Modelpublic

shanhy/xlm-roberta-base_seed42_original_kin-amh-eng_train

sourceHugging Facemitupdated 3y agoView on Hugging Face

0likes10downloads

Model Card

xlm-roberta-baseseed42originalkin-amh-engtrain

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

—Loss: 0.0167
—Spearman Corr: 0.8481

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: 128
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 64
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 30
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Spearman Corr
No log	1.95	200	0.0180	0.8098
0.0401	3.9	400	0.0152	0.8402
0.0212	5.85	600	0.0183	0.8493
0.0155	7.8	800	0.0166	0.8535
0.0116	9.76	1000	0.0206	0.8508
0.0097	11.71	1200	0.0155	0.8459
0.008	13.66	1400	0.0159	0.8481
0.0068	15.61	1600	0.0153	0.8467
0.0058	17.56	1800	0.0196	0.8481
0.0053	19.51	2000	0.0167	0.8481

Framework versions

—Transformers 4.37.2
—Pytorch 2.1.0+cu121
—Datasets 2.17.0
—Tokenizers 0.15.1