File size: 2,189 Bytes
4c52014
 
 
 
 
 
 
 
00648e1
4c52014
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- f1
- recall
- accuracy
base_model: bert-base-cased
model-index:
- name: Bert-Thesis-NonKFold
  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-Thesis-NonKFold

This model is a fine-tuned version of [bert-base-cased](https://huggingface.co/bert-base-cased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4861
- F1: 0.7464
- Recall: 0.7464
- Accuracy: 0.7464

## 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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10

### Training results

| Training Loss | Epoch | Step  | Validation Loss | F1     | Recall | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:--------:|
| 1.0891        | 1.0   | 1446  | 0.9053          | 0.7222 | 0.7222 | 0.7222   |
| 0.7239        | 2.0   | 2892  | 0.8697          | 0.7397 | 0.7397 | 0.7397   |
| 0.4902        | 3.0   | 4338  | 0.8814          | 0.7491 | 0.7491 | 0.7491   |
| 0.3287        | 4.0   | 5784  | 0.9655          | 0.7512 | 0.7512 | 0.7512   |
| 0.2156        | 5.0   | 7230  | 1.0648          | 0.7450 | 0.7450 | 0.7450   |
| 0.1473        | 6.0   | 8676  | 1.1826          | 0.7446 | 0.7446 | 0.7446   |
| 0.1071        | 7.0   | 10122 | 1.2922          | 0.7465 | 0.7465 | 0.7465   |
| 0.0692        | 8.0   | 11568 | 1.4034          | 0.7483 | 0.7483 | 0.7483   |
| 0.0511        | 9.0   | 13014 | 1.4611          | 0.7478 | 0.7478 | 0.7478   |
| 0.0386        | 10.0  | 14460 | 1.4861          | 0.7464 | 0.7464 | 0.7464   |


### Framework versions

- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3