- Article
In recent years, the expansion of large-model parameters has substantially increased storage and inference overhead. Consequently, post-training quantization has become a key technique for reducing model size and inference-time energy consumption. Ho...