48 lines
1.3 KiB
Python
48 lines
1.3 KiB
Python
from typing import Any, Optional
|
|
|
|
from langchain_openai import AzureChatOpenAI
|
|
|
|
from .base_client import BaseLLMClient
|
|
from .validators import validate_model
|
|
|
|
|
|
class AzureOpenAIClient(BaseLLMClient):
|
|
"""Client for Azure OpenAI models via AzureChatOpenAI."""
|
|
|
|
def __init__(
|
|
self,
|
|
model: str,
|
|
base_url: Optional[str] = None,
|
|
**kwargs,
|
|
):
|
|
super().__init__(model, base_url, **kwargs)
|
|
|
|
def get_llm(self) -> Any:
|
|
"""Return configured AzureChatOpenAI instance."""
|
|
llm_kwargs = {
|
|
"azure_deployment": self.model,
|
|
"model": self.model,
|
|
}
|
|
|
|
# Prefer explicit config sources and let AzureChatOpenAI resolve env fallbacks.
|
|
azure_endpoint = self.kwargs.get("azure_endpoint") or self.base_url
|
|
if azure_endpoint:
|
|
llm_kwargs["azure_endpoint"] = azure_endpoint
|
|
|
|
for key in (
|
|
"api_version",
|
|
"api_key",
|
|
"timeout",
|
|
"max_retries",
|
|
"reasoning_effort",
|
|
"callbacks",
|
|
):
|
|
if key in self.kwargs and self.kwargs[key] is not None:
|
|
llm_kwargs[key] = self.kwargs[key]
|
|
|
|
return AzureChatOpenAI(**llm_kwargs)
|
|
|
|
def validate_model(self) -> bool:
|
|
"""Validate model for Azure OpenAI."""
|
|
return validate_model("azure", self.model)
|