2026-01-28 - 2026-02-04
Overview
There has been no commit activity in this period.
158 Issues closed from 1 user
Closed
#633 [PR #631] [CLOSED] Feat/upgrade backends ci
Closed
#634 [PR #637] Create Manohar kumar
Closed
#622 [PR #524] [CLOSED] moving processing image log out of loop
Closed
#623 [PR #485] Create CNAME
Closed
#613 [PR #408] [CLOSED] Add GPU acceleration to Apple Silicon Macs
Closed
#618 [PR #428] Create SECURITY.md
Closed
#609 [PR #387] [CLOSED] Create Amit 06
Closed
#606 [PR #357] Fix Python dependency API risk issue
Closed
#600 [PR #288] very nice Graphical user interface
Closed
#593 [PR #273] Optimize image(sample) size to smaller one
Closed
#587 [PR #248] [MERGED] update replicate
Closed
#586 [PR #210] Fix basicsr losses import
Closed
#583 [PR #198] [CLOSED] Updated the link to the Baseten sample app
Closed
#576 [PR #138] changed readme
Closed
#579 [PR #158] [MERGED] Add V1.3 model
Closed
#580 [PR #171] Added link to pre-trained model GFPGANv1.3
Closed
#571 [PR #117] [bug fix] No module named 'gfpgan.version'
Closed
#572 [PR #121] Update README.md
Closed
#566 [PR #113] [CLOSED] Updated README.md to fix a typo
Closed
#561 [PR #54] Reordered requirements
Closed
#559 [PR #64] [CLOSED] Add Docker environment & web demo
Closed
#560 [PR #52] Gradio Web Demo
Closed
#558 [PR #43] Update README.md
Closed
#546 Reza
Closed
#544 make the people face clear
Closed
#540 women
Closed
#536 import os import torch import RRDBNet_arch as arch pretrained_net = torch.load('./models/RRDB_ESRGAN_x4.pth') save_path = './models/RRDB_ESRGAN_x4.pth' crt_model = arch.RRDBNet(3, 3, 64, 23, gc=32) crt_net = crt_mode)
Closed
#531 Bruno Samora
Closed
#525 NameError: name 'fused_act_ext' is not defined
Closed
#523 NameError: name 'fused_act_ext' is not defined
Closed
#518 No module named 'basicsr'
Closed
#516 done
Closed
#511 Fix a picture
Closed
#495 Photo upload problem
Closed
#491 关于数据集的问题
Closed
#488 Please help me
Closed
#484 ls: cannot access 'results/cmp': No such file or directory
Closed
#472 Gfpgan Not working on colab
Closed
#459 gfpgan file
Closed
#453 Mamun
Closed
#440 How do I use GPU for the enhancement process ?
Closed
#432 旧照片修复
Closed
#413 Photos quality
Closed
#407 hi
Closed
#405 [PR #96]: https://github.com/tqdm/tqdm/pull/96
Closed
#398 H
Closed
#377 ERROR: Could not find a version that satisfies the requirement tb-nightly (from gfpgan) (from versions: none)
Closed
#366 pip install gfpgan==1.3.8 error
Closed
#365 Add device support for mps
Closed
#362 Allow to define model_rootpath for FaceRestoreHelper
Closed
#355 Why is the effect of processing photos locally not as good as online?
Closed
#356 怎样去训练一个1.4版本的GFPGAN
Closed
#338 Mahfuz Emon
Closed
#337 How to force GFPGAN run only on GPU
Closed
#328 [Feature Request] Please make available through pip
Closed
#317 GFPGAN
Closed
#307 AI
Closed
#305 please set recompute_scale_factor=True
Closed
#303 I get a black background
Closed
#302 How do I force it to work on the CPU?
Closed
#299 [Question] How do I make GFPGAN colorize images on huggingface?
Closed
#297 Pixel image
Closed
#298 Cannot find ground truth images in lfw_cropped_faces, CelebChild-Test, and WebPhoto-Test dataset
Closed
#295 H
Closed
#281 Using 4-channels images
Closed
#280 Original trainning model for GFPGANv1.3
Closed
#269 不真实
Closed
#264 Training RuntimeError: stack expects each tensor to be equal size, but got XXX vs YYY
Closed
#260 arcface_resnet18
Closed
#261 CUDA error: out of memory
Closed
#258 (deleted)
Closed
#259 Difference of network structure between training and reference?
Closed
#256 num_batches_tracked
Closed
#251 Pretty obvious margins after pasting the face back to background. How to solve this?
Closed
#253 How to calculate FID?
Closed
#247 TypeError: got an unexpected keyword argument 'model_rootpath'
Closed
#249 identity distance
Closed
#250 close
Closed
#245 Problem with output.
Closed
#242 sampling from latent space
Closed
#240 unexpected EOF, expected 273069 more bytes.
Closed
#238 error: (-215:Assertion failed) !buf.empty() in function 'imdecode_'
Closed
#239 Unable to upload items
Closed
#236 Permission Denied when Installing setup.py
Closed
#234 HI!! Testset!!
Closed
#223 How to use GPU
Closed
#212 Discriminators loading
Closed
#204 Not accepting weights despite being in the right folder
Closed
#202 COLAB no longer working?
Closed
#200 Process with Video
Closed
#194 not
Closed
#189 Feature Request For Automatic Tile Size
Closed
#183 关于背景图像增强
Closed
#181 I want export "GFPGANCleanv1-NoCE-C2.pth" to ONNX
Closed
#156 how trans train models to one model.pth for inference
Closed
#151 How to run it on Windows?
Closed
#145 How to colorize the whole image?
Closed
#140 Inspiration
Closed
#132 running from docker on CPU - does not work
Closed
#131 pth2onnx
Closed
#122 how to get "StyleGAN2GeneratorClean" pretrained model?
Closed
#117 Missing the Code of Conduct
Closed
#115 'NoneType' object has no attribute 'load_state_dict'
Closed
#100 NameError: name 'fused_act_ext' is not defined
Closed
#96 对比 Bringing-Old-Photos-Back-to-Life 不支持裂痕修复
Closed
#94 Permission denied: '/usr/bin/yapf' on arch linux
Closed
#92 About fix decoder
Closed
#86 TypeError: get_face_landmarks_5() got an unexpected keyword argument 'eye_dist_threshold' in latest update
Closed
#79 如何训练不上色的模型
Closed
#80 重新设置图像的退化过程
Closed
#77 RuntimeError: "reflection_pad2d" not implemented for 'Half'
Closed
#75 subprocess.CalledProcessError and RuntimeError: mat1 dim 1 must match mat2 dim 0
Closed
#76 "GFPGANCleanv1-NoCE-C2.pth" file not downloading for me.
Closed
#73 训练iter数
Closed
#72 Pre trained models for transfer learning
Closed
#69 Training with clean version
Closed
#67 About training with 8 gpus
Closed
#65 如何降低美颜效果
Closed
#62 您好，请问为什么我没办法使用V1？
Closed
#59 GAN训练
Closed
#55 Train with GPU and inference without GPU. Is it possible ?
Closed
#51 Finetuning provided model
Closed
#49 Error with the new model
Closed
#50 About the rgb results from from Degradation Removal.
Closed
#47 channel_multiplier
Closed
#46 I can't use GFPGANv1.pth
Closed
#43 FaceRestoreHelper
Closed
#42 Has colorization been removed or only disabled?
Closed
#41 训练问题
Closed
#39 Training on custom dataset
Closed
#38 want to whole image in the result (without cropping)
Closed
#37 results are strange while testing on whole img
Closed
#35 using my own pretrained stylegan2 network
Closed
#34 Love it, thank you! Settings...
Closed
#32 你好大佬，请教几个环境问题
Closed
#30 处理后的整个照片 颜色发棕 , 和原图片的色调不一样 !!
Closed
#31 'BASICSR_JIT' is not recognized as an internal or external command
Closed
#28 train
Closed
#26 use autocrop instead of facexlib
Closed
#27 ImportError: No module named 'deform_conv'
Closed
#24 landmark
Closed
#22 Can we display the middle result in training so that we can know when to stop training
Closed
#23 使用checkpoint继续训练的bug
Closed
#20 训练日志异常
Closed
#18 Complile error when building extension 'deform_conv'
Closed
#19 How to run the demo on pytorch 1.4.0 or 1.1.0?
Closed
#16 Final Quality
Closed
#14 Restore face without colorizing them
Closed
#15 重新训练自己的数据集
Closed
#12 FileNotFoundError: [Errno 2] No such file or directory: 'experiments/pretrained_models/FFHQ_eye_mouth_landmarks_512.pth'
Closed
#10 while freeze the stylegan?
Closed
#11 About the facexlib
Closed
#8 RuntimeError: Could not run 'torchvision::roi_align' with arguments from the 'CUDA' backend.
Closed
#6 NameError: name 'fused_act_ext' is not defined
Closed
#7 Hugging Face Hub Integration
Closed
#4 "stylegan2_arch.py" seems to be missing
Closed
#3 where is the pretrained model FFHQ_eye_mouth_landmarks_512.pth &arcface.....
Closed
#2 Pre-trained models?
634 Issues created by 1 user
Opened
#1 "stylegan2_arch.py" seems to be missing
Opened
#2 Pre-trained models?
Opened
#3 where is the pretrained model FFHQ_eye_mouth_landmarks_512.pth &arcface.....
Opened
#4 "stylegan2_arch.py" seems to be missing
Opened
#5 RuntimeError: Could not run 'torchvision::roi_align' with arguments from the 'CUDA' backend.
Opened
#6 NameError: name 'fused_act_ext' is not defined
Opened
#7 Hugging Face Hub Integration
Opened
#8 RuntimeError: Could not run 'torchvision::roi_align' with arguments from the 'CUDA' backend.
Opened
#9 FileNotFoundError: [Errno 2] No such file or directory: 'experiments/pretrained_models/FFHQ_eye_mouth_landmarks_512.pth'
Opened
#10 while freeze the stylegan?
Opened
#11 About the facexlib
Opened
#12 FileNotFoundError: [Errno 2] No such file or directory: 'experiments/pretrained_models/FFHQ_eye_mouth_landmarks_512.pth'
Opened
#13 Final Quality
Opened
#14 Restore face without colorizing them
Opened
#15 重新训练自己的数据集
Opened
#16 Final Quality
Opened
#17 训练日志异常
Opened
#18 Complile error when building extension 'deform_conv'
Opened
#19 How to run the demo on pytorch 1.4.0 or 1.1.0?
Opened
#20 训练日志异常
Opened
#21 landmark
Opened
#22 Can we display the middle result in training so that we can know when to stop training
Opened
#23 使用checkpoint继续训练的bug
Opened
#24 landmark
Opened
#25 train
Opened
#26 use autocrop instead of facexlib
Opened
#27 ImportError: No module named 'deform_conv'
Opened
#28 train
Opened
#29 你好大佬，请教几个环境问题
Opened
#30 处理后的整个照片 颜色发棕 , 和原图片的色调不一样 !!
Opened
#31 'BASICSR_JIT' is not recognized as an internal or external command
Opened
#32 你好大佬，请教几个环境问题
Opened
#33 Is this possible, to improve?
Opened
#34 Love it, thank you! Settings...
Opened
#35 using my own pretrained stylegan2 network
Opened
#36 Training on custom dataset
Opened
#37 results are strange while testing on whole img
Opened
#38 want to whole image in the result (without cropping)
Opened
#39 Training on custom dataset
Opened
#40 FaceRestoreHelper
Opened
#41 训练问题
Opened
#42 Has colorization been removed or only disabled?
Opened
#43 FaceRestoreHelper
Opened
#44 channel_multiplier
Opened
#45 RuntimeError:"Distributed package doesn't have NCCL" ???
Opened
#46 I can't use GFPGANv1.pth
Opened
#47 channel_multiplier
Opened
#48 Finetuning provided model
Opened
#49 Error with the new model
Opened
#50 About the rgb results from from Degradation Removal.
Opened
#51 Finetuning provided model
Opened
#52 Train with GPU and inference without GPU. Is it possible ?
Opened
#53 Have better result in V1, V2 is too sharp
Opened
#54 will the degradation of real-esrgan replace into GFPGAN?
Opened
#55 Train with GPU and inference without GPU. Is it possible ?
Opened
#56 GAN训练
Opened
#57 Can't use .half() with new GFPGAN clean
Opened
#58 how can i change the face area to be cropped larger?
Opened
#59 GAN训练
Opened
#60 Expected input and output size for notebook?
Opened
#61 一个小时300张的处理速度正常吗？
Opened
#62 您好，请问为什么我没办法使用V1？
Opened
#63 How do you get a coloured image from a greyscale one?
Opened
#64 How to get original size? (no upscale)
Opened
#65 如何降低美颜效果
Opened
#66 Training with clean version
Opened
#67 About training with 8 gpus
Opened
#68 Have the same question as "Training with clean version #61"
Opened
#69 Training with clean version
Opened
#70 训练iter数
Opened
#71 RuntimeError: Error(s) in loading state_dict for GFPGANv1:
Opened
#72 Pre trained models for transfer learning
Opened
#73 训练iter数
Opened
#74 RuntimeError: "reflection_pad2d" not implemented for 'Half'
Opened
#75 subprocess.CalledProcessError and RuntimeError: mat1 dim 1 must match mat2 dim 0
Opened
#76 "GFPGANCleanv1-NoCE-C2.pth" file not downloading for me.
Opened
#77 RuntimeError: "reflection_pad2d" not implemented for 'Half'
Opened
#78 How to remove sharpening?
Opened
#79 如何训练不上色的模型
Opened
#80 重新设置图像的退化过程
Opened
#81 Currently， I have not updated the original architecture for GFPGANCleanv1-NoCE-C2.pth. So you could not finetune GFPGANCleanv1-NoCE-C2.pth.
Opened
#82 TypeError: can't multiply sequence by non-int of type 'float'
Opened
#83 IndexError: list index out of range
Opened
#84 How does the V2Model add sharpening?
Opened
#85 RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x16384 and 4096x8192)
Opened
#86 TypeError: get_face_landmarks_5() got an unexpected keyword argument 'eye_dist_threshold' in latest update
Opened
#87 can you share a bit how did you preprocess the ground truth of the image?
Opened
#88 How to run GFPGANv1.pth (paper model with colorization) on CPU or non-nVidia GPUs?
Opened
#89 Better hair enhancement
Opened
#90 What is the original intention and result of the idea that uses different latent codes for different layer?
Opened
#91 Micro faces when enhancing side face
Opened
#92 About fix decoder
Opened
#93 对比 Bringing-Old-Photos-Back-to-Life 不支持裂痕修复
Opened
#94 Permission denied: '/usr/bin/yapf' on arch linux
Opened
#95 有没有 可以部署在 移动端的NCNN 版本
Opened
#96 对比 Bringing-Old-Photos-Back-to-Life 不支持裂痕修复
Opened
#97 NameError: name 'fused_act_ext' is not defined
Opened
#98 如何生成自己数据的landmark文件？
Opened
#99 About the convergence curve of the training loss function
Opened
#100 NameError: name 'fused_act_ext' is not defined
Opened
#101 请问下数据集文件夹构成
Opened
#102 Training failed in single GPU
Opened
#103 Possible to use Real-ESRGAN on mac os AMD, and 16-bit images?
Opened
#104 NameError: name 'fused_act_ext' is not defined
Opened
#105 problem in enhancement without colorization
Opened
#106 Detail mattes and ID maps (suggestion)
Opened
#107 problem about face_ratio
Opened
#108 关于CS-SFT模块
Opened
#109 保存模型出现错误 Save the latest model calculate_psnr() missing 1 required positional argument: 'img'
Opened
#110 Run GFPGAN in docker container
Opened
#111 Data science
Opened
#112 have trouble export "GFPGANCleanv1-NoCE-C2.pth" to ONNX
Opened
#113 Error running on Hugging Face Spaces
Opened
#114 Background restoration not being applied
Opened
#115 'NoneType' object has no attribute 'load_state_dict'
Opened
#116 NameError: name 'fused_act_ext' is not defined
Opened
#117 Missing the Code of Conduct
Opened
#118 Running on CPU
Opened
#119 how to get "StyleGAN2GeneratorClean" pretrained model?
Opened
#120 Questions about the discriminator code
Opened
#121 RealESRGANer returns fully black image after upsampling
Opened
#122 how to get "StyleGAN2GeneratorClean" pretrained model?
Opened
#123 how to finetune No colorization model like 256*256
Opened
#124 .version is not found
Opened
#125 Facial enhancement
Opened
#126 Use it as software
Opened
#127 Training on custom resolution 128x128?
Opened
#128 Errors when use other models
Opened
#129 running from docker on CPU - does not work
Opened
#130 Na
Opened
#131 pth2onnx
Opened
#132 running from docker on CPU - does not work
Opened
#133 package docs
Opened
#134 Build (on ubuntu 21.10)
Opened
#135 [Errno 2] No such file or directory: '/tmp/tmpnr5j1qm4/output.png'
Opened
#136 I cant get deform_conv.so?
Opened
#137 [request] better pre-trained model?
Opened
#138 I cant train the model when set BASICSR_JIT == 'True'
Opened
#139 How to check model summary?
Opened
#140 Inspiration
Opened
#141 Not working on middle eastern - Muslim faces
Opened
#142 How to colorize the whole image?
Opened
#143 How can I enhance results with Photos from my own Photo Library?
Opened
#144 FacialComponentDiscriminator
Opened
#145 How to colorize the whole image?
Opened
#146 m1 macos12 incompatibility problem
Opened
#147 stylegan2_clean_arch pretrain model?
Opened
#148 CUDA out of memory
Opened
#149 Finetune existing gfpgan_v1 model.
Opened
#150 How we can run the GFPGAN on different GPU?
Opened
#151 How to run it on Windows?
Opened
#152 PNG with transparent background
Opened
#153 Are the decoders finetuned?
Opened
#154 How to continue the color restoration of the picture outside the face rectangle?
Opened
#155 unable to start training : NCCL library error
Opened
#156 how trans train models to one model.pth for inference
Opened
#157 offline execute on windows
Opened
#158 Trying to get ESRGAN BG Upscaling to work with CPU
Opened
#159 V1.3 model is AMAZING! But sometimes alters the face too much
Opened
#160 inference problem
Opened
#161 Using 4-channels images as a target
Opened
#162 config files for training V1.2/V1.3 models
Opened
#163 potential issue
Opened
#164 How can I use gfpgan for video?
Opened
#165 Can I force some styles?
Opened
#166 WIN11安装完所需条件后，提示错误不知道是什么原因
Opened
#167 inbetween model 1.2 and 1.3 ? PEople dont resemble themselves in 1.3 but quality is better
Opened
#168 error
Opened
#169 in Windows AssertionError: An object named 'ResNetArcFace' was already registered in 'arch' registry!
Opened
#170 请问训练出的只是模型参数吗？感觉是压缩格式的，如果想要整合成一个完整模型再进行处理，应该怎么做，有相关教程吗？
Opened
#171 Model GFPGANv1.3 does not exist
Opened
#172 Youtube
Opened
#173 Can I train the model with unaligned faces?
Opened
#174 Questions about GFPGAN 1.2 and 1.3
Opened
#175 Some colors in black and white photo
Opened
#176 The algorithm is too agressive
Opened
#177 Numpy is not available
Opened
#178 Are these models trained with Caucasian faces?
Opened
#179 关于增加额外数据训练
Opened
#180 自行训练模型，inference结果不一致
Opened
#181 I want export "GFPGANCleanv1-NoCE-C2.pth" to ONNX
Opened
#182 Issue with custom model inferencing
Opened
#183 关于背景图像增强
Opened
#184 Color gap between source and output
Opened
#185 Using facial component discriminator for eyeglass
Opened
#186 ImportError: No module named 'deform_conv'
Opened
#187 windows error NameError: name 'fused_act_ext' is not defined
Opened
#188 Higher resolution pretrained model ?
Opened
#189 Feature Request For Automatic Tile Size
Opened
#190 How to quickly run GFPGAN locally (with Docker)
Opened
#191 not
Opened
#192 Deploying to baseten
Opened
#193 mistake
Opened
#194 not
Opened
#195 RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory
Opened
#196 Gray setting ?
Opened
#197 Apple Mac M1 GPU
Opened
#198 Black Background, with enhanced face
Opened
#199 About experiments in the paper
Opened
#200 Process with Video
Opened
#201 UserWarning: The parameter 'pretrained' is deprecated since 0.13
Opened
#202 COLAB no longer working?
Opened
#203 Not accepting weights despite being in the right folder
Opened
#204 Not accepting weights despite being in the right folder
Opened
#205 No module named 'basicsr.losses.losses'
Opened
#206 when uses V1 model error
Opened
#207 Identity loss?
Opened
#208 No module named 'basicsr.losses.gan_loss'
Opened
#209 TypeError: get_face_landmarks_5() got an unexpected keyword argument 'eye_dist_threshold'
Opened
#210 Converting GFPGANv1 to GFPGANv1Clean
Opened
#211 tanh activation
Opened
#212 Discriminators loading
Opened
#213 Un-natural with 4xSR
Opened
#214 Sequence item with index 1 has a wrong type有人遇见过这种问题吗
Opened
#215 Total noob here - can't make the installation wor k
Opened
#216 arcface_resnet18.pth 这个怎么的来的呢
Opened
#217 我按照教程在mac上运行报错
Opened
#218 麻烦请教下，这个模型怎么转成paddle的
Opened
#219 numpy version
Opened
#220 How to use GPU
Opened
#221 [Documentation / Explanation] How can it work?
Opened
#222 请问如何设置才能运行结束后，不让图片被分隔成多个头像，而是一整张照片呢？
Opened
#223 How to use GPU
Opened
#224 Random Glasses
Opened
#225 打印输出问题，print problem
Opened
#226 inference_gfpgan.py. Use os.sep to search for pathes
Opened
#227 hi, bro it seems to be memory overflow ..>????
Opened
#228 自带磨皮效果？
Opened
#229 multi thread
Opened
#230 Can I use your latent code for real world-image face editing?
Opened
#231 GFPGANv1.3.pth to onnx
Opened
#232 ERROR: Could not build wheels for numpy, which is required to install pyproject.toml-based projects
Opened
#233 Permission Denied when Installing setup.py
Opened
#234 HI!! Testset!!
Opened
#235 unexpected glasses on restored images
Opened
#236 Permission Denied when Installing setup.py
Opened
#237 unexpected EOF, expected 273069 more bytes.
Opened
#238 error: (-215:Assertion failed) !buf.empty() in function 'imdecode_'
Opened
#239 Unable to upload items
Opened
#240 unexpected EOF, expected 273069 more bytes.
Opened
#241 Why does the same image in my local output file size is different with the image in the demo output file size?
Opened
#242 sampling from latent space
Opened
#243 10x speed up inferences with MobileStyleGAN
Opened
#244 TypeError: got an unexpected keyword argument 'model_rootpath'
Opened
#245 Problem with output.
Opened
#246 Is there a way to save the output in layers
Opened
#247 TypeError: got an unexpected keyword argument 'model_rootpath'
Opened
#248 Results seem to be with blending issue. ***image included***
Opened
#249 identity distance
Opened
#250 close
Opened
#251 Pretty obvious margins after pasting the face back to background. How to solve this?
Opened
#252 Hi, it have an executable for windows?
Opened
#253 How to calculate FID?
Opened
#254 why was codeformer removed
Opened
#255 How to use it in android
Opened
#256 num_batches_tracked
Opened
#257 arcface_resnet18
Opened
#258 (deleted)
Opened
#259 Difference of network structure between training and reference?
Opened
#260 arcface_resnet18
Opened
#261 CUDA error: out of memory
Opened
#262 Training RuntimeError: stack expects each tensor to be equal size, but got XXX vs YYY
Opened
#263 GFPGAN to TorchScript/TensorRT
Opened
#264 Training RuntimeError: stack expects each tensor to be equal size, but got XXX vs YYY
Opened
#265 新训练数据的eye_mouth_landmarks要如何生成
Opened
#266 styleGAN 256x256
Opened
#267 除了脸部，其他部分都是糊的，有对环境也提高清晰度的模型吗？
Opened
#268 RealESRGAN not working
Opened
#269 不真实
Opened
#270 Hneev
Opened
#271 Clear
Opened
#272 Output Result Files to only 1 folder
Opened
#273 import {
Opened
#274 Enhanced
Opened
#275 HELP PLS
Opened
#276 Issue on RealESRGAn & SSL
Opened
#277 Original trainning model for GFPGANv1.3
Opened
#278 生成的模糊块影响图像观感
Opened
#279 how to train the prior model, 请问如何训练先验网络的
Opened
#280 Original trainning model for GFPGANv1.3
Opened
#281 Using 4-channels images
Opened
#282 How this model changes can make the whole picture clear, not just the face. thanks
Opened
#283 Restore Faces are more crisp and good then restored_imags
Opened
#284 NCNN
Opened
#285 Create a Docker container
Opened
#286 CFPGAN error TypeError: argument of type 'NoneType' is not iterable
Opened
#287 Is gfpgan is safe to use for your private images?
Opened
#288 [question] the effect of param tile and tile-pad
Opened
#289 How to inference use 8 gpus
Opened
#290 photoshop plugin
Opened
#291 point GPU in distribute training?
Opened
#292 Face identification box is too large, so the fixing function breaks hair and accessories
Opened
#293 Can't Import basicsr
Opened
#294 Replicate and Huggingface Gradio Demos, different image download problem
Opened
#295 H
Opened
#296 [Question] How do I make GFPGAN colorize images on huggingface?
Opened
#297 Pixel image
Opened
#298 Cannot find ground truth images in lfw_cropped_faces, CelebChild-Test, and WebPhoto-Test dataset
Opened
#299 [Question] How do I make GFPGAN colorize images on huggingface?
Opened
#300 I get a black background
Opened
#301 Hi
Opened
#302 How do I force it to work on the CPU?
Opened
#303 I get a black background
Opened
#304 500
Opened
#305 please set recompute_scale_factor=True
Opened
#306 Photo clinic
Opened
#307 AI
Opened
#308 Improve inference time.
Opened
#309 photo restoration with gfp-gan
Opened
#310 Only restore ROI
Opened
#311 Only restoring, without resizing.
Opened
#312 Control the strength of restoration effect
Opened
#313 Fatima
Opened
#314 Following picture can not be restored, seems can not recognize the face
Opened
#315 Both versions on Replicate (Tencent and xintao) needlessly censors faces
Opened
#316 Questions about v1.3 and v1.4
Opened
#317 GFPGAN
Opened
#318 huggingface: Runtime error
Opened
#319 Unable to install GFPGAN
Opened
#320 Is there any way to improve inference speed?
Opened
#321 Slightly lowering the quality to make it realistic
Opened
#322 repeating line of code in inference_gfpgan.py
Opened
#323 GFPGAN
Opened
#324 Project dependencies may have API risk issues
Opened
#325 [Feature Request] Please make available through pip
Opened
#326 How to train the StyleGAN2Generator used as prior knowledge if i want to make some modifications
Opened
#327 pip3 install gfpgan , Cannot install on Python version 3.11.2
Opened
#328 [Feature Request] Please make available through pip
Opened
#329 How to use higher resolution images for model training？？
Opened
#330 Hair cut off
Opened
#331 add to the site replicate.com model of the first version
Opened
#332 ModuleNotFoundError: basicsr
Opened
#333 How to train V1.2&1.3？
Opened
#334 Pre-trained GFPGAN with sft_half=False
Opened
#335 Mahfuz Emon
Opened
#336 why perceptual loss is so big？
Opened
#337 How to force GFPGAN run only on GPU
Opened
#338 Mahfuz Emon
Opened
#339 Career Support Information YouTube Channel
Opened
#340 original or clean?
Opened
#341 请问gfpgan的评价指标该如何生成
Opened
#342 Jol kkyj
Opened
#343 我使用的时候发现，内存飙升，GPU却使用的很少。这是什么原因
Opened
#344 Changes eye color
Opened
#345 error from basicsr & torchvision
Opened
#346 Tencentarc damo no working
Opened
#347 what's the difference between v1.4 and v1.3? how can it's result improve so much?
Opened
#348 Run GFPGAN with V1 GFPGANv1.pth
Opened
#349 Hi
Opened
#350 Kaya Store
Opened
#351 FaceRestoreHelper
Opened
#352 Run on GPU
Opened
#353 not working
Opened
#354 Output message
Opened
#355 Why is the effect of processing photos locally not as good as online?
Opened
#356 怎样去训练一个1.4版本的GFPGAN
Opened
#357 https://t.me/exploitex/6332
Opened
#358 Error - Prediction failed for an unknown reason. It might have run out of memory? (exitcode -9)
Opened
#359 New
Opened
#360 cc
Opened
#361 How to write config.pbtxt in triton for GFPGAN
Opened
#362 Allow to define model_rootpath for FaceRestoreHelper
Opened
#363 pip install gfpgan==1.3.8 error
Opened
#364 Kk
Opened
#365 Add device support for mps
Opened
#366 pip install gfpgan==1.3.8 error
Opened
#367 为什么我在加载完cv2之后，加载basicsr时直接段错误（吐核）？
Opened
#368 How to inference with batch_size（batched）
Opened
#369 Using custom weights with .pth file (8x_NMKD-Faces_160000_G)
Opened
#372 model is run, but it wont output any image
Opened
#370 amo
Opened
#371 SSL: CERTIFICATE_VERIFY_FAILED
Opened
#373 Fails on i5-3470 / GTX 750: Hardware unsupported
Opened
#374 How to improve GPU-Util in NVIDIA GPU
Opened
#375 Low Quality Faces (blurry) after detecting and warpalign faces using Facexlib
Opened
#376 GFPGAN can't save output images
Opened
#377 ERROR: Could not find a version that satisfies the requirement tb-nightly (from gfpgan) (from versions: none)
Opened
#378 No matching distribution found for tb-nightly
Opened
#379 Weight does nothing?
Opened
#380 Do you have a plan to provide pre-training weights for StyleGan2GeneratorClean?
Opened
#381 [Guide] Get started quickly with Docker
Opened
#382 Segmentation fault (core dumped)
Opened
#383 Mddeepak111@gmail.com
Opened
#384 July
Opened
#385 为什么我修复的只有脸部？
Opened
#386 Photo editor
Opened
#387 How to optimze GFPGAN for CPU ?
Opened
#388 A general question
Opened
#389 Any tensorrt support for it?
Opened
#390 nvfuser_codegen.dll
Opened
#391 about v1.4.(Add [V1.4 model], which produces slightly more details and better identity than V1.3.)
Opened
#392 prasad
Opened
#393 Utf-8-charset support?
Opened
#394 could you offer the new train option file about v1.4? I have some questions about the degradation settings about new version...
Opened
#395 Video smooth
Opened
#396 Jbxb
Opened
#397 Realesrgam is slow on cpu we do not use it
Opened
#398 H
Opened
#399 Fix Required: Weight Loading Path in gfpgan/utils.py for facexlib
Opened
#400 Job1
Opened
#401 Videio
Opened
#402 Video
Opened
#403 framesPath = 'path/to/GFPGAN/frames/%5d.png' ffmpeg -y -r {fps} -f image2 -i {framesPath} -i {audioPath} -vcodec mpeg4 -b:v 20000k {vidOutPath}
Opened
#404 Dể
Opened
#405 [PR #96]: https://github.com/tqdm/tqdm/pull/96
Opened
#406 selva
Opened
#407 hi
Opened
#408 Hi
Opened
#409 Should remove codeformer ?
Opened
#410 如何在comfyUI里使用模型
Opened
#411 'ffhq-dataset-v2.json' file not found
Opened
#412 weight权重不起作用？
Opened
#413 Photos quality
Opened
#414 AttributeError: Caught AttributeError in DataLoader worker process 0. AttributeError: 'NoneType' object has no attribute 'astype' subprocess.CalledProcessError: Command '['/home/liu/anaconda3/envs/pytorch-CycleGAN-and-pix2pix/bin/python', '-u', 'gfpgan…
Opened
#415 Flickering in video
Opened
#416 How to find the location of the code：Tile 1/6 Tile 2/6 Tile 3/6 Tile 4/6 Tile 5/6 Tile 6/6
Opened
#417 如何加载自己训练的模型
Opened
#418 F
Opened
#419 Kamal ansari
Opened
#420 T
Opened
#421 Sai
Opened
#422 Restores face effectively, but doesn't output it correctly.
Opened
#423 How to specify the device
Opened
#424 How fast can you get GFPgan? Does it use GPU?
Opened
#425 I want to use free
Opened
#426 Mintu
Opened
#427 converting GFPGANv1.3.pth to onnx format
Opened
#428 Favourite picture
Opened
#429 When runnign GFPGan, it does not use much of GPU (graphic card), how to make it use GPU?
Opened
#430 模型是否支持FP16
Opened
#431 1
Opened
#432 旧照片修复
Opened
#433 r
Opened
#434 加载训练后的模型时modulated.weight权重参数不对应；The weight parameter modulated.weight does not correspond when loading the trained model
Opened
#435 billing problem
Opened
#436 hi
Opened
#437 Aaki
Opened
#438 rrrr
Opened
#439 Hi
Opened
#440 How do I use GPU for the enhancement process ?
Opened
#441 Add atile
Opened
#442 set up billing error
Opened
#443 gfpgan
Opened
#444 Local gradio demo
Opened
#445 Please
Opened
#446 Munnkhan
Opened
#447 cant run this app.because of
Opened
#448 ariful47
Opened
#449 PLEASE HELP ANYONE - THIS ERROR (ModuleNotFoundError: No module named 'torchvision.transforms.functional_tensor')
Opened
#450 Ytrded
Opened
#451 torchvision.transforms.functional_tensor import rgb_to_grayscale
Opened
#452 Kushwaha
Opened
#453 Mamun
Opened
#454 Love
Opened
#455 Nice
Opened
#456 I want to pull your project, can I?
Opened
#457 Surojit
Opened
#458 Ujjal Deb
Opened
#459 gfpgan file
Opened
#460 Image blending problem while caching the gfpgan model
Opened
#461 Upload Problem
Opened
#462 ফাগুনের আগুন লেগেছে নবীন হৃদয়ে,সেজেছে বসন্ত আজ পলাশের রঙে ।
Opened
#463 roop face enhancer is not working now
Opened
#464 billing problem
Opened
#465 Shskil
Opened
#466 Nice
Opened
#467 C:\Users\MARY\Downloads
Opened
#468 C:\Users\MARY\Downloads
Opened
#469 Gfpgan Not working on colab
Opened
#470 npm install replicate
Opened
#471 For Free
Opened
#472 Gfpgan Not working on colab
Opened
#473 Temporal instability with faces on continuous frames
Opened
#474 Can we only repair the lip(mouth) area？
Opened
#475 How to train on 1024 x 1024 face images?
Opened
#476 .github
Opened
#477 10
Opened
#478 28
Opened
#479 'torchvision.transforms.functional_tensor' has been depreciated to 'torchvision.transforms.functional'
Opened
#480 Anaconda Prompt Displays Error Code When Running Pip Command
Opened
#481 Jvkvfxj
Opened
#482 Training Data and configurations for GFPGAN v1.4
Opened
#483 Huggingface scaling issue
Opened
#484 ls: cannot access 'results/cmp': No such file or directory
Opened
#485 Please help me
Opened
#486 Would love to have you all as part of our company
Opened
#487 Issues with PyTorch Distributed Training on Google Colab
Opened
#488 Please help me
Opened
#489 DirectML on Windows with AMD GPUs
Opened
#490 Aaa
Opened
#491 关于数据集的问题
Opened
#492 Photo upload problem
Opened
#493 TEST GFPGAN
Opened
#494 请问推理的时候如何能进行batch操作？
Opened
#495 Photo upload problem
Opened
#496 Reconstruir esta foto
Opened
#497 Зураг сэргээх
Opened
#498 Missing Modules, Colab doesn't work anymore
Opened
#499 Examples – tencentarc/gfpgan – Replicate
Opened
#500 Gy
Opened
#501 Windows runs V1 but meets "time.sleep" problem
Opened
#502 ZALA NARENDRASINH
Opened
#503 Problem about finetuning GFPGAN v1.4
Opened
#504 Not working with pytorch 2.*
Opened
#505 Multi GPU training vs Single GPU training
Opened
#506 What upscaler is used in huggingface demo site apart from gfpgan 1.4?
Opened
#507 can not run with image
Opened
#508 Fix a picture
Opened
#509 Issue at Step 3: Cannot access 'results/cmp': No such file or directory
Opened
#510 GFPGAN推理，开进程池崩溃
Opened
#511 Fix a picture
Opened
#512 Could not find an implementation for LeakyRelu(6) in onnx runtime
Opened
#513 not running
Opened
#514 Face masked issue
Opened
#515 No module named 'basicsr'
Opened
#516 done
Opened
#517 Tot
Opened
#518 No module named 'basicsr'
Opened
#519 Solution of no module named : torchvision.transforms.functional_tensor
Opened
#520 ModuleNotFoundError: No module named 'torchvision.transforms.functional_tensor'
Opened
#521 Trying to fine-tune gfpgan
Opened
#522 Straight hair added to African-American photos
Opened
#523 NameError: name 'fused_act_ext' is not defined
Opened
#524 NameError: name 'fused_act_ext' is not defined
Opened
#525 NameError: name 'fused_act_ext' is not defined
Opened
#526 Jj
Opened
#527 NameError: name 'fused_act_ext' is not defined
Opened
#528 Broken face ( GFPGAN - Reactor - SD )
Opened
#529 Realismo
Opened
#530 Support
Opened
#531 Bruno Samora
Opened
#532 web data set not found
Opened
#533 Melhorar qualidade
Opened
#534 Melhorar a qualidade
Opened
#535 Vó
Opened
#536 import os import torch import RRDBNet_arch as arch pretrained_net = torch.load('./models/RRDB_ESRGAN_x4.pth') save_path = './models/RRDB_ESRGAN_x4.pth' crt_model = arch.RRDBNet(3, 3, 64, 23, gc=32) crt_net = crt_mode)
Opened
#537 Rakib
Opened
#538 RESOLATION NOT CREAT
Opened
#539 women
Opened
#540 women
Opened
#541 make the people face clear
Opened
#542 Make the blue photo clear
Opened
#543 can't build the whl for BasicSR
Opened
#544 make the people face clear
Opened
#545 same fps on 1080ti and 5070ti
Opened
#546 Reza
Opened
#547 Ttt
Opened
#548 ValueError: betas must be either both floats or both Tensors
Opened
#549 [PR #31] [CLOSED] Added tests and fastapi endpoint
Opened
#550 [PR #33] [MERGED] Update ReadMe
Opened
#551 [PR #32] [CLOSED] 'stable'
Opened
#552 [PR #37] [MERGED] Major revision: Support Pypi
Opened
#553 [PR #31] Added tests and fastapi endpoint
Opened
#554 [PR #43] [MERGED] Update README.md
Opened
#555 [PR #33] Update ReadMe
Opened
#556 [PR #32] 'stable'
Opened
#557 [PR #37] Major revision: Support Pypi
Opened
#558 [PR #43] Update README.md
Opened
#559 [PR #64] [CLOSED] Add Docker environment & web demo
Opened
#560 [PR #52] Gradio Web Demo
Opened
#561 [PR #54] Reordered requirements
Opened
#562 [PR #67] Add Docker environment & web demo
Opened
#563 [PR #103] chore: Run GFPGAN in docker container
Opened
#564 [PR #107] [MERGED] Update README.md
Opened
#565 [PR #112] [MERGED] Minor linguistic edits
Opened
#566 [PR #113] [CLOSED] Updated README.md to fix a typo
Opened
#567 [PR #107] Update README.md
Opened
#568 [PR #117] [MERGED] [bug fix] No module named 'gfpgan.version'
Opened
#569 [PR #112] Minor linguistic edits
Opened
#570 [PR #121] [MERGED] Update README.md
Opened
#571 [PR #117] [bug fix] No module named 'gfpgan.version'
Opened
#572 [PR #121] Update README.md
Opened
#573 [PR #138] [CLOSED] changed readme
Opened
#574 [PR #142] added an exception case for png image
Opened
#575 [PR #149] [MERGED] Add baseten.co demo to list of demos
Opened
#576 [PR #138] changed readme
Opened
#577 [PR #171] [CLOSED] Added link to pre-trained model GFPGANv1.3
Opened
#578 [PR #149] Add baseten.co demo to list of demos
Opened
#579 [PR #158] [MERGED] Add V1.3 model
Opened
#580 [PR #171] Added link to pre-trained model GFPGANv1.3
Opened
#581 [PR #177] New argument to force full processing, even when CUDA is not available.
Opened
#582 [PR #197] [CLOSED] tidy code inference_gfpgan.py
Opened
#583 [PR #198] [CLOSED] Updated the link to the Baseten sample app
Opened
#584 [PR #197] tidy code inference_gfpgan.py
Opened
#585 [PR #239] [CLOSED] update requirements.txt
Opened
#586 [PR #210] Fix basicsr losses import
Opened
#587 [PR #248] [MERGED] update replicate
Opened
#588 [PR #257] [CLOSED] Update inference_gfpgan.py
Opened
#589 [PR #265] use a stable version of tensorboard
Opened
#590 [PR #260] fix unwanted color change in processed images
Opened
#591 [PR #273] [CLOSED] Optimize image(sample) size to smaller one
Opened
#592 [PR #257] Update inference_gfpgan.py
Opened
#593 [PR #273] Optimize image(sample) size to smaller one
Opened
#594 [PR #276] 增加中文README
Opened
#595 [PR #278] [CLOSED] Fix scale
Opened
#596 [PR #288] [CLOSED] very nice Graphical user interface
Opened
#597 [PR #291] very nice Graphical user interface
Opened
#598 [PR #305] Display informational troubleshooting for CUDA valued users.
Opened
#599 [PR #278] Fix scale
Opened
#600 [PR #288] very nice Graphical user interface
Opened
#601 [PR #324] Translate English Readme into Chinese
Opened
#602 [PR #308] Delete no-response.yml
Opened
#603 [PR #349] Create Thanh
Opened
#604 [PR #357] [CLOSED] Fix Python dependency API risk issue
Opened
#605 [PR #386] Create Amit 06
Opened
#606 [PR #357] Fix Python dependency API risk issue
Opened
#607 [PR #394] consider get the models from hugging face
Opened
#608 [PR #388] Mr Anil bhai
Opened
#609 [PR #387] [CLOSED] Create Amit 06
Opened
#610 [PR #404] Create giang
Opened
#611 [PR #402] Create Manujsjj
Opened
#612 [PR #405] [CLOSED] Add device support for mps
Opened
#613 [PR #408] [CLOSED] Add GPU acceleration to Apple Silicon Macs
Opened
#614 [PR #419] Create sivakumar470
Opened
#615 [PR #428] [CLOSED] Create SECURITY.md
Opened
#616 [PR #405] Add device support for mps
Opened
#617 [PR #443] facexlib download fix
Opened
#618 [PR #428] Create SECURITY.md
Opened
#620 [PR #467] download weights once
Opened
#621 [PR #468] You can set model_root for FaceRestoreHelper
Opened
#619 [PR #485] [CLOSED] Create CNAME
Opened
#622 [PR #524] [CLOSED] moving processing image log out of loop
Opened
#623 [PR #485] Create CNAME
Opened
#624 [PR #588] Create Png
Opened
#625 [PR #563] MdCreate Siam
Opened
#626 [PR #572] Update utils.py
Opened
#627 [PR #601] basicsr version update
Opened
#628 [PR #617] [update]完成中文README的翻译
Opened
#629 [PR #625] Add model_rootpath parameter to GFPGAN __init__ for customizable paths for model weights
Opened
#630 [PR #619] Fix Python 3.13 compatibility
Opened
#631 [PR #620] v1.0
Opened
#632 [PR #637] [CLOSED] Create Manohar kumar
Opened
#633 [PR #631] [CLOSED] Feat/upgrade backends ci
Opened
#634 [PR #637] Create Manohar kumar