---
license: apache-2.0
tags:
- generated_from_trainer
base_model: crossroderick/occitan-distilgpt2-medium
model-index:
- name: occitan-distilgpt2-medium
  results: []
widget:
  - text: "Adieusatz! Ieu soi Micheu e"
    example_title: "Example 1"
  - text: "A mon vejaire"
    example_title: "Example 2"
  - text: "Mercés plan per la"
    example_title: "Example 3"
  - text: "Vòstres mots son mai"
    example_title: "Example 4"
---
# occitan-distilgpt2-medium

*Adieusatz brave monde!*

This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the Occitan language section of the Wikipedia dataset, and is part of a personal effort to study the Occitan language in an NLP environment. It achieves the following results on the evaluation set:

- Loss: 2.1597

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 12
- eval_batch_size: 12
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 48
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 2.2727        | 1.0   | 854  | 2.1682          |
| 2.2369        | 2.0   | 1708 | 2.1618          |
| 2.2173        | 3.0   | 2562 | 2.1597          |


### Framework versions

- Transformers 4.38.1
- Pytorch 2.0.1+cu117
- Datasets 2.17.1
- Tokenizers 0.15.2