MLM: Multi-linguistic LoRA Merging
Published in Efficient Reasoning Workshop at Neural Information Processing Systems (NeurIPS), 2025
We propose Multi-linguistic LoRA Merging (MLM), a method that merges language-specific LoRA adapters to efficiently transfer multilingual capability into a single model without full retraining, enabling scalable multilingual large language model deployment.
* Work done during a visiting student program at the University of Toronto in collaboration with LG Electronics Toronto AI Lab.