Modelpublic

Kuongan/xlm-roberta-base-pcm-finetuned-RC

sourceHugging Facemitupdated 2y agoView on Hugging Face

0likes10downloads

Model Card

xlm-roberta-base-pcm-finetuned-RC

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: 0.2535
—F1: 0.7752
—Roc Auc: 0.8434
—Accuracy: 0.6425

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: cosine
—lrschedulerwarmup_steps: 100
—num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.4847	1.0	187	0.4784	0.1137	0.5005	0.2160
0.4397	2.0	374	0.4546	0.2391	0.5679	0.2750
0.3848	3.0	561	0.3969	0.3759	0.6366	0.3018
0.3585	4.0	748	0.3665	0.4574	0.6711	0.3508
0.3049	5.0	935	0.3598	0.4761	0.6802	0.3823
0.295	6.0	1122	0.3421	0.5135	0.7011	0.3883
0.2641	7.0	1309	0.3280	0.5639	0.7292	0.4359
0.2289	8.0	1496	0.3214	0.6291	0.7614	0.4769
0.2105	9.0	1683	0.3074	0.6173	0.7549	0.4997
0.176	10.0	1870	0.2993	0.6614	0.7754	0.5359
0.1623	11.0	2057	0.2809	0.6977	0.7964	0.5634
0.1563	12.0	2244	0.2689	0.7166	0.8055	0.5909
0.1172	13.0	2431	0.2609	0.7312	0.8152	0.6030
0.0996	14.0	2618	0.2608	0.7422	0.8224	0.6043
0.1021	15.0	2805	0.2588	0.7606	0.8375	0.6231
0.0942	16.0	2992	0.2623	0.7595	0.8366	0.6264
0.0906	17.0	3179	0.2535	0.7752	0.8434	0.6425
0.0773	18.0	3366	0.2555	0.7702	0.8404	0.6345
0.0746	19.0	3553	0.2533	0.7726	0.8419	0.6372
0.0738	20.0	3740	0.2538	0.7739	0.8423	0.6405

Framework versions

—Transformers 4.47.0
—Pytorch 2.5.1+cu121
—Datasets 3.2.0
—Tokenizers 0.21.0