finetuned_bert-base-on-IEMOCAP_1

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

• Loss: 1.3271
• Accuracy: 0.6516

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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.2909	1.0	112	1.2207	0.5314
0.9328	2.0	224	0.9234	0.6446
0.7572	3.0	336	0.7751	0.7040
0.5686	4.0	448	0.7708	0.7175
0.3796	5.0	560	0.8258	0.7130
0.2736	6.0	672	0.8620	0.7231
0.2182	7.0	784	0.8939	0.7231
0.1755	8.0	896	1.0788	0.7220
0.1514	9.0	1008	1.0029	0.7365
0.1665	10.0	1120	1.0819	0.7152
0.2008	11.0	1232	1.1361	0.7152
0.0925	12.0	1344	1.1105	0.7253
0.162	13.0	1456	1.1379	0.7197
0.1302	14.0	1568	1.2054	0.7209
0.0701	15.0	1680	1.2240	0.7343
0.1392	16.0	1792	1.2842	0.7175
0.1419	17.0	1904	1.2614	0.7242
0.1125	18.0	2016	1.2826	0.7209
0.1128	19.0	2128	1.3033	0.7209
0.0813	20.0	2240	1.3068	0.7220

Framework versions

• Transformers 4.30.1
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3