llama.cpp
fix(context): draft model fit vs load inconsistency
#25056

Merged

fix(context): draft model fit vs load inconsistency #25056

ggerganov merged 3 commits into ggml-org:master from wadealexc:fix/draft-load-context

wadealexc requested a review 55 days ago

github-actions added server

ggerganov assigned

ggerganov 55 days ago

wadealexc force pushed from fedef144 to e9852363 54 days ago

wadealexc force pushed from e9852363 to 5765ccd3 53 days ago

wadealexc force pushed from 5765ccd3 to 3493b95c 53 days ago

wadealexc commented on 2026-06-29

wadealexc force pushed from 3493b95c to 8a028311 53 days ago

wadealexc force pushed from 8a028311 to 4ad6bdb4 51 days ago

wadealexc force pushed from 4ad6bdb4 to c257309e 49 days ago

wadealexc force pushed from c257309e to 0a0c16ec 46 days ago

fix: draft model fit vs load inconsistency

03732d5f

refactor(server): unify draft/mtp parameter initialization, model, an…

7fa470e0

wadealexc force pushed from 0a0c16ec to 7fa470e0 45 days ago

ggerganov approved these changes on 2026-07-07

fix: valign

8cdc5e14

ggerganov merged f5525f7e into master 45 days ago

Reviewers

ggerganov

Assignees

ggerganov

Labels

server

Milestone

No milestone