Skip to content

Commit 8425dde

Browse files
committed
Add prompt cache
1 parent 410ebc1 commit 8425dde

File tree

3 files changed

+6
-1
lines changed

3 files changed

+6
-1
lines changed

README.md

+2
Original file line numberDiff line numberDiff line change
@@ -97,6 +97,7 @@ You can now chat with the model:
9797
--threads 16 `
9898
--n-gpu-layers 32 `
9999
--reverse-prompt '[[USER_NAME]]:' `
100+
--prompt-cache "./cache/prompt" `
100101
--file "./vendor/llama.cpp/prompts/chat-with-vicuna-v1.txt" `
101102
--color `
102103
--interactive
@@ -141,6 +142,7 @@ To extend the context to 8k execute the following:
141142
--threads 16 `
142143
--n-gpu-layers 32 `
143144
--reverse-prompt '[[USER_NAME]]:' `
145+
--prompt-cache "./cache/prompt" `
144146
--file "./vendor/llama.cpp/prompts/chat-with-vicuna-v1.txt" `
145147
--color `
146148
--interactive

cache/.gitignore

+3
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
# Ignore everything in this directory except this file.
2+
*
3+
!.gitignore

vendor/llama.cpp

0 commit comments

Comments
 (0)