File size: 3,362 Bytes
a57fa17
 
47f7450
a57fa17
 
 
 
 
 
 
 
47f7450
a57fa17
 
 
 
 
 
47f7450
a57fa17
47f7450
a57fa17
47f7450
 
 
 
 
a57fa17
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
47f7450
a57fa17
 
 
 
 
47f7450
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a57fa17
 
 
 
 
 
 
 
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
---
library_name: transformers
base_model: monsoon-nlp/bert-base-thai
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: bert-base-thai-intent-booking
  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. -->

# bert-base-thai-intent-booking

This model is a fine-tuned version of [monsoon-nlp/bert-base-thai](https://huggingface.co/monsoon-nlp/bert-base-thai) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.1511
- Accuracy: 0.1937
- F1: 0.1641
- Precision: 0.2236
- Recall: 0.1937

## 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: 16
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 64
- num_epochs: 20

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 2.368         | 1.0   | 65   | 2.3624          | 0.0901   | 0.0149 | 0.0081    | 0.0901 |
| 2.3586        | 2.0   | 130  | 2.2374          | 0.1577   | 0.0963 | 0.1107    | 0.1577 |
| 2.2987        | 3.0   | 195  | 2.2589          | 0.1216   | 0.0626 | 0.0890    | 0.1216 |
| 2.279         | 4.0   | 260  | 2.1771          | 0.1757   | 0.1329 | 0.2163    | 0.1757 |
| 2.2326        | 5.0   | 325  | 2.2099          | 0.2027   | 0.1548 | 0.1497    | 0.2027 |
| 2.2273        | 6.0   | 390  | 2.1809          | 0.1712   | 0.1245 | 0.1127    | 0.1712 |
| 2.2303        | 7.0   | 455  | 2.2168          | 0.1486   | 0.1030 | 0.1190    | 0.1486 |
| 2.196         | 8.0   | 520  | 2.1862          | 0.1937   | 0.1478 | 0.1615    | 0.1937 |
| 2.1848        | 9.0   | 585  | 2.1320          | 0.2162   | 0.1773 | 0.2192    | 0.2162 |
| 2.183         | 10.0  | 650  | 2.1771          | 0.1712   | 0.1240 | 0.1703    | 0.1712 |
| 2.1669        | 11.0  | 715  | 2.1672          | 0.2117   | 0.1849 | 0.2453    | 0.2117 |
| 2.1586        | 12.0  | 780  | 2.1237          | 0.2162   | 0.1939 | 0.3552    | 0.2162 |
| 2.1465        | 13.0  | 845  | 2.1269          | 0.2117   | 0.1834 | 0.2440    | 0.2117 |
| 2.1454        | 14.0  | 910  | 2.1160          | 0.2162   | 0.1939 | 0.3552    | 0.2162 |
| 2.1404        | 15.0  | 975  | 2.1089          | 0.2162   | 0.1936 | 0.3561    | 0.2162 |
| 2.1293        | 16.0  | 1040 | 2.1272          | 0.2162   | 0.1947 | 0.3584    | 0.2162 |
| 2.1193        | 17.0  | 1105 | 2.1043          | 0.2117   | 0.1836 | 0.2431    | 0.2117 |
| 2.1094        | 18.0  | 1170 | 2.1053          | 0.2117   | 0.1895 | 0.2977    | 0.2117 |
| 2.1063        | 19.0  | 1235 | 2.1055          | 0.2117   | 0.1901 | 0.2989    | 0.2117 |
| 2.0888        | 20.0  | 1300 | 2.1067          | 0.2117   | 0.1895 | 0.2977    | 0.2117 |


### Framework versions

- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1