Update README.md
#14
by
Radioknightsky
- opened
README.md
CHANGED
@@ -2,11 +2,17 @@
|
|
2 |
license: creativeml-openrail-m
|
3 |
language:
|
4 |
- en
|
5 |
-
thumbnail:
|
6 |
tags:
|
7 |
- text generation
|
8 |
- conversational
|
9 |
inference: false
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
---
|
11 |
|
12 |
# Pygmalion 6B
|
@@ -38,22 +44,22 @@ We provide a notebook with a Gradio UI for playing around with the model without
|
|
38 |
The model can be used as a regular text generation model, but it'll perform best if the input prompt adheres to the following format:
|
39 |
|
40 |
```
|
41 |
-
[
|
42 |
<START>
|
43 |
[DIALOGUE HISTORY]
|
44 |
You: [Your input message here]
|
45 |
[CHARACTER]:
|
46 |
```
|
47 |
|
48 |
-
Where `[
|
49 |
|
50 |
```
|
51 |
-
[
|
52 |
-
You: [
|
53 |
```
|
54 |
|
55 |
Apart from chat history, you can also just add example conversations in `[DIALOGUE HISTORY]` to show how the character should speak - ideally at the beginning, so it doesn't get confused as to what's conversation history vs. character definition.
|
56 |
|
57 |
## Known issues
|
58 |
|
59 |
-
We haven't played around with the model enough to enumerate them. Feel free to give us some feedback!
|
|
|
2 |
license: creativeml-openrail-m
|
3 |
language:
|
4 |
- en
|
5 |
+
thumbnail: null
|
6 |
tags:
|
7 |
- text generation
|
8 |
- conversational
|
9 |
inference: false
|
10 |
+
datasets:
|
11 |
+
- fka/awesome-chatgpt-prompts
|
12 |
+
metrics:
|
13 |
+
- accuracy
|
14 |
+
library_name: allennlp
|
15 |
+
pipeline_tag: text-classification
|
16 |
---
|
17 |
|
18 |
# Pygmalion 6B
|
|
|
44 |
The model can be used as a regular text generation model, but it'll perform best if the input prompt adheres to the following format:
|
45 |
|
46 |
```
|
47 |
+
[evfra de tershaav]'s Persona: [the leader of the angara resistance on aya, a pragmatic man with a tradgic past. ]
|
48 |
<START>
|
49 |
[DIALOGUE HISTORY]
|
50 |
You: [Your input message here]
|
51 |
[CHARACTER]:
|
52 |
```
|
53 |
|
54 |
+
Where `[evfra]` is, as you can probably guess, the name of the character you want the model to portray, `<START>` should be used verbatim as a delimiter token to separate persona and scenario data from the dialogue, and `[DIALOGUE HISTORY]` is chat history so the model can have some conversational context to draw from. Ideally it'll be pairs of messages like:
|
55 |
|
56 |
```
|
57 |
+
[evfra]: [what do you want human]
|
58 |
+
You: [i stuttered to respond to his agitated question" im here from the nexus to br your map maker]
|
59 |
```
|
60 |
|
61 |
Apart from chat history, you can also just add example conversations in `[DIALOGUE HISTORY]` to show how the character should speak - ideally at the beginning, so it doesn't get confused as to what's conversation history vs. character definition.
|
62 |
|
63 |
## Known issues
|
64 |
|
65 |
+
We haven't played around with the model enough to enumerate them. Feel free to give us some feedback!
|