Hugging Face's logo

k-seungri
/
k_whisper_output

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Korean
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
k_whisper_output / README.md
k-seungri's picture
k-seungri
End of training
f7d41e9 verified
preview code
|
raw
history blame
1.71 kB
metadata
language:
  - ko
license: apache-2.0
base_model: openai/whisper-base
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - k-seungri/k_whisper_dataset
model-index:
  - name: k-seungri/k_whisper_model
    results: []

k-seungri/k_whisper_model

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

  • Loss: 1.0608
  • Cer: 37.5

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 Cer
0.0002 142.86 1000 0.9623 34.8214
0.0001 285.71 2000 1.0241 34.5238
0.0001 428.57 3000 1.0469 39.8810
0.0001 571.43 4000 1.0608 37.5

Framework versions

  • Transformers 4.38.0.dev0
  • Pytorch 2.1.0+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.1