"atomicAdd" argument types error (c10::Half *, c10::Half)

[GuoRunTang]

在尝试自己配置GLIP环境的时候报了这样的错误，请问我可以如何修改呢？
单卡3090 cuda_12.1 pytorch 2.1.0+cu121

atomicAdd(grad_im + cur_bottom_grad_pos, weight * cur_top_grad);
^
nt, int, int, int, int, int, int, int, scalar_t *) [with scalar_t=c10::Half]" at line 366

D:\GRT\GLIP-main\GLIP-main\maskrcnn_benchmark\csrc\cuda\deform_conv_kernel_cuda.cu(701): error: no instance of overloaded function "atomicAdd" matches the argument list
argument types are: (c10::Half *, c10::Half)
atomicAdd(grad_im + cur_bottom_grad_pos, weight * cur_top_grad);
^
detected during instantiation of "void modulated_deformable_col2im_gpu_kernel(int, const scalar_t *, const scalar_t *, const scalar_t *, int, int, int, int, int, int, int, int, int, int, int,
int, int, int, int, int, scalar_t *) [with scalar_t=c10::Half]" at line 826

2 errors detected in the compilation of "D:/GRT/GLIP-main/GLIP-main/maskrcnn_benchmark/csrc/cuda/deform_conv_kernel_cuda.cu".
error: command 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.1\bin\nvcc.exe' failed with exit code 1

[jayaramreddy10]

@GuoRunTang , I am getting same issue, could you resolve this?