metadata
language:
- ko
license: apache-2.0
base_model: openai/whisper-base
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- Oyounghyun/whisper_study_data
model-index:
- name: study0627
results: []
study0627
This model is a fine-tuned version of openai/whisper-base on the train dataset. It achieves the following results on the evaluation set:
- Loss: 0.0686
- Cer: 2.4264
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: 10
- training_steps: 2000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 0.0599 | 1.9608 | 100 | 0.1089 | 4.6347 |
| 0.01 | 3.9216 | 200 | 0.0720 | 2.4264 |
| 0.0033 | 5.8824 | 300 | 0.0682 | 3.0807 |
| 0.002 | 7.8431 | 400 | 0.0670 | 2.5627 |
| 0.0014 | 9.8039 | 500 | 0.0667 | 2.6718 |
| 0.0011 | 11.7647 | 600 | 0.0667 | 2.5082 |
| 0.0009 | 13.7255 | 700 | 0.0668 | 2.7263 |
| 0.0008 | 15.6863 | 800 | 0.0671 | 2.5082 |
| 0.0007 | 17.6471 | 900 | 0.0674 | 2.5082 |
| 0.0006 | 19.6078 | 1000 | 0.0675 | 2.6172 |
| 0.0005 | 21.5686 | 1100 | 0.0675 | 2.5082 |
| 0.0005 | 23.5294 | 1200 | 0.0676 | 2.3991 |
| 0.0005 | 25.4902 | 1300 | 0.0679 | 2.3991 |
| 0.0004 | 27.4510 | 1400 | 0.0682 | 2.4264 |
| 0.0004 | 29.4118 | 1500 | 0.0682 | 2.4264 |
| 0.0004 | 31.3725 | 1600 | 0.0684 | 2.4264 |
| 0.0003 | 33.3333 | 1700 | 0.0684 | 2.4264 |
| 0.0004 | 35.2941 | 1800 | 0.0685 | 2.4264 |
| 0.0004 | 37.2549 | 1900 | 0.0685 | 2.4264 |
| 0.0003 | 39.2157 | 2000 | 0.0686 | 2.4264 |
Framework versions
- Transformers 4.43.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1