metadata

license: mit
base_model: roberta-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: irony_pt_Portugal
    results: []

irony_pt_Portugal

This model is a fine-tuned version of roberta-base on part of the MultiPICo dataset. It achieves the following results on the evaluation set:

Loss: 0.0033
Accuracy: 0.5693
Precision: 0.4453
Recall: 0.7483
F1: 0.5584

Model description

The model is trained considering the annotation of annotators from Portuguese only, on instances in Portuguese (PT and BZ linguistic varieties). The annotations from these annotators are aggregated using majority voting and then used to train the model.

Training and evaluation data

The model has been trained on the annotation from annotators from Portugal from the MultiPICo dataset (instances in Portuguese). The data has been randomly split into a train and a validation set.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-06
train_batch_size: 16
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.0044	1.0	76	0.0042	0.5545	0.4358	0.7619	0.5545
0.004	2.0	152	0.0037	0.5347	0.4286	0.8367	0.5668
0.0037	3.0	228	0.0034	0.5446	0.4351	0.8435	0.5741
0.0034	4.0	304	0.0033	0.5421	0.4354	0.8707	0.5805
0.0031	5.0	380	0.0032	0.5396	0.4365	0.9116	0.5903
0.0028	6.0	456	0.0043	0.5792	0.4484	0.6803	0.5405
0.0026	7.0	532	0.0031	0.5569	0.4448	0.8776	0.5904
0.0026	8.0	608	0.0031	0.5619	0.4453	0.8299	0.5796
0.002	9.0	684	0.0033	0.5693	0.4453	0.7483	0.5584

Framework versions

Transformers 4.34.1
Pytorch 2.0.1+cu117
Datasets 2.14.5
Tokenizers 0.14.1