froggeric commited on
Commit
fc35398
1 Parent(s): 6bc8bf9

changed example of custom matrix to be a good example for preserving multilingual capabilities, coding/reasoning skills, and a bit of decensoring

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -13,7 +13,7 @@ training data. Size also matters: the bigger the model (eg: 70b vs 13b) and the
13
  the bigger the source file needs to be to make an impact. Multiple input files can be combined if needed;
14
  for example:
15
  ```
16
- cat technical.txt multilingual.txt wiki.txt >custom.matrix
17
  ```
18
  Note on **context size** when generating the matrix: in general, a small context size such as 512 is recommended, and community
19
  tests have shown it usually performs than a larger one such as 4096. However, I would argue this is is highly dependent on the
 
13
  the bigger the source file needs to be to make an impact. Multiple input files can be combined if needed;
14
  for example:
15
  ```
16
+ cat multilingual.txt code.txt badwords_multilingual.txt > custom_multilingual.matrix
17
  ```
18
  Note on **context size** when generating the matrix: in general, a small context size such as 512 is recommended, and community
19
  tests have shown it usually performs than a larger one such as 4096. However, I would argue this is is highly dependent on the