metadata
library_name: transformers
license: apache-2.0
base_model: team-lucid/hubert-base-korean
tags:
- generated_from_trainer
model-index:
- name: for_test
results: []
for_test
This model is a fine-tuned version of team-lucid/hubert-base-korean on the None dataset. It achieves the following results on the evaluation set:
- Loss: 10074.0381
- Cer: 0.8429
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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 5400.8619 | 0.6369 | 300 | 13304.7881 | 1.0451 |
| 4024.1566 | 1.2739 | 600 | 11936.0117 | 0.9169 |
| 3685.2072 | 1.9108 | 900 | 11838.9512 | 0.8402 |
| 3836.7075 | 2.5478 | 1200 | 11351.3096 | 0.8275 |
| 3289.8719 | 3.1847 | 1500 | 11245.4717 | 0.8273 |
| 3506.6528 | 3.8217 | 1800 | 11008.6963 | 0.8322 |
| 3340.1028 | 4.4586 | 2100 | 10811.1230 | 0.8335 |
| 2946.4978 | 5.0955 | 2400 | 10763.3887 | 0.8341 |
| 3180.8653 | 5.7325 | 2700 | 10414.3926 | 0.8348 |
| 3159.9134 | 6.3694 | 3000 | 10376.6455 | 0.8354 |
| 2967.3987 | 7.0064 | 3300 | 10216.6924 | 0.8394 |
| 3072.4803 | 7.6433 | 3600 | 9977.7178 | 0.8387 |
| 3011.5284 | 8.2803 | 3900 | 10170.3740 | 0.8414 |
| 3042.4953 | 8.9172 | 4200 | 10057.4072 | 0.8420 |
| 3046.5066 | 9.5541 | 4500 | 10074.0381 | 0.8429 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.5.0+cu121
- Datasets 3.1.0
- Tokenizers 0.19.1