Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
# SPDX-License-Identifier: AGPL-3.0-only
# Copyright 2026-present the Unsloth AI Inc. team. All rights reserved. See /studio/LICENSE.AGPL-3.0
"""Wait for SSE frames; if the driving task failed first, raise its exception, not a bare timeout."""
from __future__ import annotations
import asyncio
async def wait_for_frame(
event: asyncio.Event,
task: asyncio.Task,
*,
timeout: float = 20.0,
what: str = "the expected SSE frame",
) -> None:
"""The timeout is a deadlock backstop only: a timeout means the event never happened, not slow."""
waiter = asyncio.ensure_future(event.wait())
try:
done, _ = await asyncio.wait(
{waiter, task}, timeout = timeout, return_when = asyncio.FIRST_COMPLETED
)
# Check the task first: a send() that sets the event then raises would otherwise be swallowed.
if task in done:
exc = task.exception()
if exc is not None:
when = "after" if event.is_set() else "before"
raise AssertionError(f"the request failed {when} {what} was sent: {exc!r}") from exc
if waiter in done:
return
if task in done:
raise AssertionError(f"the request completed without sending {what}")
raise AssertionError(
f"timed out after {timeout}s waiting for {what}, and the request task is still "
f"running: the path is parked rather than failing"
)
finally:
waiter.cancel()