Asked about Prompt, retrieve or fine-tune
What is LoRA?
An answer that works
Parameter-efficient fine-tuning. The original weights are frozen and a small number of added parameters are trained, reaching most of the benefit at a fraction of the cost. The result is a small adapter that can be swapped per task against one base model.
Reading an answer is not being able to give one. The section that teaches this is Prompt, retrieve or fine-tune, 8 hours of Fine-tuning.
Learn this properlyAlso asked about Fine-tuning
When would you fine-tune rather than use RAG?What is the cost of fine-tuning beyond the training run?What is distillation?How much data do you need to fine-tune?Where do the examples come from?What is catastrophic forgetting?How do you know the fine-tune actually beat the prompt?You have five fine-tunes of one base model to serve. Merged or separate adapters?
All 13 questions on Fine-tuning