KoRiF/speecht5_finetuned_common_voice_be

SpeechT5 Belarusian Speaker

This model is a fine-tuned version of microsoft/speecht5_tts on the Common Voice [Belarusian] dataset. It achieves the following results on the evaluation set:

• Loss: 0.6249

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: 4
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
0.8333	0.32	100	0.7151
0.7477	0.64	200	0.6609
0.7104	0.96	300	0.6311
0.69	1.28	400	0.6275
0.6658	1.6	500	0.6249

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3