🎈 auto fixes by pre-commit hooks

weclone/core/inference/offline_infer.py

@@ -7,7 +7,6 @@ from llamafactory.model import load_tokenizer
 from pydantic import BaseModel
 from vllm import LLM, SamplingParams
 from vllm.lora.request import LoRARequest
-from vllm.sampling_params import GuidedDecodingParams
 
 from weclone.utils.config import load_config
 from weclone.utils.config_models import VllmArgs
@@ -73,7 +72,7 @@ def vllm_infer(
     template_obj.mm_plugin.expand_mm_tokens = False  # for vllm generate
 
     if guided_decoding_class:
-        json_schema = guided_decoding_class.model_json_schema() 
+        json_schema = guided_decoding_class.model_json_schema()
 
     sampling_params = SamplingParams(
         repetition_penalty=generating_args.repetition_penalty or 1.0,  # repetition_penalty must > 0
@@ -118,5 +117,5 @@ def vllm_infer(
     results = LLM(**engine_args).chat(
         messages_list, sampling_params, lora_request=lora_request, chat_template_kwargs=extra_body
     )  # type: ignore
-    
+
     return results

weclone/utils/length_cdf.py

@@ -29,17 +29,17 @@ def calculate_token_length(
     template: str = "qwen3",
 ) -> int:
     """计算指定文本的token长度
-    
+
     Args:
         text: 要计算token长度的文本
         model_name_or_path: 模型路径
         template: 模板名称
-        
+
     Returns:
         文本的token长度
     """
     logger.info(f"正在计算文本token长度: {text[:50]}...")
-    
+
     model_args, data_args, _, _, _ = get_train_args(
         {
             "stage": "sft",
@@ -50,14 +50,14 @@ def calculate_token_length(
             "do_train": True,
         }
     )
-    
+
     tokenizer_module = load_tokenizer(model_args)
     tokenizer = tokenizer_module["tokenizer"]
-    
+
     # 直接使用tokenizer编码文本
     tokens = tokenizer.encode(text, add_special_tokens=False)
     token_length = len(tokens)
-    
+
     logger.info(f"文本token长度: {token_length}")
     return token_length