metadata

language:
  - en
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - Jungwonchang/spgispeech_xs
base_model: openai/whisper-large-v2
model-index:
  - name: openai/whisper-large-v2, all the parameters updated for 5 epochs
    results:
      - task:
          type: automatic-speech-recognition
          name: Automatic Speech Recognition
        dataset:
          name: Test set for spgispeech
          type: kensho/spgispeech
          config: test
          split: test
        metrics:
          - type: wer
            value: 6.85
            name: WER
          - type: cer
            value: 2.02
            name: CER

openai/whisper-large-v2, all the parameters updated for 5 epochs

This model is a fine-tuned version of openai/whisper-large-v2 on the 2 hour dataset of SPGIspeech(custom dataset) dataset.

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: 16
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 50
training_steps: 120
mixed_precision_training: Native AMP

Training results

Framework versions

Transformers 4.36.0.dev0
Pytorch 1.12.1+cu116
Datasets 2.4.0
Tokenizers 0.15.0