Update app.py
Browse files
app.py
CHANGED
@@ -8,15 +8,17 @@ import re
|
|
8 |
import os
|
9 |
import time
|
10 |
import gc
|
|
|
11 |
from huggingface_hub import hf_hub_download
|
12 |
from llama_cpp import Llama
|
13 |
|
14 |
# Configuration variables that can be set through environment variables
|
|
|
15 |
MODEL_REPO_ID = os.getenv("MODEL_REPO_ID", "mradermacher/Llama3-Med42-8B-GGUF")
|
16 |
MODEL_FILENAME = os.getenv("MODEL_FILENAME", "Llama3-Med42-8B.Q5_K_M.gguf")
|
17 |
N_THREADS = int(os.getenv("N_THREADS", "4"))
|
18 |
|
19 |
-
#
|
20 |
class ConsultationState(Enum):
|
21 |
INITIAL = "initial"
|
22 |
GATHERING_INFO = "gathering_info"
|
@@ -33,7 +35,7 @@ class ChatResponse(BaseModel):
|
|
33 |
response: str
|
34 |
finished: bool
|
35 |
|
36 |
-
#
|
37 |
HEALTH_ASSESSMENT_QUESTIONS = [
|
38 |
"What are your current symptoms and how long have you been experiencing them?",
|
39 |
"Do you have any pre-existing medical conditions or chronic illnesses?",
|
@@ -42,7 +44,7 @@ HEALTH_ASSESSMENT_QUESTIONS = [
|
|
42 |
"Have you had any similar symptoms in the past? If yes, what treatments worked?"
|
43 |
]
|
44 |
|
45 |
-
#
|
46 |
NURSE_OGE_IDENTITY = """
|
47 |
You are Nurse Oge, a medical AI assistant focused on serving patients in Nigeria. Always be empathetic,
|
48 |
professional, and thorough in your assessments. When asked about your identity, explain that you are
|
@@ -51,32 +53,29 @@ health information before providing any medical advice.
|
|
51 |
"""
|
52 |
|
53 |
class NurseOgeAssistant:
|
|
|
|
|
|
|
54 |
def __init__(self):
|
55 |
try:
|
56 |
-
#
|
57 |
-
|
58 |
repo_id=MODEL_REPO_ID,
|
59 |
filename=MODEL_FILENAME,
|
60 |
-
resume_download=True
|
61 |
-
)
|
62 |
-
|
63 |
-
# Initialize the Llama model with appropriate parameters
|
64 |
-
self.llm = Llama(
|
65 |
-
model_path=model_path,
|
66 |
n_ctx=2048, # Context window size
|
67 |
n_threads=N_THREADS, # CPU threads to use
|
68 |
-
n_gpu_layers=0
|
69 |
-
verbose=False # Set to True for debugging
|
70 |
)
|
71 |
|
72 |
except Exception as e:
|
73 |
raise RuntimeError(f"Failed to initialize the model: {str(e)}")
|
74 |
|
75 |
-
#
|
76 |
self.consultation_states = {}
|
77 |
self.gathered_info = {}
|
78 |
|
79 |
def _is_identity_question(self, message: str) -> bool:
|
|
|
80 |
identity_patterns = [
|
81 |
r"who are you",
|
82 |
r"what are you",
|
@@ -87,6 +86,7 @@ class NurseOgeAssistant:
|
|
87 |
return any(re.search(pattern, message.lower()) for pattern in identity_patterns)
|
88 |
|
89 |
def _is_location_question(self, message: str) -> bool:
|
|
|
90 |
location_patterns = [
|
91 |
r"where are you",
|
92 |
r"which country",
|
@@ -97,6 +97,7 @@ class NurseOgeAssistant:
|
|
97 |
return any(re.search(pattern, message.lower()) for pattern in location_patterns)
|
98 |
|
99 |
def _get_next_assessment_question(self, conversation_id: str) -> Optional[str]:
|
|
|
100 |
if conversation_id not in self.gathered_info:
|
101 |
self.gathered_info[conversation_id] = []
|
102 |
|
@@ -106,6 +107,9 @@ class NurseOgeAssistant:
|
|
106 |
return None
|
107 |
|
108 |
async def process_message(self, conversation_id: str, message: str, history: List[Dict]) -> ChatResponse:
|
|
|
|
|
|
|
109 |
try:
|
110 |
# Initialize state for new conversations
|
111 |
if conversation_id not in self.consultation_states:
|
@@ -159,7 +163,7 @@ class NurseOgeAssistant:
|
|
159 |
# Prepare messages for the model
|
160 |
messages = [
|
161 |
{"role": "system", "content": NURSE_OGE_IDENTITY},
|
162 |
-
{"role": "user", "content": f"Based on the following patient information, provide
|
163 |
]
|
164 |
|
165 |
# Implement retry logic for model inference
|
@@ -200,37 +204,41 @@ class NurseOgeAssistant:
|
|
200 |
finished=True
|
201 |
)
|
202 |
|
203 |
-
#
|
204 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
205 |
|
206 |
-
#
|
207 |
-
|
208 |
|
209 |
# Add memory management middleware
|
210 |
@app.middleware("http")
|
211 |
async def add_memory_management(request: Request, call_next):
|
|
|
212 |
gc.collect()
|
213 |
response = await call_next(request)
|
214 |
gc.collect()
|
215 |
return response
|
216 |
|
217 |
-
# Initialize the assistant during startup
|
218 |
-
@app.on_event("startup")
|
219 |
-
async def startup_event():
|
220 |
-
global nurse_oge
|
221 |
-
try:
|
222 |
-
nurse_oge = NurseOgeAssistant()
|
223 |
-
except Exception as e:
|
224 |
-
print(f"Failed to initialize NurseOgeAssistant: {e}")
|
225 |
-
|
226 |
# Health check endpoint
|
227 |
@app.get("/health")
|
228 |
async def health_check():
|
|
|
229 |
return {"status": "healthy", "model_loaded": nurse_oge is not None}
|
230 |
|
231 |
# Chat endpoint
|
232 |
@app.post("/chat")
|
233 |
async def chat_endpoint(request: ChatRequest):
|
|
|
234 |
if nurse_oge is None:
|
235 |
raise HTTPException(
|
236 |
status_code=503,
|
@@ -251,14 +259,15 @@ async def chat_endpoint(request: ChatRequest):
|
|
251 |
return response
|
252 |
|
253 |
# Gradio chat interface function
|
254 |
-
def gradio_chat(message, history):
|
|
|
255 |
if nurse_oge is None:
|
256 |
return "The medical assistant is not available at the moment. Please try again later."
|
257 |
|
258 |
-
response = nurse_oge.process_message("gradio_user", message, history)
|
259 |
return response.response
|
260 |
|
261 |
-
# Create and configure Gradio interface
|
262 |
demo = gr.ChatInterface(
|
263 |
fn=gradio_chat,
|
264 |
title="Nurse Oge - Medical Assistant",
|
@@ -272,10 +281,7 @@ demo = gr.ChatInterface(
|
|
272 |
theme=gr.themes.Soft(
|
273 |
primary_hue="blue",
|
274 |
secondary_hue="purple",
|
275 |
-
)
|
276 |
-
retry_btn="Try Again",
|
277 |
-
undo_btn="Undo Last",
|
278 |
-
clear_btn="Clear Chat"
|
279 |
)
|
280 |
|
281 |
# Add custom CSS for better appearance
|
|
|
8 |
import os
|
9 |
import time
|
10 |
import gc
|
11 |
+
from contextlib import asynccontextmanager
|
12 |
from huggingface_hub import hf_hub_download
|
13 |
from llama_cpp import Llama
|
14 |
|
15 |
# Configuration variables that can be set through environment variables
|
16 |
+
# These allow for flexible deployment configuration without code changes
|
17 |
MODEL_REPO_ID = os.getenv("MODEL_REPO_ID", "mradermacher/Llama3-Med42-8B-GGUF")
|
18 |
MODEL_FILENAME = os.getenv("MODEL_FILENAME", "Llama3-Med42-8B.Q5_K_M.gguf")
|
19 |
N_THREADS = int(os.getenv("N_THREADS", "4"))
|
20 |
|
21 |
+
# Data models for API request/response handling
|
22 |
class ConsultationState(Enum):
|
23 |
INITIAL = "initial"
|
24 |
GATHERING_INFO = "gathering_info"
|
|
|
35 |
response: str
|
36 |
finished: bool
|
37 |
|
38 |
+
# Standardized health assessment questions for consistent patient evaluation
|
39 |
HEALTH_ASSESSMENT_QUESTIONS = [
|
40 |
"What are your current symptoms and how long have you been experiencing them?",
|
41 |
"Do you have any pre-existing medical conditions or chronic illnesses?",
|
|
|
44 |
"Have you had any similar symptoms in the past? If yes, what treatments worked?"
|
45 |
]
|
46 |
|
47 |
+
# AI assistant's identity and role definition
|
48 |
NURSE_OGE_IDENTITY = """
|
49 |
You are Nurse Oge, a medical AI assistant focused on serving patients in Nigeria. Always be empathetic,
|
50 |
professional, and thorough in your assessments. When asked about your identity, explain that you are
|
|
|
53 |
"""
|
54 |
|
55 |
class NurseOgeAssistant:
|
56 |
+
"""
|
57 |
+
Main assistant class that handles conversation management and medical consultations
|
58 |
+
"""
|
59 |
def __init__(self):
|
60 |
try:
|
61 |
+
# Initialize the Llama model using from_pretrained as per documentation
|
62 |
+
self.llm = Llama.from_pretrained(
|
63 |
repo_id=MODEL_REPO_ID,
|
64 |
filename=MODEL_FILENAME,
|
|
|
|
|
|
|
|
|
|
|
|
|
65 |
n_ctx=2048, # Context window size
|
66 |
n_threads=N_THREADS, # CPU threads to use
|
67 |
+
n_gpu_layers=0 # CPU-only inference
|
|
|
68 |
)
|
69 |
|
70 |
except Exception as e:
|
71 |
raise RuntimeError(f"Failed to initialize the model: {str(e)}")
|
72 |
|
73 |
+
# State management for multiple concurrent conversations
|
74 |
self.consultation_states = {}
|
75 |
self.gathered_info = {}
|
76 |
|
77 |
def _is_identity_question(self, message: str) -> bool:
|
78 |
+
"""Detect if the user is asking about the assistant's identity"""
|
79 |
identity_patterns = [
|
80 |
r"who are you",
|
81 |
r"what are you",
|
|
|
86 |
return any(re.search(pattern, message.lower()) for pattern in identity_patterns)
|
87 |
|
88 |
def _is_location_question(self, message: str) -> bool:
|
89 |
+
"""Detect if the user is asking about the assistant's location"""
|
90 |
location_patterns = [
|
91 |
r"where are you",
|
92 |
r"which country",
|
|
|
97 |
return any(re.search(pattern, message.lower()) for pattern in location_patterns)
|
98 |
|
99 |
def _get_next_assessment_question(self, conversation_id: str) -> Optional[str]:
|
100 |
+
"""Get the next health assessment question based on conversation progress"""
|
101 |
if conversation_id not in self.gathered_info:
|
102 |
self.gathered_info[conversation_id] = []
|
103 |
|
|
|
107 |
return None
|
108 |
|
109 |
async def process_message(self, conversation_id: str, message: str, history: List[Dict]) -> ChatResponse:
|
110 |
+
"""
|
111 |
+
Process incoming messages and manage the conversation flow
|
112 |
+
"""
|
113 |
try:
|
114 |
# Initialize state for new conversations
|
115 |
if conversation_id not in self.consultation_states:
|
|
|
163 |
# Prepare messages for the model
|
164 |
messages = [
|
165 |
{"role": "system", "content": NURSE_OGE_IDENTITY},
|
166 |
+
{"role": "user", "content": f"Based on the following patient information, provide thorough assessment, diagnosis and recommendations:\n\n{context}\n\nOriginal query: {message}"}
|
167 |
]
|
168 |
|
169 |
# Implement retry logic for model inference
|
|
|
204 |
finished=True
|
205 |
)
|
206 |
|
207 |
+
# Define FastAPI lifespan for startup/shutdown events
|
208 |
+
@asynccontextmanager
|
209 |
+
async def lifespan(app: FastAPI):
|
210 |
+
# Initialize on startup
|
211 |
+
global nurse_oge
|
212 |
+
try:
|
213 |
+
nurse_oge = NurseOgeAssistant()
|
214 |
+
except Exception as e:
|
215 |
+
print(f"Failed to initialize NurseOgeAssistant: {e}")
|
216 |
+
yield
|
217 |
+
# Clean up on shutdown if needed
|
218 |
+
# Add cleanup code here
|
219 |
|
220 |
+
# Initialize FastAPI with lifespan
|
221 |
+
app = FastAPI(lifespan=lifespan)
|
222 |
|
223 |
# Add memory management middleware
|
224 |
@app.middleware("http")
|
225 |
async def add_memory_management(request: Request, call_next):
|
226 |
+
"""Middleware to help manage memory usage"""
|
227 |
gc.collect()
|
228 |
response = await call_next(request)
|
229 |
gc.collect()
|
230 |
return response
|
231 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
232 |
# Health check endpoint
|
233 |
@app.get("/health")
|
234 |
async def health_check():
|
235 |
+
"""Endpoint to verify service health"""
|
236 |
return {"status": "healthy", "model_loaded": nurse_oge is not None}
|
237 |
|
238 |
# Chat endpoint
|
239 |
@app.post("/chat")
|
240 |
async def chat_endpoint(request: ChatRequest):
|
241 |
+
"""Main chat endpoint for API interactions"""
|
242 |
if nurse_oge is None:
|
243 |
raise HTTPException(
|
244 |
status_code=503,
|
|
|
259 |
return response
|
260 |
|
261 |
# Gradio chat interface function
|
262 |
+
async def gradio_chat(message, history):
|
263 |
+
"""Handler for Gradio chat interface"""
|
264 |
if nurse_oge is None:
|
265 |
return "The medical assistant is not available at the moment. Please try again later."
|
266 |
|
267 |
+
response = await nurse_oge.process_message("gradio_user", message, history)
|
268 |
return response.response
|
269 |
|
270 |
+
# Create and configure Gradio interface
|
271 |
demo = gr.ChatInterface(
|
272 |
fn=gradio_chat,
|
273 |
title="Nurse Oge - Medical Assistant",
|
|
|
281 |
theme=gr.themes.Soft(
|
282 |
primary_hue="blue",
|
283 |
secondary_hue="purple",
|
284 |
+
)
|
|
|
|
|
|
|
285 |
)
|
286 |
|
287 |
# Add custom CSS for better appearance
|