Spaces:
Running
Running
File size: 9,865 Bytes
a985b94 f93287b a985b94 f93287b 593c3ba f93287b 593c3ba f93287b 593c3ba f93287b 593c3ba | 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 | import sys
import os
import pickle
import sqlite3
import pandas as pd
from fastapi import FastAPI
from fastapi.staticfiles import StaticFiles
from fastapi.responses import FileResponse
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from dotenv import load_dotenv
from google import genai
import chromadb
from typing import List, Dict
env_path = os.path.join(os.path.dirname(__file__), '.env')
load_dotenv(env_path)
# Add parent dir to path so we can import from middleware
sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
from middleware.material_predictor import predict_material_needs
app = FastAPI(title="Wafer Defect API")
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
# Robust paths for Docker/Hosting
BASE_DIR = os.path.dirname(os.path.abspath(__file__))
DB_PATH = os.path.join(BASE_DIR, '..', 'middleware', 'wafer_control.db')
MODEL_PATH = os.path.join(BASE_DIR, '..', 'middleware', 'material_model.pkl')
CHROMA_PATH = os.path.join(BASE_DIR, 'chroma_db')
# Ensure directories exist
os.makedirs(CHROMA_PATH, exist_ok=True)
DEFECT_COLORS = {
'Center': '#ef4444', 'Donut': '#f59e0b', 'Edge-Loc': '#10b981',
'Edge-Ring': '#3b82f6', 'Loc': '#8b5cf6', 'Random': '#ec4899',
'Scratch': '#06b6d4', 'Near-full': '#f97316', 'None': '#6b7280',
'Undetected': '#374151',
}
# Globally load data so we don't block requests
df = pd.DataFrame()
if os.path.exists(DB_PATH):
print(f"Loading DB from {DB_PATH}...")
conn = sqlite3.connect(DB_PATH)
df = pd.read_sql_query("SELECT * FROM wafer_logs", conn)
conn.close()
df['scan_time'] = pd.to_datetime(df['scan_time'])
df['scan_date'] = df['scan_time'].dt.date
else:
print(f"Warning: DB not found at {DB_PATH}. Dashboard will be empty.")
# Setup Vector DB and LLM
print(f"Connecting to ChromaDB at {CHROMA_PATH}...")
try:
chroma_client = chromadb.PersistentClient(path=CHROMA_PATH)
collection = chroma_client.get_or_create_collection(name="semiconductor_knowledge")
except Exception as e:
print(f"Warning: Could not connect to ChromaDB collection. Error: {e}")
collection = None
print("Initializing Gemini API...")
gemini_client = None
if os.getenv("GEMINI_API_KEY"):
gemini_client = genai.Client(api_key=os.getenv("GEMINI_API_KEY"))
else:
print("Warning: GEMINI_API_KEY not found in environment.")
print("Loading ML model...")
model_pkg = None
if os.path.exists(MODEL_PATH):
with open(MODEL_PATH, 'rb') as f:
model_pkg = pickle.load(f)
@app.get("/api/kpi")
def get_kpis():
total_scans = len(df)
fail_df = df[df['status'] == 'FAIL']
fail_count = len(fail_df)
pass_count = len(df[df['status'] == 'PASS'])
pass_rate = round((pass_count / total_scans) * 100, 1) if total_scans else 0
scrap_count = len(df[df['action'] == 'ROUTE_TO_SCRAP'])
avg_waste = round(fail_df['material_wasted_pct'].mean(), 2) if fail_count else 0
avg_confidence = round(fail_df['confidence'].mean(), 2) if fail_count else 0
return {
"total_scans": total_scans,
"pass_count": pass_count,
"pass_rate": pass_rate,
"fail_count": fail_count,
"fail_rate": round(100 - pass_rate, 1),
"scrap_count": scrap_count,
"avg_waste": avg_waste,
"avg_confidence": avg_confidence
}
@app.get("/api/charts/defects")
def get_defects():
fail_df = df[df['status'] == 'FAIL']
defect_counts = fail_df['defect_type'].value_counts().reset_index()
defect_counts.columns = ['defect_type', 'count']
gt_counts = fail_df['ground_truth'].value_counts().reset_index()
gt_counts.columns = ['ground_truth', 'count']
return {
"predictions": defect_counts.to_dict(orient="records"),
"ground_truth": gt_counts.head(15).to_dict(orient="records")
}
@app.get("/api/charts/waste")
def get_waste():
fail_df = df[df['status'] == 'FAIL']
waste_by_type = fail_df.groupby('defect_type').agg(
total_waste=('material_wasted_pct', lambda x: x.sum() / 100.0)
).reset_index().sort_values('total_waste', ascending=True)
action_counts = df['action'].value_counts().reset_index()
action_counts.columns = ['action', 'count']
return {
"waste_by_type": waste_by_type.to_dict(orient="records"),
"actions": action_counts.to_dict(orient="records")
}
@app.get("/api/charts/trends")
def get_trends():
daily = df.groupby('scan_date').agg(
scans=('id', 'count'),
fails=('status', lambda x: (x == 'FAIL').sum()),
waste=('material_wasted_pct', lambda x: x.sum() / 100.0)
).reset_index()
daily['fail_rate'] = round((daily['fails'] / daily['scans']) * 100, 1)
return {
"dates": daily['scan_date'].astype(str).tolist(),
"fail_rate": daily['fail_rate'].tolist(),
"waste": daily['waste'].tolist()
}
@app.get("/api/model/status")
def model_status():
if not model_pkg:
return {"loaded": False}
m = model_pkg['metrics']
imp = model_pkg['metrics']['importances']
imp_df = pd.DataFrame({'feature': list(imp.keys()), 'importance': list(imp.values())})
imp_df = imp_df.sort_values('importance', ascending=True).tail(10)
return {
"loaded": True,
"metrics": {"r2": round(m['r2'], 4), "mae": round(m['mae'], 2)},
"importance": imp_df.to_dict(orient="records")
}
class PredictionRequest(BaseModel):
scans: int
fail_rate: float
@app.post("/api/predict")
def predict_waste(req: PredictionRequest):
if not model_pkg:
return {"error": "No model loaded"}
fail_df = df[df['status'] == 'FAIL']
dist = fail_df['defect_type'].value_counts(normalize=True).to_dict()
pred = predict_material_needs(model_pkg['model'], model_pkg['feature_cols'], req.scans, req.fail_rate / 100.0, dist)
pred['fail_rate'] = req.fail_rate
return pred
class ChatMessage(BaseModel):
role: str
content: str
class ChatRequest(BaseModel):
messages: List[ChatMessage]
@app.post("/api/chat")
def chat_with_bot(req: ChatRequest):
if not gemini_client:
return {"error": "Gemini API key not configured"}
user_message = req.messages[-1].content if req.messages else ""
# 1. RAG Retrieval from ChromaDB
context_docs = ""
if collection and user_message:
try:
results = collection.query(query_texts=[user_message], n_results=2)
if results and results['documents'] and results['documents'][0]:
context_docs = "\n".join(results['documents'][0])
except Exception as e:
print(f"ChromaDB Query Error: {e}")
# 2. Get Live Dashboard Context
total_scans = len(df)
fail_df = df[df['status'] == 'FAIL']
fail_count = len(fail_df)
pass_rate = round(((total_scans - fail_count) / total_scans) * 100, 1) if total_scans else 0
top_defects = fail_df['defect_type'].value_counts().head(3).to_dict()
live_kpis = f"""
Current Dashboard State:
- Total Wafers Scanned: {total_scans}
- Current Pass Rate: {pass_rate}%
- Total Defective Wafers: {fail_count}
- Top Defect Types Right Now: {top_defects}
"""
# 3. Construct System Prompt
system_instruction = f"""
You are the 'Gorilla Semiconductors Engineering Assistant', an expert semiconductor manufacturing assistant.
You help engineers understand dashboard data and troubleshoot wafer defects.
Maintain a strictly professional, analytical, and authoritative engineering tone.
Here is the LIVE DATA from the dashboard:
{live_kpis}
Here is retrieved technical context from our engineering database based on the user's query:
{context_docs if context_docs else "No specific engineering docs retrieved."}
Use the live data to answer questions about 'current status' or 'dashboard'.
Use the engineering docs to answer questions about 'why' a defect happens.
"""
try:
# Convert messages to format expected by google-genai
contents = []
for msg in req.messages:
role = "user" if msg.role == "user" else "model"
contents.append(
genai.types.Content(role=role, parts=[genai.types.Part.from_text(text=msg.content)])
)
response = gemini_client.models.generate_content(
model='gemini-2.5-flash-lite',
contents=contents,
config=genai.types.GenerateContentConfig(
system_instruction=system_instruction,
temperature=0.3
)
)
return {"response": response.text}
except Exception as e:
print(f"Gemini API Error: {e}")
return {"error": str(e)}
# --- SERVE FRONTEND ---
FRONTEND_PATH = os.path.join(BASE_DIR, "..", "frontend", "dist")
if os.path.exists(FRONTEND_PATH):
@app.get("/{full_path:path}")
async def serve_frontend(full_path: str):
# 1. Skip API routes
if full_path.startswith("api"):
return {"detail": "Not Found"}
# 2. Check if the file exists (for assets like .js, .css, .png)
file_path = os.path.join(FRONTEND_PATH, full_path)
if os.path.isfile(file_path):
return FileResponse(file_path)
# 3. Fallback to index.html for React Router
index_file = os.path.join(FRONTEND_PATH, "index.html")
if os.path.exists(index_file):
return FileResponse(index_file)
return {"detail": "Frontend build not found"}
else:
@app.get("/")
def read_root():
return {"message": "Wafer Defect API is running. Frontend build folder not found."}
|