Curiousfox/speecht5_tailo-hokkien_ver1.0.b

speecht5_tailo_Hokkien_ver1.0.b

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

• Loss: 0.3200

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: 1
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 2000.0
• training_steps: 20000

Training results

Training Loss	Epoch	Step	Validation Loss
0.7266	2.6882	2000	0.4283
0.4435	5.3763	4000	0.3808
0.4016	8.0645	6000	0.3643
0.3828	10.7527	8000	0.3472
0.3712	13.4409	10000	0.3340
0.3600	16.1290	12000	0.3275
0.3553	18.8172	14000	0.3244
0.3529	21.5054	16000	0.3237
0.3486	24.1935	18000	0.3205
0.3480	26.8817	20000	0.3200

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2