IceClear/StableSR issues and pull requests

#47 - Inability to fetch new data.

Issue - State: closed - Opened by Suummqq over 1 year ago - 2 comments

#46 - 512 vs 768

Issue - State: open - Opened by Skumball over 1 year ago - 2 comments

#45 - Question about performance gap on test set

Issue - State: open - Opened by amandaluof over 1 year ago - 10 comments

#44 - Stuck in epoch 0%

Issue - State: closed - Opened by misi0202 over 1 year ago - 4 comments

#43 - Google colab not working even with PRO and high-RAM

Issue - State: open - Opened by tralala87 over 1 year ago - 19 comments

#42 - Running out of RAM in free google colab

Issue - State: closed - Opened by tralala87 over 1 year ago - 2 comments

#41 - Why I get killed?

Issue - State: open - Opened by misi0202 over 1 year ago - 8 comments

#40 - problem when inference with 768v model

Issue - State: closed - Opened by cswry over 1 year ago - 1 comment

#39 - Sampling strategy DDPM vs DDIM

Issue - State: closed - Opened by claudiom4sir over 1 year ago - 2 comments

#38 - Is the weight of VQGAN same as the Encoder-Decoder of Stable Diffusion?

Issue - State: open - Opened by 543877815 over 1 year ago - 1 comment

#37 - RuntimeError: Error(s) in loading state_dict for EncoderUNetModelWT

Issue - State: open - Opened by Experienment-Gu over 1 year ago - 1 comment

#36 - Question for Finetuning

Issue - State: closed - Opened by xyIsHere over 1 year ago - 17 comments

#35 - How can I train the model with SD 1.5 (instead of SD 2.1)?

Issue - State: open - Opened by WanquanF over 1 year ago - 7 comments

#34 - CFW module weight didn't work as expected.

Issue - State: open - Opened by JackeyDeng over 1 year ago - 5 comments

#33 - Blurred Image

Issue - State: closed - Opened by Yu-Yang-Wang over 1 year ago - 2 comments

#32 - SFT dataset

Issue - State: open - Opened by makeyyp over 1 year ago - 4 comments

#31 - [Question] training with higher resolution

Issue - State: closed - Opened by dill-shower over 1 year ago - 2 comments

#30 - Feature Mismatch for Replication

Issue - State: closed - Opened by kravrolens over 1 year ago - 4 comments

#29 - question about using LDM or DM

Issue - State: closed - Opened by wwbwang over 1 year ago - 2 comments

#28 - Hello, author. I would like to ask how the files in the "latents" directory under "CFW_trainingdata" are generated?

Issue - State: closed - Opened by wenwenpg over 1 year ago - 1 comment

#27 - Where do hyperparameters of RealESRGAN degradation come from?

Issue - State: open - Opened by jan-pair over 1 year ago - 6 comments

#26 - Replication issue

Issue - State: open - Opened by ITBeyond1230 over 1 year ago - 10 comments

#25 - about vqgantile_size and vqgantile_stride

Issue - State: closed - Opened by zcdliuwei over 1 year ago - 2 comments

#24 - train dataset

Issue - State: closed - Opened by kia350 over 1 year ago - 10 comments

#23 - how to train a vqgan model to conduct special task?

Issue - State: closed - Opened by CunkuanLei over 1 year ago - 2 comments

#22 - Training Details

Issue - State: closed - Opened by rongakowang over 1 year ago - 5 comments

#21 - Diverse super-resolution results

Issue - State: closed - Opened by USTC-JialunPeng over 1 year ago - 2 comments

#20 - It seems that StableSR uses VAE (VAEGAN) instead of VQGAN.

Issue - State: closed - Opened by USTC-JialunPeng over 1 year ago - 1 comment
Labels: good first issue

#19 - Get the wired output image when using the pretrained model

Issue - State: closed - Opened by YangGangZhiQi over 1 year ago - 2 comments

#18 - Implementation details about inserting SFT layers

Issue - State: closed - Opened by zhenxili96 over 1 year ago - 1 comment

#17 - testing any other size image

Issue - State: closed - Opened by ouyangjiacs over 1 year ago - 9 comments

#16 - Test with multiple GPUs

Issue - State: closed - Opened by rongakowang over 1 year ago - 5 comments

#15 - StableSR distorts Chinese characters in image

Issue - State: closed - Opened by ITBeyond1230 over 1 year ago - 3 comments

#14 - finetune

Issue - State: closed - Opened by flynnamy over 1 year ago - 10 comments

#13 - batch size is 192 in training? CUDA OOM

Issue - State: closed - Opened by zhenxili96 over 1 year ago - 5 comments

#12 - How to prepare CFW_trainingdata?

Issue - State: closed - Opened by kravrolens over 1 year ago - 9 comments

#11 - Which pre-trained models need to be downloaded in your code.

Issue - State: closed - Opened by ouyangjiacs over 1 year ago - 11 comments

#10 - How to use the LR images in training?

Issue - State: closed - Opened by zhenxili96 over 1 year ago - 1 comment

#9 - OOM

Issue - State: closed - Opened by zcdliuwei over 1 year ago - 4 comments

#8 - Question about Autoencoder loss function, and training pipeline confirmation

Issue - State: closed - Opened by DZY-irene over 1 year ago - 2 comments

#7 - What's the latent and sample mean

Issue - State: closed - Opened by ligenjie over 1 year ago - 4 comments

#6 - Linked Colab Demo doesn't work.

Issue - State: closed - Opened by jarredou over 1 year ago - 12 comments

#5 - Why does this error occur?

Issue - State: closed - Opened by ARTUROSING over 1 year ago - 4 comments

#4 - Tile strategy: CUDA out of memory.

Issue - State: closed - Opened by msk7131 over 1 year ago - 2 comments
Labels: good first issue

#3 - Cuda OOM

Issue - State: closed - Opened by Inferencer over 1 year ago - 4 comments

#2 - Additional comparisons to Tiled DDPM, ControlNet Tile, Loopback Scaler and DeepFloyed.

Issue - State: closed - Opened by UIUC-Marisa3 over 1 year ago - 11 comments
Labels: good first issue

#1 - about Zoom in and add more details

Issue - State: closed - Opened by zhanghongyong123456 over 1 year ago - 3 comments

GitHub / IceClear/StableSR issues and pull requests