jukofyork commited on
Commit
783b7b7
1 Parent(s): 6dca807

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +30 -3
README.md CHANGED
@@ -1,3 +1,30 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ ---
4
+
5
+ **IMPORTANT**: Make sure you have the latest version of [llama.cpp](https://github.com/ggerganov/llama.cpp) to use these:
6
+
7
+ - The [PR that allows multiple control vectors to be loaded](https://github.com/ggerganov/llama.cpp/commit/97877eb10bd8e7f8023420b5b5300bcbdadd62dc) (without zero-padding) just got merged today (27/06/24).
8
+ - Older versions of `llama.cpp` will just ***silently*** load the first control vector and none of the others!!!
9
+
10
+ To use these control vectors effectively you will need to use the "`--control-vector-scaled`" option like this:
11
+
12
+ ```sh
13
+ llama-cli --model <model name>.gguf --control-vector-scaled <model name>-dark.gguf 1.0 --control-vector-scaled <model name>-positive.gguf -1.0 --control-vector-scaled <model name>-chaos.gguf 1.0 --control-vector-scaled <model name>-law.gguf -1.0 [the rest of your arguments...]
14
+ ```
15
+ or:
16
+ ```sh
17
+ llama-cli --model <model name>.gguf --control-vector-scaled <model name>-dark.gguf 0.5 --control-vector-scaled <model name>-positive.gguf -0.5 --control-vector-scaled <model name>-chaos.gguf 0.5 --control-vector-scaled <model name>-law.gguf -0.5 [the rest of your arguments...]
18
+ ```
19
+ or:
20
+ ```sh
21
+ llama-cli --model <model name>.gguf --control-vector-scaled <model name>-dark.gguf 1.0 --control-vector-scaled <model name>-positive.gguf -0.5 --control-vector-scaled <model name>-chaos.gguf 1.0 --control-vector-scaled <model name>-law.gguf -0.5 [the rest of your arguments...]
22
+ ```
23
+
24
+ **NOTE:**
25
+
26
+ - Use ***negative scale factors*** for traits you ***DON'T*** want and ***positive scale factors*** for traits you ***DO*** want.
27
+ - The "positive" and "law" traits likely need to use a smaller magnitude of scale factor like `-0.5` or `-0.75` (unless you like ***really*** Grim stories! :D).
28
+ - The "dark" and "chaos" traits can likely use larger magnitude scale factors like `1.5` or `2.0` before the model starts to repeat gibberish.
29
+ - You can use some or all of the 4 control vector files, or alternatively set the scale factor to `0.0` for any traits you don't want to use.
30
+ - You can use the same "`--control-vector-scaled`" command line arguments for "`llama-server`" as in the above "`llama-cli`" examples.