Update README.md
Browse files
README.md
CHANGED
@@ -8,7 +8,7 @@ tags:
|
|
8 |
---
|
9 |
|
10 |
# Limamono-7B (Mistral) v0.3
|
11 |
-
This is a very early version of a strongly NSFW roleplaying model with _extremely
|
12 |
of almost entirely synthetic data of hopefully higher quality than typical human conversations. The
|
13 |
intended model audience is straight men and lesbians.
|
14 |
|
@@ -33,7 +33,7 @@ Other formats are not supported and may conflict with the special features of th
|
|
33 |
Limamono uses the [extended Alpaca format](https://github.com/tatsu-lab/stanford_alpaca),
|
34 |
with `### Input:` immediately preceding user inputs and `### Response:` immediately preceding
|
35 |
model outputs. It's been trained with a fixed "trigger phrase" similar to that of the original
|
36 |
-
Alpaca before the `### Instruction
|
37 |
|
38 |
```
|
39 |
Below is an instruction that describes background information for a story-rich chat. Write an appropriate response for both the instruction and user input.
|
@@ -60,6 +60,14 @@ More in detail, the instruction should _preferably_ include a moderately long (a
|
|
60 |
long) character description made in the style of the various fandom wikis on the Internet, with the
|
61 |
character name as the first line.
|
62 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
### Message length control
|
64 |
Inspired by the previously named "Roleplay" preset in SillyTavern, like with LimaRP it is possible to
|
65 |
append a length modifier to the instruction sequences in this way:
|
|
|
8 |
---
|
9 |
|
10 |
# Limamono-7B (Mistral) v0.3
|
11 |
+
This is a very early version of a strongly NSFW roleplaying model trained with _extremely limited_ amounts
|
12 |
of almost entirely synthetic data of hopefully higher quality than typical human conversations. The
|
13 |
intended model audience is straight men and lesbians.
|
14 |
|
|
|
33 |
Limamono uses the [extended Alpaca format](https://github.com/tatsu-lab/stanford_alpaca),
|
34 |
with `### Input:` immediately preceding user inputs and `### Response:` immediately preceding
|
35 |
model outputs. It's been trained with a fixed "trigger phrase" similar to that of the original
|
36 |
+
Alpaca, just before the `### Instruction:` sequence, following the template below.
|
37 |
|
38 |
```
|
39 |
Below is an instruction that describes background information for a story-rich chat. Write an appropriate response for both the instruction and user input.
|
|
|
60 |
long) character description made in the style of the various fandom wikis on the Internet, with the
|
61 |
character name as the first line.
|
62 |
|
63 |
+
You can refer to the included [Charlotte model card](https://huggingface.co/lemonilia/Limamono-Mistral-7B-v0.3/blob/main/Charlotte.png)
|
64 |
+
for an example on how characer descriptions can be formatted, but another option is also taking a
|
65 |
+
hint from the semiempty model output in text-generation-webui or other text completion UIs. Note that
|
66 |
+
you'll likely need to add the _trigger phrase_ for the model to generate text as intended from scratch.
|
67 |
+
|
68 |
+
It is **strongly advised** to include a Scenario for guiding the model behavior at the beginning of the
|
69 |
+
chat.
|
70 |
+
|
71 |
### Message length control
|
72 |
Inspired by the previously named "Roleplay" preset in SillyTavern, like with LimaRP it is possible to
|
73 |
append a length modifier to the instruction sequences in this way:
|