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

[Core] Add xxHash as a high-performance hash option for accelerating prefix caching (#29163)

Browse Source 
Signed-off-by: LuminolT <lumischen01@gmail.com>
Signed-off-by: Lumis Chen <lumischen01@gmail.com>
Co-authored-by: Russell Bryant <rbryant@redhat.com>
This commit is contained in:
Lumis Chen
2025-12-04 00:06:57 +08:00
committed by GitHub
parent 5aa9b09040
commit 9bcf92295a
7 changed files with 332 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
tests/v1/engine/test_engine_args.py
View File

@@ -9,6 +9,7 @@ from vllm.config import VllmConfig
from vllm.engine.arg_utils import EngineArgs
from vllm.usage.usage_lib import UsageContext
from vllm.utils.argparse_utils import FlexibleArgumentParser
from vllm.utils.hashing import _xxhash
def test_prefix_caching_from_cli():
@@ -48,6 +49,21 @@ def test_prefix_caching_from_cli():
args = parser.parse_args(["--prefix-caching-hash-algo", "invalid"])
@pytest.mark.skipif(_xxhash is None, reason="xxhash not installed")
def test_prefix_caching_xxhash_from_cli():
parser = EngineArgs.add_cli_args(FlexibleArgumentParser())
# set hash algorithm to xxhash (pickle)
args = parser.parse_args(["--prefix-caching-hash-algo", "xxhash"])
vllm_config = EngineArgs.from_cli_args(args=args).create_engine_config()
assert vllm_config.cache_config.prefix_caching_hash_algo == "xxhash"
# set hash algorithm to xxhash_cbor
args = parser.parse_args(["--prefix-caching-hash-algo", "xxhash_cbor"])
vllm_config = EngineArgs.from_cli_args(args=args).create_engine_config()
assert vllm_config.cache_config.prefix_caching_hash_algo == "xxhash_cbor"
def test_defaults_with_usage_context():
engine_args = EngineArgs(model="facebook/opt-125m")
vllm_config: VllmConfig = engine_args.create_engine_config(UsageContext.LLM_CLASS)