[MyPy][3/N] Fix MyPy errors in test groups (part 3) (#55939)

Signed-off-by: Taimys <ffy2281693949@gmail.com>
Signed-off-by: Luyi Xiao <81631911+luyixiao95@users.noreply.github.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Luyi Xiao <81631911+luyixiao95@users.noreply.github.com>
Co-authored-by: Taimys <ffy2281693949@gmail.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Taimys
2026-09-30 11:06:07 +00:00
committed by GitHub
co-authored by Codex Luyi Xiao Taimys Harry Mellor Claude Opus 5.5
parent c4fa6f36a7
commit 5225f1bacb
118 changed files with 948 additions and 605 deletions
+29 -15
View File
@@ -4,7 +4,7 @@ import json
from argparse import Namespace
from io import BytesIO
from pathlib import Path
from typing import Any
from typing import Any, cast
import pybase64 as base64
import pytest
@@ -17,6 +17,7 @@ from vllm.benchmarks.lib.endpoint_request_func import (
_get_chat_content,
_get_chat_messages,
)
from vllm.tokenizers import TokenizerLike
pytestmark = pytest.mark.skip_global_cleanup
@@ -31,6 +32,10 @@ class _Tokenizer:
return _TokenizedPrompt(prompt)
def _tokenizer() -> TokenizerLike:
return cast(TokenizerLike, _Tokenizer())
def _write_jsonl(path: Path, rows: list[dict[str, Any]]) -> None:
with path.open("w") as f:
for row in rows:
@@ -92,7 +97,7 @@ def test_get_samples_custom_image_cli_path_supports_multi_image_and_content(
],
)
samples = get_samples(_args_for_custom_image(jsonl), _Tokenizer())
samples = get_samples(_args_for_custom_image(jsonl), _tokenizer())
assert len(samples) == 2
assert samples[0].request_id == "req-0"
@@ -106,6 +111,7 @@ def test_get_samples_custom_image_cli_path_supports_multi_image_and_content(
assert samples[1].multi_modal_data is None
assert isinstance(samples[1].prompt, list)
assert samples[1].prompt[0] == {"type": "text", "text": "Now compare "}
assert isinstance(samples[1].prompt[1], dict)
assert samples[1].prompt[1]["image_url"]["url"] == f"file://{image_c}"
@@ -126,7 +132,7 @@ def test_custom_image_dataset_uses_all_image_files(tmp_path: Path) -> None:
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
samples = dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
)
@@ -171,7 +177,7 @@ def test_custom_image_dataset_preserves_interleaved_content_order(
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
samples = dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
)
@@ -181,14 +187,18 @@ def test_custom_image_dataset_preserves_interleaved_content_order(
assert sample.multi_modal_data is None
assert sample.prompt_len == 2
assert isinstance(sample.prompt, list)
assert [part["type"] for part in sample.prompt] == [
parts = []
for part in sample.prompt:
assert isinstance(part, dict)
parts.append(part)
assert [part["type"] for part in parts] == [
"text",
"image_url",
"text",
"image_url",
]
assert sample.prompt[1]["image_url"]["url"] == f"file://{image_a}"
assert sample.prompt[3]["image_url"] == {
assert parts[1]["image_url"]["url"] == f"file://{image_a}"
assert parts[3]["image_url"] == {
"url": f"file://{image_b}",
"detail": "low",
}
@@ -223,7 +233,7 @@ def test_custom_image_dataset_wraps_interleaved_content_for_multimodal_chat(
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
samples = dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
enable_multimodal_chat=True,
@@ -289,7 +299,7 @@ def test_custom_image_dataset_encodes_image_media_when_requested(
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
samples = dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
ensure_client_side_data=True,
@@ -335,7 +345,7 @@ def test_custom_image_dataset_encodes_interleaved_image_media(
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
samples = dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
ensure_client_side_data=True,
@@ -343,9 +353,13 @@ def test_custom_image_dataset_encodes_interleaved_image_media(
sample = samples[0]
assert isinstance(sample.prompt, list)
_assert_png_data_url(sample.prompt[1]["image_url"]["url"])
_assert_png_data_url(sample.prompt[2]["image_url"]["url"])
assert sample.prompt[2]["image_url"]["detail"] == "low"
parts = []
for part in sample.prompt:
assert isinstance(part, dict)
parts.append(part)
_assert_png_data_url(parts[1]["image_url"]["url"])
_assert_png_data_url(parts[2]["image_url"]["url"])
assert parts[2]["image_url"]["detail"] == "low"
@pytest.mark.benchmark
@@ -363,7 +377,7 @@ def test_custom_image_dataset_rejects_invalid_image_media(
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
with pytest.raises(ValueError, match="Invalid image URL"):
dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
ensure_client_side_data=True,
@@ -380,7 +394,7 @@ def test_custom_image_dataset_rejects_invalid_content_part(
dataset = CustomImageDataset(dataset_path=str(jsonl), disable_shuffle=True)
with pytest.raises(ValueError, match="type 'text', 'image', or 'image_url'"):
dataset.sample(
tokenizer=_Tokenizer(),
tokenizer=_tokenizer(),
num_requests=1,
output_len=32,
)