speecht5_finetuned_mozilla_foundation_common_voice_13_german

This model is a fine-tuned version of microsoft/speecht5_tts on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4527

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: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• 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
0.5165	1.0	2193	0.4883
0.5125	2.0	4387	0.4789
0.4985	3.0	6580	0.4727
0.4963	4.0	8774	0.4701
0.4901	5.0	10967	0.4656
0.4814	6.0	13161	0.4611
0.4822	7.0	15354	0.4588
0.48	8.0	17548	0.4607
0.4676	9.0	19741	0.4564
0.4665	10.0	21935	0.4558
0.4652	11.0	24128	0.4549
0.4615	12.0	26322	0.4536
0.4716	13.0	28515	0.4541
0.462	14.0	30709	0.4526
0.4666	15.0	32902	0.4521
0.4591	16.0	35096	0.4532
0.4608	17.0	37289	0.4522
0.4605	18.0	39483	0.4513
0.4588	19.0	41676	0.4530
0.4561	20.0	43860	0.4527

Framework versions

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