Created by: ausk

Hi, thank contributors, the lastest ocr model detects and recognises so awosome.

To better understand the network structure, I want to visual the predictor of TextDetector, TextRecognizer. If I print the predictor in pdb, it echos:

(Pdb) predictor
<paddle.fluid.core_avx.AnalysisPredictor object at 0x000001B81C986A

So, how can I visual the predictor's network structure(like pytorch to print, or netron to visual)?