[CI] Reorganize entrypoints CI (#44947)
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
This commit is contained in:
@@ -1,42 +0,0 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
|
||||
from tests.utils import VLLM_PATH, RemoteOpenAIServer
|
||||
|
||||
chatml_jinja_path = VLLM_PATH / "examples/template_chatml.jinja"
|
||||
assert chatml_jinja_path.exists()
|
||||
|
||||
|
||||
def run_and_test_dummy_opt_api_server(model, tp=1):
|
||||
# the model is registered through the plugin
|
||||
server_args = [
|
||||
"--gpu-memory-utilization",
|
||||
"0.10",
|
||||
"--dtype",
|
||||
"float32",
|
||||
"--chat-template",
|
||||
str(chatml_jinja_path),
|
||||
"--load-format",
|
||||
"dummy",
|
||||
"-tp",
|
||||
f"{tp}",
|
||||
]
|
||||
with RemoteOpenAIServer(model, server_args) as server:
|
||||
client = server.get_client()
|
||||
completion = client.chat.completions.create(
|
||||
model=model,
|
||||
messages=[
|
||||
{"role": "system", "content": "You are a helpful assistant."},
|
||||
{"role": "user", "content": "Hello!"},
|
||||
],
|
||||
temperature=0,
|
||||
)
|
||||
generated_text = completion.choices[0].message.content
|
||||
assert generated_text is not None
|
||||
# make sure only the first token is generated
|
||||
rest = generated_text.replace("<s>", "")
|
||||
assert rest == ""
|
||||
|
||||
|
||||
def test_oot_registration_for_api_server(dummy_opt_path: str):
|
||||
run_and_test_dummy_opt_api_server(dummy_opt_path)
|
||||
+1
-1
@@ -5,7 +5,7 @@ import json
|
||||
import openai
|
||||
import pytest
|
||||
|
||||
from ....utils import RemoteOpenAIServer
|
||||
from tests.utils import RemoteOpenAIServer
|
||||
|
||||
MODEL = "ibm-granite/granite-4.0-h-tiny"
|
||||
|
||||
+1
-2
@@ -9,12 +9,11 @@ import pytest_asyncio
|
||||
from huggingface_hub import snapshot_download
|
||||
from typing_extensions import TypedDict
|
||||
|
||||
from tests.utils import RemoteOpenAIServer
|
||||
from vllm.tool_parsers.abstract_tool_parser import ToolParser
|
||||
from vllm.tool_parsers.granite4_tool_parser import Granite4ToolParser
|
||||
from vllm.tool_parsers.hermes_tool_parser import Hermes2ProToolParser
|
||||
|
||||
from ....utils import RemoteOpenAIServer
|
||||
|
||||
LORA_MODEL = "minpeter/LoRA-Llama-3.2-1B-tool-vllm-ci"
|
||||
|
||||
TOOLS = [
|
||||
+1
-1
@@ -9,7 +9,7 @@ import pytest
|
||||
import pytest_asyncio
|
||||
from rapidfuzz import fuzz
|
||||
|
||||
from ....utils import RemoteOpenAIServer
|
||||
from tests.utils import RemoteOpenAIServer
|
||||
|
||||
MODEL_NAME = "openai/gpt-oss-20b"
|
||||
|
||||
Reference in New Issue
Block a user