Spaces:
Sleeping
Sleeping
File size: 12,873 Bytes
fdadb3e 77ed3a7 fdadb3e 77ed3a7 fdadb3e 77ed3a7 fdadb3e 50b9481 fdadb3e 77ed3a7 |
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 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 |
import gradio as gr
from sentence_transformers import SentenceTransformer
import google.generativeai as genai
import faiss
import json
import time
import os
from dotenv import load_dotenv
import speech_recognition as sr
# Load resources
index = faiss.read_index("health_index.faiss")
with open("health_meta.json") as f:
data = json.load(f)
model = SentenceTransformer("BAAI/bge-base-en-v1.5")
# Set your Gemini API key
load_dotenv()
API_KEY = os.getenv("GEMINI_API")
genai.configure(api_key=API_KEY) # Replace with your actual API key
# Initialize the Gemini model
gen_model = genai.GenerativeModel('gemini-2.5-flash-lite-preview-06-17')
def get_health_response(user_query: str, user_query_org: str, k=3):
query_emb = model.encode([user_query], convert_to_numpy=True)
_, indices = index.search(query_emb, k=k)
context = "\n".join([f"Disease: {data[idx]['disease']}\nSymptoms: {data[idx]['symptoms']}"
f"\nDescription: {data[idx]['description']}"
for idx in indices[0]])
prompt = f"""You are a medical health assistant which answers health related queries or tells the disease based on the symptoms provided
For the given context: {context}
Answer this question in the same language as the question: {user_query_org}
and cite a trustworthy source (like healthline, WebMD, wikipedia or WHO).
Also if you recieve non-medical queries, tell the user to ask only health related queries.
Answer:"""
response = gen_model.generate_content(prompt)
return response.text.strip()
def generate_translation(text):
prompt = f"Give only the most accurate English translation of the given text if it is any other language except English, if the input is already in English return it as it is, nothing else: {text}"
try:
response = gen_model.generate_content(prompt)
return response.text.strip()
except Exception as e:
print(f"Error for {text}")
return "Description not available."
from gradio.themes.utils import fonts
from gradio.themes.base import Base
class HealthTheme(Base):
def __init__(self):
super().__init__(
font=[
fonts.GoogleFont("Asap"),
fonts.Font("ui-sans-serif"),
fonts.Font("sans-serif")
],
font_mono=[
fonts.GoogleFont("Fira Code"),
fonts.Font("ui-monospace"),
fonts.Font("monospace")
]
)
self.set(
body_background_fill="#FFFFFF",
body_background_fill_dark="linear-gradient(to right, #001027, #00112e, #001235, #00123c, #001142)",
background_fill_primary="#FFFFFF",
background_fill_primary_dark="#19191956",
background_fill_secondary="#ECF2F7",
background_fill_secondary_dark="linear-gradient(to right, #000b1a, #000b1e, #000b22, #000b26, #000b2a)",
block_background_fill="#ECF2F7",
block_background_fill_dark="linear-gradient(to right, #000b1a, #000b1e, #000b22, #000b26, #000b2a)",
block_border_color="#dce3e8",
block_border_color_dark="#000431",
button_primary_background_fill="#338AC9",
button_primary_background_fill_dark="#0c6ebd",
button_primary_background_fill_hover="#0c6ebd",
button_primary_background_fill_hover_dark="#000538",
button_primary_text_color="#ECF2F7",
button_primary_text_color_dark="#08003BFF",
button_primary_text_color_hover_dark="#ECF2F7",
input_background_fill="#dce3e8",
input_background_fill_dark="#FF0000FF",
block_label_text_color="#4EACEF",
block_label_text_color_dark="#4EACEF",
block_title_text_color="#4EACEF",
loader_color="#4EACEF",
loader_color_dark="#4EACEF",
body_text_color="#191919",
body_text_color_dark="#ECF2F7",
body_text_color_subdued="#636668",
body_text_color_subdued_dark="#c4c4c4",
body_text_size="*text_md",
body_text_weight="400",
border_color_accent="#dce3e8",
border_color_accent_dark="#242424",
border_color_accent_subdued="#dce3e867",
border_color_accent_subdued_dark="#24242467",
border_color_primary="#dce3e8",
border_color_primary_dark="#242424",
button_border_width="*input_border_width",
button_border_width_dark="*input_border_width",
button_cancel_background_fill="#dce3e8",
button_cancel_background_fill_dark="#242424",
button_cancel_background_fill_hover="#d0d7db",
button_cancel_background_fill_hover_dark="#202020",
button_cancel_border_color="#191919",
button_cancel_border_color_dark="#ECF2F7",
button_cancel_border_color_hover="#202020",
button_cancel_border_color_hover_dark="#a1c3d8",
button_cancel_text_color="#4EACEF",
button_cancel_text_color_dark="#4EACEF",
button_cancel_text_color_hover="#0c6ebd",
button_cancel_text_color_hover_dark="#0c6ebd",
button_large_padding="*spacing_lg calc(2 * *spacing_lg)",
button_large_radius="*radius_lg",
button_large_text_size="*text_lg",
button_large_text_weight="600",
button_primary_border_color="#191919",
button_primary_border_color_dark="#ECF2F7",
button_primary_border_color_hover="#202020",
button_primary_border_color_hover_dark="#a1c3d8",
button_primary_text_color_hover="#e1eaf0",
button_secondary_background_fill="#dce3e8",
button_secondary_background_fill_dark="#040052",
button_secondary_background_fill_hover="#d0d7db",
button_secondary_background_fill_hover_dark="#000644",
button_secondary_border_color="#dce3e8",
button_secondary_border_color_dark="#242424",
button_secondary_border_color_hover="#d0d7db",
button_secondary_border_color_hover_dark="#202020",
button_secondary_text_color ="#4EACEF",
button_secondary_text_color_dark="#4EACEF",
button_secondary_text_color_hover="#0c6ebd",
button_secondary_text_color_hover_dark="#d9eeff",
button_small_padding="*spacing_sm calc(2 * *spacing_sm)",
button_small_radius ="*radius_lg",
button_small_text_size="*text_md",
button_small_text_weight="400",
button_transition ="background-color 0.2s ease",
color_accent="*primary_500",
color_accent_soft="#dce3e8",
color_accent_soft_dark="#0e1834",
container_radius="*radius_lg",
embed_radius="*radius_lg",
error_background_fill="#dce3e8",
error_background_fill_dark="#242424",
error_border_color="#191919",
error_border_color_dark="#ECF2F7",
error_border_width="1px",
error_border_width_dark="1px",
error_icon_color="#b91c1c",
error_icon_color_dark="#ef4444",
error_text_color="#4EACEF",
error_text_color_dark="#4EACEF",
form_gap_width="0px",
input_background_fill_focus="#dce3e8",
input_background_fill_focus_dark="#2F2626",
input_background_fill_hover="#d0d7db",
input_background_fill_hover_dark="#202020",
input_border_color="#191919",
input_border_color_dark="#ECF2F7",
input_border_color_focus="#191919",
input_border_color_focus_dark="#ECF2F7",
input_border_color_hover="#202020",
input_border_color_hover_dark="#a1c3d8",
input_border_width="0px",
input_padding="*spacing_xl",
input_placeholder_color="#19191930",
input_placeholder_color_dark="#FFFFFF4F",
input_radius="*radius_lg",
input_shadow="#19191900",
input_shadow_dark="#ECF2F700",
input_shadow_focus="#19191900",
input_shadow_focus_dark="#ECF2F700",
input_text_size="*text_md",
input_text_weight="400",
layout_gap="*spacing_xxl",
link_text_color="#4EACEF",
link_text_color_active="#4EACEF",
link_text_color_active_dark="#4EACEF",
link_text_color_dark ="#4EACEF",
link_text_color_hover ="#0c6ebd",
link_text_color_hover_dark="#0c6ebd",
link_text_color_visited ="#4EACEF",
link_text_color_visited_dark="#4EACEF",
)
# Use the theme
custom_theme = HealthTheme()
# Add audio transcription function
def transcribe_audio(audio_path):
r = sr.Recognizer()
with sr.AudioFile(audio_path) as source:
audio = r.record(source)
try:
return r.recognize_google(audio)
except sr.UnknownValueError:
return "[Could not understand audio]"
except sr.RequestError:
return "[Audio service error]"
def print_like_dislike(x: gr.LikeData):
print(f"Message #{x.index} was {'liked' if x.liked else 'disliked'}: {x.value}")
# Modified add_message function
def add_message(history, message):
# Process text input
if message["text"]:
history.append({"role": "user", "content": message["text"]})
# Process files (including audio)
if message.get("files"):
for file in message["files"]:
# Transcribe audio files
if file.endswith(('.wav', '.mp3', '.ogg', '.flac')):
transcribed = transcribe_audio(file)
history.append({"role": "user", "content": f"[Audio]: {transcribed}"})
# Handle other file types
else:
history.append({"role": "user", "content": f"[File received: {file}]"})
return history, gr.MultimodalTextbox(value=None, interactive=False)
def bot(history):
# Context window of last N turns
N = 6
memory_context = ""
for turn in history[-N:]:
if isinstance(turn["content"], str):
role = turn["role"]
prefix = "User" if role == "user" else "Assistant"
memory_context += f"{prefix}: {turn['content']}\n"
user_input = history[-1]["content"]
translated = generate_translation(user_input)
full_prompt = f"{memory_context}User: {translated}\nAssistant:"
response = get_health_response(full_prompt, user_input)
history.append({"role": "assistant", "content": ""})
for char in response:
history[-1]['content'] += char
time.sleep(0.02)
yield history
def undo(history):
if len(history) >= 2:
return history[:-2]
return []
def retry(history):
if len(history) >= 2:
last_user = history[-2]
history = history[:-2] + [last_user]
return history
return history
# --- UI Setup ---
with gr.Blocks(theme = custom_theme) as demo:
gr.Markdown("""<h1 style='font-weight:600;'>🩺 Clinikit</h1>
<p style='color:#666;font-size:15px'>Ask health-related questions or enter symptoms below. Built with memory, streaming, multilingual text support and voice inputs(english only).</p>""")
chatbot = gr.Chatbot(
label="Assistant",
type="messages",
avatar_images=(None, "https://e7.pngegg.com/pngimages/369/865/png-clipart-physician-hospital-dr-mary-c-kirk-md-doctor-of-medicine-computer-icons-the-doctor-miscellaneous-black-thumbnail.png")
)
msg = gr.MultimodalTextbox(
interactive=True,
file_count="multiple",
placeholder="Enter symptoms or ask a health question...",
show_label=False,
sources=["microphone"]
)
with gr.Row():
retry_btn = gr.Button("🔁 Retry")
undo_btn = gr.Button("↩️ Undo")
chat_msg = msg.submit(add_message, [chatbot, msg], [chatbot, msg])
bot_msg = chat_msg.then(bot, chatbot, chatbot)
bot_msg.then(lambda: gr.MultimodalTextbox(interactive=True), None, [msg])
retry_btn.click(retry, chatbot, chatbot).then(bot, chatbot, chatbot).then(lambda h: h, chatbot, chatbot)
undo_btn.click(undo, chatbot, chatbot).then(lambda h: h, chatbot, chatbot)
chatbot.like(print_like_dislike, None, None, like_user_message=True)
gr.Markdown("""
<footer style='text-align:center; margin-top:20px; color:#aaa;'>
Built using Gradio, Hugging Face & Mistral |
<a href="https://github.com/kumardevansh/clinikit" target="_blank" style="color:#aaa; text-decoration:underline;">
View on GitHub
</a>
</footer>
""")
demo.launch(share=True, pwa=True) |