[ET-VK][qlinear] Add bias support to q4gsw and dq8ca_q4gsw quantized linear ops #10529

Job	Run time
test-metal-backend-modules / macos-job	1s
test-executorch-metal-build / macos-job	8m 18s
export-model-metal-artifact (mistralai, Voxtral-Mini-3B-2507, quantized-int4-metal) / macos-job	1s
export-model-metal-artifact (nvidia, parakeet-tdt, quantized-int4-metal) / macos-job	1s
export-model-metal-artifact (mistralai, Voxtral-Mini-3B-2507, non-quantized) / macos-job	1s
export-model-metal-artifact (mistralai, Voxtral-Mini-4B-Realtime-2602, quantized-int4-metal) / macos-job	1s
export-model-metal-artifact (openai, whisper-large-v3-turbo, non-quantized) / macos-job	1s
export-model-metal-artifact (openai, whisper-small, non-quantized) / macos-job	1s
export-model-metal-artifact (openai, whisper-small, quantized-int4-metal) / macos-job	1s
export-model-metal-artifact (openai, whisper-large-v3-turbo, quantized-int4-metal) / macos-job	1s
export-model-metal-artifact (nvidia, parakeet-tdt, non-quantized) / macos-job	1s
test-model-metal-e2e	-1s
	8m 27s

Provide feedback