metadata
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-large-xls-r-300m-korean-g
results: []
wav2vec2-large-xls-r-300m-korean-g
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9504
- Cer: 0.1673
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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 4.7279 | 3.25 | 500 | 3.5098 | 0.9857 |
| 1.2914 | 6.49 | 1000 | 0.9970 | 0.2487 |
| 0.3791 | 9.74 | 1500 | 0.9036 | 0.2060 |
| 0.2147 | 12.99 | 2000 | 0.8963 | 0.1983 |
| 0.1607 | 16.23 | 2500 | 0.9386 | 0.1848 |
| 0.1281 | 19.48 | 3000 | 0.8897 | 0.1736 |
| 0.1094 | 22.73 | 3500 | 0.9474 | 0.1754 |
| 0.0933 | 25.97 | 4000 | 0.9557 | 0.1711 |
| 0.0845 | 29.22 | 4500 | 0.9504 | 0.1673 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 1.18.3
- Tokenizers 0.13.3