vision.multibox_transform_loc
run on Cuda give wrong output .
def multibox_transform_loc(
cls_prob, loc_pred, anchor, clip=True, threshold=0.01, variances=(0.1, 0.1, 0.2, 0.2)
)
vision.multibox_transform_loc
output
-
wrong ----
cls_prob
is GPU output,loc_pred
is const expr . -
wrong ----
cls_prob
is GPU output,loc_pred
is GPU output . -
correct ----
cls_prob
is const expr,loc_pred
is const expr .
Gpu output and Data dump(from caffe blobs) diff:
-
cls_prob
------ mbox_conf_softmax 1.9020114e-06 -
loc_pred
------ mbox_loc 4.2561445e-05
Shapes
-
cls_prob
------ (1, 2, 7650) -
loc_pred
------ (1, 30600)
Three output above line 5587 is same. On image 3 (correct output) below line 5587 values is extramly small.
Any suggestions would be greatly appreciated!
This repo can reproduce the bug on tvm repo commit 4c77bae772ad68f3dc4dda009384cb65af9dfaec