llama.cpp
Command line switch to use F16 for memory_k and memory_v (refactor of #154)
#294
Merged
Merged
Command line switch to use F16 for memory_k and memory_v (refactor of #154) #294
ggerganov
merged
2 commits into
ggml-org:master
from
Green-Sky:f16_memory_cli
Use F16 for memory_k and memory_v
640b5602
add command line switch to use f16 instead of f32 for memory k+v
31edd6fa
ggerganov
approved these changes
on 2023-03-19
Login to write a write a comment.
Login via GitHub