metadata
tags:
- generated_from_trainer
model-index:
- name: khmer-trocr-base-printed
results: []
khmer-trocr-base-printed
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1980
- Cer: 0.5955
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: 4e-05
- train_batch_size: 8
- eval_batch_size: 6
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 1.1417 | 0.4876 | 1000 | 1.0816 | 0.8567 |
| 1.031 | 0.9751 | 2000 | 0.9824 | 0.8652 |
| 0.9015 | 1.4627 | 3000 | 0.8875 | 0.8421 |
| 0.7111 | 1.9503 | 4000 | 0.6645 | 0.7871 |
| 0.5049 | 2.4378 | 5000 | 0.4831 | 0.7234 |
| 0.4108 | 2.9254 | 6000 | 0.3594 | 0.6712 |
| 0.2876 | 3.4130 | 7000 | 0.3076 | 0.6458 |
| 0.2163 | 3.9005 | 8000 | 0.2418 | 0.6214 |
| 0.1862 | 4.3881 | 9000 | 0.2119 | 0.5998 |
| 0.1933 | 4.8757 | 10000 | 0.1980 | 0.5955 |
Framework versions
- Transformers 4.42.3
- Pytorch 2.1.2
- Datasets 2.20.0
- Tokenizers 0.19.1