Sanzana Lora
commited on
Update app.py
Browse files
app.py
CHANGED
@@ -2,20 +2,20 @@ import re
|
|
2 |
import gradio as gr
|
3 |
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
4 |
|
5 |
-
WHITESPACE_HANDLER = lambda k: re.sub('\s+', ' ', re.sub('\n+', ' ', k.strip()))
|
6 |
-
|
7 |
-
# Load the model and tokenizer
|
8 |
-
model_name = "csebuetnlp/mT5_m2m_crossSum"
|
9 |
-
tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
|
10 |
-
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
|
11 |
-
|
12 |
-
get_lang_id = lambda lang: tokenizer._convert_token_to_id(
|
13 |
-
model.config.task_specific_params["langid_map"][lang][1]
|
14 |
-
)
|
15 |
|
16 |
# Function for cross-lingual summarization
|
17 |
def cross_lingual_summarization(article_text, target_language):
|
18 |
target_language = ""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
input_ids = tokenizer(
|
20 |
[WHITESPACE_HANDLER(article_text)],
|
21 |
return_tensors="pt",
|
|
|
2 |
import gradio as gr
|
3 |
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
4 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
|
6 |
# Function for cross-lingual summarization
|
7 |
def cross_lingual_summarization(article_text, target_language):
|
8 |
target_language = ""
|
9 |
+
WHITESPACE_HANDLER = lambda k: re.sub('\s+', ' ', re.sub('\n+', ' ', k.strip()))
|
10 |
+
|
11 |
+
# Load the model and tokenizer
|
12 |
+
model_name = "csebuetnlp/mT5_m2m_crossSum"
|
13 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name, use_fast=False)
|
14 |
+
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
|
15 |
+
|
16 |
+
get_lang_id = lambda lang: tokenizer._convert_token_to_id(
|
17 |
+
model.config.task_specific_params["langid_map"][lang][1]
|
18 |
+
)
|
19 |
input_ids = tokenizer(
|
20 |
[WHITESPACE_HANDLER(article_text)],
|
21 |
return_tensors="pt",
|