metadata
library_name: transformers
language:
- ko
license: apache-2.0
base_model: openai/whisper-base
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- jindol/debugged_03_Whisper_datasets
model-index:
- name: repo_name
results: []
repo_name
This model is a fine-tuned version of openai/whisper-base on the debugged_03_Whisper_datasets dataset. It achieves the following results on the evaluation set:
- Loss: 1.6540
- Cer: 32.5714
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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- 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 | 1.4956 | 32.5714 |
| 0.0001 | 1000.0 | 2000 | 1.5831 | 36.0 |
| 0.0 | 1500.0 | 3000 | 1.6306 | 32.5714 |
| 0.0 | 2000.0 | 4000 | 1.6540 | 32.5714 |
Framework versions
- Transformers 4.46.2
- Pytorch 2.0.1
- Datasets 3.1.0
- Tokenizers 0.20.3