TrOCR_0208-2
This model is a fine-tuned version of microsoft/trocr-base-stage1 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2584
- Cer: 0.1211
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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Cer |
---|---|---|---|---|
1.3873 | 1.71 | 500 | 1.6813 | 0.2361 |
0.8298 | 3.42 | 1000 | 1.7390 | 0.2441 |
0.5587 | 5.14 | 1500 | 1.5896 | 0.2090 |
0.376 | 6.85 | 2000 | 1.4717 | 0.1775 |
0.2847 | 8.56 | 2500 | 1.5528 | 0.1928 |
0.2376 | 10.27 | 3000 | 1.4412 | 0.1727 |
0.2101 | 11.99 | 3500 | 1.3770 | 0.1592 |
0.2551 | 13.7 | 4000 | 1.4311 | 0.1564 |
0.226 | 15.41 | 4500 | 1.2536 | 0.1337 |
0.1365 | 17.12 | 5000 | 1.2753 | 0.1272 |
0.14 | 18.84 | 5500 | 1.2584 | 0.1211 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.1+cu121
- Datasets 2.13.0
- Tokenizers 0.15.0
- Downloads last month
- 21
Inference API (serverless) does not yet support transformers models for this pipeline type.
Model tree for yoon1000/TrOCR_0208-2
Base model
microsoft/trocr-base-stage1