mit-han-lab/gan-compression issues and pull requests

#118 - URL is not supported

Issue - State: closed - Opened by lixiaoyu0429 6 months ago - 7 comments

#117 - Request for Access to Pretrained Model for Verification and Replication Purposes.

Issue - State: closed - Opened by BigBooks 10 months ago - 3 comments

#116 - ERROR 403: Forbidden

Issue - State: closed - Opened by sr1dh4r about 1 year ago - 5 comments

#115 - Cannot access to https://hanlab.mit.edu/

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

#114 - Bump werkzeug from 1.0.0 to 2.2.3

Pull Request - State: open - Opened by dependabot[bot] over 1 year ago
Labels: dependencies

#113 - [Question] About SuperSeparableConv2d

Issue - State: closed - Opened by saduf almost 2 years ago - 2 comments

#112 - Bump certifi from 2019.11.28 to 2022.12.7

Pull Request - State: open - Opened by dependabot[bot] almost 2 years ago
Labels: dependencies

#111 - Guidance to covert pth to ptl

Issue - State: closed - Opened by spinoza1791 almost 2 years ago

#110 - TypeError: _output_padding() missing 1 required positional argument: 'num_spatial_dims'

Issue - State: closed - Opened by dianxin556 almost 2 years ago - 4 comments

#109 - Bump pillow from 7.0.0 to 9.3.0

Pull Request - State: open - Opened by dependabot[bot] almost 2 years ago
Labels: dependencies

#108 - "Once-for-all" Network Training Problem

Issue - State: closed - Opened by saijo0404 almost 2 years ago - 2 comments

#107 - fix(sec): upgrade numpy to 1.22.2

Pull Request - State: open - Opened by 645775992 about 2 years ago

#106 - Distill Problem

Issue - State: closed - Opened by saijo0404 about 2 years ago - 4 comments

#105 - distilling on higer resolution

Issue - State: closed - Opened by youjin-c about 2 years ago - 2 comments

#104 - Bump protobuf from 3.11.3 to 3.18.3

Pull Request - State: open - Opened by dependabot[bot] about 2 years ago
Labels: dependencies

#103 - About Select the Best Model (evolution_search.py)

Issue - State: closed - Opened by Zhaoyu321 about 2 years ago - 2 comments

#102 - What do these two paths mean?（--metaA_path --metaB_path）

Issue - State: closed - Opened by Zhaoyu321 about 2 years ago - 3 comments

#101 - Bump numpy from 1.18.1 to 1.22.0

Pull Request - State: open - Opened by dependabot[bot] over 2 years ago
Labels: dependencies

#100 - Gray-Scale Input Support

Issue - State: closed - Opened by ailen-wrx over 2 years ago - 1 comment

#99 - about SuperConv2d

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

#98 - Bump pillow from 7.0.0 to 9.0.1

Pull Request - State: closed - Opened by dependabot[bot] over 2 years ago - 1 comment
Labels: dependencies

#97 - Question about the budget setting

Issue - State: closed - Opened by joe8767 over 2 years ago - 2 comments

#96 - Question about testing the compressed model

Issue - State: closed - Opened by joe8767 over 2 years ago - 1 comment

#95 - Bump protobuf from 3.11.3 to 3.15.0

Pull Request - State: closed - Opened by dependabot[bot] over 2 years ago - 1 comment
Labels: dependencies

#94 - Bump pillow from 7.0.0 to 9.0.0

Pull Request - State: closed - Opened by dependabot[bot] almost 3 years ago - 1 comment
Labels: dependencies

#93 - "download_real_stat.sh" doesn't work.

Issue - State: closed - Opened by dhwi96 almost 3 years ago

#92 - where is bash scripts/cycle_gan/horse2zebra/search.sh?

Issue - State: closed - Opened by JJY0710 almost 3 years ago - 1 comment

#91 - How to generate cityscape_A.npz

Issue - State: closed - Opened by ziwenjie almost 3 years ago - 1 comment

#90 - Does this way can apply to pix2pixHD model?

Issue - State: closed - Opened by forest520 almost 3 years ago - 2 comments

#89 - Bump pillow from 7.0.0 to 8.3.2

Pull Request - State: closed - Opened by dependabot[bot] about 3 years ago - 1 comment
Labels: dependencies

#88 - Looking forward to the results of P-KDGAN on image synthesis

Issue - State: open - Opened by zzw-zwzhang about 3 years ago

#87 - Change '--student_netG' super_mobile_resnet_9blocks to super_mobile_resnet_6blocks

Issue - State: closed - Opened by PatrickHaopc over 3 years ago - 2 comments

#86 - Question of finetuning in fast gan compression

Issue - State: closed - Opened by colorjam over 3 years ago - 1 comment

#85 - U-net vs Pix2Pix?

Issue - State: closed - Opened by seekingdeep over 3 years ago

#84 - Super spectral normalization

Issue - State: closed - Opened by sunpeng1996 over 3 years ago - 1 comment

#83 - Cann't download one of pretrained model in the Tutorial

Issue - State: closed - Opened by kasim0226 over 3 years ago - 1 comment

#82 - It can be continued to train?

Issue - State: closed - Opened by PatrickHaopc over 3 years ago - 1 comment

#81 - Are there any specific options to calculate real statistics and FID?

Issue - State: closed - Opened by steechee over 3 years ago - 3 comments

#80 - Bump pillow from 7.0.0 to 8.2.0

Pull Request - State: closed - Opened by dependabot[bot] over 3 years ago - 1 comment
Labels: dependencies

#79 - About config

Issue - State: closed - Opened by vanduong0504 over 3 years ago - 1 comment

#78 - default setting for opt.phase in evolution_options.py and search_options.py in options directory

Issue - State: closed - Opened by kminsoo over 3 years ago - 1 comment

#77 - Crops, Patches and Slides

Issue - State: closed - Opened by seekingdeep over 3 years ago

#76 - How to resume model training with train_supernet.sh?

Issue - State: closed - Opened by YSL0226 over 3 years ago - 1 comment

#75 - Is it possible to load official cyclegan model in train_supernet.sh?

Issue - State: closed - Opened by YSL0226 over 3 years ago - 2 comments

#74 - Instance Segmentation using pix2pix

Issue - State: closed - Opened by seekingdeep over 3 years ago - 1 comment

#73 - No such file or directory in demo code (scripts/cycle_gan/horse2zebra_fast/evolution_search.sh)

Issue - State: closed - Opened by YSL0226 over 3 years ago - 2 comments

#72 - (param 'initial_lr' is not specified) when resuming training

Issue - State: closed - Opened by seekingdeep over 3 years ago - 1 comment

#71 - why the test result of the compressed gaugan trained by me is not as expected? but the training result of supernet is right.

Issue - State: closed - Opened by yuanlunxi over 3 years ago - 10 comments

#70 - A problem about once-for-all network

Issue - State: closed - Opened by vitamin520 over 3 years ago - 1 comment

#69 - test my compressed model, the result is not as the same in the training, but

Issue - State: closed - Opened by yuanlunxi over 3 years ago - 1 comment

#68 - Bump pillow from 7.0.0 to 8.1.1

Pull Request - State: closed - Opened by dependabot[bot] over 3 years ago - 1 comment
Labels: dependencies

#67 - Update evolution_search.sh

Pull Request - State: closed - Opened by seekingdeep over 3 years ago - 1 comment

#66 - Update evolution_search.sh

Pull Request - State: closed - Opened by seekingdeep over 3 years ago

#65 - hi, if I want to use batch norm instead of instance norm, how can I implement the once-for-all part?

Issue - State: closed - Opened by leonardodora over 3 years ago - 4 comments

#64 - Is cross-over ignored in evolution-search？

Issue - State: closed - Opened by insomnia250 almost 4 years ago - 1 comment

#63 - 2 images as input to pix2pix

Issue - State: closed - Opened by aomv almost 4 years ago - 2 comments

#62 - Compressed model inference speed is slower than original pytorch pix2pix

Issue - State: open - Opened by mpottinger almost 4 years ago - 7 comments

#61 - ValueError: Required crop size is larger than input image size

Issue - State: closed - Opened by PatrickHaopc almost 4 years ago - 5 comments

#60 - can it go below 1.6G MACs?

Issue - State: closed - Opened by seekingdeep almost 4 years ago - 2 comments

#59 - The Fast Pipeline [ESFNet & NAS-Unet]

Issue - State: closed - Opened by seekingdeep almost 4 years ago - 3 comments

#58 - Bump pillow from 7.0.0 to 7.1.0

Pull Request - State: closed - Opened by dependabot[bot] almost 4 years ago - 1 comment
Labels: dependencies

#57 - cnn directions about calc_distill_loss

Issue - State: closed - Opened by cookingbear almost 4 years ago - 2 comments

#56 - What's the differences among Full Pipeline (Paper), Full Pipeline (Retrained) and Lite Pipeline?

Issue - State: closed - Opened by zkchen95 almost 4 years ago - 3 comments

#55 - SuperConv output channels

Issue - State: closed - Opened by 0x366 almost 4 years ago - 1 comment

#54 - My Own Full model apply problem

Issue - State: closed - Opened by qiaozhuocheng almost 4 years ago - 6 comments

#53 - bug in evolution.py

Issue - State: closed - Opened by artyomnaz almost 4 years ago - 1 comment

#52 - real_stats for COCO

Issue - State: closed - Opened by deJQK about 4 years ago - 5 comments

#51 - question regarding pix2pixHD

Issue - State: closed - Opened by dearkafka about 4 years ago - 1 comment

#50 - using the compressed model on a Raspberry Pi 4

Issue - State: closed - Opened by seekingdeep about 4 years ago - 2 comments

#49 - Questions about resnet_supernet in pix2pix_model.

Issue - State: closed - Opened by pengfeiZhao1993 about 4 years ago - 2 comments

#48 - can not find opt_compressed.pkl

Issue - State: closed - Opened by dearkafka about 4 years ago - 1 comment

#47 - Why it is getting slower and slower when training the supernets?

Issue - State: closed - Opened by Feywell about 4 years ago - 6 comments

#46 - How does ‘netG_pretrained’ in MobileDistiller.py work?

Issue - State: closed - Opened by Feywell about 4 years ago - 3 comments

#45 - There is a bug in tensor2im

Issue - State: closed - Opened by Feywell about 4 years ago - 5 comments

#44 - It's not friendly for small cpu memory users to get fid.

Issue - State: closed - Opened by Feywell about 4 years ago - 3 comments

#43 - two questions about once for all

Issue - State: closed - Opened by zl-hw about 4 years ago - 3 comments

#42 - Searching is slow.How can I continue searching when it is break.

Issue - State: closed - Opened by Feywell about 4 years ago - 7 comments

#41 - Dataset mode [single] only supports direction BtoA. We will change the direction to BtoA.!

Issue - State: closed - Opened by ghost about 4 years ago - 4 comments

#40 - license

Issue - State: closed - Opened by ghost about 4 years ago - 1 comment

#39 - gan compression on tensorflow

Issue - State: open - Opened by sepideh-srj about 4 years ago - 2 comments

#38 - D_real: -0.000 D_fake: -0.000

Issue - State: closed - Opened by ghost about 4 years ago - 1 comment

#37 - NotImplementedError: Unknown module [<class 'torch.nn.modules.instancenorm.InstanceNorm2d'>]!

Issue - State: closed - Opened by edvardHua about 4 years ago - 2 comments

#36 - import bug

Pull Request - State: closed - Opened by ghost about 4 years ago - 1 comment

#35 - get_real_stat.py error

Issue - State: closed - Opened by ghost about 4 years ago - 23 comments

#34 - No module named 'metric.cityscapes_mIoU'

Issue - State: closed - Opened by ghost about 4 years ago - 1 comment

#33 - multi-gpu with forward hook when training supernet with resnet

Issue - State: closed - Opened by deJQK over 4 years ago - 4 comments

#32 - Pix2PixHD

Issue - State: closed - Opened by ghost over 4 years ago - 4 comments

#31 - Is Pre-distillation a required step?

Issue - State: closed - Opened by pkhigh over 4 years ago - 5 comments

#30 - other datasets

Issue - State: closed - Opened by techmonsterwang over 4 years ago - 3 comments

#29 - SERACH MULTI

Issue - State: closed - Opened by techmonsterwang over 4 years ago - 2 comments

#28 - MobileNet Teacher Model

Issue - State: closed - Opened by huangqiu15444 over 4 years ago - 3 comments

#27 - multi-gpu

Issue - State: closed - Opened by techmonsterwang over 4 years ago - 3 comments

#26 - Why GauGan compression does not inherit teacher's D?

Issue - State: closed - Opened by s1ddok over 4 years ago - 2 comments

#25 - How do I compress the other GAN models

Issue - State: closed - Opened by ypw1996 over 4 years ago - 3 comments

#21 - Why the FID results I got is different from the results in your paper?

Issue - State: closed - Opened by zhuzm0902 over 4 years ago - 3 comments

#18 - What kind of model compression can be used in the method proposed in the paper?

Issue - State: closed - Opened by jiaying96 over 4 years ago - 2 comments

#11 - Is it possible used in stylegan2?

Issue - State: closed - Opened by Z863058 over 4 years ago - 4 comments

GitHub / mit-han-lab/gan-compression issues and pull requests