File size: 2,429 Bytes
281263d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c072376
281263d
 
c072376
281263d
 
c072376
281263d
 
c072376
281263d
 
 
 
 
 
 
 
 
c072376
 
 
 
 
281263d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c072376
 
 
 
 
 
281263d
 
 
 
c072376
281263d
c072376
281263d
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
---
license: mit
tags:
- generated_from_trainer
datasets:
- lg-ner
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: luganda-ner-v3
  results:
  - task:
      name: Token Classification
      type: token-classification
    dataset:
      name: lg-ner
      type: lg-ner
      config: lug
      split: test
      args: lug
    metrics:
    - name: Precision
      type: precision
      value: 0.8141289437585734
    - name: Recall
      type: recall
      value: 0.7971793149764943
    - name: F1
      type: f1
      value: 0.8055649813369528
    - name: Accuracy
      type: accuracy
      value: 0.952700740525628
---

<!-- 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. -->

# luganda-ner-v3

This model is a fine-tuned version of [xlm-roberta-large](https://huggingface.co/xlm-roberta-large) on the lg-ner dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2295
- Precision: 0.8141
- Recall: 0.7972
- F1: 0.8056
- Accuracy: 0.9527

## 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: 2e-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: 6

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| No log        | 1.0   | 261  | 0.4226          | 0.6273    | 0.3606 | 0.4580 | 0.8928   |
| 0.5572        | 2.0   | 522  | 0.2835          | 0.7720    | 0.6185 | 0.6868 | 0.9219   |
| 0.5572        | 3.0   | 783  | 0.2740          | 0.7579    | 0.7401 | 0.7489 | 0.9311   |
| 0.1745        | 4.0   | 1044 | 0.2423          | 0.7895    | 0.7683 | 0.7788 | 0.9399   |
| 0.1745        | 5.0   | 1305 | 0.2273          | 0.8048    | 0.7945 | 0.7996 | 0.9498   |
| 0.086         | 6.0   | 1566 | 0.2295          | 0.8141    | 0.7972 | 0.8056 | 0.9527   |


### Framework versions

- Transformers 4.27.4
- Pytorch 1.13.1+cu116
- Datasets 2.11.0
- Tokenizers 0.13.2