speecht5_finetuned_voxpopuli_lt

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

• Loss: 0.5588

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: 4
• seed: 42
• gradient_accumulation_steps: 8
• 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: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.7332	24.8649	100	0.6417
0.658	49.8649	200	0.6113
0.603	74.8649	300	0.5863
0.5626	99.8649	400	0.5698
0.5389	124.8649	500	0.5631
0.5248	149.8649	600	0.5639
0.5105	174.8649	700	0.5564
0.5083	199.8649	800	0.5587
0.5038	224.8649	900	0.5544
0.5029	249.8649	1000	0.5588

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0