About training spec

[chenshihfang]

Hi,

Has this model been implemented without any training process? And is the only required training process the training of HQ-SAM independently? Alternatively, please provide the GPT/memory usage details for training this model.

Thank you.