Upload folder using huggingface_hub
Browse files- Dockerfile +16 -0
- app.py +101 -0
- pneumonia_resnet_model.h5 +3 -0
- requirements.txt +11 -0
Dockerfile
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
FROM python:3.9-slim
|
| 2 |
+
|
| 3 |
+
# Set the working directory inside the container
|
| 4 |
+
WORKDIR /app
|
| 5 |
+
|
| 6 |
+
# Copy all files from the current directory to the container's working directory
|
| 7 |
+
COPY . .
|
| 8 |
+
|
| 9 |
+
# Install dependencies from the requirements file without using cache to reduce image size
|
| 10 |
+
RUN pip install --no-cache-dir --upgrade -r requirements.txt
|
| 11 |
+
|
| 12 |
+
# Define the command to start the application using Gunicorn with 4 worker processes
|
| 13 |
+
# - `-w 4`: Uses 4 worker processes for handling requests
|
| 14 |
+
# - `-b 0.0.0.0:7860`: Binds the server to port 7860 on all network interfaces
|
| 15 |
+
# - `app:app`: Runs the Flask app (assuming `app.py` contains the Flask instance named `app`)
|
| 16 |
+
CMD ["gunicorn", "-w", "4", "-b", "0.0.0.0:7860", "app:pneumonia_detector_api"]
|
app.py
ADDED
|
@@ -0,0 +1,101 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Import necessary libraries
|
| 2 |
+
import numpy as np
|
| 3 |
+
import tensorflow as tf
|
| 4 |
+
from flask import Flask, request, jsonify
|
| 5 |
+
from tensorflow.keras.models import load_model
|
| 6 |
+
import os
|
| 7 |
+
|
| 8 |
+
# Initialize Flask App
|
| 9 |
+
pneumonia_detector_api = Flask("Pneumonia Detection API")
|
| 10 |
+
|
| 11 |
+
# Initialize Flask App
|
| 12 |
+
pneumonia_detector_api = Flask("Pneumonia Detection API")
|
| 13 |
+
|
| 14 |
+
# Load trained model
|
| 15 |
+
model = load_model("/content/drive/MyDrive/AIML Practice/Python Basics/deployment_files/capstone_backend_files/pneumonia_resnet_model.h5")
|
| 16 |
+
|
| 17 |
+
print("Model loaded successfully.")
|
| 18 |
+
|
| 19 |
+
IMG_SIZE = 224
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
# -------------------------------
|
| 23 |
+
# Home Endpoint
|
| 24 |
+
# -------------------------------
|
| 25 |
+
@pneumonia_detector_api.get('/')
|
| 26 |
+
def home():
|
| 27 |
+
return "Welcome to the Pneumonia Detection API"
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
# -------------------------------
|
| 31 |
+
# Image Preprocessing
|
| 32 |
+
# -------------------------------
|
| 33 |
+
def preprocess_image(img_array):
|
| 34 |
+
|
| 35 |
+
# Resize image
|
| 36 |
+
img = tf.image.resize(img_array, (IMG_SIZE, IMG_SIZE))
|
| 37 |
+
|
| 38 |
+
# Normalize
|
| 39 |
+
img = img / 255.0
|
| 40 |
+
|
| 41 |
+
# Ensure 3 channels
|
| 42 |
+
if img.shape[-1] == 1:
|
| 43 |
+
img = np.repeat(img, 3, axis=-1)
|
| 44 |
+
|
| 45 |
+
# Add batch dimension
|
| 46 |
+
img = np.expand_dims(img, axis=0)
|
| 47 |
+
|
| 48 |
+
return img
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
# -------------------------------
|
| 52 |
+
# Prediction Endpoint
|
| 53 |
+
# -------------------------------
|
| 54 |
+
@pneumonia_detector_api.post('/v1/predict')
|
| 55 |
+
def predict_pneumonia():
|
| 56 |
+
|
| 57 |
+
if 'file' not in request.files:
|
| 58 |
+
return jsonify({"error": "No file uploaded"}), 400
|
| 59 |
+
|
| 60 |
+
file = request.files['file']
|
| 61 |
+
|
| 62 |
+
try:
|
| 63 |
+
|
| 64 |
+
# Load numpy image
|
| 65 |
+
img_array = np.load(file)
|
| 66 |
+
|
| 67 |
+
# Preprocess
|
| 68 |
+
processed_img = preprocess_image(img_array)
|
| 69 |
+
|
| 70 |
+
# Predict
|
| 71 |
+
prediction = model.predict(processed_img)[0][0]
|
| 72 |
+
|
| 73 |
+
probability = float(prediction)
|
| 74 |
+
|
| 75 |
+
if probability > 0.5:
|
| 76 |
+
label = "Pneumonia"
|
| 77 |
+
else:
|
| 78 |
+
label = "Normal"
|
| 79 |
+
|
| 80 |
+
result = {
|
| 81 |
+
"Predicted_Class": label,
|
| 82 |
+
"Probability_Pneumonia": round(probability, 4)
|
| 83 |
+
}
|
| 84 |
+
|
| 85 |
+
return jsonify(result)
|
| 86 |
+
|
| 87 |
+
except Exception as e:
|
| 88 |
+
|
| 89 |
+
return jsonify({"error": str(e)}), 500
|
| 90 |
+
|
| 91 |
+
|
| 92 |
+
# -------------------------------
|
| 93 |
+
# Run Flask Server
|
| 94 |
+
# -------------------------------
|
| 95 |
+
if __name__ == "__main__":
|
| 96 |
+
|
| 97 |
+
pneumonia_detector_api.run(
|
| 98 |
+
host="0.0.0.0",
|
| 99 |
+
port=5000,
|
| 100 |
+
debug=True
|
| 101 |
+
)
|
pneumonia_resnet_model.h5
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:249c043f1fa1352615ab6d9e9b3c99fc8ca3acc5f16a2c1a656b9640c7805eee
|
| 3 |
+
size 169100952
|
requirements.txt
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
pandas==2.2.2
|
| 2 |
+
numpy==2.0.2
|
| 3 |
+
scikit-learn==1.6.1
|
| 4 |
+
xgboost==2.1.4
|
| 5 |
+
joblib==1.4.2
|
| 6 |
+
Werkzeug==2.2.2
|
| 7 |
+
flask==2.2.2
|
| 8 |
+
gunicorn==20.1.0
|
| 9 |
+
requests==2.28.1
|
| 10 |
+
uvicorn[standard]
|
| 11 |
+
streamlit==1.43.2
|