[ICML 2023] SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models
Primary LanguagePythonMIT LicenseMIT