Philip Zmushko@philip-zmushko · 1 worksResearches efficient distributed training methods for large language models.