README.md

---
library_name: transformers
tags:
- deutsch
- german
- seedbox
- mistral
- mixtral
- multilingual
license: apache-2.0
language:
- de
- en
pipeline_tag: text-generation
---

![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/645ded34a45b4182d7f5c385/9QywLGTbRrHYSq-m6fQmJ.jpeg)


# KafkaLM-8x7b-German-V0.1

**KafkaLM 8x7b** is a MoE model based on [Mistral AI´s Mixtral 8x7b](https://mistral.ai/news/mixtral-of-experts/) which was finetuned on an ensemble of popular high-quality open-source instruction sets (translated from English to German). 

KafkaLM 8x7b is a [Seedbox](https://huggingface.co/seedboxai) project trained by [Dennis Dickmann](https://huggingface.co/doubledsbv).

**Why Kafka?** 
The models are proficient, yet creative, have some tendencies to linguistically push boundaries 😊


## Model Details

The purpose of releasing the **KafkaLM series** is to contribute to the German AI community with a set of fine-tuned LLMs that are easy to use in everyday applications across a variety of tasks.

The main goal was to provide LLMs proficient in German, especially to be used in German-speaking business contexts where English alone is not sufficient.

### DPO

The model has been aligned with a german and modified version of the ultra feedback dataset from huggingface. 

### Dataset

I used a 8k filtered version of the following [seedboxai/multitask_german_examples_32k](https://huggingface.co/datasets/seedboxai/multitask_german_examples_32k)



### Inference

Getting started with the model is straightforward

```python
import transformers
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "seedboxai/KafkaLM-Mixtral-8x7B-V0.2"

model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.bfloat16)
tokenizer = transformers.AutoTokenizer.from_pretrained(model_id)

pipeline = transformers.pipeline(
    model=model, tokenizer=tokenizer,
    return_full_text=True,  
    task='text-generation',
    device="cuda",
)

messages = [
    {"role": "system", "content": "Du bist ein hilfreicher KI-Assistent."},
    {"role": "user", "content": "Wer ist eigentlich dieser Kafka?"},
]

prompt = pipeline.tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)

terminators = [
    pipeline.tokenizer.eos_token_id,
    pipeline.tokenizer.convert_tokens_to_ids("</s>")
]

outputs = pipeline(
    prompt,
    max_new_tokens=max_new_tokens,
    eos_token_id=terminators,
    do_sample=True,
    temperature=0.7,
    top_p=0.9,
)

print(outputs[0]["generated_text"][len(prompt):])

```

## Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model.