sampling : refactor init to use llama_sampling_params #3696
sampling : refactor init to use llama_sampling_params
cd1e9378
llama : combine repetition, frequency and presence penalties in 1 call
6e658765
examples : remove embd-input and gptneox-wip
84ed48b4
sampling : rename penalty params + reduce size of "prev" vector
b5265615
sampling : add llama_sampling_print helper
7e2b5fb1
sampling : hide prev behind API and apply #3661
56ba00b9
ggerganov
force pushed
to
56ba00b9
2 years ago
ggerganov
merged
d1031cf4
into master 2 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub