File size: 2,977 Bytes
5c36071
 
9fe52e4
5c36071
 
 
0675fc3
5c36071
 
 
 
 
 
0675fc3
5c36071
 
 
1758f24
 
 
 
 
 
 
5c36071
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c275879
2175e85
fe9641f
7da4557
e78019d
b566820
1758f24
5c36071
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
---
base_model: microsoft/layoutlm-base-uncased
license: mit
tags:
- generated_from_keras_callback
model-index:
- name: layoutlm-cord-tf-colab
  results: []
---

<!-- This model card has been generated automatically according to the information Keras had access to. You should
probably proofread and complete it, then remove this comment. -->

# layoutlm-cord-tf-colab

This model is a fine-tuned version of [microsoft/layoutlm-base-uncased](https://huggingface.co/microsoft/layoutlm-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 0.0426
- Validation Loss: 0.1530
- Train Overall Precision: 0.9498
- Train Overall Recall: 0.9642
- Train Overall F1: 0.9569
- Train Overall Accuracy: 0.9669
- Epoch: 7

## 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:
- optimizer: {'inner_optimizer': {'module': 'transformers.optimization_tf', 'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': 2.9999999242136255e-05, 'decay': 0.0, 'beta_1': 0.8999999761581421, 'beta_2': 0.9990000128746033, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}, 'registered_name': 'AdamWeightDecay'}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
- training_precision: mixed_float16

### Training results

| Train Loss | Validation Loss | Train Overall Precision | Train Overall Recall | Train Overall F1 | Train Overall Accuracy | Epoch |
|:----------:|:---------------:|:-----------------------:|:--------------------:|:----------------:|:----------------------:|:-----:|
| 1.3249     | 0.5218          | 0.8344                  | 0.8318               | 0.8331           | 0.8744                 | 0     |
| 0.4002     | 0.2680          | 0.9113                  | 0.9148               | 0.9130           | 0.9334                 | 1     |
| 0.2153     | 0.2180          | 0.9062                  | 0.9193               | 0.9127           | 0.9448                 | 2     |
| 0.1483     | 0.1840          | 0.9430                  | 0.9437               | 0.9433           | 0.9610                 | 3     |
| 0.0940     | 0.1687          | 0.9383                  | 0.9482               | 0.9432           | 0.9614                 | 4     |
| 0.0740     | 0.1539          | 0.9463                  | 0.9528               | 0.9496           | 0.9665                 | 5     |
| 0.0600     | 0.1795          | 0.9355                  | 0.9498               | 0.9426           | 0.9584                 | 6     |
| 0.0426     | 0.1530          | 0.9498                  | 0.9642               | 0.9569           | 0.9669                 | 7     |


### Framework versions

- Transformers 4.41.2
- TensorFlow 2.15.0
- Datasets 2.20.0
- Tokenizers 0.19.1