File size: 2,436 Bytes
7ed27e3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9e883c8
7ed27e3
9e883c8
7ed27e3
 
 
 
 
 
 
 
9e883c8
 
 
 
 
7ed27e3
 
 
9e883c8
 
 
 
 
 
7ed27e3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: medicalai/ClinicalBERT
tags:
- generated_from_trainer
metrics:
- f1
- accuracy
model-index:
- name: working
  results: []
---

<!-- 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. -->

# Herbal Multilabel Classification

This model is a fine-tuned version of [medicalai/ClinicalBERT](https://huggingface.co/medicalai/ClinicalBERT) on a custom dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0108
- F1: 0.9834
- Roc Auc: 0.9930
- Accuracy: 0.9853

## Model description

It is a multilabel classification model that deals with 10 herbal plants 
(Jackfruit, Sambong, Lemon, Jasmine, Mango, Mint, Ampalaya, Malunggay, Guava, Lagundi)
which are abundant in the Philippines.
The model classifies a herbal(s) that is/are applicable based on the input symptom
of the user.

## Intended uses & limitations

The model is created for the purpose of completing a University course.
It will be integrated to a React Native mobile application for the
project.
The model performs well when the input of the user contains a symptom that has been trained
to the model from the dataset. However, other words/inputs that do not present a significance to
the purpose of the model would generate an underwhelming and inaccurate result.

## 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: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5

### Training results

| Training Loss | Epoch | Step | Validation Loss | F1     | Roc Auc | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:------:|:-------:|:--------:|
| No log        | 1.0   | 136  | 0.0223          | 0.9834 | 0.9930  | 0.9853   |
| No log        | 2.0   | 272  | 0.0163          | 0.9881 | 0.9959  | 0.9926   |
| No log        | 3.0   | 408  | 0.0137          | 0.9834 | 0.9930  | 0.9853   |
| 0.0216        | 4.0   | 544  | 0.0120          | 0.9834 | 0.9930  | 0.9853   |
| 0.0216        | 5.0   | 680  | 0.0108          | 0.9834 | 0.9930  | 0.9853   |


### Framework versions

- Transformers 4.37.0
- Pytorch 2.1.2+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0