zhangtingyun

I just simply called the method in pdfminer to parse a pdf, but there is a problem with the coordinates of the parsed result, which is different from what I...

### Discussed in https://github.com/PaddlePaddle/PaddleOCR/discussions/13227 Originally posted by **zhangtingyun** July 1, 2024 我更换飞浆的版面分析的模型 训练中文数据集 CDLA 但ap一直上不去，一直在80%左右，但官网的准确率不应该在 94%左右吗？ 配置如下： pretrain_weights: https://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_cdla.pdparams #pretrain_weights: output/best_model.pdparams weights: output/picodet_lcnet_x1_0_fgd_layout_cdla/model_final find_unused_parameters: True use_ema: true cycle_epoch: 10 snapshot_epoch:...