You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
测试模型(en_dict.txt中的字符作为字典内容),识别一张全英文图片,抛出异常java.lang.IndexOutOfBoundsException: Index 96 out of bounds for length 96。是不是在修改检测模型shape时参数设置错误?或者字典选择错误?好像其它地方都没有太多疑问,因为运行项目中的ch_PP-OCRv3_rec_infer_onnx模型是正常的。恳请老师帮我拨开云雾,再次感谢老师的无私杰出奉献。 核心错误信息如下:
Caused by: ai.djl.translate.TranslateException: java.lang.IndexOutOfBoundsException: Index 96 out of bounds for length 96
at ai.djl.inference.Predictor.batchPredict(Predictor.java:191)
at ai.djl.inference.Predictor.predict(Predictor.java:128)
at com.aias.paddleOCR.recognition.OcrV3Recognition.predict(OcrV3Recognition.java:95)
Caused by: java.lang.IndexOutOfBoundsException: Index 96 out of bounds for length 96
at java.base/jdk.internal.util.Preconditions.outOfBounds(Preconditions.java:64)
at java.base/jdk.internal.util.Preconditions.outOfBoundsCheckIndex(Preconditions.java:70)
at java.base/jdk.internal.util.Preconditions.checkIndex(Preconditions.java:266)
at java.base/java.util.Objects.checkIndex(Objects.java:359)
at java.base/java.util.ArrayList.get(ArrayList.java:427)
at com.aias.paddleOCR.recognition.PpWordRecognitionTranslator.processOutput(PpWordRecognitionTranslator.java:70)
at com.aias.paddleOCR.recognition.PpWordRecognitionTranslator.processOutput(PpWordRecognitionTranslator.java:22)
at ai.djl.inference.Predictor.batchPredict(Predictor.java:172)
The text was updated successfully, but these errors were encountered:
2023-06-2006:30:11.0362670[W:onnxruntime:,execution_frame.cc:835onnxruntime::ExecutionFrame::VerifyOutputSizes]Expected shape from model of {-1,1,97} does not match actual shape of{1,40,97}{1,40,97}for outputs oft max_2.tmp_0
2023-06-2006:30:11.1175901[W:onnxruntime:,execution_frame.cc:835onnxruntime::ExecutionFrame::VerifyOutputSizes]Expected shape from model of {-1,1,97} does not match actual shape of{1,40,97}{1,40,97}for outputs oft max_2.tmp_0
2023-06-2006:30:11.2691185[W:onnxruntime:,execution_frame.cc:835onnxruntime::ExecutionFrame::VerifyOutputSizes]Expected shape from model of {-1,1,97} does not match actual shape of{1,40,97}for outputs oft max_2.tmp_0```
系统环境: win11/jdk17/idea2022
djl环境: djl系列模块 版本 0.22.1
模型: 官方的英文 en_PP-OCRv3 检测和识别模型(https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.6/doc/doc_en/models_list_en.md)
模型转换: 按本项目引导成功转换成onnx模型,在转换检测模型时,预先用paddle_infer_shape.py模块对检测模型的shape进行修改,关键修改参数设置为:--input_shape_dict="{'x':[-1,3,-1,-1]},然后再转换模型,并成功转换成onnx格式
应用字典(ppocr_keys_v1.txt): 官方的ppocr_keys_v1.txt,但识别出来的是乱码,后改为en_dict.txt,含95个字符的字典 (https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.6/ppocr/utils/en_dict.txt)
测试模型(en_dict.txt中的字符作为字典内容),识别一张全英文图片,抛出异常java.lang.IndexOutOfBoundsException: Index 96 out of bounds for length 96。是不是在修改检测模型shape时参数设置错误?或者字典选择错误?好像其它地方都没有太多疑问,因为运行项目中的ch_PP-OCRv3_rec_infer_onnx模型是正常的。恳请老师帮我拨开云雾,再次感谢老师的无私杰出奉献。
核心错误信息如下:
The text was updated successfully, but these errors were encountered: