license: apache-2.0
language:
- de
pipeline_tag: text-generation
tags:
- german
- deutsch
- simplification
- vereinfachung
Model Card for Model ID
We fine-tuned the LeoLM/leo-mistral-hessianai-7b with a set of ca. 2600 newspaper articles which have been simplified by the Austrian Press Agency. Our aim was to have a model which can simplify German-language text. This model has been trained with the completition-only configuration.
Model Details
Model Description
- Developed by: Members of the Public Interest AI research group, HIIG Berlin
- Model type: simplification model, text generation
- Language(s) (NLP): German
- License: Apache 2.0
- Finetuned from model: jphme/em_german_leo_mistral
Model Sources
- Repository: https://github.com/fhewett/simba
- Project website: https://publicinterest.ai/tool/simba
Uses
Direct Use
This model works best for simplifying German-language newspaper articles (news items, not commentaries or editorials). It may work for other types of texts.
Downstream Use
We have fine-tuned using only newspaper articles. We have not yet performed extensive out-of-domain testing, but believe that the model's capabilities could be improved by fine-tuning on more diverse data. Contact us if you have a dataset which you think could work (parallel texts, German standard & German simplified).
Bias, Risks, and Limitations
As with most text generation models, the model sometimes produces information that is incorrect.
Recommendations
Please check manually that your output text corresponds to the input text, as factual inconsistencies may have arisen.
How to Get Started with the Model
Use the code below to get started with the model.
[More Information Needed]
Training Details
Training Data
A sample of the data used to train our model can be found here.
Training Hyperparameters
- Training regime: [More Information Needed]
Evaluation
Summary
For now, we have manually checked the performance of our model on a small sample of texts. Whilst it seems to produce good summaries of all texts, it only seems to simplify newspaper articles (i.e. similar to our training data). We have not yet applied any large-scale metrics based evaluation.
Model Card Contact
simba -at- hiig.de