CoolFace
Modelpublic

sil-ai/gyi-LUK-audio-aligned-speecht5

sourceHugging Facemitupdated 2mo agoView on Hugging Face
1likes13downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

gyi-LUK-audio-aligned-speecht5

This model is a fine-tuned version of microsoft/speecht5_tts on the None dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.0899

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
  • —trainbatchsize: 16
  • —evalbatchsize: 16
  • —seed: 3407
  • —gradientaccumulationsteps: 4
  • —totaltrainbatch_size: 64
  • —optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • —lrschedulertype: cosine
  • —lrschedulerwarmup_steps: 2000
  • —training_steps: 20000
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation Loss
0.09350.010000.0865
0.0751100.020000.0819
0.0725150.030000.0800
0.0564200.040000.0813
0.0531250.050000.0804
0.0484300.060000.0812
0.0466350.070000.0824
0.0427400.080000.0868
0.0417450.090000.0871
0.0646500.0100000.0864
0.0387550.0110000.0881
0.0409600.0120000.0878
0.0389650.0130000.0890
0.0422700.0140000.0891
0.0428750.0150000.0894
0.0424800.0160000.0891
0.0441850.0170000.0897
0.044900.0180000.0906
0.0365950.0190000.0906
0.03591000.0200000.0899

Framework versions

  • —Transformers 4.57.1
  • —Pytorch 2.8.0+cu128
  • —Datasets 4.2.0
  • —Tokenizers 0.22.2