Update app.py
Browse files
app.py
CHANGED
@@ -85,7 +85,8 @@ def bansum(text):
|
|
85 |
|
86 |
if "photo" not in st.session_state:
|
87 |
st.session_state["photo"]="not done"
|
88 |
-
c2, c3 = st.columns([
|
|
|
89 |
def change_photo_state():
|
90 |
st.session_state["photo"]="done"
|
91 |
@st.cache
|
@@ -93,71 +94,108 @@ def save(l):
|
|
93 |
return l
|
94 |
#@st.cache
|
95 |
def main():
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
105 |
|
106 |
-
|
107 |
-
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
-
|
112 |
-
|
113 |
-
|
114 |
-
|
115 |
-
|
116 |
-
|
117 |
-
|
118 |
-
|
119 |
-
|
120 |
-
|
121 |
-
|
122 |
-
|
123 |
-
|
124 |
-
|
125 |
-
|
126 |
-
|
127 |
-
|
128 |
-
|
129 |
-
|
130 |
-
|
131 |
-
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
|
138 |
-
|
139 |
-
|
140 |
-
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
|
145 |
-
|
146 |
-
text = pytesseract.image_to_string(img, lang="ben")
|
147 |
-
st.subheader("সারাংশ/সারমর্ম")
|
148 |
-
bansum(text)
|
149 |
-
if c7.button("English"):
|
150 |
-
text=pytesseract.image_to_string(img)
|
151 |
-
st.subheader("Summarized Text")
|
152 |
-
engsum(text)
|
153 |
-
else:
|
154 |
-
text=None
|
155 |
-
text = message
|
156 |
-
c8, c9 = st.columns([1,1])
|
157 |
-
if c8.button("Bangla"):
|
158 |
-
bansum(text)
|
159 |
-
if c9.button("English"):
|
160 |
-
engsum(text)
|
161 |
# if st.button("English Text Generation"):
|
162 |
# def query(payload):
|
163 |
# response = requests.post(API_URL2, headers=headers2, json=payload)
|
|
|
85 |
|
86 |
if "photo" not in st.session_state:
|
87 |
st.session_state["photo"]="not done"
|
88 |
+
c2, c3 = st.columns([1,1])
|
89 |
+
a, b = st.columns([1, 1])
|
90 |
def change_photo_state():
|
91 |
st.session_state["photo"]="done"
|
92 |
@st.cache
|
|
|
94 |
return l
|
95 |
#@st.cache
|
96 |
def main():
|
97 |
+
with st.container():
|
98 |
+
with a:
|
99 |
+
#import torch
|
100 |
+
import streamlit as st
|
101 |
+
from streamlit_option_menu import option_menu
|
102 |
+
from streamlit_chat import message as st_message
|
103 |
+
from transformers import BlenderbotTokenizer
|
104 |
+
from transformers import BlenderbotForConditionalGeneration
|
105 |
+
st.title("Simple Chatbot for fun!")
|
106 |
+
|
107 |
+
@st.experimental_singleton
|
108 |
+
def get_models():
|
109 |
+
# it may be necessary for other frameworks to cache the model
|
110 |
+
# seems pytorch keeps an internal state of the conversation
|
111 |
+
model_name = "facebook/blenderbot-400M-distill"
|
112 |
+
tokenizer = BlenderbotTokenizer.from_pretrained(model_name)
|
113 |
+
model = BlenderbotForConditionalGeneration.from_pretrained(model_name)
|
114 |
+
return tokenizer, model
|
115 |
+
if "history" not in st.session_state:
|
116 |
+
st.session_state.history = []
|
117 |
+
st.title("Hello Chatbot")
|
118 |
+
def main():
|
119 |
+
st.text_input("Talk to the bot", key="input_text", on_change=generate_answer)
|
120 |
+
def generate_answer():
|
121 |
+
tokenizer, model = get_models()
|
122 |
+
user_message = st.session_state.input_text
|
123 |
+
inputs = tokenizer(st.session_state.input_text, return_tensors="pt")
|
124 |
+
result = model.generate(**inputs)
|
125 |
+
message_bot = tokenizer.decode(
|
126 |
+
result[0], skip_special_tokens=True
|
127 |
+
) # .replace("<s>", "").replace("</s>", "")
|
128 |
+
st.session_state.history.append({"message": user_message, "is_user": True})
|
129 |
+
st.session_state.history.append({"message": message_bot, "is_user": False})
|
130 |
+
from copyreg import clear_extension_cache
|
131 |
+
for chat in st.session_state.history:
|
132 |
+
st_message(**chat)
|
133 |
+
with b:
|
134 |
+
message = st.text_input("Type your text here!")
|
135 |
+
camera_photo = c2.camera_input("Capture a photo to summarize: ", on_change=change_photo_state)
|
136 |
+
uploaded_photo = save(c3.file_uploader("Upload your Images/PDF",type=['jpg','png','jpeg','pdf'], on_change=change_photo_state))
|
137 |
+
if st.session_state["photo"]=="done" or message:
|
138 |
+
if uploaded_photo and uploaded_photo.type=='application/pdf':
|
139 |
+
tet = read_pdf(uploaded_photo)
|
140 |
+
# with tempfile.NamedTemporaryFile(delete=False) as temp_file:
|
141 |
+
# temp_file.write(uploaded_photo.read())
|
142 |
+
# temp_file_path = temp_file.name
|
143 |
+
|
144 |
+
# loader = PyPDFLoader(temp_file_path)
|
145 |
+
# if loader:
|
146 |
+
# text.extend(loader.load())
|
147 |
+
# os.remove(temp_file_path)
|
148 |
+
# text_splitter = CharacterTextSplitter(separator="\n", chunk_size=1000, chunk_overlap=100, length_function=len)
|
149 |
+
# text_chunks = text_splitter.split_documents(text)
|
150 |
+
values = st.slider('Select a approximate number of lines to see and summarize',value=[0, len(tet)//(7*100)])
|
151 |
+
text = tet[values[0]*7*10:values[1]*10*100] if values[0]!=len(tet)//(10*100) else tet[len(tet)//(10*100):]
|
152 |
+
#st.success(type(text_chunks))
|
153 |
+
if st.button("English Pdf Summarize"):
|
154 |
+
st.subheader("Selected text for summarize: ")
|
155 |
+
st.success(text)
|
156 |
+
st.subheader("Summarized Text: ")
|
157 |
+
engsum(text)
|
158 |
|
159 |
+
elif uploaded_photo and uploaded_photo.type !='application/pdf':
|
160 |
+
text=None
|
161 |
+
img = Image.open(uploaded_photo)
|
162 |
+
img = img.save("img.png")
|
163 |
+
img = cv2.imread("img.png")
|
164 |
+
st.text("Select the summarization type:")
|
165 |
+
c4, c5 = st.columns([1,1])
|
166 |
+
if c4.button("BENGALI"):
|
167 |
+
text = pytesseract.image_to_string(img, lang="ben")
|
168 |
+
st.subheader("সারাংশ/সারমর্ম")
|
169 |
+
bansum(text)
|
170 |
+
if c5.button("ENGLISH"):
|
171 |
+
text=pytesseract.image_to_string(img)
|
172 |
+
st.subheader("Summarized Text")
|
173 |
+
engsum(text)
|
174 |
+
#st.success(text)
|
175 |
+
elif camera_photo:
|
176 |
+
text=None
|
177 |
+
img = Image.open(camera_photo)
|
178 |
+
img = img.save("img.png")
|
179 |
+
img = cv2.imread("img.png")
|
180 |
+
#text = pytesseract.image_to_string(img) if st.checkbox("Bangla") else pytesseract.image_to_string(img, lang="ben")
|
181 |
+
st.text("Select the summarization type:")
|
182 |
+
c6, c7 = st.columns([1,1])
|
183 |
+
if c6.button("Bangla"):
|
184 |
+
text = pytesseract.image_to_string(img, lang="ben")
|
185 |
+
st.subheader("সারাংশ/সারমর্ম")
|
186 |
+
bansum(text)
|
187 |
+
if c7.button("English"):
|
188 |
+
text=pytesseract.image_to_string(img)
|
189 |
+
st.subheader("Summarized Text")
|
190 |
+
engsum(text)
|
191 |
+
else:
|
192 |
+
text=None
|
193 |
+
text = message
|
194 |
+
c8, c9 = st.columns([1,1])
|
195 |
+
if c8.button("Bangla"):
|
196 |
+
bansum(text)
|
197 |
+
if c9.button("English"):
|
198 |
+
engsum(text)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
199 |
# if st.button("English Text Generation"):
|
200 |
# def query(payload):
|
201 |
# response = requests.post(API_URL2, headers=headers2, json=payload)
|