Adapter
Full Form: Adapter Module
Category: AI Techniques
📖 Definition
Adapters are small modules added to pre-trained models that enable task-specific adaptation without modifying the original model weights. They're a lightweight alternative to full fine-tuning.
🔑 Key Points
- Small modules added to existing models
- Only adapter weights are trained
- Enables multiple tasks with one base model
- Similar to LoRA, a form of parameter-efficient tuning
💡 Why It Matters
Adapters make AI customization accessible by avoiding full model retraining. They're key to efficient AI deployment.