Help on fine tuning this pretrained model
#1
by
jithendraKrishna
- opened
I’m very much intrested in finetuning the enhancement model, but whatever source i look into i couldn’t do it, can some one please help me on what am i missing and help with this issue ?
I have image to image dataset, but the model’s architecture is complex so i couldn’t figure out how to finetune.
Any help would be very much appreciated. Thank You in advance..
cc @sayakpaul
Have you made any advancements? My team wants to fine-tune MAXIM for one very specific task and I wonder what the training step should look like, how loss function should be implemented and what should the last layer look like? I don't see any fine-tuning posts or information on this model whatsoever and many demos are killed.
Thanks in advance