|
|
|
|
from llama_stack.apis.resource import Resource, ResourceType
|
|
|
|
|
|
class CommonEvalTaskFields(BaseModel):
|
|
dataset_id: str
|
|
scoring_functions: List[str]
|
|
metadata: Dict[str, Any] = Field(
|
|
default_factory=dict,
|
|
description="Metadata for this evaluation task",
|
|
)
|
|
|
|
|
|
@json_schema_type
|
|
class EvalTask(CommonEvalTaskFields, Resource):
|
|
type: Literal[ResourceType.eval_task.value] = ResourceType.eval_task.value
|
|
|
|
@property
|
|
def eval_task_id(self) -> str:
|
|
return self.identifier
|
|
|
|
@property
|
|
def provider_eval_task_id(self) -> str:
|
|
return self.provider_resource_id
|
|
|
|
|
|
class EvalTaskInput(CommonEvalTaskFields, BaseModel):
|
|
eval_task_id: str
|
|
provider_id: Optional[str] = None
|
|
provider_eval_task_id: Optional[str] = None
|
|
|
|
|
|
@runtime_checkable
|
|
class EvalTasks(Protocol):
|
|
@webmethod(route="/eval_tasks/list", method="GET")
|
|
async def list_eval_tasks(self) -> List[EvalTask]: ...
|
|
|
|
@webmethod(route="/eval_tasks/get", method="GET")
|
|
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_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: ...
|