Describe your process for model quantization and the impact it has on accuracy versus speed for Qualcomm deployments.
