Hello,
I am currently trying to train and fine-tune your model. I am training the diffusion model using only the BTCV dataset. However, even after training for 20,000 iterations, the generated results are still quite blurry.
I also tried fine-tuning your pretrained diffusion model weights on the BTCV dataset, but after fine-tuning, the generation quality seems to degrade significantly and the results are not very good.