en.sedaily

Google Legend Jeff Dean Bets on AI That Runs Its Own Experiments

Jeff Dean left Google after 27 years to found Discovery Loop, a startup automating research loops to run thousands of ...
GitHub

Bigram vs LSTM vs Transformer on Tiny Shakespeare

Midterm: comparing a bigram, an LSTM and a small transformer on next-character prediction (Tiny Shakespeare) - itsameaditya/nlp-midterm-bigram-lstm-transformer ...
GitHub

Predicting NYC Subway Arrivals Better Than the MTA's Countdown Clocks

Deep learning models (LSTM, transformer) that predict NYC subway arrival times, benchmarked against the MTA's live countdown predictions. - Yared-betsega/NYC-subway-arrival-prediction ...

Transcribing 1 hour of audio in about 1 second. A thorough investigation of the Japanese speech recognition model 'Haas-ja-v1.0'

This Japanese transcription ecosystem is already taking shape. It's highly accurate, and it seems like a better choice than ...

1시간 분량의 음성을 약 1초 만에 받아쓰기. 일본어 음성 인식 'Haas-ja-v1.0' 철저 조사

이제 생태계가 만들어지고 있는 이 일본어 받아쓰기. 정확도도 높고, whisper 계열을 사용하는 것보다 더 나은 선택인 것 같네요.PR 코멘트채용 사이트 Green의 게재에 협력하고 있습니다.IT/Web 업계의 구인 및 ...