llama.cpp
graph : support cacheless embeddings with FA and iSWA
#16528

Merged

graph : support cacheless embeddings with FA and iSWA #16528

ggerganov merged 3 commits into master from gg/cacheless-embd

ggerganov requested a review from

CISC 1 year ago

danbev approved these changes on 2025-10-12

JohannesGaessler requested a review from

JohannesGaessler 1 year ago

github-actions added Nvidia GPU

github-actions added ggml

JohannesGaessler force pushed from d8d99d1d to c3089252 1 year ago

graph : support cacheless embeddings with FA and iSWA

57345463

ggerganov force pushed from c3089252 to 57345463 1 year ago

cont : deduplicate mask creation

71a264c7

cont : fix name [no ci]

5c658319

ggerganov merged e38b7c6e into master 1 year ago

ggerganov deleted the gg/cacheless-embd branch 1 year ago

Reviewers

danbev

CISC

JohannesGaessler

Assignees

No one assigned

Labels

Nvidia GPU ggml

Milestone

No milestone