|
--- |
|
license: apache-2.0 |
|
pipeline_tag: text-generation |
|
language: |
|
- da |
|
tags: |
|
- pretrained |
|
inference: |
|
parameters: |
|
temperature: 0.7 |
|
datasets: |
|
- DDSC/partial-danish-gigaword-no-twitter |
|
base_model: mistralai/Mistral-7B-v0.1 |
|
--- |
|
|
|
# Model Card for Munin 7B Alpha |
|
|
|
The Munin 7B Alpha Large Language Model (LLM) is a pretrained generative text model with 7 billion parameters, based on [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1). |
|
|
|
It has been trained on [Danish Gigaword](https://gigaword.dk/) using [continual pretraining](https://doi.org/10.48550/arXiv.2308.04014). |
|
|
|
For full details of this model please read our [release blog post](https://foundationmodels.dk/blog/2024/01/11/releasing-munin-7b-alpha---a-danish-llm/). |
|
The code-base can be found on [our Git repo](https://github.com/centre-for-humanities-computing/danish-foundation-models). |
|
|
|
**Note:** This model is an Alpha model. We don't recommend using this model in production. If you do use the model, please let us know. |
|
|
|
## Notice |
|
|
|
Munin 7B Alpha is, like Mistral 7B, a pretrained base model and therefore does not have any moderation mechanisms. |
|
|
|
|
|
## Development |
|
|
|
The model is developed by the [Danish Foundation Models Team](https://foundationmodels.dk) |
|
|
|
## With Support From |
|
|
|
- [Danish e-infrastructure Consortium](https://www.deic.dk/) |
|
- [Acquisition and Logistics Organisation at the Danish Ministry of Defence](https://www.fmi.dk/) |
|
- Danish Ministry of Higher Education and Science under [the Digital Security, Trust |
|
and Data Ethics performance contract](https://bedreinnovation.dk/) |
