[Core][Refactor] Move KV cache planning logic to kv_cache_planning.py (#61183)
Signed-off-by: Dakai An <dakaian108@gmail.com> Signed-off-by: Yifan Qiao <yifanqiao@inferact.ai> Co-authored-by: Yifan Qiao <yifanqiao@inferact.ai> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
co-authored by
Yifan Qiao
Claude Opus 5.5
parent
c077dc663d
commit
df6c7ad031
@@ -4,7 +4,7 @@
|
||||
import pytest
|
||||
import torch
|
||||
|
||||
from vllm.v1.core.kv_cache_utils import check_enough_kv_cache_memory
|
||||
from vllm.v1.core.kv_cache_planning import check_enough_kv_cache_memory
|
||||
from vllm.v1.kv_cache_interface import FullAttentionSpec
|
||||
|
||||
|
||||
@@ -38,7 +38,7 @@ def test_kv_cache_oom_insufficient_memory(monkeypatch):
|
||||
config.parallel_config.decode_context_parallel_size = 1
|
||||
|
||||
monkeypatch.setattr(
|
||||
"vllm.v1.core.kv_cache_utils.max_memory_usage_bytes",
|
||||
"vllm.v1.core.kv_cache_planning.max_memory_usage_bytes",
|
||||
lambda c, s: 100 * 1024**3, # 100 GiB
|
||||
)
|
||||
|
||||
|
||||
Reference in New Issue
Block a user