Module lora 

LoRA (Low-Rank Adaptation)

Implements parameter-efficient fine-tuning:

• LoRA: Low-Rank Adaptation of Large Language Models
• QLoRA: Quantized LoRA for even more efficiency
• Adapter layers with low-rank decomposition
• Merge and unmerge LoRA weights

Structs

LoRAConfig

LoRA configuration

LoRALinear

LoRA layer wrapping a linear layer

QLoRAConfig

QLoRA (Quantized LoRA) configuration

QLoRALinear

QLoRA layer with quantized base weights

Enums

QuantType

Quantization type for QLoRA