Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+model.onnx.data filter=lfs diff=lfs merge=lfs -text

Dockerfile

@@ -0,0 +1,12 @@
+FROM python:3.9-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+EXPOSE 8000
+
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,103 @@
+import uvicorn
+from fastapi import FastAPI, HTTPException
+from fastapi.staticfiles import StaticFiles
+from fastapi.responses import FileResponse
+from pydantic import BaseModel
+import onnxruntime as ort
+import numpy as np
+from transformers import AutoTokenizer
+import os
+import logging
+
+# Configure logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+app = FastAPI()
+
+# Configuration
+MODEL_PATH = "model.onnx"
+TOKENIZER_NAME = "jhu-clsp/mmBERT-small"
+
+# Global variables
+ort_session = None
+tokenizer = None
+
+class PredictRequest(BaseModel):
+    text: str
+
+class PredictResponse(BaseModel):
+    label: str
+    score: float
+    is_hate: bool
+
+def load_model():
+    global ort_session, tokenizer
+    try:
+        if os.path.exists(MODEL_PATH):
+            logger.info(f"Loading ONNX model from {MODEL_PATH}...")
+            ort_session = ort.InferenceSession(MODEL_PATH)
+            logger.info("Model loaded successfully.")
+        else:
+            logger.warning(f"Model file {MODEL_PATH} not found. Inference will fail.")
+        
+        logger.info(f"Loading tokenizer {TOKENIZER_NAME}...")
+        tokenizer = AutoTokenizer.from_pretrained(TOKENIZER_NAME)
+        logger.info("Tokenizer loaded successfully.")
+    except Exception as e:
+        logger.error(f"Error loading model or tokenizer: {e}")
+
+@app.on_event("startup")
+async def startup_event():
+    load_model()
+
+@app.post("/predict", response_model=PredictResponse)
+async def predict(request: PredictRequest):
+    if ort_session is None or tokenizer is None:
+        # Mock response if model is not loaded (for testing UI)
+        logger.warning("Model not loaded, returning mock response.")
+        return PredictResponse(label="Non-Hate (Mock)", score=0.0, is_hate=False)
+        # Alternatively, raise error:
+        # raise HTTPException(status_code=503, detail="Model not loaded")
+
+    try:
+        inputs = tokenizer(
+            request.text,
+            max_length=128,
+            padding="max_length",
+            truncation=True,
+            return_tensors="np"
+        )
+        
+        ort_inputs = {
+            "input_ids": inputs["input_ids"].astype(np.int64),
+            "attention_mask": inputs["attention_mask"].astype(np.int64)
+        }
+        
+        ort_outs = ort_session.run(None, ort_inputs)
+        logits = ort_outs[0]
+        
+        # Softmax
+        probs = np.exp(logits) / np.sum(np.exp(logits), axis=1, keepdims=True)
+        pred_idx = np.argmax(probs, axis=1)[0]
+        score = float(probs[0][pred_idx])
+        
+        # 0: Non-Hate, 1: Hate
+        is_hate = bool(pred_idx == 1)
+        label = "Hate" if is_hate else "Non-Hate"
+        
+        return PredictResponse(label=label, score=score, is_hate=is_hate)
+        
+    except Exception as e:
+        logger.error(f"Prediction error: {e}")
+        raise HTTPException(status_code=500, detail=str(e))
+
+# Serve static files
+app.mount("/static", StaticFiles(directory="static"), name="static")
+
+@app.get("/")
+async def read_index():
+    return FileResponse("static/index.html")
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)

model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f885054f14e2df02c723904fe5bb7c36ada86cad113af8f02d1a4a3cdcfce047
+size 1850778

model.onnx.data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4a68e16c86f4101ef7da05e5bf5f5198113337d2dc17d06b705a35523a5f1ff
+size 566099968

requirements.txt

@@ -0,0 +1,6 @@
+fastapi
+uvicorn
+onnxruntime
+numpy
+transformers
+pydantic

static/index.html

@@ -0,0 +1,66 @@
+<!DOCTYPE html>
+<html lang="hi">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>Hindi Hate Speech Detector</title>
+    <link rel="preconnect" href="https://fonts.googleapis.com">
+    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
+    <link href="https://fonts.googleapis.com/css2?family=Outfit:wght@300;400;600;700&family=Rozha+One&display=swap" rel="stylesheet">
+    <link rel="stylesheet" href="/static/style.css">
+</head>
+<body>
+    <div class="background-shapes">
+        <div class="shape shape-1"></div>
+        <div class="shape shape-2"></div>
+        <div class="shape shape-3"></div>
+    </div>
+
+    <div class="container">
+        <header>
+            <h1>Hindi Hate Speech Detector</h1>
+            <p class="subtitle">Advanced AI to identify toxic content in Hindi text</p>
+        </header>
+
+        <main>
+            <div class="card input-card">
+                <div class="input-wrapper">
+                    <textarea id="inputText" placeholder="यहाँ अपना टेक्स्ट लिखें... (Type Hindi text here)"></textarea>
+                    <div class="char-count">0 / 500</div>
+                </div>
+                <button id="analyzeBtn" class="btn-primary">
+                    <span class="btn-text">Analyze Text</span>
+                    <div class="loader hidden"></div>
+                </button>
+            </div>
+
+            <div id="resultCard" class="card result-card hidden">
+                <div class="result-header">
+                    <h2>Analysis Result</h2>
+                    <span id="confidenceScore" class="confidence"></span>
+                </div>
+                <div class="result-content">
+                    <div id="resultIcon" class="icon"></div>
+                    <div class="result-text">
+                        <h3 id="resultLabel"></h3>
+                        <p id="resultMessage"></p>
+                    </div>
+                </div>
+            </div>
+
+            <div class="examples-section">
+                <h3>Try Examples</h3>
+                <div class="examples-grid" id="examplesGrid">
+                    <!-- Examples will be populated by JS -->
+                </div>
+            </div>
+        </main>
+
+        <footer>
+            <p>Powered by BERT-CNN Ensemble Model</p>
+        </footer>
+    </div>
+
+    <script src="/static/script.js"></script>
+</body>
+</html>

static/script.js

@@ -0,0 +1,114 @@
+document.addEventListener('DOMContentLoaded', () => {
+    const inputText = document.getElementById('inputText');
+    const analyzeBtn = document.getElementById('analyzeBtn');
+    const resultCard = document.getElementById('resultCard');
+    const resultLabel = document.getElementById('resultLabel');
+    const resultMessage = document.getElementById('resultMessage');
+    const confidenceScore = document.getElementById('confidenceScore');
+    const loader = document.querySelector('.loader');
+    const btnText = document.querySelector('.btn-text');
+    const charCount = document.querySelector('.char-count');
+    const examplesGrid = document.getElementById('examplesGrid');
+
+    // Hindi Examples
+    const examples = [
+        "बांग्लादेश की शानदार वापसी, भारत को 314 रन पर रोका #INDvBAN #CWC19",
+        "तुम सब गद्दार हो, देश छोड़ दो।",
+        "भारत एक महान देश है।",
+        "इन लोगों को मार डालना चाहिए।",
+        "हमें मिलजुल कर रहना चाहिए।",
+        "वह बहुत बुरा इंसान है।"
+    ];
+
+    // Populate Examples
+    examples.forEach(text => {
+        const chip = document.createElement('div');
+        chip.className = 'example-chip';
+        chip.textContent = text;
+        chip.onclick = () => {
+            inputText.value = text;
+            updateCharCount();
+            analyzeText();
+        };
+        examplesGrid.appendChild(chip);
+    });
+
+    // Character Count
+    inputText.addEventListener('input', updateCharCount);
+
+    function updateCharCount() {
+        const length = inputText.value.length;
+        charCount.textContent = `${length} / 500`;
+        if (length > 500) {
+            charCount.style.color = 'var(--danger)';
+        } else {
+            charCount.style.color = 'var(--text-dim)';
+        }
+    }
+
+    // Analyze Text
+    analyzeBtn.addEventListener('click', analyzeText);
+
+    async function analyzeText() {
+        const text = inputText.value.trim();
+        if (!text) return;
+
+        // UI Loading State
+        setLoading(true);
+        resultCard.classList.add('hidden');
+
+        try {
+            const response = await fetch('/predict', {
+                method: 'POST',
+                headers: {
+                    'Content-Type': 'application/json',
+                },
+                body: JSON.stringify({ text: text }),
+            });
+
+            if (!response.ok) {
+                throw new Error('API Error');
+            }
+
+            const data = await response.json();
+            showResult(data);
+
+        } catch (error) {
+            console.error('Error:', error);
+            alert('An error occurred while analyzing the text. Please try again.');
+        } finally {
+            setLoading(false);
+        }
+    }
+
+    function setLoading(isLoading) {
+        if (isLoading) {
+            analyzeBtn.disabled = true;
+            loader.classList.remove('hidden');
+            btnText.textContent = 'Analyzing...';
+        } else {
+            analyzeBtn.disabled = false;
+            loader.classList.add('hidden');
+            btnText.textContent = 'Analyze Text';
+        }
+    }
+
+    function showResult(data) {
+        resultCard.classList.remove('hidden', 'hate', 'non-hate');
+
+        // Add class based on result
+        if (data.is_hate) {
+            resultCard.classList.add('hate');
+            resultLabel.textContent = 'Hate Speech Detected';
+            resultMessage.textContent = 'This content contains toxic or hateful language.';
+        } else {
+            resultCard.classList.add('non-hate');
+            resultLabel.textContent = 'Non-Hate Speech';
+            resultMessage.textContent = 'This content appears to be safe.';
+        }
+
+        // Format confidence score
+        const scorePercent = (data.score * 100).toFixed(1);
+        confidenceScore.textContent = `Confidence: ${scorePercent}%`;
+    }
+});

static/style.css

@@ -0,0 +1,303 @@
+:root {
+    --primary: #6366f1;
+    --primary-dark: #4f46e5;
+    --secondary: #ec4899;
+    --bg-dark: #0f172a;
+    --card-bg: rgba(30, 41, 59, 0.7);
+    --text-light: #f8fafc;
+    --text-dim: #94a3b8;
+    --success: #10b981;
+    --danger: #ef4444;
+    --glass-border: rgba(255, 255, 255, 0.1);
+}
+
+* {
+    margin: 0;
+    padding: 0;
+    box-sizing: border-box;
+}
+
+body {
+    font-family: 'Outfit', sans-serif;
+    background-color: var(--bg-dark);
+    color: var(--text-light);
+    min-height: 100vh;
+    display: flex;
+    justify-content: center;
+    align-items: center;
+    overflow-x: hidden;
+    position: relative;
+}
+
+/* Background Shapes */
+.background-shapes {
+    position: fixed;
+    top: 0;
+    left: 0;
+    width: 100%;
+    height: 100%;
+    z-index: -1;
+    overflow: hidden;
+}
+
+.shape {
+    position: absolute;
+    filter: blur(100px);
+    opacity: 0.4;
+    animation: float 20s infinite ease-in-out;
+}
+
+.shape-1 {
+    width: 400px;
+    height: 400px;
+    background: var(--primary);
+    top: -100px;
+    left: -100px;
+}
+
+.shape-2 {
+    width: 300px;
+    height: 300px;
+    background: var(--secondary);
+    bottom: -50px;
+    right: -50px;
+    animation-delay: -5s;
+}
+
+.shape-3 {
+    width: 200px;
+    height: 200px;
+    background: #8b5cf6;
+    top: 40%;
+    left: 60%;
+    animation-delay: -10s;
+}
+
+@keyframes float {
+    0%, 100% { transform: translate(0, 0); }
+    50% { transform: translate(30px, -30px); }
+}
+
+.container {
+    width: 100%;
+    max-width: 800px;
+    padding: 2rem;
+    z-index: 1;
+}
+
+header {
+    text-align: center;
+    margin-bottom: 3rem;
+}
+
+h1 {
+    font-size: 3rem;
+    font-weight: 700;
+    background: linear-gradient(to right, #fff, #94a3b8);
+    -webkit-background-clip: text;
+    -webkit-text-fill-color: transparent;
+    margin-bottom: 0.5rem;
+}
+
+.subtitle {
+    color: var(--text-dim);
+    font-size: 1.1rem;
+}
+
+.card {
+    background: var(--card-bg);
+    backdrop-filter: blur(16px);
+    border: 1px solid var(--glass-border);
+    border-radius: 1.5rem;
+    padding: 2rem;
+    margin-bottom: 2rem;
+    box-shadow: 0 4px 30px rgba(0, 0, 0, 0.1);
+    transition: transform 0.3s ease;
+}
+
+.input-card:hover {
+    transform: translateY(-2px);
+}
+
+.input-wrapper {
+    position: relative;
+    margin-bottom: 1.5rem;
+}
+
+textarea {
+    width: 100%;
+    height: 150px;
+    background: rgba(15, 23, 42, 0.6);
+    border: 2px solid var(--glass-border);
+    border-radius: 1rem;
+    padding: 1rem;
+    color: var(--text-light);
+    font-family: 'Outfit', sans-serif;
+    font-size: 1.1rem;
+    resize: none;
+    transition: border-color 0.3s ease;
+}
+
+textarea:focus {
+    outline: none;
+    border-color: var(--primary);
+}
+
+.char-count {
+    position: absolute;
+    bottom: 1rem;
+    right: 1rem;
+    font-size: 0.8rem;
+    color: var(--text-dim);
+}
+
+.btn-primary {
+    width: 100%;
+    padding: 1rem;
+    background: linear-gradient(135deg, var(--primary), var(--secondary));
+    border: none;
+    border-radius: 1rem;
+    color: white;
+    font-size: 1.1rem;
+    font-weight: 600;
+    cursor: pointer;
+    transition: opacity 0.3s ease, transform 0.2s ease;
+    display: flex;
+    justify-content: center;
+    align-items: center;
+}
+
+.btn-primary:hover {
+    opacity: 0.9;
+    transform: scale(1.02);
+}
+
+.btn-primary:active {
+    transform: scale(0.98);
+}
+
+.loader {
+    width: 20px;
+    height: 20px;
+    border: 3px solid rgba(255, 255, 255, 0.3);
+    border-radius: 50%;
+    border-top-color: white;
+    animation: spin 1s ease-in-out infinite;
+    margin-left: 0.5rem;
+}
+
+@keyframes spin {
+    to { transform: rotate(360deg); }
+}
+
+.hidden {
+    display: none;
+}
+
+/* Result Card */
+.result-card {
+    border-left: 5px solid transparent;
+    animation: slideUp 0.5s ease-out;
+}
+
+.result-card.hate {
+    border-left-color: var(--danger);
+    background: linear-gradient(to right, rgba(239, 68, 68, 0.1), transparent);
+}
+
+.result-card.non-hate {
+    border-left-color: var(--success);
+    background: linear-gradient(to right, rgba(16, 185, 129, 0.1), transparent);
+}
+
+.result-header {
+    display: flex;
+    justify-content: space-between;
+    align-items: center;
+    margin-bottom: 1rem;
+}
+
+.confidence {
+    font-size: 0.9rem;
+    color: var(--text-dim);
+    background: rgba(0, 0, 0, 0.2);
+    padding: 0.25rem 0.75rem;
+    border-radius: 1rem;
+}
+
+.result-content {
+    display: flex;
+    align-items: center;
+    gap: 1rem;
+}
+
+.icon {
+    width: 50px;
+    height: 50px;
+    border-radius: 50%;
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    font-size: 1.5rem;
+}
+
+.hate .icon::before {
+    content: '⚠️';
+}
+
+.non-hate .icon::before {
+    content: '✅';
+}
+
+#resultLabel {
+    font-size: 1.5rem;
+    margin-bottom: 0.25rem;
+}
+
+.hate #resultLabel { color: var(--danger); }
+.non-hate #resultLabel { color: var(--success); }
+
+@keyframes slideUp {
+    from { opacity: 0; transform: translateY(20px); }
+    to { opacity: 1; transform: translateY(0); }
+}
+
+/* Examples */
+.examples-section {
+    margin-top: 3rem;
+}
+
+.examples-section h3 {
+    margin-bottom: 1rem;
+    color: var(--text-dim);
+    font-weight: 400;
+}
+
+.examples-grid {
+    display: grid;
+    grid-template-columns: repeat(auto-fit, minmax(250px, 1fr));
+    gap: 1rem;
+}
+
+.example-chip {
+    background: rgba(255, 255, 255, 0.05);
+    border: 1px solid var(--glass-border);
+    padding: 1rem;
+    border-radius: 0.75rem;
+    cursor: pointer;
+    transition: all 0.3s ease;
+    font-size: 0.95rem;
+    color: var(--text-light);
+}
+
+.example-chip:hover {
+    background: rgba(255, 255, 255, 0.1);
+    border-color: var(--primary);
+}
+
+footer {
+    text-align: center;
+    margin-top: 3rem;
+    color: var(--text-dim);
+    font-size: 0.9rem;
+}