Data Science

[Notes] Gradient Checkpointing with BERT

A brief analysis of huggingface’s implementation

[Paper] Adafactor: Adaptive Learning Rates with Sublinear Memory Cost

Essential for fine-tuning T5…

Generating Synthetic Tabular Data Using GAN

A case study: detecting credit fraud

[Paper] Are We Really Making Much Progress?

A Worrying Analysis of Recent Neural Recommendation…

Automatic Testing Your SQLite Database with Great Expectations

A great tool for eliminating pipeline…

[Paper] Please Stop Permuting Features

Permutation importance can be very misleading

Create a Customized Text Annotation Tool in Two Days — Part 2

Building a React Front-end Interface

Create a Customized Text Annotation Tool in Two Days — Part 1

Building a Back-end API Server with…

[Notes] “Statistical Inference Enables Bad Science; Statistical Thinking Enables Good Science”

UMAP on RAPIDS (15x Speedup)

Exploring the new GPU acceleration library from NVIDIA