File size: 3,861 Bytes
10990cf ae27b01 10990cf ae27b01 |
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 |
---
language:
- en
- de
library_name: transformers
pipeline_tag: text-generation
license: apache-2.0
---
_Hermes + Leo = Hermeo_
# Hermeo-7B
A German-English language model merged from [DPOpenHermes-7B-v2](https://huggingface.co/openaccess-ai-collective/DPOpenHermes-7B-v2) and [leo-mistral-hessianai-7b-chat](https://huggingface.co/LeoLM/leo-mistral-hessianai-7b-chat) using [mergekit](https://github.com/cg123/mergekit).
Both base models are fine-tuned versions of [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1).
### Model details
- **Merged from:** [leo-mistral-hessianai-7b-chat](https://huggingface.co/LeoLM/leo-mistral-hessianai-7b-chat) and [DPOpenHermes-7B-v2](https://huggingface.co/openaccess-ai-collective/DPOpenHermes-7B-v2)
- **Model type:** Causal decoder-only transformer language model
- **Languages:** English and German
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0.html)
### Acknowledgements
- This model release is heavily inspired by [Weyaxi/OpenHermes-2.5-neural-chat-v3-2-Slerp](https://huggingface.co/Weyaxi/OpenHermes-2.5-neural-chat-v3-2-Slerp)
- Thanks to the authors of the base models: [Mistral](https://mistral.ai/), [LAION](https://laion.ai/), [HessianAI](https://hessian.ai/), [Open Access AI Collective](https://huggingface.co/openaccess-ai-collective), [@teknium](https://huggingface.co/teknium), [@bjoernp](https://huggingface.co/bjoernp)
- The [German evaluation datasets and scripts](https://github.com/bjoernpl/lm-evaluation-harness-de/tree/mmlu_de) from [@bjoernp](https://huggingface.co/bjoernp) were used.
- The computing resources from [DFKI's PEGASUS cluster](https://pegasus.dfki.de/) were used for the evaluation.
## Evaluation
The evaluation methdology of the [Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard) is followed.
### German benchmarks
| **German tasks:** | **MMLU-DE** | **Hellaswag-DE** | **ARC-DE** |
|-------------------------------|-------------|---------------|--------------|
| **Models / Few-shots:** | _(5 shots)_ | _(10 shots)_ | _(24 shots)_ |
| _7B parameters_ | | | |
| llama-2-7b | 0.400 | 0.513 | 0.381 |
| leo-hessianai-7b | 0.400 | 0.609 | 0.429 |
| bloom-6b4-clp-german | 0.274 | 0.550 | 0.351 |
| mistral-7b | **0.524** | 0.588 | 0.473 |
| leo-mistral-hessianai-7b | 0.481 | 0.663 | 0.485 |
| leo-mistral-hessianai-7b-chat | 0.458 | 0.617 | 0.465 |
| DPOpenHermes-7B-v2 | TBA | 0.603 | 0.515 |
| hermeo-7b (this model) | 0.511 | **0.668** | **0.528** |
| _13B parameters_ | | | |
| llama-2-13b | 0.469 | 0.581 | 0.468 |
| leo-hessianai-13b | **0.486** | **0.658** | **0.509** |
| _70B parameters_ | | | |
| llama-2-70b | 0.597 | 0.674 | 0.561 |
| leo-hessianai-70b | **0.653** | **0.721** | **0.600** |
### English benchmarks
TBA
## Prompting / Prompt Template
Prompt dialogue template (ChatML format):
```
"""
<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
"""
```
The model input can contain multiple conversation turns between user and assistant, e.g.
```
<|im_start|>user
{prompt 1}<|im_end|>
<|im_start|>assistant
{reply 1}<|im_end|>
<|im_start|>user
{prompt 2}<|im_end|>
<|im_start|>assistant
(...)
```
## License
[Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0.html)
|