Partition Mutable Buffer as Core ML State#5165

Merged

cccclai merged 7 commits intopytorch:mainfrom

YifanShenSZ:coreml-state

Sep 10, 2024

Commits on Sep 7, 2024

partition mutable buffer to coreml state
yifan_shen3
committed
delegate llama mutable buffer to coreml
yifan_shen3
committed

Commits on Sep 8, 2024

fix lint
yifan_shen3
committed
support embedding quantize
yifan_shen3
committed
try fix CI: 1. pin coremltools 8.0b2; 2. refrain from defaulting stateful llama until CI machine upgraded to MacOS 15
yifan_shen3
committed
address review comments: 1. add arg help info; 2. add mutable buffer partition log
yifan_shen3
committed

Commits on Sep 9, 2024

fix CI: executorch example model test env is using older transformers, that does not support numpy 2.0
yifan_shen3
committed