vsisik/speecht5_tts_SK_v3

SpeechT5 TTS Slovak v3

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

• Loss: 0.4046

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: 32
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.5259	3.2258	500	0.4625
0.4823	6.4516	1000	0.4345
0.4702	9.6774	1500	0.4258
0.4502	12.9032	2000	0.4189
0.4579	16.1290	2500	0.4173
0.4418	19.3548	3000	0.4134
0.448	22.5806	3500	0.4117
0.4467	25.8065	4000	0.4094
0.4388	29.0323	4500	0.4084
0.4327	32.2581	5000	0.4071
0.4398	35.4839	5500	0.4069
0.4381	38.7097	6000	0.4065
0.4357	41.9355	6500	0.4053
0.4352	45.1613	7000	0.4059
0.4298	48.3871	7500	0.4050
0.4293	51.6129	8000	0.4043
0.4342	54.8387	8500	0.4050
0.4309	58.0645	9000	0.4045
0.4277	61.2903	9500	0.4047
0.4319	64.5161	10000	0.4046

Framework versions

• Transformers 4.42.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1