29 lines
680 B
Markdown
29 lines
680 B
Markdown
|
---
|
||
|
|
license: apache-2.0
|
||
|
|
---
|
||
|
|
**Blue-Orchid-2x7b**
|
||
|
|
|
||
|
|
GGUF: https://huggingface.co/nakodanei/Blue-Orchid-2x7b_GGUF
|
||
|
|
|
||
|
|
Roleplaying focused MoE Mistral model.
|
||
|
|
|
||
|
|
One expert is a merge of mostly RP models, the other is a merge of mostly storywriting models. So it should be good at both. The base model is SanjiWatsuki/Kunoichi-DPO-v2-7B.
|
||
|
|
|
||
|
|
- Expert 1 is a merge of LimaRP, Limamono, Noromaid 0.4 DPO and good-robot.
|
||
|
|
- Expert 2 is a merge of Erebus, Holodeck, Dans-AdventurousWinds-Mk2, Opus, Ashhwriter and good-robot.
|
||
|
|
|
||
|
|
## Prompt template (LimaRP):
|
||
|
|
|
||
|
|
```
|
||
|
|
### Instruction:
|
||
|
|
{system prompt}
|
||
|
|
|
||
|
|
### Input:
|
||
|
|
User: {prompt}
|
||
|
|
|
||
|
|
### Response:
|
||
|
|
Character:
|
||
|
|
|
||
|
|
```
|
||
|
|
|
||
|
|
Alpaca prompt template should work fine too.
|