ForcePilot/src/models/rerank.py

170 lines
6.4 KiB
Python
Raw Normal View History

import asyncio
import os
from abc import ABC, abstractmethod
from collections.abc import Iterable, Sequence
from typing import Any
import aiohttp
2025-02-23 16:38:56 +08:00
import numpy as np
from src import config
from src.utils import get_docker_safe_url, logger
2025-02-23 16:38:56 +08:00
def sigmoid(x):
return 1 / (1 + np.exp(-x))
class BaseReranker(ABC):
def __init__(self, model_name, api_key, base_url, **kwargs):
self.url = get_docker_safe_url(base_url)
self.model = model_name
self.api_key = api_key
self.headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
self.session: aiohttp.ClientSession | None = None
self.timeout = aiohttp.ClientTimeout(total=30)
self.parameters: dict[str, Any] = dict(kwargs.get("parameters", {}))
async def _ensure_session(self) -> None:
if self.session is None or self.session.closed:
self.session = aiohttp.ClientSession(headers=self.headers, timeout=self.timeout)
@abstractmethod
def _build_payload(self, query: str, documents: list[str], max_length: int) -> dict[str, Any]:
raise NotImplementedError
@abstractmethod
def _extract_results(self, result: dict[str, Any]) -> list[dict[str, Any]]:
raise NotImplementedError
async def acompute_score(
self,
sentence_pairs: Sequence[Sequence[str]],
batch_size: int = 32,
max_length: int = 512,
normalize: bool = True,
) -> list[float]:
if not sentence_pairs or len(sentence_pairs) < 2:
return []
2025-02-23 16:38:56 +08:00
query, sentences = sentence_pairs[0], sentence_pairs[1]
documents = [sentences] if isinstance(sentences, str) else list(sentences)
if not documents:
return []
2025-02-23 16:38:56 +08:00
await self._ensure_session()
all_scores: list[float] = []
batch_size = max(1, int(batch_size))
total_batches = (len(documents) + batch_size - 1) // batch_size
for batch_no, start in enumerate(range(0, len(documents), batch_size), start=1):
batch = documents[start : start + batch_size]
try:
scores = await self._batch_rerank(query, batch, max_length=max_length)
all_scores.extend(scores)
logger.debug(f"Reranking batch {batch_no}/{total_batches} completed")
except Exception as exc: # noqa: BLE001
logger.error(f"Reranking batch {batch_no} failed: {exc}")
all_scores.extend([0.5] * len(batch))
2025-02-23 16:38:56 +08:00
if normalize:
all_scores = [float(sigmoid(score)) for score in all_scores]
2025-02-23 16:38:56 +08:00
return all_scores
async def _batch_rerank(self, query: str, documents: Iterable[str], max_length: int) -> list[float]:
docs = list(documents)
if not docs:
return []
payload = self._build_payload(query, docs, max_length)
await self._ensure_session()
assert self.session is not None
try:
async with self.session.post(self.url, json=payload) as response:
response.raise_for_status()
result: dict[str, Any] = await response.json()
except TimeoutError as exc:
total_timeout = self.timeout.total if self.timeout else 0.0
logger.error(f"Reranking request timeout after {total_timeout:.1f}s")
raise exc
except aiohttp.ClientError as exc:
logger.error(f"Reranking request failed: {exc}")
raise exc
processed = sorted(self._extract_results(result), key=lambda item: item.get("index", 0))
return [float(entry.get("relevance_score", 0.0)) for entry in processed]
def compute_score(self, sentence_pairs, batch_size=256, max_length=512, normalize=False):
try:
_ = asyncio.get_running_loop()
except RuntimeError:
return asyncio.run(self.acompute_score(sentence_pairs, batch_size, max_length, normalize))
raise RuntimeError("compute_score cannot be used while an event loop is running. Use acompute_score instead.")
async def aclose(self) -> None:
if self.session and not self.session.closed:
await self.session.close()
def __del__(self) -> None:
if self.session and not self.session.closed:
try:
loop = asyncio.get_event_loop()
except RuntimeError:
asyncio.run(self.aclose())
return
if loop.is_closed():
asyncio.run(self.aclose())
elif not loop.is_running():
loop.run_until_complete(self.aclose())
class OpenAIReranker(BaseReranker):
def _build_payload(self, query: str, documents: list[str], max_length: int) -> dict[str, Any]:
return {
"model": self.model,
"query": query,
"documents": documents,
"max_chunks_per_doc": max_length,
}
def _extract_results(self, result: dict[str, Any]) -> list[dict[str, Any]]:
return list(result.get("results", []))
class DashscopeReranker(BaseReranker):
def _build_payload(self, query: str, documents: list[str], max_length: int) -> dict[str, Any]:
params = {"top_n": len(documents), "return_documents": False}
instruct = self.parameters.get("instruct")
if instruct:
params["instruct"] = instruct
return {
"model": self.model,
"input": {"query": query, "documents": documents},
"parameters": params,
}
def _extract_results(self, result: dict[str, Any]) -> list[dict[str, Any]]:
return list(result.get("output", {}).get("results", []))
def get_reranker(model_id, **kwargs):
2025-05-24 11:29:45 +08:00
support_rerankers = config.reranker_names.keys()
assert model_id in support_rerankers, f"Unsupported Reranker: {model_id}, only support {support_rerankers}"
model_info = config.reranker_names[model_id]
2025-10-22 11:51:32 +08:00
base_url = model_info.base_url
api_key = os.getenv(model_info.api_key) or model_info.api_key
2025-10-22 11:51:32 +08:00
assert api_key, f"{model_info.name} api_key is required"
provider = model_id.split("/", maxsplit=1)[0] if "/" in model_id else ""
if provider in {"siliconflow", "vllm"}:
return OpenAIReranker(model_name=model_info.name, api_key=api_key, base_url=base_url, **kwargs)
if provider == "dashscope":
return DashscopeReranker(model_name=model_info.name, api_key=api_key, base_url=base_url, **kwargs)
return OpenAIReranker(model_name=model_info.name, api_key=api_key, base_url=base_url, **kwargs)