47 lines
1.8 KiB
Markdown
47 lines
1.8 KiB
Markdown
---
|
|
language:
|
|
- it
|
|
thumbnail: "url to a thumbnail used in social sharing"
|
|
tags:
|
|
- psychology
|
|
- distilgpt2
|
|
widget:
|
|
- text: "Il cognitivismo"
|
|
- text: "Parliamo di attaccamento"
|
|
- text: "Le cause del disturbo d'ansia nei bambini sono"
|
|
---
|
|
|
|
# Italian Psychology DistilGPT-2
|
|
|
|
This model is a fine-tuned version of the GPT-2 language model, trained on a dataset of Italian psychology articles.
|
|
It is capable of generating human-like text on topics related to psychology and mental health in Italian language.
|
|
|
|
## Model details
|
|
|
|
The base model used for fine-tuning is the GPT-2 model, with a transformer architecture and 1250M parameters.
|
|
The fine-tuning dataset consists of approximately 10,000 Italian psychology articles.
|
|
|
|
## Example usage
|
|
|
|
```python
|
|
from transformers import pipeline
|
|
|
|
nlp = pipeline("text-generation", model="misterkilgore/distilgpt2-psy-ita")
|
|
|
|
generated_text = nlp("Le cause del disturbo d'ansia nei bambini sono", max_length=100)
|
|
print(generated_text)
|
|
```
|
|
|
|
## Limitations and bias
|
|
This model has been trained on a dataset of Italian psychology articles and may not perform well on other types of text or in other languages.
|
|
Additionally, the dataset used to fine-tune the model may contain biases and limitations, which will be reflected in the generated text.
|
|
|
|
## Dataset
|
|
The dataset used to fine-tune this model is composed of Italian psychology articles.
|
|
It contains various topics on mental health and psychology, but some limitations and biases may be present. This model is meant to be used only for research and educational purposes.
|
|
|
|
## Training data
|
|
The training data is composed of Italian psychology articles.
|
|
Fine-tuning was performed on this dataset to adapt the base GPT-2 model to the specific topic of psychology and mental health in Italian language.
|
|
|