A Walk on the Wild Side

Building LoRA Models

code

ml

huggingface

nlp

python

pyspark

databricks

Since 2021 LoRA methods and derivatives (such as QLora, IA3, etc) have become more common. Fine-tuning a BERT or decoder model, even though much better than training from…

Dr Alun ap Rhisiart

Stratified train-test-split

code

julia

data cleaning

data engineering

I previously wrote some Julia code to do a train-test-split with the train dataset also split to train/validation. However, I wanted to improve on this by making a…

Dr Alun ap Rhisiart

Dealing with multi-class ML data

code

julia

ml

data cleaning

data engineering

For the purpose of child safeguarding, one thing I do is build NLP models that classify texts according to certain risk categories (eating disorders, suicide ideation…

Dr Alun ap Rhisiart

A Walk on the Wild Side

new

In my work, I use a variety of languages, principally SQL, Julia, Python, and R, and a variety of tools, such as Databricks, dbt, Prefect, Jupyter Lab, VS Code, RStudio, and…

Alun ap Rhisiart