Blog for Vespa - The open big data serving engine
Vespa Product Updates, December 2020
Advances in Vespa features and performance include improved tensor ranking performance, Apache ZooKeeper integration, Vespa Python API for researchers and ONNX integration.
17 Dec 2020
Stateful model serving: how we accelerate inference using ONNX Runtime
There's a difference between stateless and stateful model serving.
14 Dec 2020
Fine-tuning a BERT model for search applications
How to ensure training and serving encoding compatibility.
25 Nov 2020
From research to production: scaling a state-of-the-art machine learning system
How we implemented a production-ready question-answering application and reduced response time by more than two orders of magnitude.
12 Nov 2020
Fine-tuning a BERT model with transformers
Setup a custom Dataset, fine-tune BERT with Transformers Trainer and export the model via ONNX.
12 Nov 2020
Vespa Product Updates, October 2020
Improvement to Vespa feeding APIs
30 Oct 2020
Vespa Product Updates, September 2020
Introducing ONNX-Runtime, Hamming Distance Metric, Conditional Update Performance Improvements and Compressed Transaction Log with Synced Ack
01 Oct 2020
Efficient open-domain question-answering on Vespa.ai
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.
30 Sep 2020
Vespa Product Updates, August 2020
Introducing NLP with Transformers, Grafana how-to, Improved GEO Search Support, Query Profile Variants Optimizations, & Build on Debian 10
27 Aug 2020
Vespa Product Updates, June 2020
Announcing support for approximate nearest neighbor vector search which can be combined with filters and text search with state-of-the art performance
05 Jul 2020
Introducing NLP with Transformers on Vespa
We’ve been working a lot lately on evaluating Transformer models in Vespa. Here we show how and share a bit on how we view the benefits of inference in Vespa....
02 Jul 2020
Approximate Nearest Neighbor Search in Vespa - Part 1
In this blog post we explore how the Vespa team selected HNSW (Hierarchical Navigable Small World Graphs) as the baseline approximate nearest neighbor algorithm for extension and integration in Vespa....
30 Jun 2020
Never miss a
from us, subscribe to our newsletter