metadata
base_model: microsoft/trocr-base-stage1
tags:
- generated_from_trainer
model-index:
- name: TrOCR_0219_All_data
results: []
TrOCR_0219_All_data
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: 0.8182
- Cer: 0.1603
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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 0.7188 | 0.05 | 500 | 0.8385 | 0.1740 |
| 0.9207 | 0.09 | 1000 | 0.8321 | 0.1692 |
| 1.1151 | 0.14 | 1500 | 0.8163 | 0.0989 |
| 0.9162 | 0.18 | 2000 | 0.6953 | 0.0830 |
| 0.5558 | 0.23 | 2500 | 0.8182 | 0.1603 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.1+cu121
- Datasets 2.13.0
- Tokenizers 0.15.0