CCSR

CCSR copied to clipboard

Hello, thanks a lot for your work! I currently work on grayscale CT-images super-resolution. As I went through both steps of the training process using pre-trained stable diffusion the results...

pihchikk

Does Colab Work ? Error Occured.

1

When I run [CCSR colab demo](https://colab.research.google.com/github/camenduru/CCSR-colab/blob/main/CCSR_colab.ipynb) that can be found in `readme.md` , Error occured like below. ``` ERROR: pip's dependency resolver does not currently take into account all the...

coleea

Got the following error when training stage2 `Missing key(s) in state_dict: "betas_inter", "alphas_cumprod_inter", "alphas_cumprod_prev_inter", "sqrt_alphas_cumprod_inter", "sqrt_one_minus_alphas_cumprod_inter", "log_one_minus_alphas_cumprod_inter", "sqrt_recip_alphas_cumprod_inter", "sqrt_recipm1_alphas_cumprod_inter", "posterior_variance_inter", "posterior_log_variance_clipped_inter", "posterior_mean_coef1_inter", "posterior_mean_coef2_inter", "decoder_loss.logvar", "decoder_loss.perceptual_loss.scaling_layer.shift", "decoder_loss.perceptual_loss.scaling_layer.scale", "decoder_loss.perceptual_loss.net.slice1.0.weight", "decoder_loss.perceptual_loss.net.slice1.0.bias", "decoder_loss.perceptual_loss.net.slice1.2.weight", "decoder_loss.perceptual_loss.net.slice1.2.bias...`...

yscylhy

Use case of CCSR compared to SeeSR

1

Please can you comment on when one would use CCSR vs SeeSR? Both appear to have similar objectives? How does CCSR perform vs SeeSR? Thanks

mashoutsider

trian model CUDA out of memory

7

Is there any way to train on 24G on a GTX3090, even with one batch size? torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 3; 23.69 GiB...

aoyang-hd

No module named Taming found

1

How can I fix this error? File "importlib\__init__.py", line 126, in import_module File "", line 1204, in _gcd_import File "", line 1176, in _find_and_load File "", line 1147, in _find_and_load_unlocked...

aifuzz59

How to set the value of queue_size? I'm still confused.

1


zhouyizhuo

How many gpus did you use to train the model? And how many days did it cost?

2

han265

Do you use prompt during training?

1

Great job! I am wondering do you use prompt to guide the process of restoration? why and why not?

martinliuxy

why use one-step sampling

1

很棒的工作，感谢作者分享！ 有个小问题想请教一下：请问推理过程中的x_T是纯噪声吗，x_T到x_tmax这一步的作用是什么？为什么不直接用LR加对应噪声得到x_tmax, 或者x_T逐步去噪直到x_tmin?

Harperrrr111