-
使用本仓库最新的代码RT-DETR训练自己的数据集,mAP一直为0,相同的数据集使用PP-YOLOE+的mAP能到45.7%,查看历史issue有人也遇到了相同的问题,但不知道最后解决的代码版本是?能否提供一下,谢谢!
-
how can i custom my datasets to train PureT model?
-
Hello, when I was training 1_train_xe.sh, there was no error 446093.npz, how can I solve this problem? Previous address path... /open_source_dataset/mscoco_dataset/features/up_down The data can be que…
-
Congratulations on having your work accepted by ACM MM 2023!
I have a question about the temperature, in your appendix the temperature is analyzed experimentally, but the final temperature parameter…
-
Hello
As the dataset does not provide automatically the object class annotation but object class ID, I try to get the object class name from mscoco annotations, however object class IDs does not al…
-
Is it possible to run sp-nas with its own database (unlike mscoco, pascalvoc...)?
-
hi.
I was able to complete the training phase, but I encounter this error during the testing phase after evaluating the model. Can anyone help m
[error.pdf](https://github.com/DeepRNN/image_captioni…
-
hello, can you tell me how to do ensembling? Try different initialize method or cross-validation?
I can't replicate your results in mscoco test server, and I realized that I didn't do ensembling.
Th…
-
Recently, I try to use cascade RCNN(https://github.com/guoruoqian/cascade-rcnn_Pytorch) in my own dataset. However it only has imagenet pretrained model, without pretrained model in MScoco. Can someon…
-
Thanks for this amazing work. I have a question about tokenizer.model in /data/weights dir, the tokenizer.model is not provide in this project.
```
LaVIN/
|-- lavin
|-- scripts
|-- train.p…