---
license: mit
language: es
tags:
- generated_from_trainer
model-index:
- name: poem-gen-spanish-t5-small
  results: []
---

# poem-gen-spanish-t5-small

This model is a fine-tuned version of [flax-community/spanish-t5-small](https://huggingface.co/flax-community/spanish-t5-small) on the [Spanish Poetry Dataset](https://www.kaggle.com/andreamorgar/spanish-poetry-dataset/version/1) dataset.

The model was created during the [First Spanish Hackathon](https://somosnlp.org/hackathon) organized by [Somos NLP](https://somosnlp.org/).

The team who participated was composed by:

- 🇨🇺 [Alberto Carmona Barthelemy](https://huggingface.co/milyiyo)
- 🇪🇸 [Andrea Morales Garzón](https://huggingface.co/andreamorgar)
- 🇨🇴 [Jorge Henao](https://huggingface.co/jorge-henao)
- 🇮🇳 [Drishti Sharma](https://huggingface.co/DrishtiSharma)


It achieves the following results on the evaluation set:
- Loss: 2.8586
- Perplexity: 17.43

## Model description

The model was trained to generate spanish poems attending to some parameters like style, sentiment, words to include and starting phrase.

Example:

```
poema:
  estilo: Pablo Neruda &&
  sentimiento: positivo &&
  palabras: cielo, luna, mar &&
  texto: Todos fueron a verle pasar
```

### How to use

You can use this model directly with a pipeline for masked language modeling:

```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
model_name = 'hackathon-pln-es/poem-gen-spanish-t5-small'
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)

author, sentiment, word, start_text = 'Pablo Neruda', 'positivo', 'cielo', 'Todos fueron a la plaza'
input_text = f"""poema: estilo: {author} && sentimiento: {sentiment} && palabras: {word} && texto: {start_text} """
inputs = tokenizer(input_text, return_tensors="pt")

outputs = model.generate(inputs["input_ids"],
                         do_sample = True,
                         max_length = 30,
                         repetition_penalty = 20.0,
                         top_k = 50,
                         top_p = 0.92)
detok_outputs = [tokenizer.decode(x, skip_special_tokens=True) for x in outputs]
res = detok_outputs[0]
```

## Training and evaluation data

The original dataset has the columns `author`, `content` and `title`.
For each poem we generate new examples:
- content: *line_i* , generated: *line_i+1*
- content: *concatenate(line_i, line_i+1)* , generated: *line_i+2*
- content: *concatenate(line_i, line_i+1, line_i+2)* , generated: *line_i+3*

The resulting dataset has the columns `author`, `content`, `title` and `generated`.

For each example we compute the sentiment of the generated column and the nouns. In the case of sentiment, we used the model `mrm8488/electricidad-small-finetuned-restaurant-sentiment-analysis` and for nouns extraction we used spaCy.
 

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 6
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6

### Training results

| Training Loss | Epoch | Step   | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 3.1354        | 0.73  | 30000  | 3.0147          |
| 2.9761        | 1.46  | 60000  | 2.9498          |
| 2.897         | 2.19  | 90000  | 2.9019          |
| 2.8292        | 2.93  | 120000 | 2.8792          |
| 2.7774        | 3.66  | 150000 | 2.8738          |
| 2.741         | 4.39  | 180000 | 2.8634          |
| 2.7128        | 5.12  | 210000 | 2.8666          |
| 2.7108        | 5.85  | 240000 | 2.8595          |


### Framework versions

- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6