Skip to content
ModelListedExternal
Download

Korean Kws

This model is a fine-tuned version of Kkonjeong/wav2vec2-base-korean on the None dataset. It achieves the following results on the evaluation set: Loss: 0.3157 Accuracy: 1.0

Sourcehuggingface.co/puppyAhn/korean_kws/tree/4ae884fefbfb7e4f1c620e63033e40ad0c526e05
KindOtherBase modelKkonjeong/wav2vec2-base-koreanVersionRevision 4ae884f

Model source

Kind
Other
File size
361 MB
Base model
Kkonjeong/wav2vec2-base-korean
Version
vRevision 4ae884f
Updated
Sep 19, 2026
Downloads
9
Source
Hugging Face

Source description

korean_kws

This model is a fine-tuned version of Kkonjeong/wav2vec2-base-korean on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3157
  • Accuracy: 1.0

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: 0.0001
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 32
  • optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 30
  • mixed_precision_training: Native AMP

Training results

Training LossEpochStepValidation LossAccuracy
No log1.021.96060.1538
No log2.041.91840.1538
No log3.061.88220.1538
No log4.081.73540.3846
No log5.0101.59310.3846
No log6.0121.44780.5385
No log7.0141.39070.3846
No log8.0161.21940.8462
No log9.0181.08221.0
No log10.0200.95681.0
No log11.0220.91611.0
No log12.0240.83651.0
No log13.0260.77611.0
No log14.0280.68131.0
No log15.0300.62481.0
No log16.0320.57501.0
No log17.0340.52551.0
No log18.0360.48441.0
No log19.0380.45451.0
No log20.0400.42801.0
No log21.0420.40551.0
No log22.0440.38451.0
No log23.0460.36751.0
No log24.0480.35341.0
No log25.0500.34231.0
No log26.0520.33131.0
No log27.0540.32451.0
No log28.0560.31971.0
No log29.0580.31691.0
No log30.0600.31571.0

Framework versions

  • Transformers 4.57.3
  • Pytorch 2.9.0+cu126
  • Datasets 4.0.0
  • Tokenizers 0.22.1

Sources

1 source
Hugging FacePrimaryVersion 4ae884fefbfb7e4f1c620e63033e40ad0c526e05
huggingface.co/puppyAhn/korean_kws/tree/4ae884fefbfb7e4f1c620e63033e40ad0c526e05

Verified Sep 19

Model artifacts

1 artifact

model.safetensors

safetensors · 361 MB · SHA-256 815620c788ae…a0ae · Hugging Face

Download

Source excerpts

2 excerpts
ReadmeSep 19huggingface.co/puppyAhn/korean_kws

--- library_name: transformers base_model: Kkonjeong/wav2vec2-base-korean tags: - generated_from_trainer metrics: - accuracy model-index: - name: korean_kws 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. --> # korean_kws This model is a fine-tuned version of [Kkonjeong/wav2vec2-base-korean](https://huggingface.co/Kkonjeong/wav2vec2-base-korean) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.3157 - Accuracy: 1.0 ## 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: 0.0001 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - gradient_accumulation_steps: 4 - total_train_batch_size: 32 - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments - lr_scheduler_type: linear - num_epochs: 30 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | No log | 1.0 | 2 | 1.9606 | 0.1538 | | No log | 2.0 | 4 | 1.9184 | 0.1538 | | No log | 3.0 | 6 | 1.8822 | 0.1538 | | No log | 4.0 | 8 | 1.7354 | 0.3846 | | No log | 5.0 | 10 | 1.5931 | 0.3846 | | No log | 6.0 | 12 | 1.4478 | 0.5385 | | No log | 7.0 | 14 | 1.3907 | 0.3846 | | No log | 8.0 | 16 | 1.2194 | 0.8462 | | No log | 9.0 | 18 | 1.0822 | 1.0 | | No log | 10.0 | 20 | 0.9568 | 1.0 | | No log | 11.0 | 22 | 0.9161 | 1.0 | | No log | 12.0 | 24 | 0.8365 | 1.0 | | No log | 13.0 | 26 | 0.7761 | 1.0 | | No log | 14.0 | 28 | 0.6813 | 1.0 | | No log | 15.0 | 30 | 0.6248 | 1.0 | | No log | 16.0 | 32 | 0.5750 | 1.0 | | No log | 17.0 | 34 | 0.5255 | 1.0 | | No log | 18.0 | 36 | 0.4844 | 1.0 | | No log | 19.0 | 38 | 0.4545 | 1.0 | | No log | 20.0 | 40 | 0.4280 | 1.0 | | No log | 21.0 | 42 | 0.4055 | 1.0 | | No log | 22.0 | 44 | 0.3845 | 1.0 | | No log | 23.0 | 46 | 0.3675 | 1.0 | | No log | 24.0 | 48 | 0.3534 | 1.0 | | No log | 25.0 | 50 | 0.3423 | 1.0 | | No log | 26.0 | 52 | 0.3313 | 1.0...

NoteSep 19huggingface.co/puppyAhn/korean_kws

Source context: 9 downloads · 0 likes · Pipeline audio-classification · Library transformers · Repo puppyAhn/korean_kws