Add pydocstyle to the ruff rules (#52136)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Signed-off-by: Harry Mellor <hej.mellor@gmail.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 5
parent
804e5377aa
commit
8c1557a79c
@@ -30,8 +30,7 @@ def llm(vllm_runner):
|
||||
|
||||
@pytest.mark.skip_global_cleanup
|
||||
def test_tiling_engine_basic(llm):
|
||||
"""
|
||||
Basic test with a small number of prompts (less than max_num_seqs).
|
||||
"""Basic test with a small number of prompts (less than max_num_seqs).
|
||||
No tiling should be triggered, but the engine still processes correctly.
|
||||
"""
|
||||
prompts = ["Hello", "World"]
|
||||
@@ -41,8 +40,7 @@ def test_tiling_engine_basic(llm):
|
||||
|
||||
@pytest.mark.skip_global_cleanup
|
||||
def test_tiling_engine_many_requests(llm):
|
||||
"""
|
||||
Test with a large number of prompts that exceeds max_num_seqs.
|
||||
"""Test with a large number of prompts that exceeds max_num_seqs.
|
||||
This verifies that _run_tiling_engine correctly chunks requests,
|
||||
processes all of them, and returns outputs in the correct order.
|
||||
"""
|
||||
@@ -54,8 +52,7 @@ def test_tiling_engine_many_requests(llm):
|
||||
|
||||
@pytest.mark.skip_global_cleanup
|
||||
def test_tiling_engine_with_pooling_params(llm):
|
||||
"""
|
||||
Test the tiling engine when different PoolingParams are provided.
|
||||
"""Test the tiling engine when different PoolingParams are provided.
|
||||
The engine must handle a list of params that matches the number of prompts.
|
||||
"""
|
||||
num_prompts = 10
|
||||
@@ -77,8 +74,7 @@ def test_tiling_engine_with_pooling_params(llm):
|
||||
|
||||
@pytest.mark.skip_global_cleanup
|
||||
def test_tiling_engine_abort_on_exception(llm):
|
||||
"""
|
||||
Test that abort_request IS called with the correct arguments when an
|
||||
"""Test that abort_request IS called with the correct arguments when an
|
||||
exception occurs inside the engine's step() loop.
|
||||
"""
|
||||
prompts = ["Prompt 0", "Prompt 1", "Prompt 2"]
|
||||
|
||||
@@ -111,7 +111,8 @@ def test_basic_embed(server: RemoteOpenAIServer, model_name: str):
|
||||
|
||||
def test_unsupported_input_type_rejected(server: RemoteOpenAIServer, model_name: str):
|
||||
"""An input_type not defined in the model's prompt config should be
|
||||
rejected with a 400 error."""
|
||||
rejected with a 400 error.
|
||||
"""
|
||||
body = {
|
||||
"model": model_name,
|
||||
"input_type": "nonexistent_type",
|
||||
@@ -138,7 +139,8 @@ def test_omitted_input_type_accepted(server: RemoteOpenAIServer, model_name: str
|
||||
|
||||
def test_v1_v2_parity(server: RemoteOpenAIServer, model_name: str):
|
||||
"""v1 (OpenAI) and v2 (Cohere) endpoints should produce the same
|
||||
float embeddings for a generic model."""
|
||||
float embeddings for a generic model.
|
||||
"""
|
||||
texts = ["hello world"]
|
||||
v2 = _cohere_embed(server, model_name, texts=texts, embedding_types=["float"])
|
||||
v1 = _openai_embed(server, model_name, texts)
|
||||
|
||||
@@ -76,7 +76,8 @@ def test_single_text_parity(server: RemoteOpenAIServer):
|
||||
|
||||
def test_batch_parity(server: RemoteOpenAIServer):
|
||||
"""A batch of texts should produce equivalent embeddings via both APIs,
|
||||
in the same order."""
|
||||
in the same order.
|
||||
"""
|
||||
texts = [
|
||||
"machine learning",
|
||||
"deep learning",
|
||||
|
||||
@@ -241,7 +241,8 @@ class TestResolveTruncation:
|
||||
|
||||
def test_truncate_none_with_max_tokens(self):
|
||||
"""truncate=NONE should NOT set truncate_prompt_tokens; the
|
||||
max_tokens limit is enforced separately via _check_max_tokens."""
|
||||
max_tokens limit is enforced separately via _check_max_tokens.
|
||||
"""
|
||||
req = self._make_request(truncate="NONE", max_tokens=10)
|
||||
tokens, side = EmbedIOProcessor._resolve_cohere_truncation(req)
|
||||
assert tokens is None
|
||||
|
||||
@@ -1,8 +1,6 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
"""
|
||||
Run `pytest tests/entrypoints/openai/test_embedding_dimensions.py`.
|
||||
"""
|
||||
"""Run `pytest tests/entrypoints/openai/test_embedding_dimensions.py`."""
|
||||
|
||||
import openai
|
||||
import pytest
|
||||
|
||||
@@ -1,7 +1,6 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
"""
|
||||
Test cases for long text embedding with automatic chunking mechanism.
|
||||
"""Test cases for long text embedding with automatic chunking mechanism.
|
||||
|
||||
This test suite validates vLLM's automatic chunking functionality for handling
|
||||
text inputs that exceed the model's maximum token length, specifically targeting
|
||||
@@ -241,8 +240,8 @@ async def test_long_text_embedding_1500_chars(
|
||||
client_with_chunked_processing: openai.AsyncOpenAI, model_name: str
|
||||
):
|
||||
"""Test embedding processing for ~1500 character long text
|
||||
(~1028 tokens, exceeding 512 token limit)."""
|
||||
|
||||
(~1028 tokens, exceeding 512 token limit).
|
||||
"""
|
||||
# Verify text length
|
||||
# Verify text has sufficient word count (approximately 1500 words)
|
||||
word_count = len(LONG_TEXT_1500_WORDS.split())
|
||||
@@ -290,8 +289,8 @@ async def test_long_text_embedding_2500_chars(
|
||||
client_with_chunked_processing: openai.AsyncOpenAI, model_name: str
|
||||
):
|
||||
"""Test embedding processing for ~2500 character long text
|
||||
(~2048 tokens, requiring multiple chunks)."""
|
||||
|
||||
(~2048 tokens, requiring multiple chunks).
|
||||
"""
|
||||
# Verify text length
|
||||
# Verify text has sufficient word count (approximately 2500 words)
|
||||
word_count = len(LONG_TEXT_2500_WORDS.split())
|
||||
@@ -339,7 +338,6 @@ async def test_batch_long_text_embedding(
|
||||
client_with_chunked_processing: openai.AsyncOpenAI, model_name: str
|
||||
):
|
||||
"""Test batch long text embedding processing."""
|
||||
|
||||
input_texts = [
|
||||
LONG_TEXT_1500_WORDS,
|
||||
LONG_TEXT_2500_WORDS,
|
||||
@@ -384,8 +382,8 @@ async def test_chunked_vs_normal_consistency(
|
||||
client_with_chunked_processing: openai.AsyncOpenAI, model_name: str
|
||||
):
|
||||
"""Test consistency between chunked and
|
||||
normal processing (using short text)."""
|
||||
|
||||
normal processing (using short text).
|
||||
"""
|
||||
# Use a short text within the 512 token limit
|
||||
short_text = (
|
||||
"Artificial intelligence technology is changing our world, "
|
||||
@@ -424,7 +422,6 @@ async def test_chunked_processing_response_format(
|
||||
client_with_chunked_processing: openai.AsyncOpenAI, model_name: str
|
||||
):
|
||||
"""Test response format and structure during chunked processing."""
|
||||
|
||||
# Test with long text to trigger chunking
|
||||
embedding_response = await client_with_chunked_processing.embeddings.create(
|
||||
model=model_name,
|
||||
|
||||
@@ -1,7 +1,8 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
"""Unit tests for Cohere embed protocol: build_typed_embeddings and its
|
||||
underlying packing helpers, plus Cohere-specific serving helpers."""
|
||||
underlying packing helpers, plus Cohere-specific serving helpers.
|
||||
"""
|
||||
|
||||
import struct
|
||||
|
||||
|
||||
Reference in New Issue
Block a user