Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Course Outline
Introduction to Parameter-Efficient Fine-Tuning (PEFT)
- The drivers behind and constraints of full fine-tuning
- An overview of PEFT: objectives and key advantages
- Industry applications and relevant use cases
LoRA (Low-Rank Adaptation)
- Theoretical concepts and intuition behind LoRA
- Technical implementation using Hugging Face and PyTorch
- Practical session: Fine-tuning a model utilizing LoRA
Adapter Tuning
- Mechanics of adapter modules
- Integration strategies with transformer-based architectures
- Practical session: Applying Adapter Tuning to a transformer model
Prefix Tuning
- Leveraging soft prompts for effective fine-tuning
- Analysis of strengths and limitations relative to LoRA and adapters
- Practical session: Executing Prefix Tuning on an LLM task
Evaluating and Comparing PEFT Methods
- Key metrics for assessing performance and efficiency
- Trade-offs involving training speed, memory consumption, and model accuracy
- Conducting benchmarking experiments and interpreting results
Deploying Fine-Tuned Models
- Processes for saving and loading optimized models
- Deployment considerations specific to PEFT-based architectures
- Integration into existing applications and production pipelines
Best Practices and Extensions
- Synergizing PEFT with quantization and distillation techniques
- Application in low-resource and multilingual contexts
- Exploring future trajectories and active areas of research
Requirements
- A solid grasp of machine learning fundamentals
- Practical experience in working with large language models (LLMs)
- Proficiency in Python and PyTorch
Target Audience
- Data scientists
- AI engineers
14 Hours
