search

ViTAE-Transformer / ViTDet

Unofficial implementation for [ECCV'22] "Exploring Plain Vision Transformer Backbones for Object Detection"
Apache License 2.0
530 stars 46 forks source link

size mismatch and AssertionError #22

Open xmdgaoxin opened 1 year ago

xmdgaoxin commented 1 year ago

Hello, I have the following problems during training, how should I solve them 捕获

joeyz0z commented 1 year ago

I meet the same problem, do you solve it?

xmdgaoxin commented 1 year ago

I remember that the problem was solved after modifying num_class in the definition file of vit, hope this will help you.

907218307 @.***

------------------ 原始邮件 ------------------ 发件人: "ViTAE-Transformer/ViTDet" @.>; 发送时间: 2023年6月8日(星期四) 中午11:51 @.>; @.**@.>; 主题: Re: [ViTAE-Transformer/ViTDet] size mismatch and AssertionError (Issue #22)

I meet the same problem, do you solve it?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

joeyz0z commented 1 year ago

Thanks for your prompt suggestion, it helps me a lot!