Running Yolov8 on Google Colab leveraging GCP GPUs

[frankvp11]

Hello!

I'm sure that many of you might find my question/request a trivial problem, and that I am just inexperienced, which I am. This will be my first time using GCP so I would like to be careful so that I don't get billed hundreds of dollars. 

With that out of the way, I could use some help. I have this colab file (https://colab.research.google.com/drive/13tLrfRPPc7m3V7X7Pdwjrv1w7RXC8CFR#scrollTo=JjooafMMr6zo ) and I would like to run the training portion of it. The problem is, the 100k dataset is far to big for colab's gpus. This led me to find GCP which provides essentially free GPU's with its $300 credit free trial. This leads me to the true problem, I have no idea how I would go about running this colab file (or the code in general) with GCP's GPU's. I've looked at a couple tutorials already, and many of them talk about port-forwarding jupyter things and whatnot which I have no idea how to use. Does anyone have/know about good tutorials which go really in depth (like for complete and total dummies (also preferably a video)) of how to essentially run my colab file, or is anyone willing to step by step work with me to get this done?

Thanks,

Frank 

[frankvp11]

I have found this repository. https://github.com/adhil0/colab_ssh

This shows me mostly everything that I needed as for one of the portions. The new "problem" is I don't know how to set up an instance (I will google this later). Also, will the $300 be enough (most likely, I know exact is difficult) to run my code?

Thanks again