How to load the datset :
-> Go to kaggle.com and open an account -> Go to setting and Create a new Token -> A kaggle.json file will be downloaded into your designated folder. -> Upload the .json file into google colab. And run !cp kaggle.json ~/.kaggle/ -> Then go to the dataset on kaggle , Link : https://www.kaggle.com/datasets/sartajbhuvaji/brain-tumor-classification-mri -> Copy API command from clicking on the three dots on the right of the screen. -> Paste the api commad on tou Google Colab Cell. !kaggle datasets download -d sartajbhuvaji/brain-tumor-classification-mri -> After that follow the given code to run a CNN on classifying brain tumor.
I used a basic cnn architecture to classify betwwn Glioma, Meningioma, NoTumor and Pituitary. You can tweak as much as you can with the code and dataset given. Maybe even have better result.