speecht5_finetuned_voxpopuli_fi

This model is a fine-tuned version of microsoft/speecht5_tts on the voxpopuli/fi dataset. It achieves the following results on the evaluation set:

• Loss: 0.4433

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: 10
• eval_batch_size: 2
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 250
• training_steps: 8000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.5561	0.9107	500	0.4943
0.5229	1.8215	1000	0.4739
0.5055	2.7322	1500	0.4629
0.5021	3.6430	2000	0.4598
0.4925	4.5537	2500	0.4547
0.4875	5.4645	3000	0.4523
0.4797	6.3752	3500	0.4509
0.4778	7.2860	4000	0.4488
0.4732	8.1967	4500	0.4477
0.4759	9.1075	5000	0.4466
0.4908	10.0182	5500	0.4448
0.4817	10.9290	6000	0.4442
0.4784	11.8397	6500	0.4443
0.4775	12.7505	7000	0.4437
0.4717	13.6612	7500	0.4436
0.4744	14.5719	8000	0.4433

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3