mirror of
https://github.com/meta-llama/llama-stack.git
synced 2025-08-09 19:58:29 +00:00
adding the fiddlecube provider
This commit is contained in:
parent
7ec79c0297
commit
aea5b2745d
4 changed files with 145 additions and 0 deletions
|
@ -85,4 +85,13 @@ Provider `inline::meta-reference` for API `safety` does not work with the latest
|
||||||
config_class="llama_stack.providers.remote.safety.bedrock.BedrockSafetyConfig",
|
config_class="llama_stack.providers.remote.safety.bedrock.BedrockSafetyConfig",
|
||||||
),
|
),
|
||||||
),
|
),
|
||||||
|
remote_provider_spec(
|
||||||
|
api=Api.safety,
|
||||||
|
adapter=AdapterSpec(
|
||||||
|
adapter_type="fiddlecube",
|
||||||
|
pip_packages=["httpx"],
|
||||||
|
module="llama_stack.providers.remote.safety.fiddlecube",
|
||||||
|
config_class="llama_stack.providers.remote.safety.fiddlecube.FiddlecubeSafetyConfig",
|
||||||
|
),
|
||||||
|
),
|
||||||
]
|
]
|
||||||
|
|
18
llama_stack/providers/remote/safety/fiddlecube/__init__.py
Normal file
18
llama_stack/providers/remote/safety/fiddlecube/__init__.py
Normal file
|
@ -0,0 +1,18 @@
|
||||||
|
# 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
|
||||||
|
|
||||||
|
from .config import FiddlecubeSafetyConfig
|
||||||
|
|
||||||
|
|
||||||
|
async def get_adapter_impl(config: FiddlecubeSafetyConfig, _deps) -> Any:
|
||||||
|
from .fiddlecube import FiddlecubeSafetyAdapter
|
||||||
|
|
||||||
|
impl = FiddlecubeSafetyAdapter(config)
|
||||||
|
await impl.initialize()
|
||||||
|
return impl
|
14
llama_stack/providers/remote/safety/fiddlecube/config.py
Normal file
14
llama_stack/providers/remote/safety/fiddlecube/config.py
Normal file
|
@ -0,0 +1,14 @@
|
||||||
|
# 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 pydantic import BaseModel
|
||||||
|
|
||||||
|
from llama_models.schema_utils import json_schema_type
|
||||||
|
|
||||||
|
|
||||||
|
@json_schema_type
|
||||||
|
class FiddlecubeSafetyConfig(BaseModel):
|
||||||
|
pass
|
104
llama_stack/providers/remote/safety/fiddlecube/fiddlecube.py
Normal file
104
llama_stack/providers/remote/safety/fiddlecube/fiddlecube.py
Normal file
|
@ -0,0 +1,104 @@
|
||||||
|
# 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.
|
||||||
|
|
||||||
|
import json
|
||||||
|
import logging
|
||||||
|
|
||||||
|
from typing import Any, Dict, List
|
||||||
|
|
||||||
|
from llama_stack.apis.inference import Message
|
||||||
|
|
||||||
|
from llama_stack.apis.safety import (
|
||||||
|
RunShieldResponse,
|
||||||
|
Safety,
|
||||||
|
SafetyViolation,
|
||||||
|
ViolationLevel,
|
||||||
|
)
|
||||||
|
from llama_stack.apis.shields import Shield
|
||||||
|
from llama_stack.providers.datatypes import ShieldsProtocolPrivate
|
||||||
|
|
||||||
|
from .config import FiddlecubeSafetyConfig
|
||||||
|
|
||||||
|
|
||||||
|
logger = logging.getLogger(__name__)
|
||||||
|
|
||||||
|
|
||||||
|
class FiddlecubeSafetyAdapter(Safety, ShieldsProtocolPrivate):
|
||||||
|
def __init__(self, config: FiddlecubeSafetyConfig) -> None:
|
||||||
|
self.config = config
|
||||||
|
self.registered_shields = []
|
||||||
|
|
||||||
|
async def initialize(self) -> None:
|
||||||
|
pass
|
||||||
|
|
||||||
|
async def shutdown(self) -> None:
|
||||||
|
pass
|
||||||
|
|
||||||
|
async def register_shield(self, shield: Shield) -> None:
|
||||||
|
pass
|
||||||
|
|
||||||
|
async def run_shield(
|
||||||
|
self, shield_id: str, messages: List[Message], params: Dict[str, Any] = None
|
||||||
|
) -> RunShieldResponse:
|
||||||
|
# Set up FiddleCube API using httpx
|
||||||
|
# [TBD] convert the `messages` into format FiddleCube expects
|
||||||
|
# make a call to the API for guardrails
|
||||||
|
# convert the [TBD] response into the format RunShieldResponse expects
|
||||||
|
# return the response
|
||||||
|
return RunShieldResponse()
|
||||||
|
|
||||||
|
shield = await self.shield_store.get_shield(shield_id)
|
||||||
|
if not shield:
|
||||||
|
raise ValueError(f"Shield {shield_id} not found")
|
||||||
|
|
||||||
|
"""This is the implementation for the bedrock guardrails. The input to the guardrails is to be of this format
|
||||||
|
```content = [
|
||||||
|
{
|
||||||
|
"text": {
|
||||||
|
"text": "Is the AB503 Product a better investment than the S&P 500?"
|
||||||
|
}
|
||||||
|
}
|
||||||
|
]```
|
||||||
|
However the incoming messages are of this type UserMessage(content=....) coming from
|
||||||
|
https://github.com/meta-llama/llama-models/blob/main/models/llama3/api/datatypes.py
|
||||||
|
|
||||||
|
They contain content, role . For now we will extract the content and default the "qualifiers": ["query"]
|
||||||
|
"""
|
||||||
|
|
||||||
|
shield_params = shield.params
|
||||||
|
logger.debug(f"run_shield::{shield_params}::messages={messages}")
|
||||||
|
|
||||||
|
# - convert the messages into format Bedrock expects
|
||||||
|
content_messages = []
|
||||||
|
for message in messages:
|
||||||
|
content_messages.append({"text": {"text": message.content}})
|
||||||
|
logger.debug(f"run_shield::final:messages::{json.dumps(content_messages, indent=2)}:")
|
||||||
|
|
||||||
|
response = self.bedrock_runtime_client.apply_guardrail(
|
||||||
|
guardrailIdentifier=shield.provider_resource_id,
|
||||||
|
guardrailVersion=shield_params["guardrailVersion"],
|
||||||
|
source="OUTPUT", # or 'INPUT' depending on your use case
|
||||||
|
content=content_messages,
|
||||||
|
)
|
||||||
|
if response["action"] == "GUARDRAIL_INTERVENED":
|
||||||
|
user_message = ""
|
||||||
|
metadata = {}
|
||||||
|
for output in response["outputs"]:
|
||||||
|
# guardrails returns a list - however for this implementation we will leverage the last values
|
||||||
|
user_message = output["text"]
|
||||||
|
for assessment in response["assessments"]:
|
||||||
|
# guardrails returns a list - however for this implementation we will leverage the last values
|
||||||
|
metadata = dict(assessment)
|
||||||
|
|
||||||
|
return RunShieldResponse(
|
||||||
|
violation=SafetyViolation(
|
||||||
|
user_message=user_message,
|
||||||
|
violation_level=ViolationLevel.ERROR,
|
||||||
|
metadata=metadata,
|
||||||
|
)
|
||||||
|
)
|
||||||
|
|
||||||
|
return RunShieldResponse()
|
Loading…
Add table
Add a link
Reference in a new issue