Trainable Parameters
المعاملات القابلة للتدريب
المعاملات التي تتلقى تحديثات من متجهات الميل أثناء التدريب. في LoRA، هي فقط المصفوفتان B وA.
Parameters that receive gradient updates during training. In LoRA, these are only the two matrices B and A.
Also translated asالمعاملات المتعلَّمة، المتغيرات القابلة للتحسين
First appears in this corpus in: PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation (2017)
Appears in these papers
- Training Data-Efficient Image Transformers & Distillation Through Attention2021in the sky ✦
- LoRA: Low-Rank Adaptation of Large Language Models2021in the sky ✦
- OPT: Open Pre-Trained Transformer Language Models2022in the sky ✦
- PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation2017in the sky ✦
- Squeeze-and-Excitation Networks2018in the sky ✦
- Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows2021in the sky ✦