Hi all,
I was just wondering - what do people use for computational infrastructure, for this challenge?
I'm used to Kaggle, and I'm new to Synapse.
Maybe Google/Colab Pro is enough - but I'll need to figure out a way to reconstruct the docker image for it.
Created by giovannidallolio You're welcome! Good Luck! Thank you. I tried Colab - it was struggling a bit to handle all the data and the training, so I think I'll go with Colab Pro. Hi,
Google Colab should work fine for this amount of data. Just read only the columns you need to save memory. Colab Pro should be enough for most models.
Synapse doesn’t provide a development environment. You’ll build models on your local machine or any cloud service and submit your results as a CSV file.