Update README.md
Browse files
README.md
CHANGED
@@ -15,14 +15,14 @@ language:
|
|
15 |
Pantheon Roleplay is a model that has been in the making for almost six months at this point, starting from a collection of personas but steadily having grown into a full-fledged roleplaying model that simultaneously features a smart assistant in the form of Aiva.
|
16 |
|
17 |
## Model details
|
18 |
-
This model features a diverse collection of datasets, totaling
|
19 |
|
20 |
- For general instructions I created GPT 4 and Claude Opus variations of the No-Robots dataset. I actually ended up not including NoRo itself as it made the model worse.
|
21 |
- For roleplay I have an extensive collection of GPT 4 and Claude Opus data, further augmented by the always popular LimaRP for the "human factor".
|
22 |
- The Pantheon Roleplay personas were made using Claude 1.3 data, further diversifying the outputs of this model.
|
23 |
- Aiva's persona is augmented with datasets featuring questions related to DM world building, Python coding and RSS summarization. (She summarizes my daily news every day!)
|
24 |
|
25 |
-
Roughly 25% of the
|
26 |
|
27 |
**TLDR;** Download. ChatML prompt format. Have fun! Leave feedback!
|
28 |
|
|
|
15 |
Pantheon Roleplay is a model that has been in the making for almost six months at this point, starting from a collection of personas but steadily having grown into a full-fledged roleplaying model that simultaneously features a smart assistant in the form of Aiva.
|
16 |
|
17 |
## Model details
|
18 |
+
This model features a diverse collection of datasets, totaling ~20 million tokens;
|
19 |
|
20 |
- For general instructions I created GPT 4 and Claude Opus variations of the No-Robots dataset. I actually ended up not including NoRo itself as it made the model worse.
|
21 |
- For roleplay I have an extensive collection of GPT 4 and Claude Opus data, further augmented by the always popular LimaRP for the "human factor".
|
22 |
- The Pantheon Roleplay personas were made using Claude 1.3 data, further diversifying the outputs of this model.
|
23 |
- Aiva's persona is augmented with datasets featuring questions related to DM world building, Python coding and RSS summarization. (She summarizes my daily news every day!)
|
24 |
|
25 |
+
Roughly 25% of the training data was instructional, with the rest being focused on roleplay. Each of these datasets was then carefully balanced to ensure diversity, removing less diverse examples if deemed necessary.
|
26 |
|
27 |
**TLDR;** Download. ChatML prompt format. Have fun! Leave feedback!
|
28 |
|