PR #15077 llama : add --n-cpu-moe option

llama : add --n-cpu-moe option #15077

slaren merged 2 commits into master from sl/ncmoe

llama : add --n-cpu-moe option

260e0301

better way to avoid memory leaks in tensor_buft_overrides

fd2d1f9e

slaren force pushed to fd2d1f9e 179 days ago

slaren merged ec428b02 into master 179 days ago

slaren deleted the sl/ncmoe branch 179 days ago

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Milestone

No milestone