Does the GPU I use for TAO affect model acuracy?

If I am running TAO in Google Cloud, how does the GPU I choose for the VM affect training? For example, if I am choosing between a T4, A100, or H100 GPU, how will this affect the model I deploy to my Jetson device? I know that the training will take longer depending on the GPU I choose, but does it affect how accurate my model will be when I deploy it? Are there other factors of the model it will impact?

Thanks,
Andrew

Training with different gpus will not affect model accuracy with the same training spec file .

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.