Edit model card

Whisper Small Somalia

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

  • Loss: 0.0005
  • Wer: 0.0

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
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • training_steps: 4000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.2361 6.5359 1000 0.0235 0.6506
0.0062 13.0719 2000 0.0024 0.0
0.0007 19.6078 3000 0.0006 0.0
0.0005 26.1438 4000 0.0005 0.0

Framework versions

  • Transformers 4.42.3
  • Pytorch 2.3.0
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
18
Safetensors
Model size
242M params
Tensor type
F32
·
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Finetuned from

Dataset used to train adityaedy01/whisper-small-so

Space using adityaedy01/whisper-small-so 1

Evaluation results