<!-- 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. -->
asd_pron_w2v_reg_balanced_500_79_corr
This model is a fine-tuned version of slplab/wav2vec2-xls-r-300m_phone-mfa_korean on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5603
- Spearman Correlation: 0.7313
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: 3e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Spearman Correlation |
---|---|---|---|---|
0.5646 | 1.0 | 308 | 0.5827 | 0.3993 |
0.3926 | 2.0 | 616 | 0.5692 | 0.4890 |
0.3873 | 3.0 | 924 | 0.5861 | 0.4202 |
0.3914 | 4.0 | 1232 | 0.5698 | 0.4443 |
0.3897 | 5.0 | 1540 | 0.5743 | 0.4825 |
0.3895 | 6.0 | 1848 | 0.5707 | 0.4169 |
0.3899 | 7.0 | 2156 | 0.5747 | 0.5449 |
0.3909 | 8.0 | 2464 | 0.5701 | 0.5510 |
0.3875 | 9.0 | 2772 | 0.5634 | 0.5389 |
0.3874 | 10.0 | 3080 | 0.5662 | 0.6041 |
0.3938 | 11.0 | 3388 | 0.5651 | 0.6427 |
0.3895 | 12.0 | 3696 | 0.5642 | 0.5501 |
0.3907 | 13.0 | 4004 | 0.5773 | 0.6043 |
0.389 | 14.0 | 4312 | 0.5697 | 0.6621 |
0.3887 | 15.0 | 4620 | 0.5563 | 0.6863 |
0.3882 | 16.0 | 4928 | 0.5647 | 0.6770 |
0.3907 | 17.0 | 5236 | 0.5719 | 0.6693 |
0.3903 | 18.0 | 5544 | 0.5610 | 0.7061 |
0.3905 | 19.0 | 5852 | 0.5616 | 0.6852 |
0.3877 | 20.0 | 6160 | 0.5722 | 0.6875 |
0.3874 | 21.0 | 6468 | 0.5647 | 0.6902 |
0.3901 | 22.0 | 6776 | 0.5619 | 0.7125 |
0.3913 | 23.0 | 7084 | 0.5717 | 0.6813 |
0.3857 | 24.0 | 7392 | 0.5533 | 0.7139 |
0.387 | 25.0 | 7700 | 0.5676 | 0.7143 |
0.3878 | 26.0 | 8008 | 0.5631 | 0.7118 |
0.3877 | 27.0 | 8316 | 0.5582 | 0.7276 |
0.389 | 28.0 | 8624 | 0.5660 | 0.7354 |
0.3909 | 29.0 | 8932 | 0.5623 | 0.7357 |
0.3876 | 30.0 | 9240 | 0.5603 | 0.7313 |
Framework versions
- Transformers 4.13.0
- Pytorch 2.0.0+cu118
- Datasets 2.14.4
- Tokenizers 0.10.3