speecht5_tts_voxpopuli_it_v2

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: 0.4484

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: 8
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• 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
0.5707	0.2358	100	0.5183
0.5452	0.4717	200	0.5096
0.5313	0.7075	300	0.4890
0.5229	0.9434	400	0.4807
0.5119	1.1792	500	0.4802
0.5121	1.4151	600	0.4681
0.5037	1.6509	700	0.4719
0.4996	1.8868	800	0.4691
0.4931	2.1226	900	0.4621
0.4903	2.3585	1000	0.4620
0.4949	2.5943	1100	0.4573
0.4853	2.8302	1200	0.4579
0.4826	3.0660	1300	0.4547
0.4827	3.3019	1400	0.4535
0.4835	3.5377	1500	0.4523
0.4802	3.7736	1600	0.4514
0.4777	4.0094	1700	0.4503
0.4792	4.2453	1800	0.4499
0.4779	4.4811	1900	0.4491
0.4755	4.7170	2000	0.4484

Framework versions

• Transformers 4.43.1
• Pytorch 2.2.0
• Datasets 3.0.1
• Tokenizers 0.19.1