[Refactor] [4/N] Move VLLM_SERVER_DEV endpoints into the serve directory (#30749)

Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
This commit is contained in:
Chauncey
2025-12-17 02:27:30 -08:00
committed by GitHub
parent f284d7bd0c
commit 9ad5b21710
19 changed files with 259 additions and 151 deletions
@@ -14,11 +14,10 @@ import requests
from prometheus_client.parser import text_string_to_metric_families
from transformers import AutoTokenizer
from tests.conftest import LocalAssetServer
from tests.utils import RemoteOpenAIServer
from vllm import version
from ...conftest import LocalAssetServer
from ...utils import RemoteOpenAIServer
MODELS = {
"text": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
"multimodal": "HuggingFaceTB/SmolVLM-256M-Instruct",
View File
@@ -37,7 +37,7 @@ def server():
"--max-num-seqs",
"128",
"--worker-extension-cls",
"tests.entrypoints.openai.test_collective_rpc.TestWorkerExtension",
"tests.entrypoints.rpc.test_collective_rpc.TestWorkerExtension",
]
with RemoteOpenAIServer(
MODEL_NAME,
View File
@@ -4,7 +4,7 @@
import requests
from prometheus_client.parser import text_string_to_metric_families
from ...utils import RemoteOpenAIServer
from tests.utils import RemoteOpenAIServer
MODEL_NAME = "meta-llama/Llama-3.2-1B"