AICTE Training and Learning Short Term Course on Data Science and Machine Learning
High level process steps for research to production in building deeplearning models
Introduce a image classification problem and CNNs and how does it work? https://docs.google.com/presentation/d/11d7a71i-t4Mi5i_HaHWgsI2OWfyC0OQ-7s4Gpb4IkVQ/edit?usp=sharing
Visit Google Colab at https://colab.research.google.com/ Signin with your gmail here. please ensire you will be storing data in your google drive here.
Create a new notebook and rename it with deeplearning_yourname.ipynb for eg in my case i would rename as deeplearning_puneetindal.ipynb where .ipynb is the extension of the jupyter notebook
click connect
if you are not on github then do crete a account on github its free
Pull the repository from github clone the code at https://github.com/EduwaiveFoundation/aictedeeplearning.git
Scrape images data collect data on local and uploading it or directly push to gcs and explain the scraper and how to write a new one - 1.5 hours. Show some insights on the image data by hitting gcp image analysis api. Introduction to data management. Show example of GCS and S3 if possible. Uploading the data. Show some examples of moving data from bucket to machine. Or google drive to colab machine Once data is in colab machine train the models and deploy it. Explain training and canarying process.
You can find the full notebook at below link which we will complete till the end of the sessions. You can use this notebook for reference purposes after you try all of the code https://colab.research.google.com/drive/13CARam6CSb8D0SU7845iQND8tBA92-wo