DavidAU commited on
Commit
b6ba998
1 Parent(s): 3f2d934

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +31 -0
README.md CHANGED
@@ -36,5 +36,36 @@ pipeline_tag: text-generation
36
 
37
  Two versions of Command-R-01 35B using TWO new DARK HORROR Neo Imatrix datasets.
38
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
39
  Quants uploading... Model card with examples pending updates.
40
 
 
36
 
37
  Two versions of Command-R-01 35B using TWO new DARK HORROR Neo Imatrix datasets.
38
 
39
+ Files have "V1" or "V2" in the filename.
40
+
41
+ The DARK HORROR NEO Imatrix datasets does the following:
42
+
43
+ - add a "coating of black paint" to any "Horror" prompt generation.
44
+ - add a "dark tint" to any other creative prompt.
45
+ - increase the intensity of a scene, story, or roleplay interaction.
46
+ - in some cases increase instruction following of the model.
47
+ - bring a sense of impending "horror"
48
+
49
+ The HORROR datasets were created using Grand Horror 16B using 90+ different horror prompts for generation.
50
+
51
+ This created incredibility intense horror stories in compact form - perfect for an Imatrix dataset.
52
+
53
+ These datasets have been tested on Command-R, Llama 3, Llama 3.1, Mistral Nemo and others.
54
+
55
+ In all cases this dataset moves the model into "horror territory" so to speak.
56
+
57
+ I want to be clear this process is not the same as a "fine tune", it is lighter than that.
58
+
59
+ And if you are looking for intense, visceral and NSFW horror then the "Grand Horror model" (and it's brothers and sisters)
60
+ is the way to go.
61
+
62
+ But if you want a little more horror, a little more darkness in "Command-R" then this may be your model.
63
+
64
+ Two versions of the DARK HORROR dataset were used for each quant.
65
+
66
+ I suggest downloading both (of the same quant) and test in your use cases.
67
+
68
+ - More updates coming -
69
+
70
  Quants uploading... Model card with examples pending updates.
71