EDGE OSDEFINITION
Parameter-Efficient Fine-Tuning (PEFT) is a set of methods that allows you to modify a pre-trained AI model, by only adapting a subset of its parameters or adding new adaptive parameters. Methods like LoRA, and adapters reduce the computational expense of adaptation for large models. This means that adapting large language models to bespoke uses becomes more feasible.
Why PEFT is useful
- Reduce training costs by reducing how many parameters are required to be trained.
- Decreases the overall usage of memory.
- Allows you to adapt pre-trained models to bespoke uses.
- Can help preserve individual adaptations per task.
- Makes adaptation more feasible.
When to use PEFT
- To adapt large language models to specific business vocabulary.
- Easy adaptation of various models of specific domains including technical, financial and legal.
- Fine-tune models for customer service and document classification.
- Allow task specific adaptations.
- Try out multiple different specialized models without having to retrain the base model.
See it running on your own hardware
Vanik Edge OS runs AI models, agents and apps inside your building, so your data stays there.