Hugging Face's logo

kimjimin
/
model1

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Korean
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
model1 / README.md
kimjimin's picture
kimjimin
End of training
e0fcf3e verified
preview code
|
raw
history blame contribute delete
1.66 kB
metadata
language:
  - ko
license: apache-2.0
base_model: openai/whisper-base
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
datasets:
  - kimjimin/dataset1
model-index:
  - name: model1
    results: []

model1

This model is a fine-tuned version of openai/whisper-base on the dataset1 dataset. It achieves the following results on the evaluation set:

  • Loss: 1.7967
  • Cer: 232.2430

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
0.0001 500.0 1000 1.6765 35.9813
0.0 1000.0 2000 1.7396 232.7103
0.0 1500.0 3000 1.7738 232.2430
0.0 2000.0 4000 1.7967 232.2430

Framework versions

  • Transformers 4.44.0.dev0
  • Pytorch 2.3.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1