[Chore] Replace all base64 usages with faster pybase64 package (#37290)

Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>

examples/online_serving/openai_chat_completion_client_for_multimodal.py

@@ -20,9 +20,9 @@ run the script with
 python openai_chat_completion_client_for_multimodal.py --chat-type audio
 """
 
-import base64
 import os
 
+import pybase64 as base64
 import requests
 from openai import OpenAI
 from utils import get_first_model

examples/online_serving/openai_realtime_client.py

@@ -24,11 +24,11 @@ The script:
 
 import argparse
 import asyncio
-import base64
 import json
 
 import librosa
 import numpy as np
+import pybase64 as base64
 import websockets
 
 from vllm.assets.audio import AudioAsset

examples/online_serving/openai_realtime_microphone_client.py

@@ -18,13 +18,13 @@ Requirements: websockets, numpy, gradio
 
 import argparse
 import asyncio
-import base64
 import json
 import queue
 import threading
 
 import gradio as gr
 import numpy as np
+import pybase64 as base64
 import websockets
 
 SAMPLE_RATE = 16_000

examples/pooling/embed/embedding_requests_base64_online.py

@@ -7,8 +7,8 @@ NOTE:
 """
 
 import argparse
-import base64
 
+import pybase64 as base64
 import requests
 import torch
 

examples/pooling/embed/vision_embedding_online.py

@@ -7,10 +7,10 @@ Refer to each `run_*` function for the command to run the server for that model.
 """
 
 import argparse
-import base64
 import io
 from typing import Literal
 
+import pybase64 as base64
 from openai import OpenAI
 from openai._types import NOT_GIVEN, NotGiven
 from openai.types.chat import ChatCompletionMessageParam

examples/pooling/plugin/prithvi_geospatial_mae_io_processor.py

@@ -1,8 +1,8 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-import base64
 import os
 
+import pybase64 as base64
 import torch
 
 from vllm import LLM

examples/pooling/plugin/prithvi_geospatial_mae_online.py

@@ -1,9 +1,9 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import os
 
+import pybase64 as base64
 import requests
 
 # This example shows how to perform an online inference that generates

examples/pooling/score/colqwen3_rerank_online.py

@@ -15,9 +15,9 @@ Then run this script:
     python colqwen3_rerank_online.py
 """
 
-import base64
 from io import BytesIO
 
+import pybase64 as base64
 import requests
 from PIL import Image
 

examples/pooling/token_embed/colqwen3_token_embed_online.py

@@ -21,10 +21,10 @@ Then run this script:
 """
 
 import argparse
-import base64
 from io import BytesIO
 
 import numpy as np
+import pybase64 as base64
 import requests
 from PIL import Image
 

tests/benchmarks/test_random_multimodal_dataset_video.py

@@ -1,12 +1,12 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import os
 from tempfile import NamedTemporaryFile
 from typing import Any, cast
 
 import cv2
+import pybase64 as base64
 import pytest
 from transformers import AutoTokenizer, PreTrainedTokenizerBase
 

tests/distributed/test_weight_transfer.py

@@ -6,10 +6,10 @@ Unit tests for engine classes (parsing, validation, registry).
 Integration tests for NCCL and IPC weight transfer between processes using Ray.
 """
 
-import base64
 import pickle
 from unittest.mock import MagicMock
 
+import pybase64 as base64
 import pytest
 import ray
 import torch

tests/entrypoints/openai/chat_completion/test_audio_in_video.py

@@ -1,10 +1,10 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import json
 
 import openai
+import pybase64 as base64
 import pytest
 import pytest_asyncio
 

tests/entrypoints/openai/chat_completion/test_vision_embeds.py

@@ -1,9 +1,8 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
-
 import numpy as np
+import pybase64 as base64
 import pytest
 import requests
 import torch

tests/entrypoints/openai/completion/test_completion_with_prompt_embeds.py

@@ -1,11 +1,11 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import io
 import json
 
 import openai  # use the official client for correctness check
+import pybase64 as base64
 import pytest
 import pytest_asyncio
 import torch

tests/entrypoints/openai/test_realtime_validation.py

@@ -2,12 +2,12 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 import asyncio
-import base64
 import json
 import warnings
 
 import librosa
 import numpy as np
+import pybase64 as base64
 import pytest
 import websockets
 

tests/entrypoints/pooling/embed/test_cohere_online.py

@@ -7,10 +7,10 @@ embedding models, covering text embedding, embedding type conversions,
 response structure, batching, normalisation, and semantic similarity.
 """
 
-import base64
 import struct
 
 import numpy as np
+import pybase64 as base64
 import pytest
 import requests
 

tests/entrypoints/pooling/embed/test_cohere_online_vision.py

@@ -6,11 +6,11 @@ Validates image embedding, batching, normalisation, and embedding type
 conversions through the /v2/embed endpoint.
 """
 
-import base64
 import struct
 import zlib
 
 import numpy as np
+import pybase64 as base64
 import pytest
 import requests
 

tests/entrypoints/pooling/embed/test_online.py

@@ -1,11 +1,11 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import json
 
 import numpy as np
 import openai
+import pybase64 as base64
 import pytest
 import pytest_asyncio
 import requests

tests/entrypoints/pooling/embed/test_protocol.py

@@ -3,10 +3,10 @@
 """Unit tests for Cohere embed protocol: build_typed_embeddings and its
 underlying packing helpers, plus Cohere-specific serving helpers."""
 
-import base64
 import struct
 
 import numpy as np
+import pybase64 as base64
 import pytest
 
 from vllm.entrypoints.pooling.embed.protocol import (

tests/entrypoints/pooling/pooling/test_online.py

@@ -1,10 +1,10 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import json
 
 import numpy as np
+import pybase64 as base64
 import pytest
 import requests
 import torch

tests/models/multimodal/generation/vlm_utils/model_utils.py

@@ -1235,9 +1235,9 @@ def voxtral_patch_hf_runner(hf_model: "HfRunner") -> "HfRunner":
     generated).
     """
 
-    import base64
     import io
 
+    import pybase64 as base64
     import soundfile as sf
 
     processor = hf_model.processor

tests/models/multimodal/pooling/test_colpali.py

@@ -7,9 +7,9 @@ ColPali is a multi-vector retrieval model based on PaliGemma backbone
 It produces per-token embeddings for both text and image inputs.
 """
 
-import base64
 from io import BytesIO
 
+import pybase64 as base64
 import pytest
 import torch
 from PIL import Image

tests/models/multimodal/pooling/test_colqwen3.py

@@ -7,9 +7,9 @@ ColBERT-style late interaction scoring (MaxSim). It produces per-token
 embeddings for both text and image inputs.
 """
 
-import base64
 from io import BytesIO
 
+import pybase64 as base64
 import pytest
 import torch
 from PIL import Image

tests/models/multimodal/pooling/test_llama_nemotron_vl.py

@@ -9,10 +9,10 @@ Tests for the LlamaNemotronVL model family:
 Both variants share a SigLIP vision encoder with a bidirectional LLaMA backbone.
 """
 
-import base64
 from io import BytesIO
 from pathlib import Path
 
+import pybase64 as base64
 import pytest
 import torch
 from transformers import AutoModel, AutoModelForSequenceClassification, AutoProcessor

tests/multimodal/media/test_audio.py

@@ -1,11 +1,11 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-import base64
 from pathlib import Path
 from unittest.mock import patch
 
 import librosa
 import numpy as np
+import pybase64 as base64
 import pytest
 
 from vllm.multimodal.media import AudioMediaIO

tests/multimodal/media/test_connector.py

@@ -2,13 +2,13 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 import asyncio
-import base64
 import mimetypes
 import os
 from tempfile import NamedTemporaryFile, TemporaryDirectory
 
 import aiohttp
 import numpy as np
+import pybase64 as base64
 import pytest
 import requests
 import torch

tests/plugins/prithvi_io_processor_plugin/prithvi_io_processor/prithvi_processor.py

@@ -1,7 +1,6 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import base64
 import datetime
 import os
 import tempfile
@@ -11,6 +10,7 @@ from typing import Any
 
 import albumentations
 import numpy as np
+import pybase64 as base64
 import rasterio
 import regex as re
 import torch

tests/plugins_tests/test_terratorch_io_processor_plugins.py

@@ -1,9 +1,9 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-import base64
 import io
 
 import imagehash
+import pybase64 as base64
 import pytest
 import requests
 from PIL import Image

tests/renderers/test_sparse_tensor_validation.py

@@ -5,9 +5,9 @@ Tests verify that malicious sparse tensors are rejected before they can trigger
 out-of-bounds memory writes during to_dense() operations.
 """
 
-import base64
 import io
 
+import pybase64 as base64
 import pytest
 import torch
 

tools/pre_commit/check_forbidden_imports.py

@@ -59,6 +59,14 @@ CHECK_IMPORTS = {
             "vllm/v1/serial_utils.py",
         },
     ),
+    "base64": ForbiddenImport(
+        pattern=r"^\s*(?:import\s+base64(?:$|\s|,)|from\s+base64\s+import)",
+        tip=(
+            "Replace 'import base64' with 'import pybase64' "
+            "or 'import pybase64 as base64'."
+        ),
+        allowed_pattern=re.compile(r"^\s*import\s+pybase64(\s*|\s+as\s+base64\s*)$"),
+    ),
     "re": ForbiddenImport(
         pattern=r"^\s*(?:import\s+re(?:$|\s|,)|from\s+re\s+import)",
         tip="Replace 'import re' with 'import regex as re' or 'import regex'.",

vllm/benchmarks/datasets.py

@@ -14,7 +14,6 @@ generation. Supported dataset types include:
 
 import argparse
 import ast
-import base64
 import io
 import json
 import logging
@@ -31,6 +30,7 @@ from tempfile import NamedTemporaryFile
 from typing import Any, cast
 
 import numpy as np
+import pybase64 as base64
 from huggingface_hub import snapshot_download
 from PIL import Image
 from typing_extensions import deprecated

vllm/distributed/weight_transfer/ipc_engine.py

@@ -2,12 +2,12 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 """IPC-based weight transfer engine using CUDA IPC for communication."""
 
-import base64
 import pickle
 from collections.abc import Callable, Iterator
 from dataclasses import asdict, dataclass
 from typing import Any
 
+import pybase64 as base64
 import requests
 import torch
 from torch.multiprocessing.reductions import reduce_tensor

vllm/entrypoints/openai/realtime/connection.py

@@ -2,13 +2,13 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 import asyncio
-import base64
 import json
 from collections.abc import AsyncGenerator
 from http import HTTPStatus
 from uuid import uuid4
 
 import numpy as np
+import pybase64 as base64
 from fastapi import WebSocket
 from starlette.websockets import WebSocketDisconnect
 

vllm/entrypoints/openai/run_batch.py

@@ -2,7 +2,6 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 import asyncio
-import base64
 import sys
 import tempfile
 from argparse import Namespace
@@ -13,6 +12,7 @@ from typing import Any, TypeAlias
 from urllib.parse import urlparse
 
 import aiohttp
+import pybase64 as base64
 import torch
 from fastapi import UploadFile
 from prometheus_client import start_http_server

vllm/entrypoints/pooling/embed/protocol.py

@@ -6,13 +6,13 @@ OpenAI: https://platform.openai.com/docs/api-reference/embeddings
 Cohere: https://docs.cohere.com/reference/embed
 """
 
-import base64
 import builtins
 import struct
 import time
 from collections.abc import Sequence
 from typing import Literal, TypeAlias
 
+import pybase64 as base64
 from pydantic import BaseModel, Field
 
 from vllm import PoolingParams

vllm/multimodal/media/audio.py

@@ -1,6 +1,5 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-import base64
 from io import BytesIO
 from pathlib import Path
 
@@ -138,7 +137,7 @@ class AudioMediaIO(MediaIO[tuple[npt.NDArray, float]]):
         media_type: str,
         data: str,
     ) -> tuple[npt.NDArray, float]:
-        return self.load_bytes(base64.b64decode(data))
+        return self.load_bytes(pybase64.b64decode(data))
 
     def load_file(self, filepath: Path) -> tuple[npt.NDArray, float]:
         return librosa.load(filepath, sr=None)
@@ -155,7 +154,7 @@ class AudioMediaIO(MediaIO[tuple[npt.NDArray, float]]):
             soundfile.write(buffer, audio, sr, format=audio_format)
             data = buffer.getvalue()
 
-        return base64.b64encode(data).decode("utf-8")
+        return pybase64.b64encode(data).decode("utf-8")
 
 
 class AudioEmbeddingMediaIO(MediaIO[torch.Tensor]):

vllm/multimodal/media/video.py

@@ -1,12 +1,12 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
-import base64
 from functools import partial
 from pathlib import Path
 from typing import Any
 
 import numpy as np
 import numpy.typing as npt
+import pybase64
 from PIL import Image
 
 from vllm import envs
@@ -84,7 +84,7 @@ class VideoMediaIO(MediaIO[tuple[npt.NDArray, dict[str, Any]]]):
                 [np.asarray(load_frame(frame_data)) for frame_data in data.split(",")]
             ), {}
 
-        return self.load_bytes(base64.b64decode(data))
+        return self.load_bytes(pybase64.b64decode(data))
 
     def load_file(self, filepath: Path) -> tuple[npt.NDArray, dict[str, Any]]:
         with filepath.open("rb") as f: