CoolFace
Modelpublic

Aakali/whisper-small-hi

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face
0likes3downloads
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 Hi - Aa

This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.4719
  • —Wer: 32.2907

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 500
  • —training_steps: 5000
  • —mixedprecisiontraining: Native AMP

Training results

Training LossEpochStepValidation LossWer
0.09212.445010000.299235.0504
0.02174.890020000.354333.5901
0.00217.335030000.420832.4896
0.00039.780040000.457032.2568
0.000212.224950000.471932.2907

Framework versions

  • —Transformers 4.41.0
  • —Pytorch 2.2.1+cu121
  • —Datasets 2.19.1
  • —Tokenizers 0.19.1