File size: 2,946 Bytes
139d41e
9c8d61e
 
139d41e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9c8d61e
139d41e
 
9c8d61e
139d41e
 
9c8d61e
139d41e
 
9c8d61e
139d41e
 
 
 
 
 
 
9c8d61e
139d41e
9c8d61e
 
 
 
 
139d41e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
aaae25c
9c8d61e
 
139d41e
 
 
2412186
fa7ac00
75acce3
139d41e
 
 
4eca7d5
 
9c8d61e
 
 
 
 
 
 
 
 
139d41e
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: distilbert/distilbert-base-multilingual-cased
tags:
- generated_from_trainer
datasets:
- cnec
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: CNEC2_0_Supertypes_xlm-roberta-large
  results:
  - task:
      name: Token Classification
      type: token-classification
    dataset:
      name: cnec
      type: cnec
      config: default
      split: validation
      args: default
    metrics:
    - name: Precision
      type: precision
      value: 0.7557829181494662
    - name: Recall
      type: recall
      value: 0.819980694980695
    - name: F1
      type: f1
      value: 0.7865740740740742
    - name: Accuracy
      type: accuracy
      value: 0.9568269568269568
---

<!-- 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. -->

# CNEC2_0_Supertypes_xlm-roberta-large

This model is a fine-tuned version of [distilbert/distilbert-base-multilingual-cased](https://huggingface.co/distilbert/distilbert-base-multilingual-cased) on the cnec dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2049
- Precision: 0.7558
- Recall: 0.8200
- F1: 0.7866
- Accuracy: 0.9568

## 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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.01
- lr_scheduler_warmup_steps: 1000
- num_epochs: 10

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1     | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.7025        | 1.11  | 500  | 0.2950          | 0.5066    | 0.5927 | 0.5463 | 0.9128   |
| 0.2152        | 2.22  | 1000 | 0.2057          | 0.6733    | 0.7539 | 0.7113 | 0.9425   |
| 0.1366        | 3.33  | 1500 | 0.1680          | 0.7228    | 0.7891 | 0.7545 | 0.9525   |
| 0.0849        | 4.44  | 2000 | 0.1710          | 0.7246    | 0.7987 | 0.7599 | 0.9540   |
| 0.0574        | 5.56  | 2500 | 0.1725          | 0.7309    | 0.8166 | 0.7714 | 0.9558   |
| 0.0384        | 6.67  | 3000 | 0.1855          | 0.7327    | 0.8243 | 0.7758 | 0.9554   |
| 0.0292        | 7.78  | 3500 | 0.1944          | 0.7557    | 0.8287 | 0.7905 | 0.9573   |
| 0.0208        | 8.89  | 4000 | 0.2053          | 0.7486    | 0.8118 | 0.7789 | 0.9555   |
| 0.0164        | 10.0  | 4500 | 0.2049          | 0.7558    | 0.8200 | 0.7866 | 0.9568   |


### Framework versions

- Transformers 4.36.2
- Pytorch 2.1.2+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0