add `.isort.cfg` (#378)

.isort.cfg

+[settings]
+profile=black
+known_first_party=sglang
\ No newline at end of file

python/sglang/api.py

@@ -3,11 +3,7 @@
 import re
 from typing import Callable, List, Optional, Union
 
-from sglang.backend.anthropic import Anthropic
 from sglang.backend.base_backend import BaseBackend
-from sglang.backend.openai import OpenAI
-from sglang.backend.runtime_endpoint import RuntimeEndpoint
-from sglang.backend.vertexai import VertexAI
 from sglang.global_config import global_config
 from sglang.lang.ir import (
     SglExpr,

python/sglang/backend/anthropic.py

-from typing import List, Optional, Union
-
-import numpy as np
 from sglang.backend.base_backend import BaseBackend
 from sglang.lang.chat_template import get_chat_template
 from sglang.lang.interpreter import StreamExecutor

python/sglang/backend/base_backend.py

-from typing import Callable, List, Optional, Union
+from typing import List, Optional, Union
 
 from sglang.lang.chat_template import get_chat_template
 from sglang.lang.interpreter import StreamExecutor

python/sglang/backend/openai.py

 import logging
 import time
-from typing import Callable, List, Optional, Union
+from typing import List, Optional
 
 import numpy as np
+
 from sglang.backend.base_backend import BaseBackend
 from sglang.lang.chat_template import ChatTemplate, get_chat_template_by_model_path
 from sglang.lang.interpreter import StreamExecutor

python/sglang/backend/runtime_endpoint.py

 import json
-from typing import Callable, List, Optional, Union
+from typing import List, Optional
 
 import numpy as np
-import requests
+
 from sglang.backend.base_backend import BaseBackend
 from sglang.global_config import global_config
 from sglang.lang.chat_template import get_chat_template_by_model_path
 from sglang.lang.interpreter import StreamExecutor
-from sglang.lang.ir import SglArgument, SglSamplingParams
-from sglang.utils import encode_image_base64, find_printable_text, http_request
+from sglang.lang.ir import SglSamplingParams
+from sglang.utils import find_printable_text, http_request
 
 
 class RuntimeEndpoint(BaseBackend):

python/sglang/backend/vertexai.py

 import os
 import warnings
-from typing import List, Optional, Union
 
-import numpy as np
 from sglang.backend.base_backend import BaseBackend
 from sglang.lang.chat_template import get_chat_template
 from sglang.lang.interpreter import StreamExecutor

python/sglang/lang/chat_template.py

-from dataclasses import dataclass, field
+from dataclasses import dataclass
 from enum import Enum, auto
-from typing import Callable, Dict, List, Optional, Tuple
+from typing import Callable, Dict, List, Tuple
 
 
 class ChatTemplateStyle(Enum):

python/sglang/lang/compiler.py

@@ -5,13 +5,7 @@ from typing import List, Union
 
 from sglang.global_config import global_config
 from sglang.lang.interpreter import ProgramState, StreamExecutor, pin_program
-from sglang.lang.ir import (
-    SglArgument,
-    SglConstantText,
-    SglExpr,
-    SglSamplingParams,
-    SglVariable,
-)
+from sglang.lang.ir import SglArgument, SglExpr, SglSamplingParams, SglVariable
 
 
 def compile_func(function, backend):

python/sglang/lang/interpreter.py

@@ -7,9 +7,10 @@ import threading
 import uuid
 from concurrent.futures import ThreadPoolExecutor
 from contextlib import contextmanager
-from typing import Any, Callable, Dict, List, Optional, Union
+from typing import Any, Callable, Dict, List, Optional
 
 import tqdm
+
 from sglang.global_config import global_config
 from sglang.lang.ir import (
     SglCommitLazy,
@@ -17,7 +18,6 @@ from sglang.lang.ir import (
     SglConstantText,
     SglExpr,
     SglExprList,
-    SglFunction,
     SglGen,
     SglImage,
     SglRoleBegin,

python/sglang/lang/ir.py

@@ -472,4 +472,4 @@ class SglCommitLazy(SglExpr):
         super().__init__()
 
     def __repr__(self):
-        return f"CommitLazy()"
+        return "CommitLazy()"

python/sglang/lang/tracer.py

 """Tracing a program."""
 
 import uuid
-from typing import Any, Callable, Dict, List, Optional, Union
+from typing import Any, Dict, List, Optional
 
 from sglang.backend.base_backend import BaseBackend
-from sglang.global_config import global_config
 from sglang.lang.interpreter import ProgramState, ProgramStateGroup
 from sglang.lang.ir import (
     SglArgument,
-    SglCommitLazy,
-    SglConcateAndAppend,
     SglConstantText,
     SglExpr,
     SglExprList,
     SglFork,
-    SglFunction,
     SglGen,
     SglGetForkItem,
     SglRoleBegin,

python/sglang/srt/constrained/jump_forward.py

 import interegular
+
 from sglang.srt.constrained import FSMInfo, disk_cache, make_deterministic_fsm
 from sglang.srt.constrained.base_cache import BaseCache
 

python/sglang/srt/hf_transformers_utils.py

@@ -3,10 +3,9 @@
 import json
 import os
 import warnings
-from typing import List, Optional, Tuple, Union
+from typing import Optional, Union
 
 from huggingface_hub import snapshot_download
-from sglang.srt.utils import is_multimodal_model
 from transformers import (
     AutoConfig,
     AutoProcessor,
@@ -15,6 +14,8 @@ from transformers import (
     PreTrainedTokenizerFast,
 )
 
+from sglang.srt.utils import is_multimodal_model
+
 
 def download_from_hf(model_path: str):
     if os.path.exists(model_path):

python/sglang/srt/layers/context_flashattention_nopad.py

@@ -3,6 +3,7 @@
 import torch
 import triton
 import triton.language as tl
+
 from sglang.srt.utils import wrap_kernel_launcher
 
 CUDA_CAPABILITY = torch.cuda.get_device_capability()

python/sglang/srt/layers/extend_attention.py

 import torch
 import triton
 import triton.language as tl
+
 from sglang.srt.layers.context_flashattention_nopad import context_attention_fwd
 from sglang.srt.utils import wrap_kernel_launcher
 

python/sglang/srt/layers/logits_processor.py

 import torch
-from sglang.srt.managers.router.model_runner import ForwardMode, InputMetadata
 from torch import nn
 from vllm.model_executor.parallel_utils.communication_op import (
     get_tensor_model_parallel_world_size,
     tensor_model_parallel_all_gather,
 )
 
+from sglang.srt.managers.router.model_runner import ForwardMode, InputMetadata
+
 
 class LogitsProcessor(nn.Module):
     def __init__(self, config):

python/sglang/srt/layers/radix_attention.py

 import torch
+from torch import nn
+
 from sglang.srt.layers.context_flashattention_nopad import context_attention_fwd
 from sglang.srt.layers.extend_attention import extend_attention_fwd
 from sglang.srt.layers.token_attention import token_attention_fwd
 from sglang.srt.managers.router.model_runner import ForwardMode, InputMetadata
-from torch import nn
 
 
 class RadixAttention(nn.Module):

python/sglang/srt/layers/token_attention.py

@@ -4,6 +4,7 @@
 import torch
 import triton
 import triton.language as tl
+
 from sglang.srt.managers.router.model_runner import global_server_args_dict
 from sglang.srt.utils import wrap_kernel_launcher
 

python/sglang/srt/managers/detokenizer_manager.py

@@ -3,6 +3,7 @@ import asyncio
 import uvloop
 import zmq
 import zmq.asyncio
+
 from sglang.srt.hf_transformers_utils import get_tokenizer
 from sglang.srt.managers.io_struct import BatchStrOut, BatchTokenIDOut
 from sglang.srt.server_args import PortArgs, ServerArgs
@@ -83,7 +84,7 @@ def start_detokenizer_process(
 ):
     try:
         manager = DetokenizerManager(server_args, port_args)
-    except Exception as e:
+    except Exception:
         pipe_writer.send(get_exception_traceback())
         raise
     pipe_writer.send("init ok")