---
library_name: transformers
language:
- ko
license: apache-2.0
base_model: openai/whisper-base
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- hyk000/phonekor
model-index:
- name: korpw
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# korpw

This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the korphoneh dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5903
- Cer: 18.4448

## 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 OptimizerNames.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     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.0341        | 10.0  | 1000 | 0.4909          | 70.6834 |
| 0.0021        | 20.0  | 2000 | 0.5542          | 18.7048 |
| 0.0008        | 30.0  | 3000 | 0.5810          | 18.7454 |
| 0.0007        | 40.0  | 4000 | 0.5903          | 18.4448 |


### Framework versions

- Transformers 4.47.0.dev0
- Pytorch 2.5.0
- Datasets 3.0.2
- Tokenizers 0.20.1