Hello, I am Ayaka 👋🏻
Here are my featured projects:
Deep Learning and Natural Language Processing
- google/jax: Autograd and XLA: differentiate, vectorize, JIT to GPU/TPU and more
- ayaka14732/tpu-starter: Everything you want to know about Google Cloud TPU
- ayaka14732/jax-smi: JAX Synergistic Memory Inspector
- huggingface/transformers: Library for state-of-the-art Transformers
- ayaka14732/bart-base-jax: JAX implementation of the BART-base model from scratch
- ayaka14732/bart-base-cantonese: The pre-trained Cantonese BART model
- ayaka14732/TransCan: English-Cantonese translation model
- ayaka14732/bart-base-jax#twblg: Mandarin-Taiwanese Hokkien translation model
- ayaka14732/TrAVis: Visualise BERT attention in-browser
- ayaka14732/wakong: A rigorous and robust masking algorithm for model pre-training
- ayaka14732/bert-tokenizer-cantonese: BERT Tokenizer with vocabulary tailored for Cantonese
- ayaka14732/cantoseg: Cantonese segmentation tool
- ayaka14732/lihkg-scraper: A Python script for scraping the LIHKG forum
- ayaka14732/yue-cmn-classification-task: Cantonese/Mandarin Classification Task
- CanCLID/awesome-cantonese-nlp: A curated list of Cantonese NLP resources
Historical Linguistics
- nk2028/qieyun-js: A JavaScript library for the Tshet-Uinh phonological system
- nk2028/qieyun-python: A Python library for the Tshet-Uinh phonological system
- nk2028/purescript-qieyun: A PureScript library for the Tshet-Uinh phonological system
- nk2028/qieyun-examples: Sample scripts for the Qieyun Autoderiver
- nk2028/qieyun-autoderiver: Extrapolation tool for the Tshet-Uinh phonological system
- nk2028/tshet-uinh-flashcard: Flashcards for the Tshet-Uinh phonological system
- nk2028/rime-tupa: A Multi-platform keyboard for the Tshet-Uinh phonological system
- nk2028/pyanxchet: An automated and explanatory puonqtshet tool
Others
- BYVoid/OpenCC: Conversion between Traditional and Simplified Chinese
- StarCC0/starcc-py: Python implementation of StarCC, the next generation of OpenCC
- StarCC0/dict: Dictionaries for StarCC
- ayaka14732/FanWunMing: A serif Simplified-to-Traditional-Chinese font
- ayaka14732/FanWunHak: A sans-serif Simplified-to-Traditional-Chinese font
- rime/rime-cantonese: A Multi-platform Cantonese Keyboard
- CanCLID/jyutping.org: Introduction to Jyutping, the romanisation system for Cantonese
- CanCLID/jyutping.net: Jyutping Input Method Website
- CanCLID/inject-jyutping: A browser extension that adds Cantonese on Chinese characters
- CanCLID/rime-loengfan: The Cantonese version of the Liang Fen input method
- nushu-script/nushu-script.github.io: Online Nüshu Dictionary
- nushu-script/rime-nushu: A Multi-platform Nüshu Keyboard
- ayaka14732/awesome-rime: A curated list of Rime IME schemata
- nk2028/commonly-used-chinese-characters-and-words: Description same as title
- nk2028/putonghua-ipa-converter: A Putonghua to IPA converter
- ayaka14732/inject-xdi8: A browser extension that adds Shidinn on Chinese characters
- nk2028/ipa-practice: Online IPA Practice System
- ayaka14732/tibetan-practice: Online Tibetan Practice System
- ayaka14732/SNHakkaNews: List of daily news videos in Shin Neng Hakka
- ayaka14732/VunsioNewsList: List of daily news videos in Vunsio Hainanese
- ayaka14732/ByteVid: Say goodbye to long and boring videos
- ayaka14732/TinyPE-on-Win10: Minimal (268 bytes) 64-bit PE file on Windows
- ayaka14732/nya-calendar: Implementation of the Nya Calendar
- ayaka14732/cs224n-a4: A decent solution to CS224n Assignment #4 (Cherokee NMT)
- ayaka14732/hls-simple: HLS loop streaming server in Haskell
- ztjhz/graphviz-editor: Generates Graphviz image URL that can be used directly on any website
- ayaka14732/graphviz-server: Web API for rendering Graphviz images
- ayaka14732/basehangul-online: Online BaseHangul Encoder And Decoder
- ayaka14732/telegram-translate-bot: Telegram translation bot @suginatransbot
- ayaka14732/shieldy: Telegram anti-spam bot @shieldydustbot