File size: 3,832 Bytes
361a2f3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
---
license: mit
base_model: nielsr/lilt-xlm-roberta-base
tags:
- generated_from_trainer
datasets:
- xfun
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: LiLT-SER-DE
  results:
  - task:
      name: Token Classification
      type: token-classification
    dataset:
      name: xfun
      type: xfun
      config: xfun.de
      split: validation
      args: xfun.de
    metrics:
    - name: Precision
      type: precision
      value: 0.7268232385661311
    - name: Recall
      type: recall
      value: 0.7853962600178095
    - name: F1
      type: f1
      value: 0.7549753905414082
    - name: Accuracy
      type: accuracy
      value: 0.7816669203063968
---

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

# LiLT-SER-DE

This model is a fine-tuned version of [nielsr/lilt-xlm-roberta-base](https://huggingface.co/nielsr/lilt-xlm-roberta-base) on the xfun dataset.
It achieves the following results on the evaluation set:
- Loss: 2.1833
- Precision: 0.7268
- Recall: 0.7854
- F1: 0.7550
- Accuracy: 0.7817

## 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: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 10000

### Training results

| Training Loss | Epoch  | Step  | Accuracy | F1     | Validation Loss | Precision | Recall |
|:-------------:|:------:|:-----:|:--------:|:------:|:---------------:|:---------:|:------:|
| 0.2776        | 10.42  | 500   | 0.7098   | 0.6660 | 1.4820          | 0.6266    | 0.7106 |
| 0.0386        | 20.83  | 1000  | 0.7884   | 0.7195 | 1.3364          | 0.6868    | 0.7556 |
| 0.002         | 31.25  | 1500  | 0.8102   | 0.7350 | 1.4865          | 0.7000    | 0.7738 |
| 0.0043        | 41.67  | 2000  | 0.7965   | 0.7167 | 1.5473          | 0.7050    | 0.7289 |
| 0.0009        | 52.08  | 2500  | 0.7797   | 0.7357 | 1.8408          | 0.7371    | 0.7342 |
| 0.0003        | 62.5   | 3000  | 0.7841   | 0.7279 | 1.9387          | 0.7021    | 0.7556 |
| 0.0044        | 72.92  | 3500  | 0.7900   | 0.7402 | 1.7595          | 0.7292    | 0.7516 |
| 0.0005        | 83.33  | 4000  | 0.7677   | 0.7370 | 2.0830          | 0.7084    | 0.7680 |
| 0.0001        | 93.75  | 4500  | 0.7746   | 0.7555 | 2.0764          | 0.7301    | 0.7827 |
| 0.0001        | 104.17 | 5000  | 0.7716   | 0.7441 | 2.0912          | 0.7158    | 0.7747 |
| 0.0           | 114.58 | 5500  | 0.7764   | 0.7572 | 2.1803          | 0.7275    | 0.7894 |
| 0.0           | 125.0  | 6000  | 0.7809   | 0.7576 | 2.1028          | 0.7384    | 0.7778 |
| 0.0001        | 135.42 | 6500  | 0.7812   | 0.7422 | 2.0825          | 0.7240    | 0.7614 |
| 0.0001        | 145.83 | 7000  | 0.7882   | 0.7481 | 2.0649          | 0.7244    | 0.7734 |
| 0.0001        | 156.25 | 7500  | 0.7789   | 0.7536 | 2.1535          | 0.7324    | 0.7760 |
| 0.0           | 166.67 | 8000  | 0.7760   | 0.7491 | 2.2120          | 0.7307    | 0.7685 |
| 0.0           | 177.08 | 8500  | 0.7941   | 0.7615 | 1.9997          | 0.75      | 0.7734 |
| 0.0           | 187.5  | 9000  | 0.7854   | 0.7588 | 2.0939          | 0.7355    | 0.7836 |
| 0.0           | 197.92 | 9500  | 2.1707   | 0.7262 | 0.7805          | 0.7524    | 0.7825 |
| 0.0           | 208.33 | 10000 | 2.1833   | 0.7268 | 0.7854          | 0.7550    | 0.7817 |


### Framework versions

- Transformers 4.39.1
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.1