Papers with iterative extreme learning machine
Extreme Fine-tuning: A Novel and Fast Fine-tuning Approach for Text Classification (2024.eacl-short)
Copied to clipboard
| Challenge: | Existing methods for fine-tuning pre-trained models require massive computational resources and time. |
| Approach: | They propose a novel approach for fine-tuning a pre-trained model using backpropagation and an iterative extreme learning machine for training a classifier. |
| Outcome: | The proposed approach outperforms state-of-the-art approaches in training-time measurement and performance with comparable model performance. |