wehub-resource-sync
49 lines
1.5 KiB
Python
49 lines
1.5 KiB
Python
"""
|
|
Multi-provider LLM client with OpenAI Responses API interface.
|
|
|
|
Usage::
|
|
|
|
from omnigent.llms import Client
|
|
|
|
client = Client()
|
|
resp = client.responses.create(
|
|
input=[{"role": "user", "content": "Hello"}],
|
|
instructions="You are a helpful assistant.",
|
|
model="anthropic/claude-sonnet-4-20250514",
|
|
)
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
from typing import TYPE_CHECKING, Any
|
|
|
|
if TYPE_CHECKING:
|
|
from omnigent.llms.client import Client
|
|
from omnigent.llms.context_window import get_model_context_window
|
|
|
|
__all__ = ["Client", "get_model_context_window"]
|
|
|
|
|
|
def __getattr__(name: str) -> Any:
|
|
"""Lazy attribute access to avoid a circular import.
|
|
|
|
Eager top-level imports here pull in ``omnigent.llms.client``,
|
|
which imports ``omnigent.reasoning_effort``, which imports
|
|
``omnigent.llms.errors`` -- a submodule load that re-triggers
|
|
this ``__init__`` mid-initialisation and explodes with
|
|
``cannot import name 'OPENAI_EFFORTS' from partially initialized
|
|
module``. Lazy resolution at attribute access time defers the
|
|
expensive imports until something actually needs ``Client`` or
|
|
``get_model_context_window``, after both modules have finished
|
|
initialising.
|
|
"""
|
|
if name == "Client":
|
|
from omnigent.llms.client import Client
|
|
|
|
return Client
|
|
if name == "get_model_context_window":
|
|
from omnigent.llms.context_window import get_model_context_window
|
|
|
|
return get_model_context_window
|
|
raise AttributeError(f"module 'omnigent.llms' has no attribute {name!r}")
|