Medium

Implement a LoRA-wrapped linear layer (frozen WW ++ trainable BAB\cdot A scaled by $

SFT, Instruction Tuning, Data & PEFT · Problem 2 of 6

Chapter 08SFT, Instruction Tuning, Data & PEFT

Implement a LoRA-wrapped linear layer (frozen WW ++ trainable BAB\cdot A scaled by $

MediumProblem 2 / 6

Implement a LoRA-wrapped linear layer (frozen WW ++ trainable BAB\cdot A scaled by α/r\alpha/r). [HuggingFace]

Implement the function/class skeleton in the editor. Any correct approach is accepted.

Hints