File size: 1,904 Bytes
e98995e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e4474d1
 
 
 
 
e98995e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9df9460
 
 
 
 
e98995e
 
 
 
 
 
 
 
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
---
base_model: dmis-lab/biobert-v1.1
tags:
- generated_from_trainer
metrics:
- precision
- recall
- accuracy
- f1
model-index:
- name: biobert-v1.1-text-classifier
  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. -->

# biobert-v1.1-text-classifier

This model is a fine-tuned version of [dmis-lab/biobert-v1.1](https://huggingface.co/dmis-lab/biobert-v1.1) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2669
- Precision: 0.9098
- Recall: 0.9091
- Accuracy: 0.9089
- F1: 0.9089

## 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: 5

### Training results

| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | Accuracy | F1     |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:--------:|:------:|
| No log        | 1.0   | 154  | 0.3413          | 0.8822    | 0.8813 | 0.8804   | 0.8808 |
| No log        | 2.0   | 308  | 0.2918          | 0.8945    | 0.8836 | 0.8845   | 0.8848 |
| No log        | 3.0   | 462  | 0.2669          | 0.9098    | 0.9091 | 0.9089   | 0.9089 |
| 0.3597        | 4.0   | 616  | 0.2781          | 0.9175    | 0.9174 | 0.9170   | 0.9170 |
| 0.3597        | 5.0   | 770  | 0.2797          | 0.9203    | 0.9206 | 0.9203   | 0.9204 |


### Framework versions

- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2