Mnist-on-EvaDB-and-DeepLake

Install Requirements

pip install -r EvaDB_requirements.txt

Run EvaDB mnist

python EvaDB_mnist.py

Input Kaggle key

Will have to input kaggle username and key to download the dataset Can find it at the settings in your kaggle profile Click create new token, then use the info in the downloaded kaggle.json file

Note

The code will run for a while ...

Run DeepLake mnist

pip install -r DeepLake_requirements.txt

Input Kaggle key

Will have to input kaggle username and key to download the dataset Can find it at the settings in your kaggle profile Click create new token, then use the info in the downloaded kaggle.json file

Run DeepLake mnist

In order to run the file on different datasets might have to change it manually

So I pprovided the outputs of the cProfile and the timing stats in the timing and profile folder

This will show the results after getting the predictions using MnistImage Classifier

python DeepLake_mnist.py

This runs the TQuery using local dataset

python Create_DeepLake_Dataset.py

This runs the TQuery using activeloop dataset that is stored in https://app.activeloop.ai/yuchenjohnlin/home

But the performance will be effected by the network

python DeepLake_Query.py

This runs the User defined function using local dataset

python DeepLake_UDFilter.py