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

Convert formatting to use ruff instead of yapf + isort (#26247)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 15:06:22 +01:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115244 additions and 94146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
tests/reasoning/utils.py
View File

@@ -3,14 +3,12 @@
from typing import Optional, Union
from vllm.entrypoints.openai.protocol import (ChatCompletionRequest,
DeltaMessage)
from vllm.entrypoints.openai.protocol import ChatCompletionRequest, DeltaMessage
from vllm.reasoning import ReasoningParser
from vllm.transformers_utils.tokenizers.mistral import MistralTokenizer
class StreamingReasoningReconstructor:
def __init__(self):
self.reasoning_content = None
self.other_content = None
@@ -19,8 +17,8 @@ class StreamingReasoningReconstructor:
# content and the reasoning content should not be present
# at the same time
assert delta.content is None or delta.reasoning_content is None, (
"Both content and reasoning content are present in the "
"delta message")
"Both content and reasoning content are present in the delta message"
)
if delta.content is not None:
if self.other_content is None:
self.other_content = delta.content
@@ -51,7 +49,8 @@ def run_reasoning_extraction(
)
else:
reasoning, content = run_reasoning_extraction_nonstreaming(
reasoning_parser, model_output, request)
reasoning_parser, model_output, request
)
return reasoning, content
@@ -61,8 +60,9 @@ def run_reasoning_extraction_mistral(
request: Union[ChatCompletionRequest, None] = None,
streaming: bool = False,
) -> tuple[Optional[str], Optional[str]]:
assert isinstance(reasoning_parser.model_tokenizer,
MistralTokenizer), type(reasoning_parser.model_tokenizer)
assert isinstance(reasoning_parser.model_tokenizer, MistralTokenizer), type(
reasoning_parser.model_tokenizer
)
if streaming:
reconstructor = run_reasoning_extraction_streaming_mistral(
reasoning_parser,
@@ -75,9 +75,11 @@ def run_reasoning_extraction_mistral(
)
else:
str_output = reasoning_parser.model_tokenizer.convert_ids_to_tokens(
model_output)
model_output
)
reasoning, content = run_reasoning_extraction_nonstreaming(
reasoning_parser, str_output, request)
reasoning_parser, str_output, request
)
return reasoning, content
@@ -88,7 +90,8 @@ def run_reasoning_extraction_nonstreaming(
) -> tuple[Optional[str], Optional[str]]:
request = request or ChatCompletionRequest(messages=[], model="test-model")
return reasoning_parser.extract_reasoning_content(
model_output=''.join(model_output), request=request)
model_output="".join(model_output), request=request
)
def run_reasoning_extraction_streaming(
@@ -128,16 +131,16 @@ def run_reasoning_extraction_streaming_mistral(
model_deltas: list[int],
request: Union[ChatCompletionRequest, None] = None,
) -> StreamingReasoningReconstructor:
assert isinstance(reasoning_parser.model_tokenizer,
MistralTokenizer), type(reasoning_parser.model_tokenizer)
assert isinstance(reasoning_parser.model_tokenizer, MistralTokenizer), type(
reasoning_parser.model_tokenizer
)
request = request or ChatCompletionRequest(messages=[], model="test-model")
reconstructor = StreamingReasoningReconstructor()
previous_text = ""
previous_tokens: list[int] = []
for model_delta in model_deltas:
token_delta = [model_delta]
delta = reasoning_parser.model_tokenizer.convert_ids_to_tokens(
[model_delta])[0]
delta = reasoning_parser.model_tokenizer.convert_ids_to_tokens([model_delta])[0]
current_text = previous_text + delta
current_tokens = previous_tokens + token_delta
delta_message = reasoning_parser.extract_reasoning_content_streaming(