wav2vec2-base-EMOPIA

This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.1691
• Accuracy: 0.6338

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: 1e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 3
• total_train_batch_size: 3
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8716	1.0	269	0.9822	0.6197
0.8143	2.0	538	1.2324	0.5352
0.7584	3.0	807	1.0226	0.6479
0.6715	4.0	1076	0.9550	0.6620
0.6471	5.0	1345	1.1272	0.6761
0.5759	6.0	1614	1.2193	0.6761
0.4963	7.0	1883	1.2214	0.7183
0.4053	8.0	2152	1.3083	0.7465
0.3344	9.0	2421	1.6391	0.6620
0.3216	10.0	2690	1.7224	0.6479
0.2248	11.0	2959	1.7973	0.6761
0.1982	12.0	3228	2.0241	0.6479
0.1362	13.0	3497	1.9933	0.6479
0.0879	14.0	3766	2.0865	0.6479
0.0712	15.0	4035	2.1691	0.6338

Framework versions

• Transformers 4.42.2
• Pytorch 2.3.1+cu118
• Datasets 2.20.0
• Tokenizers 0.19.1