Ecosyste.ms: Issues

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

GitHub / pengzhiliang/MAE-pytorch issues and pull requests

#101 - typo error local-rank

Issue - State: open - Opened by Jalilnkh 8 months ago - 1 comment

#100 - pil_loader slowly

Issue - State: open - Opened by maichm 10 months ago

#99 - Is Mixup necessary for MAE fine-tuning?

Issue - State: open - Opened by SCU-zly 11 months ago

#98 - The import accimage cannot be parsed

Issue - State: open - Opened by zhanglaoban-kk about 1 year ago

#95 - How to implement Layer-wise learning rate decay on ResNet?

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

#94 - Visual loading model error

Issue - State: closed - Opened by mouxinyue1 almost 2 years ago - 6 comments

#91 - SimMIM test

Issue - State: closed - Opened by Margrate over 2 years ago

#90 - How to resume from the checkpoint?

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

#89 - Visualize Problems

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

#88 - add area encoding

Pull Request - State: closed - Opened by hello-jinwoo over 2 years ago

#87 - learning rate curve

Issue - State: open - Opened by together0 over 2 years ago

#86 - Patch size for pretraining

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

#85 - Pretrained weight of vit-S

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

#85 - Pretrained weight of vit-S

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

#84 - A warning when pretraining

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

#83 - Which dataset is used for the released pretrained model?

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

#81 - Do I need to specify the value of mask_ratio before finetune?

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

#80 - Question upon MSE loss

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

#79 - A little suggestion about data

Issue - State: closed - Opened by wangning7149 almost 3 years ago - 1 comment

#78 - Is the finetune script doing an end2end finetune, not just linear-porbing?

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

#76 - ILSVRC2012-dataset

Issue - State: closed - Opened by xiyufeng2 almost 3 years ago - 1 comment

#75 - how to use resnet50

Issue - State: closed - Opened by 781458112 almost 3 years ago - 2 comments

#74 - Could you add a fine-tuned ViT-B model?

Issue - State: closed - Opened by Freder-chen almost 3 years ago - 5 comments

#73 - can't load the weight when using the fine-tuned model to visualization

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

#72 - About the positional embedding.

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

#71 - The setting of your 400-epoch pre-training.

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

#70 - where is the code of freezing the blocks that you don't want to finetune?

Issue - State: closed - Opened by A-zhudong almost 3 years ago - 6 comments

#69 - Mask's format

Issue - State: closed - Opened by DISAPPEARED13 almost 3 years ago - 2 comments

#68 - mask是如何生成的？

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

#67 - could you please make the large model weights public?

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

#66 - how much time be consumed in your pretraining???

Issue - State: closed - Opened by chenxl124578 almost 3 years ago - 2 comments

#65 - Low Linear-Prob accuracy

Issue - State: open - Opened by launchauto almost 3 years ago - 6 comments

#64 - Cannot distributed running?

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

#63 - Can you create this model to timm?

Issue - State: closed - Opened by Lukangkang123 almost 3 years ago - 1 comment

#62 - 请问你们提供的预训练weights是用多大的batch size训练的

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

#61 - Could you provide a one-dimensional MAE implementation in pytorch?

Issue - State: closed - Opened by bigdaronlee163 almost 3 years ago - 1 comment

#60 - TypeError: __init__() got an unexpected keyword argument 'use_mean_pooling'

Issue - State: closed - Opened by Lukangkang123 almost 3 years ago - 3 comments

#59 - Results about different pretrain epochs.

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

#58 - There is only one GPU and the weight problem of visiontransformer

Issue - State: closed - Opened by Erik267 almost 3 years ago - 3 comments

#57 - Where to place the ImageNet2012 images tar file and anotations tar file?

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

#56 - how do you do shuffle?

Issue - State: closed - Opened by Patricia1019 almost 3 years ago - 1 comment

#55 - fine tune to segmentation model not match

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

#54 - How about change encoder and decoder to transformer variants

Issue - State: closed - Opened by Dreamer312 almost 3 years ago - 3 comments

#53 - How to process the dataset...

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

#52 - Why don't train the k_bias

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

#51 - Weights & Biases logger for MAE-pytorch

Issue - State: open - Opened by morganmcg1 almost 3 years ago - 2 comments
Labels: enhancement

#50 - Do you ever use batch_size larger than 128 per GPU

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

#49 - Grad Norm Becomes Inf

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

#47 - How do you test the accuracy of your model?

Issue - State: closed - Opened by Patricia1019 almost 3 years ago - 1 comment

#46 - A little question about the visualization

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

#45 - the problem of result

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

#44 - help

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

#43 - Why use ema in finetuning?

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

#42 - the problem of imagenet datasets

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

#41 - hello, a little question about this code.

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

#40 - [Attention] Information leak in visualization

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

#39 - Need to interpolate positional embedding to work at higher resolutions

Issue - State: open - Opened by atonderski almost 3 years ago
Labels: enhancement

#38 - Why do you use cosine weight decay?

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

#37 - Can you provide the organization of the dataset

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

#36 - forward() takes 2 positional arguments but 3 were given

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

#35 - A question about pretrained model

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

#34 - get an error

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

#33 - 3张3090跑得了4096batchsize的嘛？

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

#32 - TypeError: __init__() got an unexpected keyword argument 'use_mean_pooling'

Issue - State: closed - Opened by DDxk369 almost 3 years ago - 10 comments

#31 - EMA evaluation does not use the exponential moving avera

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

#30 - timm model中找不到pretrain_mae_base_patch16_224 模型registry

Issue - State: closed - Opened by maoxiaofei99 almost 3 years ago - 3 comments

#29 - About next plan.

Issue - State: closed - Opened by zongzi13545329 almost 3 years ago - 1 comment

#27 - Reported Result for Vit-large

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

#25 - Update the requirements

Pull Request - State: closed - Opened by Celestial-Bai almost 3 years ago - 2 comments

#24 - TypeError: __init__() got an unexpected keyword argument 'in_chans'

Issue - State: closed - Opened by Celestial-Bai almost 3 years ago - 7 comments

#23 - Update run_mae_vis.py

Pull Request - State: closed - Opened by flishwang almost 3 years ago - 1 comment

#22 - Managing 512*512 input size

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

#21 - Could you provide a vpn-free link of pretrained weights rather than Google Drive?

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

#20 - normalize

Issue - State: closed - Opened by go2sea almost 3 years ago - 1 comment

#19 - finetune code about cls token

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

#18 - Info about the pretrained models

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

#17 - a small issue in modeling_finetune.py

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

#16 - Positional embedding interpolation

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

#15 - Run with pretrain_mae_vit_base_mask_0.75_400e.pth get an model initialization error

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

#14 - How is fine-tuning performed actually?

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

#13 - 请问你们用的是什么算力呀？

Issue - State: closed - Opened by Tonyboy999 almost 3 years ago - 7 comments

#12 - Several typos and missing items

Issue - State: closed - Opened by Dolores2333 almost 3 years ago - 1 comment

#11 - Training Equipment

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

#10 - problem of fine-tuning's learning rate and batch-size

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

#9 - Create run_mae_vis

Pull Request - State: closed - Opened by tikboaHIT almost 3 years ago - 1 comment

#8 - the config

Issue - State: closed - Opened by yyl199655 almost 3 years ago - 3 comments

#7 - ViT-Small

Issue - State: closed - Opened by DRSY almost 3 years ago - 5 comments

#6 - knn eval of MAE

Issue - State: open - Opened by Dongshengjiang almost 3 years ago - 13 comments
Labels: enhancement

#5 - MAE prediction visualization code

Issue - State: open - Opened by tikboaHIT almost 3 years ago - 9 comments
Labels: enhancement

#4 - Pretrained models

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

#3 - finetuning部分代码

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