Parameter-Efficient Tuning

Tensorized Clustered LoRA Merging for Multi-Task Interference

Tensorized clustered LoRA merging to reduce multi-task interference in adapter-based LLM fine-tuning.