Simon GsponerAbsurd or acceptable? Flask dev server for prod trafficA silly experiment with some insightful twistsJul 29Jul 29
Simon GsponerThe Token: Key Aspects of Large Language Models, Part 1Disclaimer: This article belongs to my personal blog. The viewpoints expressed herein are exclusively mine and do not reflect the…May 15May 15
Simon GsponerA comprehensive guide II: Finetuning a BERT LLM with LoRA and make it pipeline compatibleIn the previous gudnput text way larger than the…Feb 201Feb 201
Simon GsponerA comprehensive guide: using a BERT LLM on texts exceeding the maximum input sizeAlthough powerful, language models from the BERT family can only process about 1 page of text data at once. Nevertheless, applying “small”…Feb 10Feb 10
Simon GsponerI’ve beaten >80% of data scientists without data science, here’s howRecently, I have participated in a Kaggle competition aiming at predicting sales data of the Ecuadorian supermarket chain ‘Favorita’…Dec 4, 20231Dec 4, 20231
Simon GsponerCOVID-19 Meets Switzerland: Panic-buying and Lakeside LoafingCan behavioural decision-making tell us something about why some people hoard toilet paper, while others enjoy the lovely weather outdoors?Mar 22, 20201Mar 22, 20201