Multi Labels

Dockerfile

@@ -0,0 +1,27 @@
+# Use the official Python image as a base image
+FROM python:3.10.0-slim-buster
+
+# Set environment variables
+ENV PYTHONDONTWRITEBYTECODE 1
+ENV PYTHONUNBUFFERED 1
+
+# Set the working directory in the container
+WORKDIR /app
+
+# Copy the dependencies file to the working directory
+COPY requirements.txt .
+
+# Install any dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy the content of the local src directory to the working directory
+COPY . .
+
+# Make port 7860 available to the world outside this container
+EXPOSE 7860
+
+# Define environment variable (corrected the syntax for environment variable name and value)
+# ENV FLASK_APP=app.py
+
+# Command to run the application
+CMD ["gunicorn", "--bind", "0.0.0.0:7860", "app:app"]

app.py

@@ -0,0 +1,79 @@
+from flask import Flask, request, render_template, jsonify
+from transformers import BertTokenizer, BertModel
+import torch
+import numpy as np
+
+app = Flask(__name__)
+
+# Initialize BERT model and tokenizer
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+class BERTClass(torch.nn.Module):
+    def __init__(self):
+        super(BERTClass, self).__init__()
+        self.bert_model = BertModel.from_pretrained('bert-base-uncased', return_dict=True)
+        self.dropout = torch.nn.Dropout(0.3)
+        self.linear = torch.nn.Linear(768, 8)
+
+    def forward(self, input_ids, attn_mask, token_type_ids):
+        output = self.bert_model(
+            input_ids,
+            attention_mask=attn_mask,
+            token_type_ids=token_type_ids
+        )
+        output_dropout = self.dropout(output.pooler_output)
+        output = self.linear(output_dropout)
+        return output
+
+# Load tokenizer and model
+tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
+model = BERTClass()
+model.load_state_dict(torch.load("model/MLTC_model_state.bin", map_location=device))
+model = model.to(device)
+model.eval()
+
+# Hyperparameters
+MAX_LEN = 256
+THRESHOLD = 0.5
+
+target_list = ['price', 'packaging', 'product', 'rider', 'delivery', 'shelf', 'service', 'seller']
+
+@app.route('/', methods=['GET', 'POST'])
+def index():
+    if request.method == 'POST':
+        raw_text = request.form['text']
+        
+        if not raw_text:
+            return jsonify({'error': 'Please enter some text'}), 400
+        
+        # Tokenize and encode text
+        encoded_text = tokenizer.encode_plus(
+            raw_text,
+            max_length=MAX_LEN,
+            add_special_tokens=True,
+            return_token_type_ids=True,
+            pad_to_max_length=True,
+            return_attention_mask=True,
+            return_tensors='pt',
+        )
+        
+        input_ids = encoded_text['input_ids'].to(device)
+        attention_mask = encoded_text['attention_mask'].to(device)
+        token_type_ids = encoded_text['token_type_ids'].to(device)
+        
+        # Make predictions
+        with torch.no_grad():
+            output = model(input_ids, attention_mask, token_type_ids)
+            output = torch.sigmoid(output).detach().cpu()
+            output = output.flatten().round().numpy()
+        
+        # Determine predicted labels based on threshold
+        predictions = [target_list[idx] for idx, p in enumerate(output) if p == 1]
+        
+        return render_template('index.html', text=raw_text, predictions=predictions)
+    
+    # For GET request or initial page load
+    return render_template('index.html')
+
+if __name__ == '__main__':
+    app.run(debug=True)

main.py

@@ -0,0 +1,63 @@
+from transformers import BertTokenizer, BertModel
+import torch
+import numpy as np
+
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+class BERTClass(torch.nn.Module):
+    def __init__(self):
+        super(BERTClass, self).__init__()
+        self.bert_model = BertModel.from_pretrained('bert-base-uncased', return_dict=True)
+        self.dropout = torch.nn.Dropout(0.3)
+        self.linear = torch.nn.Linear(768, 8)
+
+    def forward(self, input_ids, attn_mask, token_type_ids):
+        output = self.bert_model(
+            input_ids,
+            attention_mask=attn_mask,
+            token_type_ids=token_type_ids
+        )
+        output_dropout = self.dropout(output.pooler_output)
+        output = self.linear(output_dropout)
+        return output
+
+tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
+
+# Hyperparameters
+MAX_LEN = 256
+THRESHOLD = 0.5
+
+# Loading pretrained model (best model)
+model = BERTClass()
+model.load_state_dict(torch.load(r"model\MLTC_model_state.bin", map_location=device))
+model = model.to(device)
+
+# raw text
+raw_text = """
+১০০% আসল প্রোডাক্ট। সিলেটের মধ্যে ৮ দিনের মধ্যে ডেলিভারি হয়েছে। বিক্রেতা খুবই সহানুভূতিশীল এবং ভালো ছিলেন। এই প্রোডাক্টটি এই বিক্রেতার কাছ থেকে কেনার জন্য অত্যন্ত সুপারিশ করছি।
+"""
+
+encoded_text = tokenizer.encode_plus(
+    raw_text,
+    max_length=MAX_LEN,
+    add_special_tokens=True,
+    return_token_type_ids=True,
+    pad_to_max_length=True,
+    return_attention_mask=True,
+    return_tensors='pt',
+)
+
+input_ids = encoded_text['input_ids'].to(device)
+attention_mask = encoded_text['attention_mask'].to(device)
+token_type_ids = encoded_text['token_type_ids'].to(device)
+
+output = model(input_ids, attention_mask, token_type_ids)
+output = torch.sigmoid(output).detach().cpu()
+output = output.flatten().round().numpy()
+
+target_list = ['price', 'packaging', 'product', 'rider', 'delivery', 'shelf', 'service', 'seller']
+
+print(f"Title: {raw_text}")
+for idx, p in enumerate(output):
+    if p == 1:
+        print(f"Label: {target_list[idx]}")

model/MLTC_model_state.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77b15f2713f2c8962d0a90d893c0eff711d9b18009ec90a68469f36951b9b228
+size 438039573

requirements.txt

@@ -0,0 +1,6 @@
+flask
+gunicorn
+torch
+transformers
+mymodule
+numpy==1.26.4

templates/index.html

@@ -0,0 +1,157 @@
+<!DOCTYPE html>
+<html>
+<head>
+    <title>BERT Predictions</title>
+    <style>
+        body {
+            background-color: #f0f0f0; /* Light grey background */
+            font-family: Arial, sans-serif;
+            margin: 0; /* Remove default margin */
+            display: flex;
+            flex-direction: column;
+            min-height: 100vh; /* Ensure the body takes up at least the full viewport height */
+            position: relative; /* Ensure relative positioning for absolute elements */
+        }
+        h2 {
+            color: #333;
+        }
+        .content-wrapper {
+            flex: 1; /* Allow content to expand within the flex container */
+            display: flex;
+            flex-direction: column;
+            align-items: center;
+            justify-content: center; /* Center content vertically and horizontally */
+        }
+        .input-container {
+            max-width: 600px;
+            width: 100%; /* Take up full width within the flex container */
+            background-color: #fff;
+            padding: 20px;
+            border-radius: 8px;
+            box-shadow: 0 0 10px rgba(0, 0, 0, 0.1);
+            text-align: left; /* Adjust text alignment as needed */
+        }
+        textarea {
+            width: 100%;
+            padding: 10px;
+            font-size: 16px;
+            margin-bottom: 10px;
+        }
+        input[type="submit"] {
+            padding: 10px 20px;
+            font-size: 16px;
+            background-color: #4CAF50; /* Green submit button */
+            color: white;
+            border: none;
+            cursor: pointer;
+            display: block; /* Ensure block display for centering */
+            margin: 0 auto; /* Center align button */
+            transition: background-color 0.3s ease; /* Smooth transition for hover effect */
+        }
+        input[type="submit"]:hover {
+            background-color: #d62728; /* Red on hover */
+        }
+        .predictions {
+            text-align: center; /* Center align predictions */
+            margin-top: 20px;
+        }
+        .prediction-item {
+            margin-bottom: 10px;
+            padding-left: 20px;
+            font-size: 18px;
+            display: inline-block; /* Ensure items stay inline */
+            text-align: left; /* Align text left within each item */
+        }
+        .icon {
+            font-size: 20px;
+            margin-right: 10px;
+            vertical-align: middle;
+        }
+        .category-price {
+            color: #1f77b4; /* Blue */
+        }
+        .category-packaging {
+            color: #ff7f0e; /* Orange */
+        }
+        .category-product {
+            color: #2ca02c; /* Green */
+        }
+        .category-rider {
+            color: #d62728; /* Red */
+        }
+        .category-delivery {
+            color: #9467bd; /* Purple */
+        }
+        .category-shelf {
+            color: #8c564b; /* Brown */
+        }
+        .category-service {
+            color: #e377c2; /* Pink */
+        }
+        .category-seller {
+            color: #7f7f7f; /* Grey */
+        }
+        /* Logo style */
+        .logo {
+            position: absolute;
+            top: 20px;
+            right: 20px;
+            width: 150px; /* Increased width of the logo */
+            height: 80px; /* Increased height of the logo */
+        }
+        /* Footer style */
+        footer {
+            width: 100%;
+            background-color: #ccc; /* Light grey background */
+            padding: 10px 0; /* Padding for content within footer */
+            text-align: center;
+            color: #666;
+            position: relative;
+        }
+    </style>
+    <script>
+        function validateForm() {
+            var text = document.forms["predictForm"]["text"].value.trim();
+            if (text == "") {
+                alert("Please enter some text");
+                return false;
+            }
+            return true;
+        }
+
+        function clearPredictions() {
+            var predictionsDiv = document.querySelector('.predictions');
+            if (predictionsDiv) {
+                predictionsDiv.innerHTML = ''; // Clear previous predictions
+            }
+        }
+    </script>
+</head>
+<body>
+    <div class="content-wrapper">
+        <img src="static/static_logo.png" alt="Company Logo" class="logo">
+        <div class="input-container">
+            <h2>Enter Text to Predict Multi-Label Classification</h2>
+            <form name="predictForm" method="post" onsubmit="clearPredictions(); return validateForm()">
+                <textarea name="text" rows="4" cols="50"></textarea><br><br>
+                <input type="submit" value="Predict">
+            </form>
+        </div>
+        
+        {% if predictions %}
+            <div class="predictions">
+                <h3>Predicted Labels:</h3>
+                {% for prediction in predictions %}
+                    <div class="prediction-item category-{{ prediction }}">
+                        <span class="icon">✓</span>{{ prediction }}
+                    </div>
+                {% endfor %}
+            </div>
+        {% endif %}
+    </div>
+
+    <footer>
+        <p>&copy; 2024 Bikasuzzaman. Machine Learning Engineer.</p>
+    </footer>
+</body>
+</html>

templates/result.html

@@ -0,0 +1,120 @@
+<!DOCTYPE html>
+<html>
+<head>
+    <title>BERT Predictions</title>
+    <style>
+        body {
+            background-color: #f0f0f0; /* Light grey background */
+            font-family: Arial, sans-serif;
+            padding: 20px;
+            text-align: center; /* Center align content */
+        }
+        h2 {
+            color: #333;
+        }
+        .input-container {
+            max-width: 600px;
+            margin: 0 auto; /* Center align container */
+            background-color: #fff;
+            padding: 20px;
+            border-radius: 8px;
+            box-shadow: 0 0 10px rgba(0, 0, 0, 0.1);
+            margin-bottom: 20px;
+        }
+        textarea {
+            width: 100%;
+            padding: 10px;
+            font-size: 16px;
+            margin-bottom: 10px;
+        }
+        input[type="submit"] {
+            padding: 10px 20px;
+            font-size: 16px;
+            background-color: #4CAF50; /* Green submit button */
+            color: white;
+            border: none;
+            cursor: pointer;
+        }
+        input[type="submit"]:hover {
+            background-color: #45a049; /* Darker green on hover */
+        }
+        .predictions {
+            text-align: center; /* Center align predictions */
+            margin-top: 20px;
+        }
+        .prediction-item {
+            margin-bottom: 10px;
+            padding-left: 20px;
+            font-size: 18px;
+            display: inline-block; /* Ensure items stay inline */
+            text-align: left; /* Align text left within each item */
+        }
+        .icon {
+            font-size: 20px;
+            margin-right: 10px;
+            vertical-align: middle;
+        }
+        .category-price {
+            color: #1f77b4; /* Blue */
+        }
+        .category-packaging {
+            color: #ff7f0e; /* Orange */
+        }
+        .category-product {
+            color: #2ca02c; /* Green */
+        }
+        .category-rider {
+            color: #d62728; /* Red */
+        }
+        .category-delivery {
+            color: #9467bd; /* Purple */
+        }
+        .category-shelf {
+            color: #8c564b; /* Brown */
+        }
+        .category-service {
+            color: #e377c2; /* Pink */
+        }
+        .category-seller {
+            color: #7f7f7f; /* Grey */
+        }
+    </style>
+    <script>
+        function validateForm() {
+            var text = document.forms["predictForm"]["text"].value.trim();
+            if (text == "") {
+                alert("Please enter some text");
+                return false;
+            }
+            return true;
+        }
+
+        function clearPredictions() {
+            var predictionsDiv = document.querySelector('.predictions');
+            if (predictionsDiv) {
+                predictionsDiv.innerHTML = ''; // Clear previous predictions
+            }
+        }
+    </script>
+</head>
+<body>
+    <div class="input-container">
+        <h2>Enter Text to Predict Multi-Label Classification</h2>
+        <form name="predictForm" method="post" onsubmit="clearPredictions(); return validateForm()">
+            <textarea name="text" rows="4" cols="50"></textarea><br><br>
+            <input type="submit" value="Predict">
+        </form>
+    </div>
+    
+    {% if predictions %}
+        <div class="predictions">
+            <h3>Predicted Categories:</h3>
+            {% for prediction in predictions %}
+                <div class="prediction-item category-{{ prediction }}">
+                    <span class="icon">✓</span>{{ prediction }}
+                </div>
+            {% endfor %}
+        </div>
+    {% endif %}
+</body>
+</html>