LoRA-Flow: Dynamic LoRA Fusion for Large Language Models in Generative Tasks
This work addresses the need for more flexible and efficient adaptation of large language models in generative tasks, particularly for scenarios with limited annotated data, though it is incremental in building upon existing LoRA combination methods.
The paper tackles the problem of combining multiple LoRA modules for generative tasks by introducing dynamic fusion weights that adjust per token, rather than using static task-level weights, and demonstrates consistent performance improvements across six generative tasks with minimal training data.
LoRA employs lightweight modules to customize large language models (LLMs) for each downstream task or domain, where different learned additional modules represent diverse skills. Combining existing LoRAs to address new tasks can enhance the reusability of learned LoRAs, particularly beneficial for tasks with limited annotated data. Most prior works on LoRA combination primarily rely on task-level weights for each involved LoRA, making different examples and tokens share the same LoRA weights. However, in generative tasks, different tokens may necessitate diverse skills to manage. Taking the Chinese math task as an example, understanding the problem description may depend more on the Chinese LoRA, while the calculation part may rely more on the math LoRA. To this end, we propose LoRA-Flow, which utilizes dynamic weights to adjust the impact of different LoRAs. The weights at each step are determined by a fusion gate with extremely few parameters, which can be learned with only 200 training examples. Experiments across six generative tasks demonstrate that our method consistently outperforms baselines with task-level fusion weights. This underscores the necessity of introducing dynamic fusion weights for LoRA combination.