Modelpublic

enoubi/XLM-RoBERTa-Twitter-Indonesian-Sarcastic-Few-Shot

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes6downloads

Model Card

XLM-RoBERTa-Twitter-Indonesian-Sarcastic-Few-Shot

This model is a fine-tuned version of xlm-roberta-large on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.3513
—Accuracy: 0.8717
—F1: 0.7677
—Precision: 0.6994
—Recall: 0.8507

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: 1e-05
—trainbatchsize: 32
—evalbatchsize: 64
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: cosine
—num_epochs: 100
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.5833	1.0	31	0.5356	0.75	0.0	0.0	0.0
0.526	2.0	62	0.4851	0.75	0.0	0.0	0.0
0.4795	3.0	93	0.4745	0.7724	0.1644	1.0	0.0896
0.3989	4.0	124	0.3300	0.8657	0.6667	0.8780	0.5373
0.2827	5.0	155	0.3112	0.8657	0.7391	0.7183	0.7612
0.2006	6.0	186	0.2641	0.8955	0.7705	0.8545	0.7015
0.1357	7.0	217	0.3315	0.8881	0.7917	0.7403	0.8507
0.1251	8.0	248	0.4118	0.8433	0.7308	0.6404	0.8507
0.0643	9.0	279	0.4539	0.8918	0.7642	0.8393	0.7015
0.046	10.0	310	0.5066	0.8694	0.7518	0.7162	0.7910

Framework versions

—Transformers 4.51.1
—Pytorch 2.5.1+cu124
—Datasets 3.5.0
—Tokenizers 0.21.0