[CI/Build] Skip test_use_trtllm_attention on non-CUDA platforms (#44687)
Signed-off-by: Dan Blanaru <48605845+DanBlanaru@users.noreply.github.com> Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
This commit is contained in:
co-authored by
Wentao Ye
mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
parent
16282a9c4e
commit
82d6b59f04
@@ -6,12 +6,19 @@ from unittest.mock import patch
|
||||
import pytest
|
||||
import torch
|
||||
|
||||
from vllm.platforms import current_platform
|
||||
from vllm.utils.flashinfer import (
|
||||
can_use_trtllm_attention,
|
||||
supports_trtllm_attention,
|
||||
use_trtllm_attention,
|
||||
)
|
||||
|
||||
if not current_platform.is_cuda():
|
||||
pytest.skip(
|
||||
"TRTLLM attention is only supported on CUDA platforms.",
|
||||
allow_module_level=True,
|
||||
)
|
||||
|
||||
MODEL_CONFIGS = {
|
||||
"Llama-3-70B": dict(num_qo_heads=64, num_kv_heads=8),
|
||||
"Llama-3-8B": dict(num_qo_heads=32, num_kv_heads=8),
|
||||
|
||||
Reference in New Issue
Block a user