Modelpublic

HCKLab/BiBert-Subjectivity

sourceHugging Facemitupdated 4y agoView on Hugging Face

1likes13downloads

Model Card

BiBert-Subjectivity

This model is a fine-tuned version of nlptown/bert-base-multilingual-uncased-sentiment on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.1481
—Accuracy: 0.9583
—F1: 0.9581
—Mae: 0.0417
—Accuracy 2: 0.9583

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

—learning_rate: 2e-05
—trainbatchsize: 16
—evalbatchsize: 16
—seed: 42
—gradientaccumulationsteps: 4
—totaltrainbatch_size: 64
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Mae	Accuracy 2
No log	1.0	112	0.1333	0.95	0.9508	0.05	0.95
No log	2.0	224	0.1517	0.953	0.9531	0.047	0.953
No log	3.0	336	0.2219	0.951	0.9505	0.049	0.951
No log	4.0	448	0.2327	0.947	0.9479	0.053	0.947
0.0865	5.0	560	0.2557	0.953	0.9528	0.047	0.953

Framework versions

—Transformers 4.25.1
—Pytorch 1.13.0+cu116
—Datasets 2.8.0
—Tokenizers 0.13.2