Masternlp/tts

speecht5_feniks

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: 0.5319

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: 4
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• 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: 200
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.4478	3.0418	100	0.4543
0.4534	6.0837	200	0.4621
0.4373	9.1255	300	0.4543
0.4224	12.1673	400	0.4494
0.4127	15.2091	500	0.4657
0.4134	18.2510	600	0.4529
0.4047	21.2928	700	0.4724
0.3932	24.3346	800	0.4777
0.3907	27.3764	900	0.4942
0.3855	30.4183	1000	0.4870
0.3783	33.4601	1100	0.4860
0.3794	36.5019	1200	0.4867
0.3704	39.5437	1300	0.4965
0.3687	42.5856	1400	0.5151
0.3674	45.6274	1500	0.5165
0.3618	48.6692	1600	0.5377
0.3536	51.7110	1700	0.5206
0.3621	54.7529	1800	0.5419
0.3533	57.7947	1900	0.5337
0.3513	60.8365	2000	0.5319

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1