File size: 2,252 Bytes
e072175
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fcc681d
ead7c7a
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
88
89
90
import gradio as gr
import re

from gradio.mix import Parallel
from transformers import (
    AutoTokenizer,
    AutoModelForSeq2SeqLM,
)

def clean_text(text):
    text = text.encode("ascii", errors="ignore").decode(
        "ascii"
    )  # remove non-ascii, Chinese characters
    text = re.sub(r"\n", " ", text)
    text = re.sub(r"\n\n", " ", text)
    text = re.sub(r"\t", " ", text)
    text = text.strip(" ")
    text = re.sub(
        " +", " ", text
    ).strip()  # get rid of multiple spaces and replace with a single
    return text


modchoice_1 = "chinhon/pegasus-newsroom-malay_headlines"

def headline_writer1(text):
    input_text = clean_text(text)

    tokenizer_1 = AutoTokenizer.from_pretrained(modchoice_1)

    model_1 = AutoModelForSeq2SeqLM.from_pretrained(modchoice_1)

    with tokenizer_1.as_target_tokenizer():
        batch = tokenizer_1(
            input_text, truncation=True, padding="longest", return_tensors="pt"
        )

    translated = model_1.generate(**batch)

    summary_1 = tokenizer_1.batch_decode(translated, skip_special_tokens=True)

    return summary_1[0]


headline1 = gr.Interface(
    fn=headline_writer1,
    inputs=gr.inputs.Textbox(),
    outputs=gr.outputs.Textbox(label=""),
)


modchoice_2 = "chinhon/pegasus-multi_news-malay_headlines_02"

def headline_writer2(text):
    input_text = clean_text(text)

    tokenizer_2 = AutoTokenizer.from_pretrained(modchoice_2)

    model_2 = AutoModelForSeq2SeqLM.from_pretrained(modchoice_2)

    with tokenizer_2.as_target_tokenizer():
        batch = tokenizer_2(
            input_text, truncation=True, padding="longest", return_tensors="pt"
        )

    translated = model_2.generate(**batch)

    summary_2 = tokenizer_2.batch_decode(translated, skip_special_tokens=True)

    return summary_2[0]


headline2 = gr.Interface(
    fn=headline_writer2,
    inputs=gr.inputs.Textbox(),
    outputs=gr.outputs.Textbox(label=""),
)


Parallel(
    headline1,
    headline2,
    title="Malay News Headlines Generator",
    inputs=gr.inputs.Textbox(
        lines=20,
        label="Paste the first few paragraphs of a Malay language news story here, and choose from 2 suggested headlines",
    ),
    theme="darkdefault",
).launch(enable_queue=True)