oyemade/speecht5_tts_cv_16_1_yoruba

SpeechT5 TTS Yoruba

This model is a fine-tuned version of microsoft/speecht5_tts on the common_voice_16_1_yor dataset. It achieves the following results on the evaluation set:

• Loss: 0.4717

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: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.6107	8.55	500	0.5211
0.5458	17.09	1000	0.4882
0.5229	25.64	1500	0.4787
0.5088	34.19	2000	0.4723
0.5026	42.74	2500	0.4691
0.4978	51.28	3000	0.4706
0.509	59.83	3500	0.4712
0.4902	68.38	4000	0.4717

Framework versions

• Transformers 4.39.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2