Praelexis AI (@praelexis) header
Praelexis AI (@praelexis) profile picture

We craft elegant, data-driven solutions for significant problems.

Stellenbosch, South Africa
Joined October 2012
Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)Photo by Praelexis AI (@praelexis)
One of our wonderful interns, Tristan, has written an overview of the 6 steps required for data collection automation through web scraping. If you'd like an overview, it's a great place to start: praelexis.com/the-6-steps-to… #blog #webscraping #programming
1
1
Large Language Models are Zero-Shot Reasoners -- buff.ly/3LEwwTj Incredible what language models are capable of with the right prompt: "increasing the accuracy on MultiArith from 17.7% to 78.7% and GSM8K from 10.4% to 40.7% with an off-the-shelf 175B para model"
Are you in legal tech? Do you have a document drafting solution, if not please help us discover your needs by completing this questionnaire? buff.ly/2BeIVPK
[1912.01865] StarGAN v2: Diverse Image Synthesis for Multiple Domains buff.ly/2VW0Cex
Quickstart — ananke-causal documentation buff.ly/2Ah8UVw
[2005.00743] Synthesizer: Rethinking Self-Attention in Transformer Models buff.ly/3c9ZkRV
Why graph databases are the right choice for many data-centric organizations buff.ly/2UPus3q
SP 800-53 Rev. 5 (Draft), Security and Privacy Controls for Info Systems and Organizations | CSRC buff.ly/3b1N5Xj
Why software engineering processes and tools don’t work for machine learning – Comet buff.ly/33Ks8Of
(58) Deep Learning State of the Art (2020) | MIT Deep Learning Series buff.ly/38GRspl
Interested in machine learning? Better learn PyTorch buff.ly/38ii6W0
Faster Neural Networks Straight from JPEG buff.ly/2LbAOnA
A Doomed Marriage of Machine Learning and Agile - KDnuggets buff.ly/2QYD4Dq
(58) GOTO 2019 • WebAssembly Beyond the Browser • Dan Callahan buff.ly/3axO2XM
New State of the Art AI Optimizer: Rectified Adam (RAdam). Improve your AI accuracy instantly versus Adam, and why it works. buff.ly/2z4RZSm
GitHub - ICLRandD/Blackstone: A spaCy pipeline and model for NLP on unstructured legal text. buff.ly/2nquMbd