why there are no structured output loss ? #2
Description
In the paper 《DEEP STRUCTURED OUTPUT LEARNING FORUNCONSTRAINED TEXT RECOGNITION》， there is a structured output loss. why this process don't appear in this code?
{{ message }}
In the paper 《DEEP STRUCTURED OUTPUT LEARNING FORUNCONSTRAINED TEXT RECOGNITION》， there is a structured output loss. why this process don't appear in this code?