151
New
The latest blog posts to take off on Hacker News. 371 posts.
152
Which vector database should I use? A comparison cheatsheet
Semantic search and retrieval-augmented generation (RAG) applications require systems to be able to save lots of embedding vectors…
153
So, you want to deploy on the edge?
Application developers often deploy their apps into a single area, generally represented by a handful of AZs in a single region. No matter where their …
154
Gzip beats BERT? Part 2: dataset issues, improved speed, and results
Summary : Update to my last post analyzing the gzip+knn paper. Some of the datasets have heavily contaminated train/test tests. I show some ways to significantly speed-up the gzip method. I present more results of my "fair" reimplementation .
155
Ways to shoot yourself in the foot with Redis
Personal website of Phil Booth
156
LLaMA2 Chat 70B outperformed ChatGPT
AlpacaEval an LLM-based automatic evaluation that is fast, cheap, and reliable. It is based on the AlpacaFarm evaluation set, which tests the ability of models to follow general user instructions. These responses are then compared to reference responses (Davinci003 for AlpacaEval, GPT-4 Preview for AlpacaEval 2.0) by…
157
SQLite Begin Concurrent
Usually, SQLite allows at most one writer to proceed concurrently. The BEGIN CONCURRENT enhancement allows multiple writers to process write transactions simultanously if the database is in "wal" or "wal2" mode, although the system still serializes COMMIT commands.
158
Data diffs: Algorithms for explaining what changed in a dataset (2022)
tl;dr: part 1 explains what an explanation algorithm is, and part 2 describes an open source SQL data differ.
159
160
WordPress Core to start using SQLite
This post is an update to the proposal for officially support SQLite in WordPress. The initial implementation was included in the Performance Lab plugin and then released as a stand-alone plugin. T…
161
Visualizing every job in the world
A visualization of every single job in the world (or at least in the EU)
162
Teach your LLM to answer with facts, not fiction
Learn how to train LLM for accurate answers with VectorSQL which allows you to execute finely-grained vector searches to target and retrieve the required information
163
Why SQLite does not use Git (2018)
The document you seek is not available. Please consider one of the links below or use the Search feature on the right-hand side of the menu bar above.
164
JPEG XL: How it started, how it’s going
Significance of Apple's support for JPEG XL and what this means for the widespread adoption of this next-generation image compression format.
165
Full-text search engine with PostgreSQL (part 2): Postgres vs. Elasticsearch
Using PostgreSQL as a full-text search engine is tempting because it requires less infrastructure. But is its set of search-related features enough to compete with the Lucene based alternatives?
166
Citus 12: Schema-based sharding for PostgreSQL
Release blog post for the 12.0 release of the Citus extension to Postgres. Dive into new Citus 12 features such as schema-based sharding for multi-tenant SaaS applications (and microservices) as well improvements in the Citus support for the Postgres 15 MERGE feature.
167
Statistical Arbitrage – An Easy Walkthrough
Statistical arbitrage (stat arb) is a pillar of quantitative trading that relies on mean reversion to predict the future returns of an asset. Mean reversion believes that if a stock has risen higher it’s more likely to revert in the short term which is the opposite of a momentum strategy that believes if a stock has…
168
Bad numbers in the “gzip beats BERT” paper?
The recent paper, “Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors by Jiang et al. [ link ] recently received a lot of attention on twitter.
169
PostgreSQL: No More Vacuum, No More Bloat
PostgreSQL, a powerful open-source object-relational database system, has been lauded for its robustness, functionality, and flexibility. However, it is not without its challenges – one of which is the notorious VACUUM process. However, the dawn of a new era is upon us with OrioleDB, a novel engine designed for…
170
There is no data engineering roadmap
Data Engineering is booming and many are looking for a cheat sheet to get started. Do you really need to learn 200 tools before you get your first job?
171
Understanding Kafka with Factorio (2019)
While playing Factorio the other day, I was struck by the many similarities with Apache Kafka.
172
Faster neural networks straight from JPEG (2018)
Uber AI Labs introduces a method for making neural networks that process images faster and more accurately by leveraging JPEG representations.
173
Creating a search engine with PostgreSQL
PostgreSQL provides the necessary building blocks for you to combine and create your own search engine for full-text search. Let's see how far we can take it.
174
20x faster than pgvector: HNSW index in Postgres with pg_embedding
We’re excited to announce the release of our pg_embedding extension for Postgres and LangChain! The new pg_embedding extension brings 20x the speed for 99% accuracy to graph-based approximate nearest neighbor search to your Postgres databases. While the pgvector extension with IVFFlat indexing has been a popular…
175
176
UCLA professor refuses to cover for Dan Ariely in issue of data provenance
In one of the more bizarre twists of the Dan Ariely saga, there is a huge issue of provenance in one of Dan Ariely’s studies. The study in question is Dan Ariely’s Ten Commandments study at JMR. The hugely impactful paper has over 3,000 citations but unfortunately failed to replicate in a large-scale systematic…
177
HTTP vs. WebSockets: Which one is the fastest for Postgres queries at the edge?
Faster is always better, especially when executing SQL queries. We recently introduced SQL-over-HTTP to our driver, which previously only supported WebSockets. Why did we do that? And which approach is faster? Our journey developing the WebSockets serverless driver We first developed the Neon driver to help developers…
178
179
ScyllaDB is Moving to a New Replication Algorithm: Tablets
How moving from Vnode-based replication to tablets helps dynamically distribute data across the cluster, ultimately increasing ScyllaDB’s elasticity.
180
We put a distributed database in the browser and made a game of it
TigerBeetle is a distributed financial transactions database, designed for mission critical safety. How do you test something so critical? Well, we could tell you (and we will), but why not show you?