ssProp: Energy-Efficient Training for Convolutional Neural Networks with Scheduled Sparse Back Propagation

Documenti analoghi

• PropMEND: Hypernetworks for Knowledge Propagation in LLMs
  di: Liu, Zeyu Leo, et al.
  Pubblicazione: (2025)
• FairSample: Training Fair and Accurate Graph Convolutional Neural Networks Efficiently
  di: Cong, Zicun, et al.
  Pubblicazione: (2024)
• Dense-Sparse Deep Convolutional Neural Networks Training for Image Denoising
  di: Alawode, Basit O., et al.
  Pubblicazione: (2021)
• Advancing On-Device Neural Network Training with TinyPropv2: Dynamic, Sparse, and Efficient Backpropagation
  di: Rüb, Marcus, et al.
  Pubblicazione: (2024)
• GraphProp: Training the Graph Foundation Models using Graph Properties
  di: Sun, Ziheng, et al.
  Pubblicazione: (2025)