metadata

language:
  - ko
license: apache-2.0
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - google/fleurs
metrics:
  - wer
model-index:
  - name: Whisper Tiny Ko - TJ
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Google Fleures
          type: google/fleurs
          config: clean
          split: None
          args: 'config:ko, split: test'
        metrics:
          - name: Wer
            type: wer
            value: 233.42796309439316

Whisper Tiny Ko - TJ

This model is a fine-tuned version of openai/whisper-tiny on the Google Fleures dataset. It achieves the following results on the evaluation set:

Loss: 0.6558
Wer: 233.4280

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
0.1157	6.29	1000	0.5599	58.7828
0.0174	12.58	2000	0.6095	143.0979
0.0072	18.87	3000	0.6457	214.4074
0.005	25.16	4000	0.6558	233.4280

Framework versions

Transformers 4.27.0.dev0
Pytorch 1.13.1+cu116
Datasets 2.10.1
Tokenizers 0.13.2