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 14, 05:30 UTC
[CI Bugfix] Pin `openai<1.100` to unblock CI (#23118)
Signed-off-by: mgoin <mgoin64@gmail.com>
Michael Goin committed
Aug 18, 2025 at 15:14 UTC
bf756321c72340466911b64602e88013d0210c1c
2 files changed
+2
-2
requirements/common.txt
+1
-1
@@ -12,7 +12,7 @@ tokenizers >= 0.21.1 # Required for fast incremental detokenization.
12
protobuf # Required by LlamaTokenizer.
13
fastapi[standard] >= 0.115.0 # Required by FastAPI's form models in the OpenAI API server's audio transcriptions endpoint.
14
aiohttp
15
-openai >= 1.99.1 # For Responses API with reasoning content
15
+openai >= 1.99.1, < 1.100.0 # For Responses API with reasoning content
16
pydantic >= 2.10
17
prometheus_client >= 0.18.0
18
pillow # Required for image processing
requirements/docs.txt
+1
-1
@@ -18,7 +18,7 @@ cbor2
18
cloudpickle
19
fastapi
20
msgspec
21
-openai
21
+openai < 1.100.0
22
openai-harmony
23
partial-json-parser
24
pillow