Add wait-for-model header when sending request to Inference API #6117

Job	Run time
build-windows (3.8)	5m 23s
build-ubuntu (3.8, Repository only)	1m 40s
build-ubuntu (3.8, Everything else)	2m 37s
build-ubuntu (3.8, torch)	56s
build-ubuntu (3.11, Repository only)	1m 41s
build-ubuntu (3.11, Everything else)	3m 10s
build-ubuntu (3.11, torch)	51s
build-ubuntu (3.11, lfs)	38s
build-ubuntu (3.8, fastai)	49s
build-ubuntu (3.10, fastai)	1m 27s
build-ubuntu (3.8, tensorflow)	59s
build-ubuntu (3.10, tensorflow)	50s
	21m 1s

Provide feedback