llama.cpp
server : use common_token_to_piece instead of common_detokenize
#11740

Merged

server : use common_token_to_piece instead of common_detokenize #11740

ngxson merged 2 commits into ggml-org:master from danbev:server-logprobs-11728

github-actions added examples

github-actions added server

ngxson commented on 2025-02-07

CyberShadow commented on 2025-02-08

danbev changed the title ~~server : use text_to_send instead of detokenized token~~ llama : add remove_space_prefix to llama_detokenize 309 days ago

CyberShadow approved these changes on 2025-02-10

danbev force pushed 309 days ago

danbev marked this pull request as ready for review 309 days ago

ggerganov commented on 2025-02-10

server : use common_token_to_piece instead of common_detokenize

b70fd3a0

danbev force pushed to b70fd3a0 309 days ago

ggerganov approved these changes on 2025-02-10

squash! server : use common_token_to_piece instead of common_detokenize

5deee0a0

danbev changed the title ~~llama : add remove_space_prefix to llama_detokenize~~ server : use common_token_to_piece instead of common_detokenize 308 days ago

ngxson approved these changes on 2025-02-11

ngxson merged a18f481f into master 308 days ago

danbev deleted the server-logprobs-11728 branch 125 days ago

Reviewers

ngxson

ggerganov

CyberShadow

Assignees

No one assigned

Labels

examples server

Milestone

No milestone

llama.cpp server : use common_token_to_piece instead of common_detokenize #11740 Merged

server : use common_token_to_piece instead of common_detokenize #11740

llama.cpp
server : use common_token_to_piece instead of common_detokenize
#11740

Merged