diff --git a/.buildkite/test-amd.yaml b/.buildkite/test-amd.yaml index 2a76fb54b8..327f00ba34 100644 --- a/.buildkite/test-amd.yaml +++ b/.buildkite/test-amd.yaml @@ -1374,8 +1374,6 @@ steps: - tests/lora - vllm/platforms/rocm.py commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - export PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True - pytest -v -s -x lora/test_chatglm3_tp.py - pytest -v -s -x lora/test_llama_tp.py - pytest -v -s -x lora/test_qwen3_with_multi_loras.py diff --git a/tests/lora/test_gptoss_tp.py b/tests/lora/test_gptoss_tp.py index 838c3ab7dd..0cf778f3b0 100644 --- a/tests/lora/test_gptoss_tp.py +++ b/tests/lora/test_gptoss_tp.py @@ -1,7 +1,12 @@ # SPDX-License-Identifier: Apache-2.0 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project +import importlib.metadata +from importlib.util import find_spec + import pytest +import torch +from packaging import version import vllm from vllm.lora.request import LoRARequest @@ -9,6 +14,22 @@ from vllm.platforms import current_platform from ..utils import multi_gpu_test +# Require amd-quark >= 0.12 on torch >= 2.11. +# Earlier torch releases work with older quark versions. See +# https://github.com/amd/Quark/issues/34 +# TODO: Remove once amd-quark>=0.12.0 +QUARK_TORCH_COMPATIBLE = find_spec("quark") is not None and ( + version.parse(importlib.metadata.version("amd-quark")) >= version.parse("0.12.0") + if version.parse(torch.__version__.split("+")[0]) >= version.parse("2.11") + else True +) + +if current_platform.is_rocm() and not QUARK_TORCH_COMPATIBLE: + pytest.skip( + "This test requires amd-quark >= 0.12 on torch >= 2.11.", + allow_module_level=True, + ) + MODEL_PATH = "openai/gpt-oss-20b" PROMPT_TEMPLATE = """<|start|>system<|message|>You are ChatGPT, a large language model trained by OpenAI.