Closed emjay73 closed 1 year ago
Thank you for sharing the code!
Can you also share the hardware environment required to run this code? (minimum memory requirements, GPU spec.. etc)
Hi, @emjay73, I think 24GB should be enough for training with a small batch size according to my preliminary experiments. The released model checkpoints are trained with one A100 (~81GB).
Oh I get it. Thanks!
Thank you for sharing the code!
Can you also share the hardware environment required to run this code? (minimum memory requirements, GPU spec.. etc)