llama.cpp
server: avoid unnecessary checkpoint invalidation for recurrent / hybrid models
#24035

Open

server: avoid unnecessary checkpoint invalidation for recurrent / hybrid models #24035

Regrad wants to merge 1 commit into ggml-org:master from Regrad:fix/qwen-hybrid-checkpoint-reuse

server: improve checkpoint reuse heuristics for recurrent/hybrid models

fc6a7e0d

Regrad requested a review 8 days ago

github-actions added examples

github-actions added server

Regrad closed this 8 days ago

Regrad reopened this 8 days ago

Regrad marked this pull request as draft 6 days ago

Reviewers

No reviews

Assignees

No one assigned

Labels

examples server

Milestone

No milestone