Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
| 3 |
+
|
| 4 |
+
languages = {
|
| 5 |
+
'Afrikaans': 'afr_Latn',
|
| 6 |
+
'Albanian': 'als_Latn',
|
| 7 |
+
'Basque': 'eus_Latn',
|
| 8 |
+
'Belarusian': 'bel_Cyrl',
|
| 9 |
+
'Bosnian': 'bos_Latn',
|
| 10 |
+
'Bulgarian': 'bul_Cyrl',
|
| 11 |
+
'Catalan': 'cat_Latn',
|
| 12 |
+
'Croatian': 'hrv_Latn',
|
| 13 |
+
'Czech': 'ces_Latn',
|
| 14 |
+
'Danish': 'dan_Latn',
|
| 15 |
+
'Dutch': 'nld_Latn',
|
| 16 |
+
'English': 'eng_Latn',
|
| 17 |
+
'Estonian': 'est_Latn',
|
| 18 |
+
'Finnish': 'fin_Latn',
|
| 19 |
+
'French': 'fra_Latn',
|
| 20 |
+
'Galician': 'glg_Latn',
|
| 21 |
+
'German': 'deu_Latn',
|
| 22 |
+
'Greek': 'ell_Grek',
|
| 23 |
+
'Hungarian': 'hun_Latn',
|
| 24 |
+
'Icelandic': 'isl_Latn',
|
| 25 |
+
'Irish': 'gle_Latn',
|
| 26 |
+
'Italian': 'ita_Latn',
|
| 27 |
+
'Lithuanian': 'lit_Latn',
|
| 28 |
+
'Luxembourgish': 'ltz_Latn',
|
| 29 |
+
'Macedonian': 'mkd_Cyrl',
|
| 30 |
+
'Maltese': 'mlt_Latn',
|
| 31 |
+
'Norwegian Bokmål': 'nob_Latn',
|
| 32 |
+
'Norwegian Nynorsk': 'nno_Latn',
|
| 33 |
+
'Polish': 'pol_Latn',
|
| 34 |
+
'Portuguese': 'por_Latn',
|
| 35 |
+
'Romanian': 'ron_Latn',
|
| 36 |
+
'Russian': 'rus_Cyrl',
|
| 37 |
+
'Serbian': 'srp_Cyrl',
|
| 38 |
+
'Slovak': 'slk_Latn',
|
| 39 |
+
'Slovenian': 'slv_Latn',
|
| 40 |
+
'Spanish': 'spa_Latn',
|
| 41 |
+
'Swedish': 'swe_Latn',
|
| 42 |
+
'Ukrainian': 'ukr_Cyrl',
|
| 43 |
+
'Welsh': 'cym_Latn'
|
| 44 |
+
}
|
| 45 |
+
|
| 46 |
+
def translate_article(article, language):
|
| 47 |
+
tokenizer = AutoTokenizer.from_pretrained("facebook/nllb-200-distilled-600M")
|
| 48 |
+
model = AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M")
|
| 49 |
+
inputs = tokenizer(article, return_tensors="pt")
|
| 50 |
+
lang_code = languages[language]
|
| 51 |
+
translated_tokens = model.generate(**inputs, forced_bos_token_id=tokenizer.lang_code_to_id[lang_code])
|
| 52 |
+
result = tokenizer.batch_decode(translated_tokens, skip_special_tokens=True)[0]
|
| 53 |
+
return result
|
| 54 |
+
|
| 55 |
+
language_choices = list(languages.keys())
|
| 56 |
+
|
| 57 |
+
iface = gr.Interface(
|
| 58 |
+
fn=translate_article,
|
| 59 |
+
inputs=["text",gr.Dropdown(
|
| 60 |
+
language_choices, value=["English"], multiselect=False, label="Choose the language.")],
|
| 61 |
+
outputs="text",
|
| 62 |
+
title="Article Translator"
|
| 63 |
+
)
|
| 64 |
+
iface.launch()
|