Modelpublic

Kuongan/Halxlm-roberta-base_finetuned_merge_cp

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes11downloads

Model Card

Halxlm-roberta-basefinetunedmerge_cp

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

—Loss: 0.7640
—Accuracy: 0.7521
—F1 Macro: 0.7521

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: 32
—evalbatchsize: 32
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro
No log	1.0	175	1.0980	0.3293	0.1651
No log	2.0	350	0.7998	0.6586	0.6472
0.948	3.0	525	0.7057	0.7321	0.7346
0.948	4.0	700	0.7473	0.7214	0.7230
0.948	5.0	875	0.7695	0.7229	0.7256
0.5834	6.0	1050	0.7640	0.7521	0.7521
0.5834	7.0	1225	0.8045	0.745	0.7469
0.5834	8.0	1400	0.9001	0.7271	0.7297
0.4061	9.0	1575	0.9467	0.7314	0.7317

Framework versions

—Transformers 4.52.4
—Pytorch 2.6.0+cu124
—Datasets 3.6.0
—Tokenizers 0.21.2