Create app.py

app.py

@@ -1,8 +1,9 @@
 # app.py — HuggingFace Space (FastAPI)
-# Serves our fine-tuned sentiment model as a REST API
-# Deploy this to: huggingface.co/spaces/YOUR_USERNAME/creatorpulse-api
+# FREE CPU tier hosting for CreatorPulse sentiment model
+# Space URL: https://ningaraddi-creatorpulse-api.hf.space
 
 from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
 from typing import List
 from transformers import pipeline
@@ -10,55 +11,67 @@ import os
 
 app = FastAPI(title="CreatorPulse Sentiment API")
 
+# ── CORS — allow calls from any origin (our React app) ─────────────────────────
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_methods=["POST", "GET"],
+    allow_headers=["*"],
+)
+
 # ── Load model once on startup ─────────────────────────────────────────────────
-HF_USERNAME = os.getenv("HF_USERNAME", "ningaraddi")
-MODEL_REPO  = f"{HF_USERNAME}/creatorpulse-sentiment"
+MODEL_REPO = "ningaraddi/creatorpulse-sentiment"
 
-print(f"Loading model from {MODEL_REPO}...")
+print(f"Loading model: {MODEL_REPO}")
 
 classifier = pipeline(
-    "sentiment-analysis",
+    "text-classification",
     model=MODEL_REPO,
     tokenizer=MODEL_REPO,
     truncation=True,
     max_length=128,
-    device=-1        # CPU on free tier — fast enough for our use case
+    device=-1,        # CPU — free tier
+    top_k=None,       # Return all labels with scores
 )
 
-print("✅ Model loaded and ready.")
+print("Model loaded!")
 
-# ── Request / Response schemas ──────────────────────────────────────────────────
+# ── Schemas ────────────────────────────────────────────────────────────────────
 class ClassifyRequest(BaseModel):
-    texts: List[str]      # up to 100 comments at once
+    inputs: List[str]
 
-class PredictionResult(BaseModel):
-    text:       str
-    label:      str       # POSITIVE | NEUTRAL | NEGATIVE
+class Prediction(BaseModel):
+    label:      str
     confidence: float
+    text:       str
 
 class ClassifyResponse(BaseModel):
-    predictions: List[PredictionResult]
+    predictions: List[Prediction]
 
-# ── Health check ────────────────────────────────────────────────────────────────
+# ── Health check ───────────────────────────────────────────────────────────────
 @app.get("/")
 def health():
     return {"status": "ok", "model": MODEL_REPO}
 
-# ── Main classify endpoint ──────────────────────────────────────────────────────
+# ── Classify endpoint ──────────────────────────────────────────────────────────
 @app.post("/classify", response_model=ClassifyResponse)
 def classify(request: ClassifyRequest):
-    # Limit to 100 texts per call to avoid timeout
-    texts = request.texts[:100]
+    texts = request.inputs[:100]   # Max 100 per call
+
+    results = classifier(texts, batch_size=8)
+
+    predictions = []
+    for text, label_list in zip(texts, results):
+        top = max(label_list, key=lambda x: x["score"])
 
-    results = classifier(texts, batch_size=16)
+        # Normalize label names
+        label_map = {"LABEL_0": "NEGATIVE", "LABEL_1": "NEUTRAL", "LABEL_2": "POSITIVE"}
+        label = label_map.get(top["label"], top["label"])
 
-    predictions = [
-        PredictionResult(
+        predictions.append(Prediction(
             text=text,
-            label=result["label"],          # POSITIVE / NEUTRAL / NEGATIVE
-            confidence=round(result["score"], 4)
-        )
-        for text, result in zip(texts, results)
-    ]
+            label=label,
+            confidence=round(top["score"], 4),
+        ))
 
     return ClassifyResponse(predictions=predictions)