“Understand more on ourselves through analysis”
The reason why I choose WhatsApp as my analysis is because I always wanted to know more about my chat room behaviour. And after reading some of the notebooks from Prajwal Prashanth at Jovian.ml. It really inspired me to start this analysis as my Final Course project.
In this project, I will attempt to find out what normally I will do in a group chat with my friends such as the active hours we usually talk and the number of emoji we use in the chat. …
So today our objective is to see the difference in the value of accuracy using a different method to train our Image Classification model:
Feel free to check out my notebook: https://jovian.ml/edsenmichaelcy/flower-classification/v/12
Step 1: Import the file we needed and put the dataset into data_dir
Step 2: Data augmentation & normalization
Step 3: Check the dataset classes and label them
Step 4: Functions to show a single picture and batch picture
Step 5:Split the training data and the validity data
Step 6: Choose the batch size, put in DataLoader and show the batch
Step 7: Get GPU up on running
Step 8: Training the Image Classification using basic CNN
Step 9:Training and Validation Datasets
Step 10: Training the model with CNN
Step 11: Predict and test the model
Step Resnet50: Transfer Learning method using Resnet50 (Pre-trained)
Step Resnet18: Transfer Learning method using Resnet18 (Pre-trained)
Final Step: Conclusion comparing the…
So what is PyTorch why everyone is so heaping about? PyTorch is an open-source machine learning framework that mainly focuses on fast tracks the path from research prototyping to production deployment. PyTorch was primarily developed by the Facebook artificial intelligence research team. Moreover, PyTorch is used for applications such as computer vision and natural language processing.
Let stop wasting time and start!