|
--- |
|
datasets: |
|
- flozi00/conversations |
|
language: |
|
- en |
|
- de |
|
--- |
|
|
|
## This project is sponsored by [  ](https://www.primeline-solutions.com/de/server/nach-einsatzzweck/gpu-rendering-hpc/) |
|
|
|
### Please Use V4 of this model instead |
|
|
|
# Model Card |
|
|
|
This model is an finetuned version for german instructions and conversations in style of Open Assistant tokens. "<|prompter|>" "<|endoftext|>" "<|assistant|>" |
|
|
|
The dataset used is deduplicated and cleaned, with no codes inside. The focus is on instruction following and conversational tasks. |
|
|
|
The model archictecture is based on Llama version 2 with 13B parameters, trained on 100% renewable energy powered hardware. |
|
|
|
This work is contributed by private research of [flozi00](https://huggingface.co/flozi00) |
|
|
|
|
|
Join discussions about german llm research, and plan larger training runs together: https://join.slack.com/t/slack-dtc7771/shared_invite/zt-219keplqu-hLwjm0xcFAOX7enERfBz0Q |
