enable isort

pyproject.toml

@@ -86,6 +86,7 @@ lint.select = [
     "F",  # see: https://pypi.org/project/pyflakes
     "N",  # see: https://pypi.org/project/pep8-naming
 #    "D",  # see: https://pypi.org/project/pydocstyle
+    "I", # implementation for isort
 ]
 lint.extend-select = [
     "C4",  # see: https://pypi.org/project/flake8-comprehensions
@@ -108,7 +109,6 @@ exclude = [
     "dist",
     "docs"
 ]
-lint.ignore-init-module-imports = true
 
 [tool.ruff.lint.per-file-ignores]
 "setup.py" = ["D100", "SIM115"]

src/litserve/__init__.py

@@ -11,12 +11,12 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+from litserve import test_examples
 from litserve.__about__ import *  # noqa: F403
 from litserve.api import LitAPI
-from litserve.server import LitServer, Request, Response
-from litserve import test_examples
-from litserve.specs.openai import OpenAISpec
 from litserve.callbacks import Callback
 from litserve.loggers import Logger
+from litserve.server import LitServer, Request, Response
+from litserve.specs.openai import OpenAISpec
 
-__all__ = ["LitAPI", "LitServer", "Request", "Response", "test_examples", "OpenAISpec", "Callback", "Logger"]
+__all__ = ["LitAPI", "LitServer", "Request", "Response", "OpenAISpec", "test_examples", "Callback", "Logger"]

src/litserve/__main__.py

@@ -11,7 +11,8 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from argparse import ArgumentParser, ArgumentDefaultsHelpFormatter, RawTextHelpFormatter
+from argparse import ArgumentDefaultsHelpFormatter, ArgumentParser, RawTextHelpFormatter
+
 from litserve.docker_builder import dockerize
 
 

src/litserve/api.py

@@ -15,8 +15,8 @@
 import json
 import warnings
 from abc import ABC, abstractmethod
-from typing import Optional
 from queue import Queue
+from typing import Optional
 
 from pydantic import BaseModel
 

src/litserve/callbacks/__init__.py

@@ -1,3 +1,3 @@
-from .base import Callback, CallbackRunner, EventTypes, NoopCallback
+from litserve.callbacks.base import Callback, CallbackRunner, EventTypes, NoopCallback
 
 __all__ = ["Callback", "CallbackRunner", "EventTypes", "NoopCallback"]

src/litserve/callbacks/defaults/metric_callback.py

@@ -2,7 +2,7 @@
 import typing
 from logging import getLogger
 
-from ..base import Callback
+from litserve.callbacks.base import Callback
 
 if typing.TYPE_CHECKING:
     from litserve import LitAPI

src/litserve/docker_builder.py

@@ -13,9 +13,9 @@
 # limitations under the License.
 import logging
 import os
+import warnings
 from pathlib import Path
 
-import warnings
 import litserve as ls
 
 logger = logging.getLogger(__name__)

src/litserve/loggers.py

@@ -12,13 +12,13 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 import functools
+import logging
 import multiprocessing as mp
 import pickle
 from abc import ABC, abstractmethod
-from typing import List, Optional, Union, TYPE_CHECKING
+from typing import TYPE_CHECKING, List, Optional, Union
 
 from starlette.types import ASGIApp
-import logging
 
 module_logger = logging.getLogger(__name__)
 

src/litserve/specs/openai.py

@@ -26,8 +26,8 @@
 from fastapi.responses import StreamingResponse
 from pydantic import BaseModel, Field
 
-from ..utils import LitAPIStatus, azip, load_and_raise
-from .base import LitSpec
+from litserve.specs.base import LitSpec
+from litserve.utils import LitAPIStatus, azip, load_and_raise
 
 if typing.TYPE_CHECKING:
     from litserve import LitServer

src/litserve/test_examples/__init__.py

@@ -1,11 +1,11 @@
-from .simple_example import SimpleBatchedAPI, SimpleLitAPI, SimpleTorchAPI, SimpleStreamAPI
-from .openai_spec_example import (
+from litserve.test_examples.openai_spec_example import (
+    OpenAIBatchContext,
     TestAPI,
     TestAPIWithCustomEncode,
     TestAPIWithStructuredOutput,
     TestAPIWithToolCalls,
-    OpenAIBatchContext,
 )
+from litserve.test_examples.simple_example import SimpleBatchedAPI, SimpleLitAPI, SimpleStreamAPI, SimpleTorchAPI
 
 __all__ = [
     "SimpleLitAPI",

src/litserve/test_examples/openai_spec_example.py

@@ -13,11 +13,11 @@
 # limitations under the License.
 import time
 
-import litserve as ls
+from litserve.api import LitAPI
 from litserve.specs.openai import ChatMessage
 
 
-class TestAPI(ls.LitAPI):
+class TestAPI(LitAPI):
     def setup(self, device):
         self.model = None
 
@@ -53,7 +53,7 @@ def encode_response(self, output):
         )
 
 
-class OpenAIBatchContext(ls.LitAPI):
+class OpenAIBatchContext(LitAPI):
     def setup(self, device: str) -> None:
         self.model = None
 
@@ -82,7 +82,7 @@ def encode_response(self, output_stream, context):
             assert ctx["temperature"] == 1.0, f"context {ctx} is not 1.0"
 
 
-class OpenAIWithUsage(ls.LitAPI):
+class OpenAIWithUsage(LitAPI):
     def setup(self, device):
         self.model = None
 
@@ -96,7 +96,7 @@ def predict(self, x):
         }
 
 
-class OpenAIWithUsageEncodeResponse(ls.LitAPI):
+class OpenAIWithUsageEncodeResponse(LitAPI):
     def setup(self, device):
         self.model = None
 

src/litserve/test_examples/simple_example.py

@@ -1,7 +1,7 @@
-import litserve as ls
+from litserve.api import LitAPI
 
 
-class SimpleLitAPI(ls.LitAPI):
+class SimpleLitAPI(LitAPI):
     def setup(self, device):
         # Set up the model, so it can be called in `predict`.
         self.model = lambda x: x**2
@@ -19,7 +19,7 @@ def encode_response(self, output):
         return {"output": output}
 
 
-class SimpleBatchedAPI(ls.LitAPI):
+class SimpleBatchedAPI(LitAPI):
     def setup(self, device) -> None:
         self.model = lambda x: x**2
 
@@ -35,7 +35,7 @@ def encode_response(self, output):
         return {"output": output}
 
 
-class SimpleTorchAPI(ls.LitAPI):
+class SimpleTorchAPI(LitAPI):
     def setup(self, device):
         # move the model to the correct device
         # keep track of the device for moving data accordingly
@@ -69,7 +69,7 @@ def encode_response(self, output):
         return {"output": float(output)}
 
 
-class SimpleStreamAPI(ls.LitAPI):
+class SimpleStreamAPI(LitAPI):
     """
     Run as:
         ```

src/litserve/utils.py

@@ -19,7 +19,6 @@
 
 from fastapi import HTTPException
 
-
 if TYPE_CHECKING:
     from litserve.server import LitServer
 

tests/conftest.py

@@ -12,16 +12,17 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 import time
-import psutil
 from typing import Generator
 
-from litserve.server import LitServer
+import psutil
 import pytest
-from litserve.api import LitAPI
-from litserve.utils import wrap_litserve_start
 from fastapi import Request, Response
 from fastapi.testclient import TestClient
 
+from litserve.api import LitAPI
+from litserve.server import LitServer
+from litserve.utils import wrap_litserve_start
+
 
 class SimpleLitAPI(LitAPI):
     def setup(self, device):

tests/e2e/default_openai_with_batching.py

@@ -1,6 +1,5 @@
-from litserve.test_examples.openai_spec_example import OpenAIBatchContext
-
 import litserve as ls
+from litserve.test_examples.openai_spec_example import OpenAIBatchContext
 
 if __name__ == "__main__":
     api = OpenAIBatchContext()

tests/e2e/default_openaispec.py

@@ -1,6 +1,6 @@
+import litserve as ls
 from litserve import OpenAISpec
 from litserve.test_examples.openai_spec_example import TestAPI
-import litserve as ls
 
 if __name__ == "__main__":
     server = ls.LitServer(TestAPI(), spec=OpenAISpec())

tests/e2e/default_openaispec_tools.py

@@ -1,7 +1,7 @@
 import litserve as ls
 from litserve import OpenAISpec
-from litserve.test_examples.openai_spec_example import TestAPI
 from litserve.specs.openai import ChatMessage
+from litserve.test_examples.openai_spec_example import TestAPI
 
 
 class TestAPIWithToolCalls(TestAPI):

tests/e2e/default_single_streaming.py

@@ -1,4 +1,4 @@
-from litserve import LitServer, LitAPI
+from litserve import LitAPI, LitServer
 
 
 class SimpleStreamingAPI(LitAPI):

tests/e2e/default_spec.py

@@ -11,10 +11,9 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from litserve.test_examples.openai_spec_example import TestAPI
-from litserve.specs.openai import OpenAISpec
-
 import litserve as ls
+from litserve.specs.openai import OpenAISpec
+from litserve.test_examples.openai_spec_example import TestAPI
 
 if __name__ == "__main__":
     server = ls.LitServer(TestAPI(), spec=OpenAISpec())

tests/e2e/test_e2e.py

@@ -13,13 +13,13 @@
 # limitations under the License.
 import json
 import os
-import psutil
-import requests
 import subprocess
 import time
+from functools import wraps
 
+import psutil
+import requests
 from openai import OpenAI
-from functools import wraps
 
 
 def e2e_from_file(filename):

tests/minimal_run.py

@@ -11,12 +11,13 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+import json
 import subprocess
-import psutil
 import time
-import json
 import urllib.request
 
+import psutil
+
 
 def main():
     process = subprocess.Popen(

tests/parity_fastapi/benchmark.py

@@ -3,10 +3,10 @@
 import random
 import time
 
+import numpy as np
 import requests
 import torch
 from PIL import Image
-import numpy as np
 
 device = "cuda" if torch.cuda.is_available() else "cpu"
 device = "mps" if torch.backends.mps.is_available() else device

tests/parity_fastapi/ls-server.py

@@ -9,7 +9,6 @@
 
 import litserve as ls
 
-
 device = "cuda" if torch.cuda.is_available() else "cpu"
 device = "mps" if torch.backends.mps.is_available() else device
 conf = {

tests/parity_fastapi/main.py

@@ -1,13 +1,12 @@
-import torch
-import requests
-
-from benchmark import run_bench
-import psutil
 import subprocess
 import time
-
 from functools import wraps
 
+import psutil
+import requests
+import torch
+from benchmark import run_bench
+
 CONF = {
     "cpu": {"num_requests": 50},
     "mps": {"num_requests": 50},

tests/perf_test/bert/benchmark.py

@@ -1,9 +1,9 @@
-import time
 import logging
+import time
 
 import requests
-from utils import benchmark
 from tenacity import retry, stop_after_attempt
+from utils import benchmark
 
 # Configuration
 SERVER_URL = "http://0.0.0.0:8000/predict"

tests/perf_test/bert/server.py

@@ -3,6 +3,7 @@
 import torch
 from jsonargparse import CLI
 from transformers import AutoModelForSequenceClassification, AutoTokenizer, BertConfig
+
 import litserve as ls
 
 # Set float32 matrix multiplication precision if GPU is available and capable

tests/simple_server.py

@@ -11,10 +11,11 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from litserve.server import LitServer
-from litserve.api import LitAPI
 from fastapi import Request, Response
 
+from litserve.api import LitAPI
+from litserve.server import LitServer
+
 
 class SimpleLitAPI(LitAPI):
     def setup(self, device):

tests/simple_server_diff_port.py

@@ -11,8 +11,8 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from litserve.server import LitServer
 from litserve.api import LitAPI
+from litserve.server import LitServer
 
 
 class SimpleLitAPI(LitAPI):