Text this: Optimizing Deep Learning Models for Resource‐Constrained Environments With Cluster‐Quantized Knowledge Distillation