Modelpublic

FarhanAmin0068/xlm-r-base-shared-task

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes4downloads

Model Card

xlm-r-base-shared-task

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

—Loss: 0.5637
—Precision: 0.7041
—Recall: 0.7064
—Accuracy: 0.7826
—F1: 0.7052

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 adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 286
—num_epochs: 10
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	Accuracy	F1
0.7008	1.0	96	0.6943	0.3790	0.5	0.7581	0.4312
0.6792	2.0	192	0.6550	0.6653	0.6377	0.7620	0.6474
0.6411	3.0	288	0.6031	0.6614	0.7021	0.7218	0.6687
0.5855	4.0	384	0.5637	0.7041	0.7064	0.7826	0.7052
0.5165	5.0	480	0.6010	0.6694	0.7180	0.7209	0.6749
0.4367	6.0	576	0.6018	0.6625	0.6977	0.7297	0.6711
0.3874	7.0	672	1.0209	0.7059	0.6566	0.7865	0.6721
0.33	8.0	768	0.8338	0.6908	0.6833	0.7747	0.6868
0.2592	9.0	864	0.8379	0.6822	0.6739	0.7689	0.6777
0.2231	10.0	960	0.8859	0.6844	0.6787	0.7698	0.6814

Framework versions

—Transformers 4.47.0
—Pytorch 2.5.1+cu121
—Datasets 3.3.1
—Tokenizers 0.21.0