SamuelM0422/speecht5_finetuned_cml-tts_pt

SpeechT5 Pt - Samuel Silva

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

• Loss: 0.413100

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: 8
• total_train_batch_size: 32
• optimizer: Use adamw_torch 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: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Step	Validation Loss
0.571400	250	0.513316
0.475900	500	0.436505
0.459000	750	0.422115
0.446800	1000	0.412104
0.439400	1250	0.403607
0.438600	1500	0.400913
0.437600	1750	0.398770
0.420900	2000	0.395658
0.427000	2250	0.393763
0.428000	2500	0.392377
0.425100	2750	0.392968

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0