I am trying to retrain yolo from scratch using the "retrain_yolo.py" script. All went well up to the yolo_eval function call. After some debugging, it seems to fail on the nms_index = tf.image.non_max_suppression(...) whoch is right after the comment "TODO: Something must be done about this ugly hack!"
The exact error message is
ValueError: Tensor conversion requested dtype float32 for Tensor with dtype int32: 'Tensor("Non-maximal-suppression/iou_threshold:0", shape=(), dtype=int32)'
Hi All,
I am trying to retrain yolo from scratch using the "retrain_yolo.py" script. All went well up to the
yolo_eval
function call. After some debugging, it seems to fail on thenms_index = tf.image.non_max_suppression(...)
whoch is right after the comment "TODO: Something must be done about this ugly hack!"The exact error message is
Can somebody help me here?