Sieun Park – Medium

Sieun Park

A late review of OpenAI’s “Training Verifiers to Solve Math Word Problems”

Solving math problems with GPT. Verifiers, GSM-8K dataset, mathematical reasoning

6 min readMar 14, 2024

--

A late review of OpenAI’s “Training Verifiers to Solve Math Word Problems”

--

Sieun Park
in
CodeX

Sentencepiece: A simple and language independent subword tokenizer and detokenizer for neural text…

The tokenization method for Alpaca, LLaMA, T5, XLNet

9 min readMay 19, 2023

--

1

Sentencepiece: A simple and language independent subword tokenizer and detokenizer for neural text…

--

1

Sieun Park
in
CodeX

An Introduction to Multiprocessing Using Python

Is multithreading or multiprocessing faster than another?

10 min readApr 16, 2023

--

An Introduction to Multiprocessing Using Python

--

Sieun Park

Subword Regularization

Prerequisite for understanding the commonly used `SentencePiece` tokenizer. Improve your tokenization using subword regularization!

5 min readMar 5, 2023

--

Subword Regularization

--

Sieun Park

Details of Faster, Mask, Cascade R-CNN 🔥

WHY should you care? Faster R-CNN is an actively cited benchmark for comparing object detection performance of modern network…

19 min readJun 7, 2022

--

Details of Faster, Mask, Cascade R-CNN 🔥

--

Sieun Park
in
CodeX

ConvMixer: Patches Are All You Need? Overview and thoughts 🤷

CNNs don’t always have to progressively decrease resolution. A revolutionary idea that might shape the next-gen architectures for Computer…

8 min readNov 1, 2021

--

1

ConvMixer: Patches Are All You Need? Overview and thoughts 🤷

--

1

Sieun Park
in
CodeX

7 Different Convolutions for designing CNNs that will Level-up your Computer Vision project

In-depth review on Basic, Transposed, Dilated, Separable, Depthwise, and Pointwise convolutions and their applications.

8 min readOct 29, 2021

--

7 Different Convolutions for designing CNNs that will Level-up your Computer Vision project

--

Sieun Park
in
CodeX

R-Drop, a simple trick to improve DropOut

The method is simple, but results are significant. R-Drop achieves SOTA on NMT tasks even with the VANILLA Transformer architecture!

8 min readOct 27, 2021

--

1

R-Drop, a simple trick to improve DropOut

--

1

Sieun Park
in
CodeX

Making VGG-style convnets great again with RepVGG

VGGs are great! While they suck at training time, they are fast, flexible, and memory-efficient compared to modern CNNs. Let’s improve them.

8 min readOct 26, 2021

--

Making VGG-style convnets great again with RepVGG

--

Sieun Park
in
CodeX

Implementing R-CNN object detection on VOC2012 with PyTorch

How do we implement object detection from scratch? Let’s start from the basic R-CNN.

12 min readOct 22, 2021

--

2

Implementing R-CNN object detection on VOC2012 with PyTorch

--

2

Sieun Park

Sieun Park

Who knew AI could be so hot? I am an 18 years old Korean student. I started working in ML at 16, but that's it. LinkedIn: https://bit.ly/2VTkth7 😀

Following

Help
Status
About
Careers
Blog
Privacy
Terms
Text to speech
Teams