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

[Chore] Remove redundant RequestPrompt (#30612)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2025-12-14 17:22:37 +08:00
committed by GitHub
parent f569c654e1
commit dcb31196da
12 changed files with 188 additions and 253 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
tests/entrypoints/openai/test_chat_error.py
View File

@@ -80,10 +80,9 @@ def _build_serving_chat(engine: AsyncLLM) -> OpenAIServingChat:
return dict(engine_prompt), {}
async def _fake_preprocess_chat(*args, **kwargs):
# return conversation, request_prompts, engine_prompts
# return conversation, engine_prompts
return (
[{"role": "user", "content": "Test"}],
[[1, 2, 3]],
[{"prompt_token_ids": [1, 2, 3]}],
)

26
tests/entrypoints/openai/test_serving_chat.py
View File

@@ -877,7 +877,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the first turn's input
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,
[
@@ -905,7 +905,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the second turn's input
req_2 = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages_2, _, _ = serving_chat._make_request_with_harmony(req_2)
input_messages_2, _ = serving_chat._make_request_with_harmony(req_2)
verify_harmony_messages(
input_messages_2,
[
@@ -927,7 +927,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the first turn's input
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages, tools=tools)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,
[
@@ -971,7 +971,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the second turn's input
req_2 = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages_2, _, _ = serving_chat._make_request_with_harmony(req_2)
input_messages_2, _ = serving_chat._make_request_with_harmony(req_2)
verify_harmony_messages(
input_messages_2,
[
@@ -1008,7 +1008,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the first turn's input
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages, tools=tools)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,
[
@@ -1052,7 +1052,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the second turn's input
req_2 = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages_2, _, _ = serving_chat._make_request_with_harmony(req_2)
input_messages_2, _ = serving_chat._make_request_with_harmony(req_2)
verify_harmony_messages(
input_messages_2,
[
@@ -1089,7 +1089,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the first turn's input
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages, tools=tools)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,
[
@@ -1133,7 +1133,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the second turn's input
req_2 = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages_2, _, _ = serving_chat._make_request_with_harmony(req_2)
input_messages_2, _ = serving_chat._make_request_with_harmony(req_2)
verify_harmony_messages(
input_messages_2,
[
@@ -1183,7 +1183,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the third turn's input
req_3 = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages_3, _, _ = serving_chat._make_request_with_harmony(req_3)
input_messages_3, _ = serving_chat._make_request_with_harmony(req_3)
verify_harmony_messages(
input_messages_3,
[
@@ -1246,7 +1246,7 @@ class TestServingChatWithHarmony:
# Test the Harmony messages for the fourth turn's input
req_4 = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages_4, _, _ = serving_chat._make_request_with_harmony(req_4)
input_messages_4, _ = serving_chat._make_request_with_harmony(req_4)
verify_harmony_messages(
input_messages_4,
[
@@ -1295,7 +1295,7 @@ class TestServingChatWithHarmony:
},
]
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,
@@ -1327,7 +1327,7 @@ class TestServingChatWithHarmony:
},
]
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,
@@ -1357,7 +1357,7 @@ class TestServingChatWithHarmony:
},
]
req = ChatCompletionRequest(model=MODEL_NAME, messages=messages)
input_messages, _, _ = serving_chat._make_request_with_harmony(req)
input_messages, _ = serving_chat._make_request_with_harmony(req)
verify_harmony_messages(
input_messages,

6
tests/entrypoints/openai/test_serving_responses.py
View File

@@ -21,7 +21,7 @@ from vllm.entrypoints.openai.serving_responses import (
extract_tool_types,
)
from vllm.entrypoints.tool_server import ToolServer
from vllm.inputs.data import TokensPrompt as EngineTokensPrompt
from vllm.inputs.data import TokensPrompt
class MockConversationContext(ConversationContext):
@@ -237,7 +237,7 @@ class TestValidateGeneratorInput:
"""Test _validate_generator_input with valid prompt length"""
# Create an engine prompt with valid length (less than max_model_len)
valid_prompt_token_ids = list(range(5)) # 5 tokens < 100 max_model_len
engine_prompt = EngineTokensPrompt(prompt_token_ids=valid_prompt_token_ids)
engine_prompt = TokensPrompt(prompt_token_ids=valid_prompt_token_ids)
# Call the method
result = serving_responses_instance._validate_generator_input(engine_prompt)
@@ -247,7 +247,7 @@ class TestValidateGeneratorInput:
# create an invalid engine prompt
invalid_prompt_token_ids = list(range(200)) # 100 tokens >= 100 max_model_len
engine_prompt = EngineTokensPrompt(prompt_token_ids=invalid_prompt_token_ids)
engine_prompt = TokensPrompt(prompt_token_ids=invalid_prompt_token_ids)
# Call the method
result = serving_responses_instance._validate_generator_input(engine_prompt)