Exploring Post-Training Quantization of Protein Language Models