File size: 4,954 Bytes
87b6654 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 |
---
tags:
- merge
- mergekit
- lazymergekit
- DiscoResearch/DiscoLM_German_7b_v1
- DRXD1000/Phoenix
- VAGOsolutions/SauerkrautLM-7b-v1-mistral
- malteos/hermeo-7b
base_model:
- DiscoResearch/DiscoLM_German_7b_v1
- DRXD1000/Phoenix
- VAGOsolutions/SauerkrautLM-7b-v1-mistral
- malteos/hermeo-7b
license: apache-2.0
language:
- de
- en
---
# Wiedervereinigung-7b-dpo
This is a dpo aligned merge of our favourite german models, scoring 7.11 on the mt-bench-de average.
Since the original models based on mistral - three of them on the brilliant german LeoLM/leo-mistral-hessianai-7b - they are reunited in this merged model.
Therefore the name, no nationalist ideas involved :-).
To improve result quality they are dpo-trained with a german translation of slimorca dpo using hermeo-7B for reject results.
If you are gpu-poor like me you can now use [LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory) to train with german datasets.
Kudos to the authors of the original models at [DiscoResearch](https://huggingface.co/DiscoResearch) and [VAGOsolutions](https://huggingface.co/VAGOsolutions), [Malte Ostendorff](https://huggingface.co/malteos)
and [Matthias Uhlig](https://huggingface.co/DRXD1000). We are your fan club.
This model was brought to you and the nvidia bill was paid by [Mayflower GmbH](https://mayflower.de/).
## Benchmark results: mt-bench-de
Is the merged model alone already good? Well, of course. But it is even better with the help of some dpo tuning.
```json
{
"first_turn": 7.3,
"second_turn": 6.925,
"categories": {
"writing": 8.425,
"roleplay": 8.6,
"reasoning": 5.4,
"math": 4.35,
"coding": 4.3,
"extraction": 7.975,
"stem": 8.5,
"humanities": 9.35
},
"average": 7.1125
}
```
## Other Versions
A big thank you to [LoneStriker](https://huggingface.co/LoneStriker) for the quantized models.
| Name | Quant method | Bits |
| ---- | ---- | ---- |
[Wiedervereinigung-7b-dpo](https://huggingface.co/mayflowergmbh/Wiedervereinigung-7b-dpo)| Unquantized | 16 |
[Wiedervereinigung-7b-dpo-GPTQ](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-GPTQ)| GPTQ | 4 |
[Wiedervereinigung-7b-dpo-AWQ](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-AWQ)| AWQ | 4 |
[Wiedervereinigung-7b-dpo-GGUF](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-GGUF)| GGUF | 3-8 |
[Wiedervereinigung-7b-dpo-8.0bpw-h8-exl2](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-8.0bpw-h8-exl2)| EXL2 | 8 |
[Wiedervereinigung-7b-dpo-6.0bpw-h6-exl2](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-6.0bpw-h6-exl2)| EXL2 | 6 |
[Wiedervereinigung-7b-dpo-5.0bpw-h6-exl2](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-5.0bpw-h6-exl2)| EXL2 | 5 |
[Wiedervereinigung-7b-dpo-4.0bpw-h6-exl2](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-4.0bpw-h6-exl2)| EXL2 | 4 |
[Wiedervereinigung-7b-dpo-3.0bpw-h6-exl2](https://huggingface.co/LoneStriker/Wiedervereinigung-7b-dpo-3.0bpw-h6-exl2)| EXL2 | 3 |
Wiedervereinigung-7b is a [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing) merge of:
* [DiscoResearch/DiscoLM_German_7b_v1](https://huggingface.co/DiscoResearch/DiscoLM_German_7b_v1)
* [DRXD1000/Phoenix](https://huggingface.co/DRXD1000/Phoenix)
* [VAGOsolutions/SauerkrautLM-7b-v1-mistral](https://huggingface.co/VAGOsolutions/SauerkrautLM-7b-v1-mistral)
* [malteos/hermeo-7b](https://huggingface.co/malteos/hermeo-7b)
## 🧩 Configuration
```yaml
models:
- model: LeoLM/leo-mistral-hessianai-7b
# No parameters necessary for base model
- model: DiscoResearch/DiscoLM_German_7b_v1
parameters:
density: 0.6
weight: 0.25
- model: DRXD1000/Phoenix
parameters:
density: 0.6
weight: 0.25
- model: VAGOsolutions/SauerkrautLM-7b-v1-mistral
parameters:
density: 0.6
weight: 0.25
- model: malteos/hermeo-7b
parameters:
density: 0.6
weight: 0.25
merge_method: dare_ties
base_model: LeoLM/leo-mistral-hessianai-7b
parameters:
int8_mask: true
dtype: bfloat16
```
## 💻 Usage
```python
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "mayflowergmbh/Wiedervereinigung-7b-dpo"
messages = [{"role": "user", "content": "Was ist ein deutsches Large Language Model?"}]
tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
"text-generation",
model=model,
torch_dtype=torch.float16,
device_map="auto",
)
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
``` |