YANG BO

icon indicating a website link https://en.nuist.edu.cn/mainm.psp

icon company Nanjing University of Information Science and Technology icon location China icon location A third-year master's student from Nanjing University of Information Science and Technology in China

Results 3 comments of YANG BO

Training/Fine-tuning img2img

> I looked at it, but it seems that the code is not for fine-tuning. I have the same question about fintune/train img2img. Maybe img2img is untrainable?

Will you provide the training config file for ViT-L (66.0 AP)?

> Hi, @zhangchbin, we have no plan to do this now. But we have updated the arxiv paper to release more details about this large model. 请问ViT-L (66.0 AP)的模型的backbone是用的eva02给出的eva02_L_pt_m38m_p14to16 |...

ViT-L (66.0 AP) pre-training & config file

> pretrain: `eva02_L_pt_m38m_medft_in21k_ft_in1k_p14` The model is [Co-DINO](https://github.com/Sense-X/Co-DETR/blob/main/projects/configs/co_dino/co_dino_5scale_swin_large_16e_o365tococo.py) and more details about this large model are presented in the paper appendix. 感谢您的回答，[Co-DINO](https://github.com/Sense-X/Co-DETR/blob/main/projects/configs/co_dino/co_dino_5scale_swin_large_16e_o365tococo.py) 这个配置文件似乎是基于Swin Transformer的64.1性能的模型，如果要复现ViT-L (66.0 AP)的话，是否是需要修改其中的backbone等部分，如果方便您可以分享对应的config file 和 backbone file嘛，再次表示感谢！