A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 17:15 UTC
Refactor error handling for multiple exceptions in preprocessing (#15650)
Signed-off-by: JasonZhu1313 <jasonchu13@outlook.com>
Jason (Siyu) Zhu committed
Mar 27, 2025 at 20:27 UTC
cec8c7d7f8753d13737427ceb5cebe987f5f0549
4 files changed
+5
-28
vllm/entrypoints/openai/serving_chat.py
+2
-10
@@ -197,16 +197,8 @@ class OpenAIServingChat(OpenAIServing):
197
truncate_prompt_tokens=request.truncate_prompt_tokens,
198
add_special_tokens=request.add_special_tokens,
199
)
200
- except ValueError as e:
201
- logger.exception("Error in preprocessing prompt inputs")
202
- return self.create_error_response(str(e))
203
- except TypeError as e:
204
- logger.exception("Error in preprocessing prompt inputs")
205
- return self.create_error_response(str(e))
206
- except RuntimeError as e:
207
- logger.exception("Error in preprocessing prompt inputs")
208
- return self.create_error_response(str(e))
209
- except jinja2.TemplateError as e:
200
+ except (ValueError, TypeError, RuntimeError,
201
+ jinja2.TemplateError) as e:
202
logger.exception("Error in preprocessing prompt inputs")
203
return self.create_error_response(str(e))
204
vllm/entrypoints/openai/serving_embedding.py
+1
-4
@@ -139,10 +139,7 @@ class OpenAIServingEmbedding(OpenAIServing):
139
truncate_prompt_tokens=truncate_prompt_tokens,
140
add_special_tokens=request.add_special_tokens,
141
)
142
- except ValueError as e:
143
- logger.exception("Error in preprocessing prompt inputs")
144
- return self.create_error_response(str(e))
145
- except TypeError as e:
142
+ except (ValueError, TypeError) as e:
143
logger.exception("Error in preprocessing prompt inputs")
144
return self.create_error_response(str(e))
145
vllm/entrypoints/openai/serving_pooling.py
+1
-7
@@ -136,13 +136,7 @@ class OpenAIServingPooling(OpenAIServing):
136
truncate_prompt_tokens=truncate_prompt_tokens,
137
add_special_tokens=request.add_special_tokens,
138
)
139
- except ValueError as e:
140
- logger.exception("Error in preprocessing prompt inputs")
141
- return self.create_error_response(str(e))
142
- except TypeError as e:
143
- logger.exception("Error in preprocessing prompt inputs")
144
- return self.create_error_response(str(e))
145
- except jinja2.TemplateError as e:
139
+ except (ValueError, TypeError, jinja2.TemplateError) as e:
140
logger.exception("Error in preprocessing prompt inputs")
141
return self.create_error_response(str(e))
142
vllm/entrypoints/openai/serving_tokenization.py
+1
-7
@@ -89,13 +89,7 @@ class OpenAIServingTokenization(OpenAIServing):
89
request.prompt,
90
add_special_tokens=request.add_special_tokens,
91
)
92
- except ValueError as e:
93
- logger.exception("Error in preprocessing prompt inputs")
94
- return self.create_error_response(str(e))
95
- except TypeError as e:
96
- logger.exception("Error in preprocessing prompt inputs")
97
- return self.create_error_response(str(e))
98
- except jinja2.TemplateError as e:
92
+ except (ValueError, TypeError, jinja2.TemplateError) as e:
93
logger.exception("Error in preprocessing prompt inputs")
94
return self.create_error_response(str(e))
95