Spaces:
Sleeping
Sleeping
File size: 2,144 Bytes
69601d4 443f81f 69601d4 443f81f 69601d4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 |
from langchain_core.output_parsers import JsonOutputParser
from langchain_core.prompts import PromptTemplate
from ..utils.functions import readYaml, getConfig
from ..utils.exceptions import CustomException
from pydantic import Field, BaseModel
from langchain_groq import ChatGroq
from ..utils.logger import logger
import os
class QueryRephraseOutput(BaseModel):
rephrasedOutput: str | None = Field(
description="A clear and concise rephrased version of the user's query. If the query is unclear, invalid, or requires clarification, this will be `None`."
)
doubt: str | None = Field(
description="A message indicating any doubt, required clarification, or reason why the input query is invalid. If the query is successfully rephrased, this will be `None`."
)
class QueryRephaser:
def __init__(self):
logger.info("Initializing QueryRephaser.")
self.yamlPath = os.path.join(os.getcwd(), "params.yaml")
self.config = getConfig(os.path.join(os.getcwd(), "config.ini"))
def getQueryRephraserChain(self):
try:
logger.info("Constructing query rephraser chain.")
queryRephraseParser = JsonOutputParser(pydantic_object = QueryRephraseOutput)
queryRephrasePrompt = PromptTemplate(
template = readYaml(self.yamlPath)["queryRephraserAgentPrompt"],
input_variables = ["metadata", "query"],
partial_variables = {"format_instructions": queryRephraseParser.get_format_instructions()}
)
llm = ChatGroq(
model=self.config.get("QUERYREPHRASER", "model"),
temperature=self.config.getfloat("QUERYREPHRASER", "temperature"),
max_tokens=self.config.getint("QUERYREPHRASER", "maxTokens")
)
queryRephraseChain = queryRephrasePrompt | llm | queryRephraseParser
logger.info("Query rephraser chain constructed successfully.")
return queryRephraseChain
except Exception as e:
logger.error(f"Error constructing query rephraser chain: {e}")
raise CustomException(e) |