audio_cls_jhc

This model is a fine-tuned version of Kkonjeong/wav2vec2-base-korean on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.1076
• Accuracy: 0.7731

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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	15	2.6736	0.0672
No log	2.0	30	2.5343	0.0924
No log	3.0	45	2.2010	0.3866
No log	4.0	60	1.8560	0.5126
No log	5.0	75	1.6494	0.5630
No log	6.0	90	1.4823	0.6555
No log	7.0	105	1.3294	0.7227
No log	8.0	120	1.1878	0.7815
No log	9.0	135	1.1212	0.7899
No log	10.0	150	1.1076	0.7731

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2