import joblib
import html
import re
from fastapi import FastAPI
from pydantic import BaseModel
from fastapi.middleware.cors import CORSMiddleware
from bs4 import BeautifulSoup
app = FastAPI()
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
# Load the models
model = joblib.load('model.joblib')
tfidf = joblib.load('tfidf.joblib')
def clean_text(text):
if not isinstance(text, str):
return ""
text = html.unescape(text)
text = BeautifulSoup(text, "html.parser").get_text()
text = re.sub(r'http\S+|www\S+|https\S+', '', text, flags=re.MULTILINE)
text = re.sub(r'\S*@\S*\s?', '', text)
text = re.sub(r'[^a-zA-Z\s]', '', text)
text = re.sub(r'\s+', ' ', text).strip()
return text.lower()
class ReviewRequest(BaseModel):
text: str
@app.post("/predict")
def predict(request: ReviewRequest):
cleaned = clean_text(request.text)
# Very short edge cases check
if len(cleaned.split()) < 2:
return {"sentiment": "Neutral", "confidence": 0.5, "cleaned": cleaned}
vector = tfidf.transform([cleaned])
prediction = model.predict(vector)[0]
probabilities = model.predict_proba(vector)[0]
confidence = float(max(probabilities))
sentiment = "Positive" if prediction == 1 else "Negative"
return {"sentiment": sentiment, "confidence": confidence, "cleaned": cleaned}
from fastapi.responses import HTMLResponse
import markdown2
@app.get("/", response_class=HTMLResponse)
def read_root():
try:
with open("README.md", "r") as f:
content = f.read()
metadata = {}
if content.startswith("---"):
parts = content.split("---", 2)
if len(parts) >= 3:
import yaml
metadata = yaml.safe_load(parts[1])
content = parts[2]
html_content = markdown2.markdown(content.strip())
tags = [
("Sentiment Analysis", "#3b82f6"),
("Logistic Regression", "#6366f1"),
("TF-IDF", "#8b5cf6"),
("Docker", "#10b981"),
("FastAPI", "#f59e0b")
]
tags_html = "".join([f'{name}' for name, color in tags])
return f"""
{metadata.get('title', 'Model Card')}
"""
except Exception as e:
return f"Failed to render Model Card: {str(e)}
"