Differentiable Product Quantization for End-to-End Embedding Compression

Abstract

Embedding layers are commonly used to map discrete symbols into continuous embedding vectors that reflect their semantic meanings. Despite their effectiveness, the number of parameters in an embedding layer increases linearly with the number of symbols and poses a critical challenge on memory and storage constraints. In this work, we propose a generic and end-to-end learnable compression framework termed differentiable product quantization (DPQ). We present two instantiations of DPQ that leverage different approximation techniques to enable differentiability in end-to-end learning. Our method can readily serve as a drop-in alternative for any existing embedding layer. Empirically, DPQ offers significant compression ratios (14-238$\times$) at negligible or no performance cost on 10 datasets across three different language tasks.

arXiv commentICML'2020. Code at https://github.com/chentingpc/dpq_embedding_compression

Cited by in corpus (4)

• Matching-oriented Product Quantization For Ad-hoc Retrieval
• Jointly Optimizing Query Encoder and Product Quantization to Improve Retrieval Performance
• Learning Elastic Embeddings for Customizing On-Device Recommenders
• Learning Multi-granular Quantized Embeddings for Large-Vocab Categorical Features in Recommender Systems