CoolFace
Modelpublic

olaysco/whisper-small-yoruba-lora

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face
0likes4downloads
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-yoruba-lora

This model is a fine-tuned version of openai/whisper-small on an unknown dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.2035

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: 0.001
  • —trainbatchsize: 32
  • —evalbatchsize: 16
  • —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_ratio: 0.1
  • —num_epochs: 8
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation Loss
0.65450.898920000.6119
0.57711.797840000.5680
0.51562.696660000.4931
0.38133.595580000.4293
0.31194.4944100000.3719
0.21775.3933120000.3084
0.13056.2921140000.2485
0.05427.1910160000.2035

Framework versions

  • —PEFT 0.14.0
  • —Transformers 4.50.3
  • —Pytorch 2.11.0+cu128
  • —Datasets 4.0.0
  • —Tokenizers 0.21.4