mirror of
https://github.com/danswer-ai/danswer.git
synced 2025-05-04 17:00:24 +02:00
48 lines
1.3 KiB
Python
48 lines
1.3 KiB
Python
import json
|
|
from typing import Any
|
|
|
|
from langchain_core.messages.ai import AIMessage
|
|
from langchain_core.messages.tool import ToolCall
|
|
from langchain_core.messages.tool import ToolMessage
|
|
from pydantic import BaseModel
|
|
|
|
from onyx.natural_language_processing.utils import BaseTokenizer
|
|
|
|
# Langchain has their own version of pydantic which is version 1
|
|
|
|
|
|
def build_tool_message(
|
|
tool_call: ToolCall, tool_content: str | list[str | dict[str, Any]]
|
|
) -> ToolMessage:
|
|
return ToolMessage(
|
|
tool_call_id=tool_call["id"] or "",
|
|
name=tool_call["name"],
|
|
content=tool_content,
|
|
)
|
|
|
|
|
|
# TODO: does this NEED to be BaseModel__v1?
|
|
class ToolCallSummary(BaseModel):
|
|
tool_call_request: AIMessage
|
|
tool_call_result: ToolMessage
|
|
|
|
# This is a workaround to allow arbitrary types in the model
|
|
# TODO: Remove this once we have a better solution
|
|
class Config:
|
|
arbitrary_types_allowed = True
|
|
|
|
|
|
def tool_call_tokens(
|
|
tool_call_summary: ToolCallSummary, llm_tokenizer: BaseTokenizer
|
|
) -> int:
|
|
request_tokens = len(
|
|
llm_tokenizer.encode(
|
|
json.dumps(tool_call_summary.tool_call_request.tool_calls[0]["args"])
|
|
)
|
|
)
|
|
result_tokens = len(
|
|
llm_tokenizer.encode(json.dumps(tool_call_summary.tool_call_result.content))
|
|
)
|
|
|
|
return request_tokens + result_tokens
|