I had tinkered with the course a couple times before but didn’t get started in earnest. Before, it seemed like the course strongly suggested Google Colab, but now with the recent version of the course, it seems like the standard set up uses Kaggle notebooks instead? Is that correct?
in lesson 3 ,I am trying to launch gradio app after predicting the model in kaggle and see it in browser .
It is showing me “ERROR: Exception in ASGI application”.
I’m excited for this course and book! One thing I’m curious about is why the course is written with examples in Kaggle notebooks, but the book is written in Colab notebooks. It looks like both are Jupyter notebooks. What are the main differences between Kaggle and Colab?
The book was written a couple of years ago when (I presume) Colab provided the best free performance, and/or best multi-notebook folder structure - at the time. An instructor led online lesson could more easily utilise individual/separate Kaggle notebooks. btw, the “current best” service recommended within the 2022 course videos wsa Paperspace.
Also often people ask about “how to get started in Kaggle competitions?”.
So it is a good way to get started and break down that ‘getting started on Kaggle’ barrier at the same time, whilst introducing a place to practice and upskill. win/win
Hello! I have a very silly question and am embarrassed to ask .
Can someone tell me how do I restart a kernel in Kaggle?
I tried pip installing the fast.ai library and it showed the following:
I don’t think you should ever be embarrassed to ask questions … that’s the only way to learn
I believe if you tap the “Stop Session” option in the menu where you had the “Restart & Clear Cell Outputs” option, it stops your current Kaggle session and then the button will turn to a “Start Session” and I believe once you do that, the kernel is restarted … You can try that and see but I don’t use Kaggle much and so am going by what I think is how it should work …
If that doesn’t work, please let us know and I’m sure somebody will help … Or I’ll do a bit of digging
You don’t need to import fastbook on kaggle. You can just import what you need directly. BTW in your screenshot it doesn’t look like you pip installed fastbook?