wav2vec2-base-timit-demo-check32dataset

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

• Loss: 1.9876
• Wer: 0.8095
• Cer: 0.4179

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: 32
• 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
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
3.5272	5.0	5	2.3416	0.8333	0.4030
2.5892	10.0	10	2.2228	0.8333	0.3955
1.5311	15.0	15	2.1193	0.8095	0.4030
0.964	20.0	20	2.0500	0.7857	0.4179
0.8757	25.0	25	2.0115	0.7619	0.4104
0.7654	30.0	30	1.9876	0.8095	0.4179

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cpu
• Datasets 2.19.2
• Tokenizers 0.19.1