biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

refactor hard coded device string in test files under tests/v1 and tests/lora (#37566)

Browse Source 
Signed-off-by: Liao, Wei <wei.liao@intel.com>
This commit is contained in:
wliao2
2026-04-02 20:21:47 -07:00
committed by GitHub
parent 4a06e1246e
commit 32e0c0bfa2
28 changed files with 239 additions and 146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
tests/v1/worker/test_gpu_input_batch.py
View File

@@ -22,10 +22,8 @@ from vllm.v1.worker.gpu_input_batch import CachedRequestState, InputBatch
VOCAB_SIZE = 1024
NUM_OUTPUT_TOKENS = 20
MAX_PROMPT_SIZE = 100
CUDA_DEVICES = [
f"{current_platform.device_type}:{i}"
for i in range(min(current_platform.device_count(), 2))
]
DEVICE_TYPE = current_platform.device_type
DEVICES = [f"{DEVICE_TYPE}:{i}" for i in range(min(current_platform.device_count(), 2))]
MAX_NUM_PROMPT_TOKENS = 64
@@ -219,7 +217,7 @@ def _construct_cached_request_state(req_id_suffix: int):
)
@pytest.mark.parametrize("device", CUDA_DEVICES)
@pytest.mark.parametrize("device", DEVICES)
@pytest.mark.parametrize("batch_size", [1, 2, 32, 64])
def test_sampling_metadata_in_input_batch(device: str, batch_size: int):
"""
@@ -313,7 +311,7 @@ def test_sampling_metadata_in_input_batch(device: str, batch_size: int):
)
@pytest.mark.parametrize("device", CUDA_DEVICES)
@pytest.mark.parametrize("device", DEVICES)
@pytest.mark.parametrize("batch_size", [32])
@pytest.mark.parametrize("swap_list", [((0, 1),)])
def test_swap_states_in_input_batch(device: str, batch_size: int, swap_list: list):
@@ -400,7 +398,7 @@ def _construct_pooling_request(req_id_suffix: int, pooling_params=None):
)
@pytest.mark.parametrize("device", CUDA_DEVICES)
@pytest.mark.parametrize("device", DEVICES)
def test_pooling_prompt_lens_not_aliased(device: str):
"""Verify that prompt_lens in PoolingMetadata does not share memory
with the internal num_prompt_tokens pinned buffer. Guards against possible