| --- |
| license: cc0-1.0 |
| language: |
| - fi |
| pipeline_tag: text-classification |
| thumbnail: https://raw.githubusercontent.com/NatLibFi/FintoAI/main/ai.finto.fi/static/img/finto-ai-social.png |
| tags: |
| - glam |
| - lam |
| - subject indexing |
| - annif |
| --- |
| # FintoAI-data-KAUNO |
| This repository is for the Annif projects with the |
| [KAUNO vocabulary](https://finto.fi/kauno) |
| used at the [Finto AI service](https://ai.finto.fi/). |
| The current models were published there 2023-12-04. |
| The models have been trained on Python 3.8.10 with [Annif](https://annif.org) version 1.0.0. |
| See [projects.toml](projects.toml) for the configurations of the models. |
|
|
| This repository is mirrored from GitHub to the 🤗 Hugging Face Hub; |
| the GitHub repository does not contain the model files, but only the configurations for the projects and the DVC pipeline, see below. |
|
|
| The training corpora that are public can be found from the [Annif-corpora repository](https://github.com/NatLibFi/Annif-corpora/). |
|
|
| ## Models |
| The downloadable directories for projects and vocabularies are stored in the |
| [`/data`](https://huggingface.co/juhoinkinen/FintoAI-data-KAUNO/tree/main/data) |
| directory of this repository in the 🤗 Hugging Face Hub. |
|
|
| ## DVC pipeline |
| The projects are trained and evaluated using a [DVC (Data Version Control) pipeline](https://dvc.org/doc/start/data-management/data-pipelines) defined in [dvc.yaml](./dvc.yaml). |
|
|
| The pipeline takes care of |
|
|
| 1. installing Annif in a venv, |
| 2. loading the vocabulary, |
| 3. training the projects, |
| 4. evaluating the projects. |
|
|
| When the necessary vocabulary and training corpora are in place the pipeline can be run using the command |
|
|
| dvc repro |
| |
| For more information about using DVC with Annif projects see the [DVC exercise of Annif tutorial](https://github.com/NatLibFi/Annif-tutorial/blob/master/exercises/OPT_dvc.md). |
|
|
