metadata

library_name: transformers
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
  - generated_from_trainer
datasets:
  - honzapucalek/monika_asr
metrics:
  - wer
model-index:
  - name: monika_asr
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: honzapucalek/monika_asr cs
          type: honzapucalek/monika_asr
          config: cs
          split: test
          args: cs
        metrics:
          - name: Wer
            type: wer
            value: 0.26058233423048693

monika_asr

This model is a fine-tuned version of openai/whisper-large-v3 on the honzapucalek/monika_asr cs dataset. It achieves the following results on the evaluation set:

Loss: 0.7834
Wer: 0.2606

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: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
training_steps: 5000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0001	45.4545	1000	0.6877	0.2625
0.0001	90.9091	2000	0.6910	0.2589
0.0	136.3636	3000	0.7108	0.2591
0.0	181.8182	4000	0.7377	0.2618
0.0	227.2727	5000	0.7669	0.2606

Framework versions

Transformers 4.49.0
Pytorch 2.1.2+cu121
Datasets 3.1.0
Tokenizers 0.21.0