TopicBERT for Energy Efficient Document Classification
This addresses energy and cost issues for researchers and practitioners using BERT for long-sequence tasks like document classification, though it is incremental as it builds on existing efficiency methods.
The paper tackles the computational inefficiency of BERT in fine-tuning for document classification by proposing TopicBERT, which combines topic and language models to reduce self-attention operations. This results in a 1.4x speedup, 40% reduction in CO₂ emissions, and retains 99.9% performance across 5 datasets.
Prior research notes that BERT's computational cost grows quadratically with sequence length thus leading to longer training times, higher GPU memory constraints and carbon emissions. While recent work seeks to address these scalability issues at pre-training, these issues are also prominent in fine-tuning especially for long sequence tasks like document classification. Our work thus focuses on optimizing the computational cost of fine-tuning for document classification. We achieve this by complementary learning of both topic and language models in a unified framework, named TopicBERT. This significantly reduces the number of self-attention operations - a main performance bottleneck. Consequently, our model achieves a 1.4x ($\sim40\%$) speedup with $\sim40\%$ reduction in $CO_2$ emission while retaining $99.9\%$ performance over 5 datasets.