File size: 1,261 Bytes
5840657
 
a572b3d
 
 
 
 
 
336f4ff
4382ddc
 
1cf6297
 
8b678e0
1cf6297
8b678e0
4f60dd3
 
 
 
8b678e0
 
 
 
 
 
 
 
 
4f60dd3
8b678e0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
datasets:
- webis/tldr-17
language:
- en
library_name: transformers
pipeline_tag: text-classification
widget:
- text: "Biden says US is at tipping point on gun control: We will ban assault weapons in this country" 
  example_title: "classification"
---

## Reddit post classification

This model predicts the subreddit of a provided post  
The transformers library is required
```
pip install 'transformers[torch]'
```

```py
from transformers import pipeline
pipe = pipeline('text-classification', model='traberph/RedBERT')
pipe("Biden says US is at tipping point on gun control: We will ban assault weapons in this country")
```

## Class Labels

To translate the labels back to subreddit names you need to download the `subreddits.json` file from this repo manually

```py
import json
s_count = 0
s_data = []
with open('subreddits.json', 'r') as file:
    s_data = json.load(file)
    s_count = len(s_data)
labels = list(s_data.keys()) 

def translate(d):
    d['label'] = s_data[ labels[ int( d['label'].split('_')[1]) ]]
    return d
```

Now the class labels can be translated back to subreddits 

```py
list(map(translate, pipe("Biden says US is at tipping point on gun control: We will ban assault weapons in this country")))
```