Chuyang

@asmorkalov Sorry for the build error, I have fixed it. Thanks @VadimLevin for the supplyment. Truly, the error comes from the procedure of copying a `Mat` to `Matx` using `copyTo`....

Hi @kaingwade, thanks for the advice. I have added the license and add documentation for imgaug module. Junk codes in header files have also been cleaned.

Hi @asenyaev, I have commited the latest changes.

Sorry for the late reply. I didn't find any existing methods in OpenCV that takes Vec/Matx as the default argument. I was wondering if I could create a dummy function...

Hi @asmorkalov @VadimLevin, I have add some tests in a new PR #22594. I searched the generated python bindings in `/modules/python_bindings_generator/pyopencv_generated_funcs.h`, and I hadn't found any existing functions in OpenCV...

Print from `sys.argv`: ```sh ['train.py', '--local-rank=0', '--model_name_or_path', './checkpoints/vicuna-7b-v1.5', ...] ``` other arguments have the format 'key', 'value', but `locak_rank` is not properly parsed. In the above example, `local_rank=0` is treated...

Hi, sorry for the delay, I have been caught up in a DDL. I'll fix them ASAP.

Meet the same problem, solved by load model in single GPU in 8bit. 4bit will cause the problem.

你好～推理的时候也是设置成 True。目前代码里推理的时候 one-to-many 分支还是 forward 了，但是 one-to-many 分支的预测最终没有用到，如果不想 forward 这个分支可以修改下代码（比如把 ms_detr 设成 False 并把 attention layer 的顺序调整为 cross_attn -> self_attn)

Hi, it cannot, but this case is rare. The o2m matcher initially computes the matching score between all GTs and all predictions, then assigns each prediction to the GT with...