create a new file

Dockerfile

@@ -0,0 +1,22 @@
+FROM python:3.9
+
+WORKDIR /app
+COPY . /app
+
+ENV HF_HOME=/app/.cache
+
+RUN mkdir -p /app/.cache/huggingface/hub && \
+    chmod -R 777 /app/.cache && \
+    chmod -R 777 /app/.cache/huggingface
+
+
+
+RUN pip install --upgrade pip
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY --chown=user ./requirements.txt requirements.txt
+RUN pip install --no-cache-dir --upgrade -r requirements.txt
+
+EXPOSE 7860
+
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

main.py

@@ -0,0 +1,13 @@
+from fastapi import FastAPI
+from routers import predict
+
+app = FastAPI(title="Logistic Regression API")
+
+# Include API router
+app.include_router(predict.router)
+
+@app.get("/")
+def home():
+    return {"message": "Welcome to the Logistic Regression API"}
+
+# Run using: uvicorn main:app --reload

models/train_model.py

@@ -0,0 +1,30 @@
+import pandas as pd
+from sklearn.model_selection import train_test_split
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.linear_model import LogisticRegression
+import pickle
+
+# Load dataset
+file_path = "data/sms_process_data_main.xlsx"
+df = pd.read_excel(file_path)
+
+# Prepare training data
+X_train, X_test, y_train, y_test = train_test_split(df['MessageText'], df['label'], test_size=0.2, random_state=42)
+
+# Convert text into numerical vectors
+vectorizer = TfidfVectorizer()
+X_train_vec = vectorizer.fit_transform(X_train)
+X_test_vec = vectorizer.transform(X_test)
+
+# Train model
+model = LogisticRegression(max_iter=1000)
+model.fit(X_train_vec, y_train)
+
+# Save model and vectorizer
+with open("models/logistic_regression.pkl", "wb") as model_file:
+    pickle.dump(model, model_file)
+
+with open("models/vectorizer.pkl", "wb") as vec_file:
+    pickle.dump(vectorizer, vec_file)
+
+print("Model trained and saved successfully!")

requirements.txt

@@ -0,0 +1,6 @@
+fastapi
+uvicorn
+scikit-learn
+pandas
+sentence-transformers
+openpyxl

routers/predict.py

@@ -0,0 +1,13 @@
+from fastapi import APIRouter
+from schema.schemas import PredictionInput, PredictionOutput
+from service.classifier import load_model, predict
+
+router = APIRouter(prefix="/predict", tags=["Prediction"])
+
+# Load the model once
+model, vectorizer = load_model()
+
+@router.post("/", response_model=PredictionOutput)
+def make_prediction(input_data: PredictionInput):
+    prediction = predict(input_data.text, model, vectorizer)
+    return {"prediction": prediction}

schema/schemas.py

@@ -0,0 +1,7 @@
+from pydantic import BaseModel
+
+class PredictionInput(BaseModel):
+    text: str
+
+class PredictionOutput(BaseModel):
+    prediction: str

service/classifier.py

@@ -0,0 +1,23 @@
+import pickle
+import os
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.linear_model import LogisticRegression
+
+MODEL_PATH = "models/logistic_regression.pkl"
+VECTORIZER_PATH = "models/vectorizer.pkl"
+
+def load_model():
+    """Load trained model and vectorizer from disk."""
+    if os.path.exists(MODEL_PATH) and os.path.exists(VECTORIZER_PATH):
+        with open(MODEL_PATH, "rb") as model_file, open(VECTORIZER_PATH, "rb") as vec_file:
+            model = pickle.load(model_file)
+            vectorizer = pickle.load(vec_file)
+        return model, vectorizer
+    else:
+        raise FileNotFoundError("Model or vectorizer not found!")
+
+def predict(text, model, vectorizer):
+    """Make predictions using the trained model."""
+    text_vectorized = vectorizer.transform([text])
+    prediction = model.predict(text_vectorized)[0]
+    return prediction