Xwinter-120b / README.md
llmixer's picture
Update README.md
390acf0 verified
|
raw
history blame
No virus
1.44 kB
---
license: llama2
language:
- en
pipeline_tag: conversational
tags:
- Xwin
- WinterGoddess
- frankenmerge
- 120b
---
# Xwinter 120B
<img src="https://cdn-uploads.huggingface.co/production/uploads/65a6db055c58475cf9e6def1/gs7KMVn4__G_JuksiuJIm.png" width=600>
A Goliath-120b style frankenmerge of Xwin-LM-70b-v0.1 and WinterGoddess-1.4x-70b. Meant as a slight update to Goliath by using the successor model to Euryale.
There's a similar merge called [WinterGoliath-123b](https://huggingface.co/ChuckMcSneed/WinterGoliath-123b) by [@ChuckMcSneed](https://huggingface.co/ChuckMcSneed).
# Prompting Format
Vicuna and Alpaca.
# Merge process
The models used in the merge are [Xwin-LM-70b-v0.1](https://huggingface.co/Xwin-LM/Xwin-LM-70B-V0.1) and [WinterGoddess-1.4x-70b](https://huggingface.co/Sao10K/WinterGoddess-1.4x-70B-L2).
The layer mix:
```yaml
- range 0, 16
Xwin
- range 8, 24
WinterGoddess
- range 17, 32
Xwin
- range 25, 40
WinterGoddess
- range 33, 48
Xwin
- range 41, 56
WinterGoddess
- range 49, 64
Xwin
- range 57, 72
WinterGoddess
- range 65, 80
Xwin
```
# Acknowledgements
[@Xwin-LM](https://huggingface.co/Xwin-LM) For creating Xwin
[@Sao10K](https://huggingface.co/Sao10K) For creating WinterGoddess
[@alpindale](https://huggingface.co/alpindale) For creating the original Goliath
[@chargoddard](https://huggingface.co/chargoddard) For developing [mergekit](https://github.com/cg123/mergekit).