An under-communicated point is that the default tuning for most libraries and platforms for evaluating machine-learned models is unsuitable for serving at scale.
In this post we explore a text-to-image search application on Vespa using approximate nearest neighbor search on vector representations of text and images.
It has become increasingly important to efficiently evaluate machine-learned models in the stateless container cluster. We have recently added accelerated model evaluation here, which opens...
In this post, we reproduce the state-of-the-art baseline for retrieval-based question-answering systems within a single, scalable production ready application on Vespa.ai.
Never miss a story from us, subscribe to our newsletter