better usage of UV_PRERELEASE=allow #38482

Job	Run time
check_code_quality	32s
check_repository_consistency	29s
LoRA tests with PEFT main	7m 38s
Hub tests for models, schedulers, and pipelines	2m 48s
Fast PyTorch Models & Schedulers CPU tests	8m 44s
PyTorch Example CPU tests	20m 21s
Fast PyTorch Pipeline CPU tests	15m 48s
	56m 20s

Provide feedback