Indonesian RoBERTa Base POSP Tagger is a part-of-speech token-classification model based on the RoBERTa model. The model was originally the pre-trained Indonesian RoBERTa Base model, which is then fine-tuned on
POSP dataset consisting of tag-labelled news.
After training, the model achieved an evaluation F1-macro of 95.34%. On the benchmark test set, the model achieved an accuracy of 93.99% and F1-macro of 88.93%.
Trainer class from the Transformers library was used to train the model. PyTorch was used as the backend framework during training, but the model remains compatible with other frameworks nonetheless.
|Model||#params||Arch.||Training/Validation data (text)|
The model was trained for 10 epochs and the best model was loaded at the end.
|Epoch||Training Loss||Validation Loss||Precision||Recall||F1||Accuracy|
from transformers import pipeline pretrained_name = "w11wo/indonesian-roberta-base-posp-tagger" nlp = pipeline( "token-classification", model=pretrained_name, tokenizer=pretrained_name ) nlp("Budi sedang pergi ke pasar.")
Do consider the biases which come from both the pre-trained RoBERTa model and the
POSP dataset that may be carried over into the results of this model.
Indonesian RoBERTa Base POSP Tagger was trained and evaluated by Wilson Wongso. All computation and development are done on Google Colaboratory using their free GPU access.
- Downloads last month