Dr$^2$Net: Dynamic Reversible Dual-Residual Networks for Memory-Efficient Finetuning

Documenti analoghi

• Adaptive Human Trajectory Prediction via Latent Corridors
  di: Thakkar, Neerja, et al.
  Pubblicazione: (2023)
• $β$-CLIP: Text-Conditioned Contrastive Learning for Multi-Granular Vision-Language Alignment
  di: Zohra, Fatimah, et al.
  Pubblicazione: (2025)
• xT: Nested Tokenization for Larger Context in Large Images
  di: Gupta, Ritwik, et al.
  Pubblicazione: (2024)
• Pix4Point: Image Pretrained Standard Transformers for 3D Point Cloud Understanding
  di: Qian, Guocheng, et al.
  Pubblicazione: (2022)
• BOLT: Boost Large Vision-Language Model Without Training for Long-form Video Understanding
  di: Liu, Shuming, et al.
  Pubblicazione: (2025)