|
--- |
|
language: su |
|
tags: |
|
- sundanese-gpt2-base |
|
license: mit |
|
datasets: |
|
- mc4 |
|
- cc100 |
|
- oscar |
|
- wikipedia |
|
widget: |
|
- text: "Nami abdi Budi, ti Indonésia" |
|
--- |
|
|
|
## Sundanese GPT-2 Base |
|
|
|
Sundanese GPT-2 Base is a causal language model based on the [OpenAI GPT-2](https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf) model. It was trained on four datasets: [OSCAR](https://hf.co/datasets/oscar)'s `unshuffled_deduplicated_su` subset, the Sundanese [mC4](https://hf.co/datasets/mc4) subset, the Sundanese [CC100](https://hf.co/datasets/cc100) subset, and Sundanese [Wikipedia](https://su.wikipedia.org/). |
|
|
|
10% of the dataset is kept for evaluation purposes. The model was trained from scratch and achieved an evaluation loss of 3.61 and an evaluation perplexity of 36.97. |
|
|
|
This model was trained using HuggingFace's Flax framework. All necessary scripts used for training could be found in the [Files and versions](https://hf.co/w11wo/sundanese-gpt2-base/tree/main) tab, as well as the [Training metrics](https://hf.co/w11wo/sundanese-gpt2-base/tensorboard) logged via Tensorboard. |
|
|
|
## Model |
|
|
|
| Model | #params | Arch. | Training/Validation data (text) | |
|
| --------------------- | ------- | ----- | ------------------------------------- | |
|
| `sundanese-gpt2-base` | 124M | GPT-2 | OSCAR, mC4, CC100, Wikipedia (758 MB) | |
|
|
|
## Evaluation Results |
|
|
|
The model was trained for 50 epochs and the following is the final result once the training ended. |
|
|
|
| train loss | valid loss | valid PPL | total time | |
|
| ---------- | ---------- | --------- | ---------- | |
|
| 2.436 | 3.61 | 36.97 | 7:1:54 | |
|
|
|
## How to Use |
|
|
|
### As Causal Language Model |
|
|
|
```python |
|
from transformers import pipeline |
|
|
|
pretrained_name = "w11wo/sundanese-gpt2-base" |
|
|
|
nlp = pipeline( |
|
"text-generation", |
|
model=pretrained_name, |
|
tokenizer=pretrained_name |
|
) |
|
|
|
nlp("Nami abdi Budi, ti Indonésia") |
|
``` |
|
|
|
### Feature Extraction in PyTorch |
|
|
|
```python |
|
from transformers import GPT2Model, GPT2TokenizerFast |
|
|
|
pretrained_name = "w11wo/sundanese-gpt2-base" |
|
model = GPT2Model.from_pretrained(pretrained_name) |
|
tokenizer = GPT2TokenizerFast.from_pretrained(pretrained_name) |
|
|
|
prompt = "Nami abdi Budi, ti Indonésia" |
|
encoded_input = tokenizer(prompt, return_tensors='pt') |
|
output = model(**encoded_input) |
|
``` |
|
|
|
## Disclaimer |
|
|
|
Do consider the biases which came from all four datasets that may be carried over into the results of this model. |
|
|
|
## Author |
|
|
|
Sundanese GPT-2 Base was trained and evaluated by [Wilson Wongso](https://w11wo.github.io/). |
|
|
|
## Citation Information |
|
|
|
```bib |
|
@article{rs-907893, |
|
author = {Wongso, Wilson |
|
and Lucky, Henry |
|
and Suhartono, Derwin}, |
|
journal = {Journal of Big Data}, |
|
year = {2022}, |
|
month = {Feb}, |
|
day = {26}, |
|
abstract = {The Sundanese language has over 32 million speakers worldwide, but the language has reaped little to no benefits from the recent advances in natural language understanding. Like other low-resource languages, the only alternative is to fine-tune existing multilingual models. In this paper, we pre-trained three monolingual Transformer-based language models on Sundanese data. When evaluated on a downstream text classification task, we found that most of our monolingual models outperformed larger multilingual models despite the smaller overall pre-training data. In the subsequent analyses, our models benefited strongly from the Sundanese pre-training corpus size and do not exhibit socially biased behavior. We released our models for other researchers and practitioners to use.}, |
|
issn = {2693-5015}, |
|
doi = {10.21203/rs.3.rs-907893/v1}, |
|
url = {https://doi.org/10.21203/rs.3.rs-907893/v1} |
|
} |
|
``` |
