GitLab

Co-authored-by: Jacob Schein <jacobschein@Jacobs-MacBook-Pro-2.local>

Co-authored-by: Antoni Baum <antoni.baum@protonmail.com>

Co-authored-by: Swapnil Parekh <swapnilp@ibm.com>
Co-authored-by: Joe G <joseph.granados@h2o.ai>
Co-authored-by: Antoni Baum <antoni.baum@protonmail.com>

Co-authored-by: Robert Shaw <rshaw@neuralmagic.com>
Co-authored-by: ZX <zx@lbx.dev>

Co-authored-by: sang <sangcho@anyscale.com>

Co-authored-by: Antoni Baum <antoni.baum@protonmail.com>

Currently we need to call rotary embedding kernel for each LoRA, which makes it hard to serve multiple long context length LoRA. Add batched rotary embedding kernel and pipe it through.

It replaces the rotary embedding layer to the one that is aware of multiple cos-sin-cache per scaling factors.

Follow up of https://github.com/vllm-project/vllm/pull/3095/files

Co-authored-by: Antoni Baum <antoni.baum@protonmail.com>

Co-authored-by: Danny Guinther <dguinther@neuralmagic.com>

Co-authored-by: simon-mo <simon.mo@hey.com>

Co-authored-by: SangBin Cho <sangcho@sangcho-LT93GQWG9C.local>

[WIP][Core][Refactor] move vllm/model_executor/parallel_utils into vllm/distributed and vllm/device_communicators (#3950)