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

[Frontend] Support for chat completions input in the tokenize endpoint (#5923)

Browse Source
This commit is contained in:
sasha0552
2024-07-16 12:18:09 +00:00
committed by GitHub
parent d97011512e
commit 7a3d2a5b95
9 changed files with 386 additions and 244 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
tests/async_engine/test_chat_template.py
View File

@@ -4,8 +4,8 @@ from dataclasses import dataclass
import pytest
from vllm.entrypoints.openai.chat_utils import load_chat_template
from vllm.entrypoints.openai.protocol import ChatCompletionRequest
from vllm.entrypoints.openai.serving_chat import OpenAIServingChat
from vllm.transformers_utils.tokenizer import get_tokenizer
chatml_jinja_path = pathlib.Path(os.path.dirname(os.path.abspath(
@@ -64,8 +64,7 @@ def test_load_chat_template():
# Testing chatml template
tokenizer = MockTokenizer()
mock_serving_chat = MockServingChat(tokenizer)
OpenAIServingChat._load_chat_template(mock_serving_chat,
chat_template=chatml_jinja_path)
load_chat_template(mock_serving_chat, chat_template=chatml_jinja_path)
template_content = tokenizer.chat_template
@@ -84,8 +83,7 @@ def test_no_load_chat_template_filelike():
mock_serving_chat = MockServingChat(tokenizer)
with pytest.raises(ValueError, match="looks like a file path"):
OpenAIServingChat._load_chat_template(mock_serving_chat,
chat_template=template)
load_chat_template(mock_serving_chat, chat_template=template)
def test_no_load_chat_template_literallike():
@@ -94,8 +92,7 @@ def test_no_load_chat_template_literallike():
tokenizer = MockTokenizer()
mock_serving_chat = MockServingChat(tokenizer)
OpenAIServingChat._load_chat_template(mock_serving_chat,
chat_template=template)
load_chat_template(mock_serving_chat, chat_template=template)
template_content = tokenizer.chat_template
assert template_content == template
@@ -109,8 +106,7 @@ def test_get_gen_prompt(model, template, add_generation_prompt,
# Initialize the tokenizer
tokenizer = get_tokenizer(tokenizer_name=model)
mock_serving_chat = MockServingChat(tokenizer)
OpenAIServingChat._load_chat_template(mock_serving_chat,
chat_template=template)
load_chat_template(mock_serving_chat, chat_template=template)
# Create a mock request object using keyword arguments
mock_request = ChatCompletionRequest(