I cannot insist that this is always correct, but in my case, the code runs well in pytorch 2.10 environment after modifying it as follows
- ./FaceBoxes/utils/nms/cpu_nms.pyx:27
cdef np.ndarray[np.int_t, ndim=1] order = np.ndarray(scores.argsort()[::-1], dtype=np.int32)
--> cdef np.ndarray[np.intp_t, ndim=1] order = scores.argsort()[::-1]
- ./FaceBoxes/utils/nms/cpu_nms.pyx:31
cdef np.ndarray[np.int_t, ndim=1] suppressed =
--> cdef np.ndarray[np.intp_t, ndim=1] suppressed = \
I cannot insist that this is always correct, but in my case, the code runs well in pytorch 2.10 environment after modifying it as follows
cdef np.ndarray[np.int_t, ndim=1] order = np.ndarray(scores.argsort()[::-1], dtype=np.int32)
--> cdef np.ndarray[np.intp_t, ndim=1] order = scores.argsort()[::-1]
cdef np.ndarray[np.int_t, ndim=1] suppressed =
--> cdef np.ndarray[np.intp_t, ndim=1] suppressed = \