Team Ai
Modelpublic

OwLim/whisper-sundanese-java-finetune

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face
0likes16downloads
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. -->

Whisper Small Sundanese Java

This model is a fine-tuned version of OwLim/whisper-java-SLR41-SLR35 on the SLR44 Augmented Sundanese dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.1298
  • —Wer: 11.3509

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
  • —trainbatchsize: 16
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 100
  • —training_steps: 1000
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
0.59190.42921000.281523.4139
0.3950.85842000.178215.8826
0.17881.28763000.155413.0773
0.16541.71674000.144511.4372
0.05812.14595000.133711.5667
0.05722.57516000.133511.6746
0.0573.00437000.130411.1135
0.02413.43358000.131711.1135
0.02173.86279000.129711.5019
0.01474.291810000.129811.3509

Framework versions

  • —Transformers 4.51.3
  • —Pytorch 2.6.0+cu124
  • —Datasets 3.6.0
  • —Tokenizers 0.21.1