Modelpublic

Yuvrajg2107/roberta-base-cpp-final

sourceHugging Facemitupdated 9mo agoView on Hugging Face

0likes14downloads

Model Card

roberta-base-cpp

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

—Loss: 0.0844
—Accuracy: 0.9550

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: 2
—evalbatchsize: 16
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 4
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 1
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.1085	0.05	1000	0.1080	0.9183
0.0716	0.1	2000	0.2524	0.8473
0.0615	0.15	3000	0.1182	0.9299
0.0648	0.2	4000	0.0757	0.9498
0.0522	0.25	5000	0.1201	0.9273
0.0377	0.3	6000	0.0846	0.9555
0.0447	0.35	7000	0.1036	0.9323
0.0421	0.4	8000	0.1804	0.8914
0.0364	0.45	9000	0.0494	0.9628
0.0301	0.5	10000	0.0583	0.9689
0.0281	0.55	11000	0.0554	0.9689
0.0362	0.6	12000	0.0898	0.9428
0.022	0.65	13000	0.0772	0.9687
0.0221	0.7	14000	0.0706	0.9613
0.0256	0.75	15000	0.0487	0.9719
0.0215	0.8	16000	0.0427	0.9765
0.0162	0.85	17000	0.0437	0.9742
0.0186	0.9	18000	0.0613	0.9680
0.0211	0.95	19000	0.0950	0.9514
0.0136	1.0	20000	0.0844	0.9550

Framework versions

—Transformers 4.57.3
—Pytorch 2.8.0+cu126
—Datasets 4.4.2
—Tokenizers 0.22.1