Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,82 +1,108 @@
|
|
1 |
import gradio as gr
|
2 |
-
import
|
3 |
import os
|
4 |
-
import
|
|
|
5 |
import pdfplumber
|
6 |
import docx
|
7 |
import pandas as pd
|
8 |
from PIL import Image
|
9 |
|
10 |
-
# Load Whisper model
|
11 |
whisper_model = whisper.load_model("base")
|
12 |
|
13 |
-
|
14 |
-
|
15 |
-
return "⚠️ No audio file uploaded."
|
16 |
|
17 |
-
|
18 |
-
|
19 |
-
return result["text"]
|
20 |
-
except Exception as e:
|
21 |
-
return f"❌ Whisper error: {str(e)}"
|
22 |
|
|
|
23 |
def extract_text_from_file(file):
|
24 |
if file is None:
|
25 |
-
return "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
|
27 |
-
|
28 |
-
|
|
|
29 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
try:
|
31 |
-
|
32 |
-
|
33 |
-
text = "\n".join(page.extract_text() or "" for page in pdf.pages)
|
34 |
-
elif ext in [".docx"]:
|
35 |
-
doc = docx.Document(file)
|
36 |
-
text = "\n".join(paragraph.text for paragraph in doc.paragraphs)
|
37 |
-
elif ext in [".xlsx", ".xls"]:
|
38 |
-
df = pd.read_excel(file)
|
39 |
-
text = df.to_string(index=False)
|
40 |
-
elif ext in [".png", ".jpg", ".jpeg"]:
|
41 |
-
image = Image.open(file)
|
42 |
-
text = "🖼️ Image uploaded. Please describe what you'd like me to do with it."
|
43 |
-
else:
|
44 |
-
text = "⚠️ Unsupported file type."
|
45 |
-
return text
|
46 |
except Exception as e:
|
47 |
-
return f"❌
|
48 |
-
|
49 |
-
def chatbot_response(message, chat_history):
|
50 |
-
# Echo-style placeholder response
|
51 |
-
bot_reply = f"🤖 You said: {message}"
|
52 |
-
chat_history.append((message, bot_reply))
|
53 |
-
return "", chat_history
|
54 |
-
|
55 |
-
with gr.Blocks(css="""
|
56 |
-
#chatbox { height: 500px; overflow: auto; }
|
57 |
-
.message-input { height: 40px; border-radius: 6px; }
|
58 |
-
.sidebar { width: 25%; overflow-y: auto; }
|
59 |
-
.main { width: 75%; }
|
60 |
-
""") as demo:
|
61 |
-
gr.Markdown("""<h2 style="text-align:center;">🤖 Neobot - Always Listening</h2>""")
|
62 |
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
68 |
|
69 |
with gr.Row():
|
70 |
-
|
71 |
-
|
72 |
|
73 |
with gr.Row():
|
74 |
-
|
75 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
76 |
|
77 |
-
|
78 |
-
|
79 |
-
record_btn.change(fn=transcribe_audio, inputs=record_btn, outputs=msg)
|
80 |
-
upload_btn.change(fn=extract_text_from_file, inputs=upload_btn, outputs=msg)
|
81 |
|
|
|
82 |
demo.launch()
|
|
|
1 |
import gradio as gr
|
2 |
+
import openai
|
3 |
import os
|
4 |
+
import tempfile
|
5 |
+
import whisper
|
6 |
import pdfplumber
|
7 |
import docx
|
8 |
import pandas as pd
|
9 |
from PIL import Image
|
10 |
|
11 |
+
# Load Whisper model
|
12 |
whisper_model = whisper.load_model("base")
|
13 |
|
14 |
+
# Set OpenAI API key
|
15 |
+
openai.api_key = os.getenv("OPENAI_API_KEY")
|
|
|
16 |
|
17 |
+
# Global store for chat messages
|
18 |
+
chat_history = []
|
|
|
|
|
|
|
19 |
|
20 |
+
# --- File Processor ---
|
21 |
def extract_text_from_file(file):
|
22 |
if file is None:
|
23 |
+
return ""
|
24 |
+
filename = file.name
|
25 |
+
ext = os.path.splitext(filename)[-1].lower()
|
26 |
+
|
27 |
+
if ext == ".pdf":
|
28 |
+
with pdfplumber.open(file) as pdf:
|
29 |
+
return "\n".join(page.extract_text() for page in pdf.pages if page.extract_text())
|
30 |
+
|
31 |
+
elif ext in [".doc", ".docx"]:
|
32 |
+
doc = docx.Document(file)
|
33 |
+
return "\n".join([para.text for para in doc.paragraphs])
|
34 |
|
35 |
+
elif ext in [".xls", ".xlsx"]:
|
36 |
+
df = pd.read_excel(file)
|
37 |
+
return df.to_string(index=False)
|
38 |
|
39 |
+
elif ext in [".png", ".jpg", ".jpeg"]:
|
40 |
+
image = Image.open(file)
|
41 |
+
return "Image uploaded. Please ask a question about the image."
|
42 |
+
|
43 |
+
else:
|
44 |
+
return file.read().decode("utf-8")
|
45 |
+
|
46 |
+
|
47 |
+
# --- Transcription ---
|
48 |
+
def transcribe_audio(audio_file):
|
49 |
+
if audio_file is None:
|
50 |
+
return ""
|
51 |
try:
|
52 |
+
result = whisper_model.transcribe(audio_file)
|
53 |
+
return result['text']
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
54 |
except Exception as e:
|
55 |
+
return f"❌ Transcription error: {str(e)}"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
|
57 |
+
|
58 |
+
# --- Chatbot Logic ---
|
59 |
+
def chat_with_gpt(message, uploaded_file):
|
60 |
+
file_text = extract_text_from_file(uploaded_file)
|
61 |
+
|
62 |
+
system_prompt = """
|
63 |
+
You are Neobot, an intelligent assistant that can answer questions, transcribe audio, and analyze uploaded documents.
|
64 |
+
"""
|
65 |
+
|
66 |
+
user_prompt = message
|
67 |
+
if file_text:
|
68 |
+
user_prompt = f"This is the content of the uploaded file:\n{file_text}\n\nUser's question: {message}"
|
69 |
+
|
70 |
+
chat_history.append({"role": "user", "content": user_prompt})
|
71 |
+
|
72 |
+
try:
|
73 |
+
response = openai.ChatCompletion.create(
|
74 |
+
model="gpt-4o",
|
75 |
+
messages=[{"role": "system", "content": system_prompt}] + chat_history
|
76 |
+
)
|
77 |
+
reply = response.choices[0].message.content
|
78 |
+
chat_history.append({"role": "assistant", "content": reply})
|
79 |
+
return [(msg['content'], None) if msg['role'] == 'user' else (None, msg['content']) for msg in chat_history]
|
80 |
+
except Exception as e:
|
81 |
+
return [[(None, f"❌ Error: {str(e)}")]]
|
82 |
+
|
83 |
+
|
84 |
+
# --- Gradio UI ---
|
85 |
+
with gr.Blocks(css="body { background-color: white; color: black; }") as demo:
|
86 |
+
gr.Markdown("""<h1 style='text-align: center;'>Neobot</h1>""")
|
87 |
+
chatbot = gr.Chatbot(type="messages", elem_id="chatbox")
|
88 |
|
89 |
with gr.Row():
|
90 |
+
msg_box = gr.Textbox(placeholder="Type a message...", show_label=False, scale=8)
|
91 |
+
record_btn = gr.Audio(type="filepath", label="Record Voice", show_label=False, scale=2)
|
92 |
|
93 |
with gr.Row():
|
94 |
+
upload_file = gr.File(label="Upload File", file_types=[".pdf", ".docx", ".xlsx", ".txt", ".png", ".jpg"])
|
95 |
+
|
96 |
+
# --- Callbacks ---
|
97 |
+
def handle_voice_input(audio):
|
98 |
+
text = transcribe_audio(audio)
|
99 |
+
return text
|
100 |
+
|
101 |
+
def respond(message, file):
|
102 |
+
return chat_with_gpt(message, file)
|
103 |
|
104 |
+
record_btn.change(handle_voice_input, inputs=record_btn, outputs=msg_box)
|
105 |
+
msg_box.submit(respond, [msg_box, upload_file], chatbot)
|
|
|
|
|
106 |
|
107 |
+
# --- Launch ---
|
108 |
demo.launch()
|