PoTPTQ: A Two-step Power-of-Two Post-training for LLMs

Documenti analoghi

• AdpQ: A Zero-shot Calibration Free Adaptive Post Training Quantization Method for LLMs
  di: Ghaffari, Alireza, et al.
  Pubblicazione: (2024)
• OAC: Output-adaptive Calibration for Accurate Post-training Quantization
  di: Edalati, Ali, et al.
  Pubblicazione: (2024)
• Resona: Improving Context Copying in Linear Recurrence Models with Retrieval
  di: Wang, Xinyu, et al.
  Pubblicazione: (2025)
• Mitigating Outlier Activations in Low-Precision Fine-Tuning of Language Models
  di: Ghaffari, Alireza, et al.
  Pubblicazione: (2023)
• Rethinking Post-Training Quantization: Introducing a Statistical Pre-Calibration Approach
  di: Ghaffari, Alireza, et al.
  Pubblicazione: (2025)