fix lint

Author: littlegy

autotest/interface/pipeline/test_pipeline_sleep_wakeup.py

@@ -7,10 +7,6 @@
 
 import pytest
 import torch
-from lmdeploy import GenerationConfig, PytorchEngineConfig, TurbomindEngineConfig, pipeline
-from lmdeploy.messages import Response
-from lmdeploy.serve.openai.protocol import UpdateParamsRequest
-from lmdeploy.utils import is_bf16_supported
 from utils.config_utils import get_parallel_config
 from utils.constant import SLEEP_WAKEUP_BACKENDS, SLEEP_WAKEUP_MODEL_LIST
 from utils.sleep_utils import (
@@ -25,6 +21,11 @@
     resolve_hf_checkpoint_dir,
 )
 
+from lmdeploy import GenerationConfig, PytorchEngineConfig, TurbomindEngineConfig, pipeline
+from lmdeploy.messages import Response
+from lmdeploy.serve.openai.protocol import UpdateParamsRequest
+from lmdeploy.utils import is_bf16_supported
+
 _SLEEP_PIPELINE_BACKEND_CLASS = {
     'pytorch': PytorchEngineConfig,
     'turbomind': TurbomindEngineConfig,

autotest/interface/restful/test_restful_abort_request.py

@@ -5,10 +5,11 @@
 
 import pytest
 import requests
-from lmdeploy.serve.openai.api_client import APIClient
 from utils.constant import BACKEND_LIST, DEFAULT_PORT, DEFAULT_SERVER, RESTFUL_MODEL_LIST
 from utils.restful_return_check import assert_chat_completions_batch_return
 
+from lmdeploy.serve.openai.api_client import APIClient
+
 BASE_URL = f'http://{DEFAULT_SERVER}:{DEFAULT_PORT}'
 JSON_HEADERS = {'Content-Type': 'application/json'}
 _REQUEST_TIMEOUT = 300

autotest/interface/restful/test_restful_sleep_wakeup.py

@@ -4,7 +4,6 @@
 import pytest
 import requests
 import torch
-from lmdeploy.serve.openai.api_client import APIClient
 from utils.constant import (
     DEFAULT_PORT,
     DEFAULT_SERVER,
@@ -25,6 +24,8 @@
     resolve_hf_checkpoint_dir,
 )
 
+from lmdeploy.serve.openai.api_client import APIClient
+
 BASE_URL = f'http://{DEFAULT_SERVER}:{DEFAULT_PORT}'
 JSON_HEADERS = {'Content-Type': 'application/json'}
 _REQUEST_TIMEOUT = 120

autotest/tools/pipeline/mllm_case.py

@@ -4,10 +4,11 @@
 
 import fire  # noqa: E402
 import numpy as np  # noqa: E402
+from PIL import Image  # noqa: E402
+
 from lmdeploy import GenerationConfig, PytorchEngineConfig, TurbomindEngineConfig, pipeline  # noqa: E402
 from lmdeploy.vl import encode_image_base64, load_image, load_video  # noqa: E402
 from lmdeploy.vl.constants import IMAGE_TOKEN  # noqa: E402
-from PIL import Image  # noqa: E402
 
 gen_config = GenerationConfig(max_new_tokens=500, min_new_tokens=10)
 

autotest/utils/run_restful_chat.py

@@ -7,7 +7,6 @@
 import allure
 import psutil
 import requests
-from lmdeploy.serve.openai.api_client import APIClient
 from openai import APIStatusError, BadRequestError, OpenAI
 from pytest_assume.plugin import assume
 from utils.config_utils import (
@@ -21,6 +20,8 @@
 from utils.restful_return_check import assert_chat_completions_batch_return
 from utils.rule_condition_assert import assert_result
 
+from lmdeploy.serve.openai.api_client import APIClient
+
 BASE_HTTP_URL = f'http://{DEFAULT_SERVER}'
 
 

autotest/utils/sleep_utils.py

@@ -8,10 +8,11 @@
 from typing import Any
 
 import torch
-from lmdeploy.utils import serialize_state_dict
 from safetensors.torch import safe_open
 from transformers.utils import SAFE_WEIGHTS_INDEX_NAME, SAFE_WEIGHTS_NAME, WEIGHTS_INDEX_NAME, WEIGHTS_NAME
 
+from lmdeploy.utils import serialize_state_dict
+
 UPDATE_WEIGHTS_CUDA_DEVICE_ENV = 'LMDEPLOY_UPDATE_WEIGHTS_CUDA_DEVICE'
 
 LEVEL2_GREEDY_MESSAGES = [{'role': 'user', 'content': '424242'}]