modified code

Routes/classify_sms.py

@@ -1,9 +0,0 @@
-from fastapi import APIRouter
-from schemas.sms_schema import SMSRequest, SMSResponse
-from services.sms_service import classify_sms
-
-classify_sms_router = APIRouter()
-
-@classify_sms_router.post("/classify_sms", response_model=SMSResponse)
-def classify(request: SMSRequest):
-    return classify_sms(request.text)

main.py

@@ -1,16 +1,32 @@
 from fastapi import FastAPI
-from routes.embedding import embedding_router
-from routes.cosine_similarity import similarity_router
-from routes.classify_sms import classify_sms_router
+from pydantic import BaseModel
+from typing import List, Tuple
+import numpy as np
+from sentence_transformers import SentenceTransformer
+
+# Load the pre-trained model
+model = SentenceTransformer('Alibaba-NLP/gte-base-en-v1.5', trust_remote_code=True)
+
+# Define request model
+class MessageRequest(BaseModel):
+    messages: List[str]
+
+# Define response model
+class EmbeddingResponse(BaseModel):
+    dimensions: int  # Only return embedding size
+    numeric_values: List[List[float]]
 
 # Initialize FastAPI app
-app = FastAPI(title="SMS Classification API", description="Classifies SMS messages into categories.")
+app = FastAPI()
 
 @app.get("/")
-def home():
-    return {"Message": "Welcome! Use /docs to test the API"}
+def home ():
+    return {"Message":"Welcome to homepage, kindly proceed by giving /docs in the URL" }
 
-# Include API routes
-app.include_router(embedding_router)
-app.include_router(similarity_router)
-app.include_router(classify_sms_router)
+@app.post("/embed", response_model=EmbeddingResponse)
+def embed(request: MessageRequest):
+    new_embeddings = model.encode(request.messages, convert_to_tensor=True)
+    return EmbeddingResponse(
+        dimensions=new_embeddings.shape[1],  # Return only the embedding dimension
+        numeric_values=new_embeddings.tolist()
+    )

models/train_models.py

@@ -1,30 +0,0 @@
-import pandas as pd
-import numpy as np
-import joblib
-from sentence_transformers import SentenceTransformer
-from sklearn.model_selection import train_test_split
-from sklearn.linear_model import LogisticRegression
-
-# Load dataset
-df = pd.read_excel("data/sms_process_data_main.xlsx")
-
-# Load SentenceTransformer model
-encoder_model = SentenceTransformer("Alibaba-NLP/gte-base-en-v1.5", trust_remote_code=True)
-
-# Generate embeddings
-embeddings = encoder_model.encode(df["MessageText"].tolist(), convert_to_numpy=True)
-
-# Encode labels
-label_map = {"Offer": 0, "Transaction": 1}
-df["label"] = df["label"].map(label_map)
-
-# Split dataset
-X_train, X_test, y_train, y_test = train_test_split(embeddings, df["label"], test_size=0.2, random_state=42)
-
-# Train model
-classifier = LogisticRegression()
-classifier.fit(X_train, y_train)
-
-# Save trained model
-joblib.dump(classifier, "models/sms_classifier.pkl")
-print("Model saved as 'sms_classifier.pkl'")

schemas/sms_schema.py

@@ -1,7 +0,0 @@
-from pydantic import BaseModel
-
-class SMSRequest(BaseModel):
-    text: str
-
-class SMSResponse(BaseModel):
-    category: str

services/sms_service.py

@@ -1,14 +0,0 @@
-import joblib
-from sentence_transformers import SentenceTransformer
-import numpy as np
-
-# Load the trained model
-classifier = joblib.load("models/sms_classifier.pkl")
-encoder_model = SentenceTransformer("Alibaba-NLP/gte-base-en-v1.5", trust_remote_code=True)
-
-def classify_sms(text: str):
-    embedding = encoder_model.encode([text], convert_to_numpy=True)
-    prediction = classifier.predict(embedding)
-    
-    category = "Offer" if prediction[0] == 0 else "Transaction"
-    return {"category": category}