PACSET (Packed Serialized Trees): Reducing Inference Latency for Tree Ensemble Deployment
This addresses latency issues for tree ensemble deployment in resource-constrained environments like IoT and web micro-services, representing an incremental improvement in optimization techniques.
The paper tackles the problem of high inference latency for tree ensembles when models are not pre-loaded into memory, such as on low-resource devices, by introducing PACSET, a method that optimizes serialization to improve data locality and I/O efficiency, resulting in a 2-6 times reduction in classification latency.
We present methods to serialize and deserialize tree ensembles that optimize inference latency when models are not already loaded into memory. This arises whenever models are larger than memory, but also systematically when models are deployed on low-resource devices, such as in the Internet of Things, or run as Web micro-services where resources are allocated on demand. Our packed serialized trees (PACSET) encode reference locality in the layout of a tree ensemble using principles from external memory algorithms. The layout interleaves correlated nodes across multiple trees, uses leaf cardinality to collocate the nodes on the most popular paths and is optimized for the I/O blocksize. The result is that each I/O yields a higher fraction of useful data, leading to a 2-6 times reduction in classification latency for interactive workloads.