Differences between neuralnetwork and mlprogram

## ❓Question

The 160-size data set is trained with 8, 16 and 32 times downsampling. Under pt weight, the target can be classified and identified by adding post-processing to the model. When neuralnetwork conversion mode is adopted, the target can be recognized normally, but when mlprogram is adopted, the target cannot be recognized normally and a bunch of detection boxes will be given. Almost all the detection images are covered.

neuralnetwork == coremltools==6.0
mlprogram == coremltools==7.0

No source file or test is identified; begin by collecting the model, conversion commands, and post-processing used with coremltools 6.0 and 7.0. Compare neuralnetwork and mlprogram outputs on the 160-size dataset, with completion defined as a reproducible explanation of the discrepancy or a confirmed fix.

Written by the indexing model from the issue text.