metadata
tags:
- generated_from_trainer
model-index:
- name: deit-hoogberta
results: []
deit-hoogberta
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 3.5649
- Cer: 0.9892
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 3.9434 | 0.28 | 500 | 4.0407 | 1.0311 |
| 4.0651 | 0.57 | 1000 | 3.8605 | 1.1336 |
| 3.8945 | 0.85 | 1500 | 3.7821 | 1.0140 |
| 3.5253 | 1.14 | 2000 | 3.7052 | 0.9804 |
| 3.5323 | 1.42 | 2500 | 3.6638 | 1.0365 |
| 3.3077 | 1.71 | 3000 | 3.6237 | 0.9716 |
| 3.3064 | 1.99 | 3500 | 3.5834 | 0.9648 |
| 3.2921 | 2.28 | 4000 | 3.5971 | 0.9872 |
| 3.0653 | 2.56 | 4500 | 3.5830 | 1.0193 |
| 3.1912 | 2.85 | 5000 | 3.5649 | 0.9892 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3