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, 22:15 UTC
Fix `/v1/audio/transcriptions ` Bad Request Error (#13811)
Wen Sun committed
Feb 25, 2025 at 22:03 UTC
6522d55b6f0336fe1de573659d97e29d2378be9b
2 files changed
+2
-3
requirements-common.txt
+1
-2
@@ -9,8 +9,7 @@ py-cpuinfo
9
transformers >= 4.48.2 # Required for Bamba model and Transformers backend.
10
tokenizers >= 0.19.1 # Required for Llama 3.
11
protobuf # Required by LlamaTokenizer.
12
-fastapi[standard] >= 0.107.0, < 0.113.0; python_version < '3.9'
13
-fastapi[standard] >= 0.107.0, != 0.113.*, != 0.114.0; python_version >= '3.9'
12
+fastapi[standard] >= 0.115.0 # Required by FastAPI's form models in the OpenAI API server's audio transcriptions endpoint.
13
aiohttp
14
openai >= 1.52.0 # Ensure modern openai package (ensure types module present and max_completion_tokens field support)
15
pydantic >= 2.9
vllm/entrypoints/openai/protocol.py
+1
-1
@@ -1448,7 +1448,7 @@ AudioResponseFormat: TypeAlias = Literal["json", "text", "srt", "verbose_json",
1448
1449
class TranscriptionRequest(OpenAIBaseModel):
1450
# Ordered by official OpenAI API documentation
1451
- #https://platform.openai.com/docs/api-reference/audio/createTranscription
1451
+ # https://platform.openai.com/docs/api-reference/audio/createTranscription
1452
1453
file: UploadFile
1454
"""