Skip to content

CUDA-graph-compatible releasing and resuming KV cache and model weight memory#2630

Merged
merrymercy merged 148 commits intosgl-project:mainfrom fzyzcjy:feat/memory_saverJan 13, 2025

Commits

Commits on Dec 26, 2024

Commits on Dec 27, 2024

Commits on Dec 28, 2024

Commits on Dec 29, 2024

Commits on Dec 30, 2024

Commits on Dec 31, 2024

Commits on Jan 13, 2025