Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,34 @@
|
|
| 1 |
---
|
| 2 |
license: cc-by-nc-4.0
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: cc-by-nc-4.0
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
I'm happy with where this model currently is so I am releasing the 7b for testing. If I get good feedback I'll be scaling up to larger models!
|
| 6 |
+
Thank you!
|
| 7 |
+
|
| 8 |
+
Datasets Used:
|
| 9 |
+
|
| 10 |
+
* Dolphin
|
| 11 |
+
* Ultrachat
|
| 12 |
+
* Capybara
|
| 13 |
+
* Augmental
|
| 14 |
+
* ToxicQA
|
| 15 |
+
* Magiccoder-Evol-Instruct-110k
|
| 16 |
+
* Yahoo Answers
|
| 17 |
+
* OpenOrca
|
| 18 |
+
* Airoboros 3.1
|
| 19 |
+
* grimulkan/physical-reasoning and theory-of-mind
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
## Prompt Format: Chat ( The default Ooba template and Silly Tavern Template )
|
| 23 |
+
```
|
| 24 |
+
{System Prompt}
|
| 25 |
+
|
| 26 |
+
Username: {Input}
|
| 27 |
+
BotName: {Response}
|
| 28 |
+
Username: {Input}
|
| 29 |
+
BotName: {Response}
|
| 30 |
+
|
| 31 |
+
```
|
| 32 |
+
|
| 33 |
+
## Recommended Silly Tavern Preset -> (Temp: 1.25, MinP: 0.1, RepPen: 1.03)
|
| 34 |
+
## Recommended Kobold Horde Preset -> MinP
|