Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / SwinTransformer/Video-Swin-Transformer issues and pull requests

#94 - Failed to run demo.

Issue - State: open - Opened by fengjingchehu 6 months ago - 1 comment

#93 - how to use pretrained model in simplest way?

Issue - State: open - Opened by Uljibuh 7 months ago - 1 comment

#90 - Model weights for Swin-L

Issue - State: open - Opened by abdullahjamal 9 months ago - 1 comment

#89 - Steps to convert mmaction's video-swin-transformer to ONNX successfully

Issue - State: open - Opened by gigasurgeon 10 months ago - 4 comments

#87 - TypeError: 'str' object is not callable

Issue - State: open - Opened by tami64 over 1 year ago - 2 comments

#82 - Pretrained Swin-L on Kinects 600 dataset

Issue - State: open - Opened by ylt5b over 1 year ago

#81 - Welcome update to OpenMMLab 2.0

Issue - State: open - Opened by vansin over 1 year ago

#80 - append changelog to fulfill the license obligation

Pull Request - State: closed - Opened by SilverSteven over 1 year ago

#79 - Code for duplicate the weights pretrained on Imagenet

Issue - State: open - Opened by xujianglan over 1 year ago

#78 - Pretraining for SSv2

Issue - State: open - Opened by TalalWasim over 1 year ago

#76 - Error found in the code about shift_size calculation

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

#75 - is there a priority between drop out and patch norm?

Issue - State: open - Opened by emergencyd almost 2 years ago

#74 - Can't export ONNX transformer

Issue - State: open - Opened by dmenig about 2 years ago - 3 comments

#73 - Swin-L pretrain

Issue - State: open - Opened by wanghao15536870732 about 2 years ago

#72 - Error: av_read_frame failed with 1094995529

Issue - State: open - Opened by DWCTOD about 2 years ago - 1 comment

#71 - Embeddings

Issue - State: open - Opened by spunknic about 2 years ago

#69 - About the input image shape

Issue - State: closed - Opened by Chen-Bo-Yang about 2 years ago - 2 comments

#68 - How to train MyDataset

Issue - State: open - Opened by ljfgrace about 2 years ago - 1 comment

#67 - How to train/test on custom dataset like HMDB51

Issue - State: closed - Opened by Lee-daeho about 2 years ago - 6 comments

#67 - How to train/test on custom dataset like HMDB51

Issue - State: closed - Opened by Lee-daeho about 2 years ago - 6 comments

#66 - Question about the pretrained model used in Sthv2

Issue - State: open - Opened by PeiqinZhuang about 2 years ago

#64 - Results cannot be reproduced through training.

Issue - State: open - Opened by geek12138 over 2 years ago

#60 - Error about pretrained model

Issue - State: open - Opened by caihaunqai over 2 years ago - 2 comments

#59 - About learning rate and batch size

Issue - State: closed - Opened by geek12138 over 2 years ago - 2 comments

#58 - About config file

Issue - State: open - Opened by TitaniumOne over 2 years ago

#57 - how to add my own loss function

Issue - State: open - Opened by huangjch526 over 2 years ago - 1 comment

#56 - OSError: Only py/yml/yaml/json type are supported now!

Issue - State: open - Opened by LeiYiNuist over 2 years ago

#55 - Do it support using my own dataset?

Issue - State: open - Opened by 847001315 over 2 years ago

#54 - Is there pre-trained 3D Swin-transformer for medical image classification

Issue - State: open - Opened by almahdir1 over 2 years ago - 1 comment

#52 - ERROR: pytorch2onnx.py unpack error

Issue - State: open - Opened by ggccccc over 2 years ago - 4 comments

#51 - Severely overfitting occurred.

Issue - State: open - Opened by dragen1860 over 2 years ago - 1 comment

#50 - lr is very small??

Issue - State: open - Opened by dragen1860 over 2 years ago - 1 comment

#49 - training from scratch? without imagenet1k/21k pretrained?

Issue - State: open - Opened by dragen1860 over 2 years ago - 1 comment

#48 - Resume from Checkpoint 'meta' key error

Issue - State: open - Opened by emilyzjin over 2 years ago - 1 comment

#47 - area_range in RandomResizedCrop?

Issue - State: open - Opened by dragen1860 over 2 years ago

#46 - Abour factorized spatiotemporal model as in Table 4

Issue - State: open - Opened by youthHan over 2 years ago - 1 comment

#44 - AttributeError: Module tools/data/kinetics/label_map_k400.txt not found

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

#43 - Any plan for spatial temporal localization?

Issue - State: open - Opened by suilin0432 over 2 years ago

#42 - the training iteration is abnormally large

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

#41 - the input is the video?

Issue - State: open - Opened by henbucuoshanghai over 2 years ago

#39 - demo.py is not the newest, please update.

Issue - State: open - Opened by wsl1014 almost 3 years ago

#37 - RuntimeError: Unable to find a valid cuDNN algorithm to run convolution

Issue - State: open - Opened by PoPdark almost 3 years ago - 1 comment

#35 - The download path of checkpoint

Issue - State: open - Opened by MiShengyang almost 3 years ago - 1 comment

#34 - Details about input frames

Issue - State: open - Opened by luohwu almost 3 years ago - 1 comment

#33 - the common dataset setting

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

#32 - Inaccessible Download Links

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

#31 - skeleton based

Issue - State: open - Opened by henbucuoshanghai almost 3 years ago

#30 - INSTALL

Issue - State: open - Opened by wenjun90 almost 3 years ago - 2 comments

#29 - Do you have any result on other video dataset, like Charades?

Issue - State: open - Opened by visaVita almost 3 years ago

#28 - ib2to3.pgen2.parse.parseerror

Issue - State: closed - Opened by Weed-eat-meat almost 3 years ago

#27 - How to get parameters and FLOPs values in video swin transformer model？

Issue - State: open - Opened by yangyangxu0 almost 3 years ago - 2 comments

#26 - KeyError: 'filename'

Issue - State: open - Opened by clannadcl almost 3 years ago - 3 comments

#25 - 0

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

#23 - Official Pytorch API or model?

Issue - State: open - Opened by cjh9 about 3 years ago

#22 - KeyError: 'patch_embed.proj.weight'

Issue - State: open - Opened by r-kellerm about 3 years ago - 4 comments

#21 - KeyError: "Recognizer3D: 'SwinTransformer3D is not in the models registry'"

Issue - State: open - Opened by yugrocks about 3 years ago - 8 comments

#20 - THUMOS14 fetch_tag_proposal.sh doesn't work

Issue - State: open - Opened by yumion about 3 years ago - 1 comment

#19 - Can "demo/demo.py" run on MAC without GPU?

Issue - State: open - Opened by erwangccc about 3 years ago - 1 comment

#18 - Reproducing results

Issue - State: open - Opened by zehzhang about 3 years ago - 4 comments

#17 - can't run Video-Swin-Transformer

Issue - State: closed - Opened by bl6594 about 3 years ago

#16 - About the 3D relative position bias

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

#15 - The model's behavior is different from the picture in the paper.

Issue - State: closed - Opened by junsang7777 about 3 years ago - 1 comment

#14 - How long does it take to train an epoch with SWIN-B?

Issue - State: closed - Opened by Christinepan881 about 3 years ago - 1 comment

#13 - Where can I find the <PRETRAIN_MODEL>?

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

#12 - Keeping the temporal dimension

Issue - State: closed - Opened by hongsukchoi about 3 years ago - 1 comment

#11 - What head to use ?

Issue - State: closed - Opened by Scienceseb about 3 years ago - 1 comment

#10 - Drop path rate

Issue - State: closed - Opened by minostauros about 3 years ago - 1 comment

#9 - swin config file work_dir = work_dir = ..?

Issue - State: closed - Opened by g1910 about 3 years ago - 1 comment

#8 - Which version of Kinetics400 do you use?

Issue - State: closed - Opened by yxchng about 3 years ago - 1 comment

#7 - Removed redundancy in data_preparation.md

Pull Request - State: closed - Opened by krag-harsh about 3 years ago

#6 - Are you interested in creating a PR under MMAction2?

Issue - State: open - Opened by kennymckormick about 3 years ago

#5 - Performance Reproducing of Swin-S

Issue - State: closed - Opened by JaminFong about 3 years ago - 1 comment

#4 - Swin-L weight

Issue - State: open - Opened by OValery16 about 3 years ago

#3 - if results['frame_inds'].ndim != 1: KeyError: 'frame_inds'

Issue - State: open - Opened by Note-Liu about 3 years ago - 1 comment

#2 - modify the config file

Issue - State: closed - Opened by Note-Liu about 3 years ago - 1 comment

#1 - KeyError: "Recognizer3D: 'SwinTransformer3D is not in the models registry'"

Issue - State: closed - Opened by Note-Liu about 3 years ago - 9 comments