forked from phoenix-oss/llama-stack-mirror
ea1faae50e
# What does this PR do? - Fully deprecate eval/tasks [//]: # (If resolving an issue, uncomment and update the line below) Closes #1088 NOTE: this will be a breaking change. We have introduced the new API in 0.1.3 . Notebook has been updated to use the new endpoints. ## Test Plan ``` pytest -v -s --nbval-lax ./docs/notebooks/Llama_Stack_Benchmark_Evals.ipynb ``` <img width="611" alt="image" src="https://github.com/user-attachments/assets/79f6efe1-81ba-494e-bf36-1fc0c2b9bc6f" /> cc @SLR722 for awareness [//]: # (## Documentation)
85 lines
2.9 KiB
Python
85 lines
2.9 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, Union
|
|
|
|
from pydantic import BaseModel, Field
|
|
from typing_extensions import Annotated
|
|
|
|
from llama_stack.apis.agents import AgentConfig
|
|
from llama_stack.apis.common.job_types import Job, JobStatus
|
|
from llama_stack.apis.inference import SamplingParams, SystemMessage
|
|
from llama_stack.apis.scoring import ScoringResult
|
|
from llama_stack.apis.scoring_functions import ScoringFnParams
|
|
from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
|
|
|
|
|
|
@json_schema_type
|
|
class ModelCandidate(BaseModel):
|
|
type: Literal["model"] = "model"
|
|
model: str
|
|
sampling_params: SamplingParams
|
|
system_message: Optional[SystemMessage] = None
|
|
|
|
|
|
@json_schema_type
|
|
class AgentCandidate(BaseModel):
|
|
type: Literal["agent"] = "agent"
|
|
config: AgentConfig
|
|
|
|
|
|
EvalCandidate = register_schema(
|
|
Annotated[Union[ModelCandidate, AgentCandidate], Field(discriminator="type")],
|
|
name="EvalCandidate",
|
|
)
|
|
|
|
|
|
@json_schema_type
|
|
class BenchmarkConfig(BaseModel):
|
|
eval_candidate: EvalCandidate
|
|
scoring_params: Dict[str, ScoringFnParams] = Field(
|
|
description="Map between scoring function id and parameters for each scoring function you want to run",
|
|
default_factory=dict,
|
|
)
|
|
num_examples: Optional[int] = Field(
|
|
description="Number of examples to evaluate (useful for testing), if not provided, all examples in the dataset will be evaluated",
|
|
default=None,
|
|
)
|
|
# we could optinally add any specific dataset config here
|
|
|
|
|
|
@json_schema_type
|
|
class EvaluateResponse(BaseModel):
|
|
generations: List[Dict[str, Any]]
|
|
# each key in the dict is a scoring function name
|
|
scores: Dict[str, ScoringResult]
|
|
|
|
|
|
class Eval(Protocol):
|
|
@webmethod(route="/eval/benchmarks/{benchmark_id}/jobs", method="POST")
|
|
async def run_eval(
|
|
self,
|
|
benchmark_id: str,
|
|
task_config: BenchmarkConfig,
|
|
) -> Job: ...
|
|
|
|
@webmethod(route="/eval/benchmarks/{benchmark_id}/evaluations", method="POST")
|
|
async def evaluate_rows(
|
|
self,
|
|
benchmark_id: str,
|
|
input_rows: List[Dict[str, Any]],
|
|
scoring_functions: List[str],
|
|
task_config: BenchmarkConfig,
|
|
) -> EvaluateResponse: ...
|
|
|
|
@webmethod(route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}", method="GET")
|
|
async def job_status(self, benchmark_id: str, job_id: str) -> Optional[JobStatus]: ...
|
|
|
|
@webmethod(route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}", method="DELETE")
|
|
async def job_cancel(self, benchmark_id: str, job_id: str) -> None: ...
|
|
|
|
@webmethod(route="/eval/benchmarks/{benchmark_id}/jobs/{job_id}/result", method="GET")
|
|
async def job_result(self, benchmark_id: str, job_id: str) -> EvaluateResponse: ...
|