Initial changes for Recurrent state rollback for nemotron for cuda - #26623

Initial changes for Recurrent state rollback for nemotron for cuda#26623

ggerganov merged 17 commits into

ggml-org:masterggml-org/llama.cpp:masterfrom

lnigam:lnigam/recurrent-state-rollbacklnigam/llama.cpp:lnigam/recurrent-state-rollbackCopy head branch name to clipboard