speecht5_finetuned_facebook_voxpopuli_spanish

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: 0.4355

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• 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: 100
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
0.484	1.0	1074	0.4623
0.4724	2.0	2148	0.4542
0.4646	3.0	3222	0.4474
0.463	4.0	4296	0.4451
0.4538	5.0	5370	0.4425
0.4509	6.0	6444	0.4408
0.4548	7.0	7518	0.4412
0.4465	8.0	8592	0.4389
0.4509	9.0	9666	0.4387
0.4424	10.0	10740	0.4369
0.4415	11.0	11814	0.4367
0.4386	12.0	12888	0.4366
0.4383	13.0	13962	0.4357
0.4396	14.0	15036	0.4348
0.4432	15.0	16110	0.4358
0.4378	16.0	17184	0.4356
0.4321	17.0	18258	0.4353
0.4381	18.0	19332	0.4353
0.4325	19.0	20406	0.4347
0.4336	20.0	21480	0.4355

Framework versions

• Transformers 4.30.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.13.1
• Tokenizers 0.13.3