Bibhuti5 / repository

Potato-Disease-Classification

Potato Disease Classification Setup for Python: Install Python (Setup instructions) Install Python packages pip3 install -r training/requirements.txt pip3 install -r api/requirements.txt Install Tensorflow Serving (Setup instructions) Setup for ReactJS Install Nodejs (Setup instructions) Install NPM (Setup instructions) Install dependencies cd frontend npm install --from-lock-json npm audit fix Copy .env.example as .env. Change API url in .env. Setup for React-Native app Initial setup for React-Native app(Setup instructions) Install dependencies cd mobile-app yarn install cd ios && pod install && cd ../ Copy .env.example as .env. Change API url in .env. Training the Model Download the data from kaggle. Only keep folders related to Potatoes. Run Jupyter Notebook in Browser. jupyter notebook Open training/potato-disease-training.ipynb in Jupyter Notebook. In cell #2, update the path to dataset. Run all the Cells one by one. Copy the model generated and save it with the version number in the models folder. Running the API Using FastAPI Get inside api folder cd api Run the FastAPI Server using uvicorn uvicorn main:app --reload --host 0.0.0.0 Your API is now running at 0.0.0.0:8000 Using FastAPI & TF Serve Get inside api folder cd api Copy the models.config.example as models.config and update the paths in file. Run the TF Serve (Update config file path below) docker run -t --rm -p 8501:8501 -v C:/Code/potato-disease-classification:/potato-disease-classification tensorflow/serving --rest_api_port=8501 --model_config_file=/potato-disease-classification/models.config Run the FastAPI Server using uvicorn For this you can directly run it from your main.py or main-tf-serving.py using pycharm run option (as shown in the video tutorial) OR you can run it from command prompt as shown below, uvicorn main-tf-serving:app --reload --host 0.0.0.0 Your API is now running at 0.0.0.0:8000 Running the Frontend Get inside api folder cd frontend Copy the .env.example as .env and update REACT_APP_API_URL to API URL if needed. Run the frontend npm run start Running the app Get inside mobile-app folder cd mobile-app Copy the .env.example as .env and update URL to API URL if needed. Run the app (android/iOS) npm run android or npm run ios Creating the TF Lite Model Run Jupyter Notebook in Browser. jupyter notebook Open training/tf-lite-converter.ipynb in Jupyter Notebook. In cell #2, update the path to dataset. Run all the Cells one by one. Model would be saved in tf-lite-models folder. Deploying the TF Lite on GCP Create a GCP account. Create a Project on GCP (Keep note of the project id). Create a GCP bucket. Upload the tf-lite model generate in the bucket in the path models/potato-model.tflite. Install Google Cloud SDK (Setup instructions). Authenticate with Google Cloud SDK. gcloud auth login Run the deployment script. cd gcp gcloud functions deploy predict_lite --runtime python38 --trigger-http --memory 512 --project project_id Your model is now deployed. Use Postman to test the GCF using the Trigger URL. Inspiration: https://cloud.google.com/blog/products/ai-machine-learning/how-to-serve-deep-learning-models-using-tensorflow-2-0-with-cloud-functions Deploying the TF Model (.h5) on GCP Create a GCP account. Create a Project on GCP (Keep note of the project id). Create a GCP bucket. Upload the tf .h5 model generate in the bucket in the path models/potato-model.h5. Install Google Cloud SDK (Setup instructions). Authenticate with Google Cloud SDK. gcloud auth login Run the deployment script. cd gcp gcloud functions deploy predict --runtime python38 --trigger-http --memory 512 --project project_id Your model is now deployed. Use Postman to test the GCF using the Trigger URL. Inspiration: https://cloud.google.com/blog/products/ai-machine-learning/how-to-serve-deep-learning-models-using-tensorflow-2-0-with-cloud-functions