Upload 4 files

Dockerfile.txt

@@ -0,0 +1,48 @@
+# === Stage 1: Build Dependencies ===
+FROM python:3.9-slim AS builder
+
+# Set environment variables
+ENV PYTHONUNBUFFERED=1 \
+    PIP_NO_CACHE_DIR=1 \
+    PIP_DISABLE_PIP_VERSION_CHECK=1 \
+    PATH="/root/.local/bin:$PATH"
+
+# Install only necessary system dependencies
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    libgl1-mesa-glx \
+    libglib2.0-0 \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install Python dependencies
+WORKDIR /app
+COPY requirements.txt . 
+RUN pip install --user --no-cache-dir -r requirements.txt
+
+# Copy model preparation script and run it
+COPY prepare.py .  
+RUN python3 prepare.py
+
+# === Stage 2: Minimal Runtime Image ===
+FROM python:3.9-slim
+
+# Set environment variables
+ENV PATH="/root/.local/bin:$PATH"
+
+# Copy only required dependencies from builder
+COPY --from=builder /root/.local /root/.local
+
+# Set work directory
+WORKDIR /app
+
+# Copy the model and class mapping
+COPY --from=builder /app/model.h5 /app/model.h5
+COPY --from=builder /app/class.json /app/class.json
+
+# Copy only necessary application files
+COPY app.py .  
+
+# Expose Streamlit port
+EXPOSE 7860
+
+# Run Streamlit app
+CMD ["streamlit", "run", "app.py", "--server.port=7860", "--server.address=0.0.0.0"]

app.py

@@ -0,0 +1,54 @@
+import streamlit as st
+import tensorflow as tf
+import numpy as np
+from PIL import Image
+import json
+
+# Load the trained CNN model
+@st.cache_resource
+def load_model():
+    return tf.keras.models.load_model("model.h5")
+
+model = load_model()
+
+# Function to preprocess a single image
+def preprocess_single_image(pil_img):
+    """
+    Preprocesses a Pillow image for model inference.
+
+    Args:
+        pil_img (PIL.Image.Image): A Pillow image object.
+
+    Returns:
+        preprocessed_img (tf.Tensor): Preprocessed image tensor.
+    """
+    img = pil_img.convert("RGB")  # Convert to RGB
+    img = img.resize((224, 224))  # Resize
+    img = np.array(img)  # Convert to NumPy array
+    img = tf.keras.applications.efficientnet.preprocess_input(img)  # Apply EfficientNet preprocessing
+    img = tf.expand_dims(img, axis=0)  # Add batch dimension
+    return img
+
+# Load class labels
+CLASS_NAMES = json.load(open("class.json", "r"))
+
+st.title("🃏 Card Classification with CNN")
+st.write("Upload an image to classify and visualize the top predictions.")
+
+# Upload image
+uploaded_file = st.file_uploader("📂 Choose an image...", type=["jpg", "png", "jpeg"])
+
+if uploaded_file is not None:
+    image = Image.open(uploaded_file)
+    st.image(image, caption="🖼 Uploaded Image", use_container_width=True)
+
+    # Preprocess image
+    img = preprocess_single_image(image)
+
+    # Predict
+    predictions = model.predict(img)
+    predicted_class_index = np.argmax(predictions)  # Get highest probability index
+    predicted_class = CLASS_NAMES[str(predicted_class_index)]  # Get class label
+
+    # Display predictions
+    st.write(f"Predictions Card : { predicted_class }")

prepare.py

@@ -0,0 +1,30 @@
+from huggingface_hub import hf_hub_download
+import shutil
+import os
+
+# Hugging Face model details
+REPO_ID = "xcurv/kcv-vanguard-day3-cnn-card-classification"
+MODEL_FILENAME = "model.h5"
+LABEL_JSON = "class.json"
+
+def download_file(repo_id, filename, dest_filename):
+    file_path = hf_hub_download(repo_id=repo_id, filename=filename)
+
+    # Get the absolute path of the actual file (resolves symlinks)
+    real_file_path = os.path.realpath(file_path)
+
+    # Ensure the resolved file exists before copying
+    if not os.path.exists(real_file_path):
+        raise FileNotFoundError(f"Resolved file path does not exist: {real_file_path}")
+
+    # Copy the actual file to the destination
+    shutil.copy2(real_file_path, dest_filename)
+    os.remove(real_file_path)
+
+# Download model if not present
+if not os.path.exists(MODEL_FILENAME):
+    download_file(REPO_ID, MODEL_FILENAME, MODEL_FILENAME)
+
+# Download label JSON if not present
+if not os.path.exists(LABEL_JSON):
+    download_file(REPO_ID, LABEL_JSON, LABEL_JSON)

requirements.txt

@@ -0,0 +1,5 @@
+streamlit
+tensorflow
+numpy
+pillow
+huggingface_hub