README.md

---
tags:
- generated_from_trainer
- TrOCR
model-index:
- name: trocr-large-printed-e13b_tesseract_MICR_ocr
  results: []
license: bsd-3-clause
language:
- en
metrics:
- cer
---

# trocr-large-printed-e13b_tesseract_MICR_ocr

This model is a fine-tuned version of [microsoft/trocr-large-printed](https://huggingface.co/microsoft/trocr-large-printed).

It achieves the following results on the evaluation set:
- Loss: 0.2432
- CER: 0.0036

## Model description

For more information on how it was created, check out the following link: https://github.com/DunnBC22/Vision_Audio_and_Multimodal_Projects/blob/main/Optical%20Character%20Recognition%20(OCR)/Tesseract%20MICR%20(E15B%20Dataset)/TrOCR-e13b%20-%20tesseractMICR.ipynb

## Intended uses & limitations

This model is intended to demonstrate my ability to solve a complex problem using technology.

## Training and evaluation data

Dataset Source: https://github.com/DoubangoTelecom/tesseractMICR/tree/master/datasets/e13b

__Histogram of Label Character Lengths__

![Histogram of Label Character Lengths](https://raw.githubusercontent.com/DunnBC22/Vision_Audio_and_Multimodal_Projects/main/Optical%20Character%20Recognition%20(OCR)/Tesseract%20MICR%20(E15B%20Dataset)/Images/Histogram%20of%20Label%20Character%20Length.png)


## 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: 2

### Training results

| Training Loss | Epoch | Step | Validation Loss | CER    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.486         | 1.0   | 841  | 0.5168          | 0.0428 |
| 0.2187        | 2.0   | 1682 | 0.2432          | 0.0036 |


### Framework versions

- Transformers 4.28.1
- Pytorch 2.0.1
- Datasets 2.13.1
- Tokenizers 0.13.3