Training on t4 gpu

[Lenos500]

Hello, I'm trying to train a model using a dataset of 1k images on a T4 gpu using colab free trial. I can keep it running for 12 hours daily then resume the training the following day. How much time do you think I need to train in order to get convincing results? Should I increase the dataset?

[lucidrains]

really depends on the resolution

but 1k images is not enough, that's for sure

[Lenos500]

really depends on the resolution

but 1k images is not enough, that's for sure

What if it's on the 512*512 resolution using around 2-3k images then how much time do you estimate is needed?