Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO

arXiv:2311.04951v2 Announce Type: replace
Abstract: Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregressive sampling. This can be used together with model-based optimizations (e.g. quantization) to provide an optimized solution. Both sampling methods make use of KV caching. A Jupyter notebook and some sample executions are …

abstract article arxiv cache consumption costs cs.ai cs.lg cs.pf dynamic experience form generative improving inference infrastructure latency openvino power power consumption reduce sampling standard text text generation together type