SpeechT5_finetuned_kha

This model is a fine-tuned version of microsoft/speecht5_vc on the audiofolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.4733

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 512
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 300
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.544	36.8664	1000	0.5145
0.5013	73.7327	2000	0.4800
0.4754	110.5991	3000	0.4705
0.4651	147.4654	4000	0.4710
0.456	184.3318	5000	0.4699
0.446	221.1982	6000	0.4702
0.443	258.0645	7000	0.4714
0.4437	294.9309	8000	0.4733

Framework versions

• Transformers 4.43.3
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.19.1