Cap FastAPI based on upstream PR 45594 (#566)

FastAPI 0.137.x is causing following failure in lm eval job:
https://github.com/neuralmagic/nm-cicd/actions/runs/27622052672/job/81674564991

`AttributeError: '_IncludedRouter' object has no attribute 'path'
File ".../prometheus_fastapi_instrumentator/routing.py", line 55, in
_get_route_name
    route_name = route.path`

Capping FastAPI<0.137 based on vLLM upstream PR
vllm-project/vllm#45594

Kicked off accept sync for the PR here:
https://github.com/neuralmagic/nm-cicd/actions/runs/27631372894/job/81707670895

Author: dhuangnm

requirements/common.txt

@@ -10,7 +10,7 @@ py-cpuinfo
 transformers >= 4.56.0, != 5.0.*, != 5.1.*, != 5.2.*, != 5.3.*, != 5.4.*, != 5.5.0
 tokenizers >= 0.21.1  # Required for fast incremental detokenization.
 protobuf >= 5.29.6, !=6.30.*, !=6.31.*, !=6.32.*, !=6.33.0.*, !=6.33.1.*, !=6.33.2.*, !=6.33.3.*, !=6.33.4.* # Required by LlamaTokenizer, gRPC. CVE-2026-0994
-fastapi[standard] >= 0.115.0 # Required by FastAPI's form models in the OpenAI API server's audio transcriptions endpoint.
+fastapi[standard] >= 0.115.0, < 0.137 # Required by FastAPI's form models in the OpenAI API server's audio transcriptions endpoint. Cap FastAPI<0.137 based on PR 45594.
 aiohttp >= 3.13.3
 openai >= 2.0.0  # For Responses API with reasoning content
 pydantic >= 2.12.0