metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - common_voice_11_0
metrics:
  - wer
model-index:
  - name: openai/whisper-large-v2
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: common_voice_11_0
          type: common_voice_11_0
          config: lv
          split: test
          args: lv
        metrics:
          - name: Wer
            type: wer
            value: 27.504743833017077

openai/whisper-large-v2

This model is a fine-tuned version of openai/whisper-large-v2 on the common_voice_11_0 dataset. It achieves the following results on the evaluation set:

Loss: 0.3198
Wer: 27.5047

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: 3e-07
train_batch_size: 64
eval_batch_size: 32
seed: 42
distributed_type: multi-GPU
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 200
training_steps: 1500

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.5148	3.01	200	0.4189	39.3454
0.3041	6.03	400	0.3335	29.5731
0.1961	9.04	600	0.3186	27.7799
0.2579	13.01	800	0.3167	27.5712
0.2034	16.03	1000	0.3179	27.4763
0.1478	19.04	1200	0.3193	27.5237
0.2169	23.01	1400	0.3198	27.5047

Framework versions

Transformers 4.26.0.dev0
Pytorch 2.0.0.dev20221218+cu116
Datasets 2.7.1
Tokenizers 0.13.2