6:30 PM – AI / ML – Shawn

Today We Did

1. Started our Cats Vs Dogs project
   1. Learned about Fastai, file paths, dataloaders, and machine learning models
2. https://colab.research.google.com/drive/1YxHrVsv_qDhywniQqR0RAKTAdR8q3Hno?usp=sharing

Homework

1. You might have to re-run all the cells to use your model if you leave the page and come back to it.
   1. Go to Runtime -> Change Runtime and make sure T4 GPU is selected
   2. 
   3. Go to Runtime -> Run All
2. Try predicting some images using learn.predict(files[0]). Play around with the 0 to change around the images as well!
3. Run learn.show_results() a couple times. See how your learner does.
4. Add the following code:

   from google.colab import files
   uploaded_file = files.upload()

   img_content = None
   for file_name in uploaded_file.keys():
       img_content = uploaded_file[file_name]

   1. Above code will provide an upload button. Click the upload button and upload a cat or dog image of your own. You can find a cat or dog image from google.
   2. Try passing in the img_content you upload into learn.predict, and see the result. If you upload a cat image then, learn.predict(img_content) will return True. If not will return False.

Email me at szhuang@ayclogic.com if you have any questions!