Files
vLLM/tests/models/multimodal/conftest.py
T
Andreas KaratzasandOpenAI Codex 1ab2801dde [ROCm] Remove stale SDPA and skinny GEMM workarounds (#50907)
Signed-off-by: Andreas Karatzas <Andreas.Karatzas@amd.com>
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
2026-08-11 14:19:16 -04:00

14 lines
599 B
Python

# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
"""Pytest configuration for vLLM multimodal tests."""
import torch.nn.functional as F
# openbmb/MiniCPM-Llama3-V-2_5's remote ``resampler.py`` annotates with ``List``
# without importing it, relying on the alias leaking from
# ``from torch.nn.functional import *``. Recent torch uses builtin generics and no
# longer re-exports the typing aliases, so executing that module raises NameError.
# TODO: Remove once the checkpoint fixes its imports.
if not hasattr(F, "List"):
F.List = list