FlowModels

FlowModels copied to clipboard

So many to learn! Thanks for your hard work! Good start for me

xiao2mo

student_loss.backward() in LADD

3

student_loss.backward() torch.nn.utils.clip_grad_norm_(student_unet.parameters(), 1.0) student_optimizer.step() student_scheduler.step() student_optimizer.zero_grad() Wouldn' t above code generate gradients on the discriminator as well? Then in the next training iter, those gradients on the discriminator will be...

jzhang38

Hello, thank you very much for open-sourcing this project for us to learn from. I also came across Nitro（https://github.com/AMD-AGI/Nitro-1/blob/main/train.py） and noticed that in your training process, you first train the...

Run-19