jzhang38 / EasyContext

Memory optimization and training recipes to extrapolate language models' context length to 1 million tokens, with minimal hardware.
Apache License 2.0
650 stars 47 forks source link

feat: usp (unified sequence parallelism) #50

Closed feifeibear closed 2 months ago