Quick snapshot about me~💫

• Focused on AI Research & Application
• Currently, I’m majoring in AI Engineering at Sepuluh Nopember Institute of Technology
• Right now I’m learning about Generation Model (Diffuser / Image Gen, LLMs, and Music too ૮₍ ´ ꒳ `₎ა ), and RVC
• I mostly code in Python, and sometimes JavaScript
• I’m usually listening to classical music 𝄞

I’ve been vibing with things like~✨

• Diffusers / Image Generation
• Music Generation Model (like DiT or any architecture)
• Retrieval-based Voice Conversion (RVC)
• Language Models
• Data Science or something like that...
• AI engineering under tight compute (LoRA/LoRA+/PEFT and other “please fit in VRAM” rituals)

I also have a soft spot for anime (アニメ) datasets — images, audio, text…
if it’s about anime, I will probably try it at least once (๑'ᵕ'๑)⸝*

On the side, I’m usually:
turning anime data into small multimodal prototypes, hunting for more efficient (and consistently functional) architectures,
reading biology/science stuff and going “wait… can ML do that?” 🧬,
and training while listening to classical / romance music 🎻

Occasionally, I touch grass too (ごくまれに) (￣▽￣)ゞ

• Generative Models

  • Frieren Diffusion LoRA -> In this project, I fine-tune a lightweight LoRA adapter for the Waifu Diffusion text-to-image model (hakurei/waifu-diffusion) using the CyberHarem Frieren dataset
  • Frieren Diffusion LoRA w/ Dreambooth -> In this project, I fine-tune waifu-diffusion model using DreamBooth with a lightweight LoRA adapter to learn Frieren’s identity

• LLMs

  • X-LoRA for Cross Lingual Task in Qwen2.5-0.5B -> Attempting to implement the X-LoRA architecture in a Bilingual (English-Indonesian) task. The datasets used were CendolCollectionv2 for Indonesian and OpenOrca for English
  • Qwen3 Astrophysics LoRA -> Finetuning Qwen3 - 1.7B using LoRA for astronomy QA

• Computer Vision

  • FGVC in Nakano Siblings Image Classification -> Implementation of Fine-Grained Visual Categorization (FGVC) on The Quintessential Quintuplets Images Dataset using TransFG is a project that trains a fine-grained image classifier to distinguish between the five visually similar Nakano sisters (Ichika, Nino, Miku, Yotsuba, and Itsuki) using The Quintessential Quintuplets Images dataset
  • Chibi Style Detection -> Chibi Style Detection is a computer-vision project that determines whether an input image looks “chibi” or “non-chibi.” It’s built with Anomalib’s PatchCore, a memory-bank anomaly detection approach that compares patch-level CNN features against a stored set of “normal” features to produce an anomaly score and a localization heatmap.

• Tools & Framework

  • MoeScraper -> Python toolkit/library to help retrieve and collect image data from anime image websites like danbooru, safebooru, and zerochan. It also supports searches based on desired tags and filters for NSFW images
  • Frameko -> Python toolkit to convert video (animation or else) into multiple frames and collect them into an image dataset. This tool is can be used to add more data or collect a dataset that will be used to train a model

• Web Application

  • Cross Lingual Waguri AI -> Waguri AI is a bilingual chatbot web app built as a demonstration of fine-tuning Qwen2.5-0.5B using Mixture of LoRA Experts (X-LoRA) for the English–Indonesian pair. This web app is built using Next.js, Typescript, and Tailwind CSS for the frontend and FastAPI for the backend

• Game Development
  • Quantum Lite Chess -> A Pygame-based chess game that adds a “quantum-lite” layer on top of standard chess rules: pieces can split into two destinations, the game can maintain multiple classical branches at once, and some events behave like probabilistic measurement/collapse

❄️ご訪問ありがとうございます。よい一日を！ (๑>؂•̀๑)