Jetfire: Efficient and Accurate Transformer Pretraining with INT8 Data Flow and Per-Block Quantization

Documenti analoghi

• Accurate INT8 Training Through Dynamic Block-Level Fallback
  di: Zhang, Pengle, et al.
  Pubblicazione: (2025)
• Oscillation-Reduced MXFP4 Training for Vision Transformers
  di: Chen, Yuxiang, et al.
  Pubblicazione: (2025)
• SageAttention2: Efficient Attention with Thorough Outlier Smoothing and Per-thread INT4 Quantization
  di: Zhang, Jintao, et al.
  Pubblicazione: (2024)
• INT-FlashAttention: Enabling Flash Attention for INT8 Quantization
  di: Chen, Shimao, et al.
  Pubblicazione: (2024)
• SpargeAttention: Accurate and Training-free Sparse Attention Accelerating Any Model Inference
  di: Zhang, Jintao, et al.
  Pubblicazione: (2025)