训练自定义身份证数据集v4报错 v3没有问题

[Zomcxj]

请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem

• 系统环境/System Environment：
• 版本号/Version：Paddle： PaddleOCR： 问题相关组件/Related components：
• 运行指令/Command Code：
• 完整报错/Complete Error Message：

请尽量不要包含图片在问题中/Please try to not include the image in the issue.

paddleocr 2.6
paddlepaddle 2.5.2
ubuntu18.04 and win11
PaddleOCR2.7.1
配置文件ch_PP-OCRv3_student.yml能够训练
ch_PP-OCRv4_student.yml加载标签报错 而且两种错误
同一套源码难道加载标签数据还有区别吗
[2024/03/06 08:04:41] ppocr ERROR: When parsing line train/A_00017.jpg [{"transcription": "姓名廉氮烹", "points": [[109.0, 161.0], [433.0, 153.0], [435.0, 217.0], [110.0, 225.0]], "difficult": false}, {"transcription": "性别女", "points": [[114.0, 273.0], [315.0, 267.0], [317.0, 317.0], [116.0, 323.0]], "difficult": false}, {"transcription": "民族哈萨克", "points": [[398.0, 266.0], [691.0, 261.0], [692.0, 314.0], [399.0, 319.0]], "difficult": false}, {"transcription": "出生2024年4月27日", "points": [[119.0, 374.0], [690.0, 370.0], [690.0, 412.0], [120.0, 417.0]], "difficult": false}, {"transcription": "住址 佳木斯刷叛石夸神均娜卯", "points": [[121.0, 481.0], [806.0, 476.0], [807.0, 520.0], [121.0, 525.0]], "difficult": false}, {"transcription": "猖藻箫冰谚387县", "points": [[264.0, 549.0], [647.0, 546.0], [647.0, 593.0], [265.0, 596.0]], "difficult": false}, {"transcription": "公民身份号码", "points": [[129.0, 745.0], [420.0, 745.0], [420.0, 789.0], [129.0, 789.0]], "difficult": false}, {"transcription": "14998788868828512X", "points": [[479.0, 744.0], [1101.0, 739.0], [1101.0, 779.0], [479.0, 784.0]], "difficult": false}]
, error happened with msg: Traceback (most recent call last):
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/simple_dataset.py", line 159, in getitem
outs = transform(data, self.ops)
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/imaug/init.py", line 56, in transform
data = op(data)
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/imaug/make_border_map.py", line 64, in call
self.draw_border_map(text_polys[i], canvas, mask=mask)
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/imaug/make_border_map.py", line 85, in draw_border_map
padded_polygon = np.array(padding.Execute(distance)[0])
IndexError: list index out of range

[2024/03/06 08:04:43] ppocr INFO: epoch: [3/200], global_step: 12, lr: 0.000688, loss: 3.590737, loss_shrink_maps: 0.000000, loss_threshold_maps: 2.590737, loss_binary_maps: 1.000000, loss_cbn: 0.000000, avg_reader_cost: 0.00444 s, avg_batch_cost: 0.00701 s, avg_samples: 0.054, ips: 7.69795 samples/s, eta: 0:13:32
[2024/03/06 08:04:44] ppocr INFO: save model in ./output/ch_PP-OCRv4_det_sfz/latest
[2024/03/06 08:04:44] ppocr ERROR: When parsing line train/A_00024.jpg [{"transcription": "姓名寿橄懊", "points": [[108.0, 147.0], [435.0, 143.0], [436.0, 207.0], [108.0, 211.0]], "difficult": false}, {"transcription": "性别男", "points": [[114.0, 259.0], [310.0, 259.0], [310.0, 309.0], [114.0, 309.0]], "difficult": false}, {"transcription": "民族", "points": [[402.0, 259.0], [640.0, 259.0], [640.0, 307.0], [402.0, 307.0]], "difficult": false}, {"transcription": "出生1964年3月27日", "points": [[117.0, 361.0], [689.0, 364.0], [688.0, 407.0], [116.0, 404.0]], "difficult": false}, {"transcription": "住址金昌军些猴哮聂乏止甚玛", "points": [[115.0, 468.0], [805.0, 469.0], [805.0, 513.0], [115.0, 512.0]], "difficult": false}, {"transcription": "坤驰要捞螺792省", "points": [[260.0, 536.0], [640.0, 539.0], [639.0, 587.0], [260.0, 584.0]], "difficult": false}, {"transcription": "公民身份号码", "points": [[121.0, 732.0], [416.0, 735.0], [415.0, 779.0], [121.0, 776.0]], "difficult": false}, {"transcription": "397532146909880753", "points": [[473.0, 737.0], [1092.0, 737.0], [1092.0, 775.0], [473.0, 775.0]], "difficult": false}]
, error happened with msg: Traceback (most recent call last):
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/simple_dataset.py", line 159, in getitem
outs = transform(data, self.ops)
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/imaug/init.py", line 56, in transform
data = op(data)
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/imaug/make_border_map.py", line 64, in call
self.draw_border_map(text_polys[i], canvas, mask=mask)
File "/home/daway/cxj/PaddleOCR/PaddleOCR-2.7.1/ppocr/data/imaug/make_border_map.py", line 117, in draw_border_map
canvas[ymin_valid:ymax_valid + 1, xmin_valid:xmax_valid + 1] = np.fmax(
ValueError: could not broadcast input array from shape (0,214) into shape (1,214)

[xiaowen0407]

请问您 解决了吗 我现在也出现同样的错误

[hardlipay]

+1.更诡异的是我v4训起来了，重启机器后v4又不能训了...