Ada Jiang

converted to a static site and got rid of the landing page. lowk had so many errors with paths and stuff but it works!! :))

I fixed a lot of frontend zoom in zoom out UI stuff, like large gaps and clunky fitting. I also added a music player so you can play music. Now I’m working on hosting it online

Basically I got all my image data from fishbase, cleaned it, and downloaded it into a JSON file. Then I focused more on the UX, and I added some floating jellyfish animations and tried to give it a scrapbook aesthetic. Now I want to work on a landing page and transition to the main page.

So far I’ve built a minimum viable product with a basic webpage that displays a fish image and the scientific + common name. I’ve also built out the backend by calling APIs like iNaturalist for fish images and I’ve downloaded data from rfishbase into a json. Now I’m working on downloading images from wikimedia commons into a json file.

I compiled it into an app.exe so it can be shipped. lol this took so long for no reason because it wouldn’t load

I decided to work on making it more experiencable so I added a data augmentation explorer alongside the previous linear regression visualizer thingy. Lowk ts was such a struggle because for some reason it wouldn’t deploy and i was lowk pylling my hair out but its ok guys we made it thru. I hosted it on huggingface

I turned custom data into dataloaders for training, worked on data augmentation and how that can increase the accuracy of my models by producing more datasets, I created train and test loop functions, and then trained and evaluated my baseline model which I’m now going to try and improve!

I just made this project lolol. I’m mostly following this tutorial: https://youtu.be/V_xro1bcAuA?si=u1-f8I-bVn8POAMF to learn PyTorch. So far I’ve learned basic PyTorch datasets + how to use them, built a basic linear regression model + created a linear regression visualizer for it, worked on classification + nonlinear models, learned about computer vision libraries + created model for FashionMNIST data, now im starting to work on a custom data classification model. So far, I’ve loaded in the new data! After this, I hope to make my models more experienceable (if thats a word) for reviewers!

I added a ms-paint themed art window that the user can draw on to reveal my art!!

I’m working on my first project! This is so exciting. I can’t wait to share more updates as I build.