metadata

language:
  - ru
license: apache-2.0
tags:
  - generated_from_trainer
base_model: openai/whisper-tiny
datasets:
  - bond005/podlodka_speech
metrics:
  - wer
model-index:
  - name: whisper-tiny-ru
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: Podlodka Speech
          type: bond005/podlodka_speech
          args: 'config: ru, split: test'
        metrics:
          - type: wer
            value: 88.10148731408573
            name: Wer

whisper-tiny-ru

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

Loss: 1.2591
Wer: 88.1015

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: 8
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 80
training_steps: 200

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.0155	4.4444	40	1.2150	56.6929
0.5695	8.8889	80	1.1441	72.5284
0.313	13.3333	120	1.1928	79.6150
0.1649	17.7778	160	1.2400	84.5144
0.1266	22.2222	200	1.2591	88.1015

Framework versions

Transformers 4.40.2
Pytorch 2.2.1+cu121
Datasets 2.19.1
Tokenizers 0.19.1