65 lines
1.7 KiB
Markdown
65 lines
1.7 KiB
Markdown
---
|
|
base_model: []
|
|
tags:
|
|
- mergekit
|
|
- merge
|
|
- mistral
|
|
- german
|
|
- deutsch
|
|
- english
|
|
- roleplay
|
|
- chatml
|
|
language:
|
|
- de
|
|
- en
|
|
---
|
|
# merge
|
|
|
|
This is a experimental merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
|
|
|
## Merge Details
|
|
### Merge Method
|
|
|
|
This model was merged using the SLERP merge method.
|
|
|
|
### Models Merged
|
|
|
|
The following models were included in the merge:
|
|
* [SanjiWatsuki/Kunoichi-DPO-v2-7B](https://huggingface.co/SanjiWatsuki/Kunoichi-DPO-v2-7B)
|
|
* [DiscoResearch/DiscoLM_German_7b_v1](https://huggingface.co/DiscoResearch/DiscoLM_German_7b_v1)
|
|
|
|
#### Why this two models?
|
|
|
|
DiscoLM German 7B is is up to this date (01/21/2024) by far the best German model and makes far fewer grammatical errors and his German generally sounds good. But it is finetuned on Mistral V0.2 or even V0.1.
|
|
|
|
Kunoichi DPO v2 7B is a model that is already solid on german, but makes some more grammar errors. This model is trained especially for roleplay.
|
|
|
|
The ulterior motive was now combining this two models to get a even better German model, especially for German roleplay. A short testing showed already good results.
|
|
|
|
|
|
The last two AI responses above were 100% correct.
|
|
|
|
|
|
### Configuration
|
|
|
|
The following YAML configuration was used to produce this model:
|
|
|
|
```yaml
|
|
|
|
slices:
|
|
- sources:
|
|
- model: SanjiWatsuki/Kunoichi-DPO-v2-7B
|
|
layer_range: [0, 32]
|
|
- model: DiscoResearch/DiscoLM_German_7b_v1
|
|
layer_range: [0, 32]
|
|
merge_method: slerp
|
|
base_model: SanjiWatsuki/Kunoichi-DPO-v2-7B
|
|
parameters:
|
|
t:
|
|
- value: [0.5, 0.9]
|
|
dtype: bfloat16
|
|
|
|
|
|
```
|
|
|
|
This settings are from the model [oshizo/japanese-e5-mistral-7b_slerp](https://huggingface.co/oshizo/japanese-e5-mistral-7b_slerp). |