From 5f01ec25ff4abf6d4214c9e864a04faf92c5ae5a Mon Sep 17 00:00:00 2001 From: "rshaw@neuralmagic.com" Date: Sun, 7 Jul 2024 19:31:08 +0000 Subject: [PATCH] skip lm head test --- tests/distributed/test_pipeline_parallel.py | 1 + tests/quantization/test_lm_head.py | 5 +++++ 2 files changed, 6 insertions(+) diff --git a/tests/distributed/test_pipeline_parallel.py b/tests/distributed/test_pipeline_parallel.py index e0f14d6568183..47bbdc83e2274 100644 --- a/tests/distributed/test_pipeline_parallel.py +++ b/tests/distributed/test_pipeline_parallel.py @@ -7,6 +7,7 @@ import ray from tests.nm_utils.utils_skip import should_skip_test_group + from ..utils import VLLM_PATH, RemoteOpenAIServer if should_skip_test_group(group_name="TEST_DISTRIBUTED"): diff --git a/tests/quantization/test_lm_head.py b/tests/quantization/test_lm_head.py index dd9a016807df9..c6e2b3cb094bf 100644 --- a/tests/quantization/test_lm_head.py +++ b/tests/quantization/test_lm_head.py @@ -7,12 +7,17 @@ import pytest import torch +from tests.nm_utils.utils_skip import should_skip_test_group from vllm.model_executor.layers.linear import UnquantizedLinearMethod from vllm.model_executor.layers.quantization.gptq import GPTQLinearMethod from vllm.model_executor.layers.quantization.gptq_marlin import ( GPTQMarlinLinearMethod) from vllm.model_executor.layers.quantization.marlin import MarlinLinearMethod +if should_skip_test_group(group_name="TEST_QUANTIZATION"): + pytest.skip("TEST_QUANTIZATION=DISABLE, skipping quantization test group", + allow_module_level=True) + PROMPT = "On the surface of Mars, we found" MODELS_QUANT = [(