finetuned_roberta-base-on-IEMOCAP_2

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

• Loss: 0.8093
• Accuracy: 0.7283

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
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.9868	1.0	113	0.9133	0.6493
0.7065	2.0	226	0.8576	0.6803
0.4773	3.0	339	0.8821	0.6947
0.4002	4.0	452	0.9283	0.7024
0.3338	5.0	565	1.0637	0.6792
0.3128	6.0	678	1.0653	0.6892
0.2552	7.0	791	1.1852	0.6814
0.244	8.0	904	1.2298	0.6825
0.2178	9.0	1017	1.2453	0.6925
0.2083	10.0	1130	1.3690	0.6858
0.1516	11.0	1243	1.3877	0.6914
0.1186	12.0	1356	1.4572	0.6903
0.1681	13.0	1469	1.4289	0.6925
0.1175	14.0	1582	1.4813	0.6869
0.1068	15.0	1695	1.4878	0.6869

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1