speecht5_fine_tuned_dhivehi_tts_V004

This model is a fine-tuned version of microsoft/speecht5_tts on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3868

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_steps: 500
• training_steps: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.527	3.6232	1000	0.4848
0.4907	7.2464	2000	0.4537
0.4697	10.8696	3000	0.4408
0.4528	14.4928	4000	0.4219
0.4357	18.1159	5000	0.4128
0.4271	21.7391	6000	0.4050
0.4171	25.3623	7000	0.3946
0.4112	28.9855	8000	0.3907
0.4076	32.6087	9000	0.3868
0.4085	36.2319	10000	0.3868

Framework versions

• Transformers 4.49.0.dev0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0