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

[Platform] Deprecate seed_everything (#31659)

Browse Source 
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2026-01-05 10:34:04 +08:00
committed by GitHub
parent 367856de14
commit bb4337b34c
77 changed files with 219 additions and 171 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/kernels/attention/test_aiter_flash_attn.py
View File

@@ -8,6 +8,7 @@ import torch
import vllm.v1.attention.backends.rocm_aiter_fa # noqa: F401
from vllm.attention.utils.fa_utils import is_flash_attn_varlen_func_available
from vllm.platforms import current_platform
from vllm.utils.torch_utils import set_random_seed
NUM_HEADS = [(4, 4), (8, 2)]
HEAD_SIZES = [128, 256]
@@ -104,7 +105,7 @@ def test_varlen_with_paged_kv(
if not is_flash_attn_varlen_func_available():
pytest.skip("flash_attn_varlen_func required to run this test.")
torch.set_default_device("cuda")
current_platform.seed_everything(0)
set_random_seed(0)
num_seqs = len(seq_lens)
query_lens = [x[0] for x in seq_lens]
kv_lens = [x[1] for x in seq_lens]