错误 layer load_param 106 611 failed

[wode2016501]

wode@xxx-NBLK-WAX9X:~/src/OcrLiteNcnn$ bash run-test.sh Setting the Number of Threads=8 Using an OpenMP Environment Variable Gpu版本测试前请先安装Vulkan SDK v1.2.162.0，https://vulkan.lunarg.com/sdk/home 请输入测试选项并回车: 1)CPU, 2)GPU 2 1.8.1 [0 AMD RADV RAVEN] queueC=1[4] queueG=0[1] queueT=0[1] [0 AMD RADV RAVEN] bugsbn1=0 bugbilz=0 bugcopc=0 bugihfa=0 [0 AMD RADV RAVEN] fp16-p/s/a=1/1/1 int8-p/s/a=1/1/1 [0 AMD RADV RAVEN] subgroup=64 basic=1 vote=1 ballot=1 shuffle=1 [1 llvmpipe (LLVM 13.0.1, 256 bits)] queueC=0[1] queueG=0[1] queueT=0[1] [1 llvmpipe (LLVM 13.0.1, 256 bits)] bugsbn1=0 bugbilz=0 bugcopc=0 bugihfa=0 [1 llvmpipe (LLVM 13.0.1, 256 bits)] fp16-p/s/a=1/1/1 int8-p/s/a=1/1/1 [1 llvmpipe (LLVM 13.0.1, 256 bits)] subgroup=8 basic=1 vote=1 ballot=1 shuffle=0 This device has 2 GPUs modelsPath=models model dbnet path=models/dbnet_op model angle path=models/angle_op model crnn path=models/crnn_lite_op keys path=models/keys.txt imgDir=images/, imgName=1.jpg resultTxtPath(images/1.jpg-result.txt) [0 AMD RADV RAVEN] queueC=1[4] queueG=0[1] queueT=0[1] [0 AMD RADV RAVEN] bugsbn1=0 bugbilz=0 bugcopc=0 bugihfa=0 [0 AMD RADV RAVEN] fp16-p/s/a=1/1/1 int8-p/s/a=1/1/1 [0 AMD RADV RAVEN] subgroup=64 basic=1 vote=1 ballot=1 shuffle=1 [1 llvmpipe (LLVM 13.0.1, 256 bits)] queueC=0[1] queueG=0[1] queueT=0[1] [1 llvmpipe (LLVM 13.0.1, 256 bits)] bugsbn1=0 bugbilz=0 bugcopc=0 bugihfa=0 [1 llvmpipe (LLVM 13.0.1, 256 bits)] fp16-p/s/a=1/1/1 int8-p/s/a=1/1/1 [1 llvmpipe (LLVM 13.0.1, 256 bits)] subgroup=8 basic=1 vote=1 ballot=1 shuffle=0 dbNet try to use Gpu0 AngleNet use Cpu CrnnNet use Cpu =====Input Params===== numThread(8),padding(50),maxSideLen(0),boxScoreThresh(0.600000),boxThresh(0.300000),unClipRatio(2.000000),doAngle(1),mostAngle(1),GPU(0) =====Init Models===== --- Init DbNet --- --- Init AngleNet --- param is too old, please regenerate! layer load_param 106 611 failed run-test.sh：行 57: 25190 段错误 （核心已转储） ./${EXE_PATH}/OcrLiteNcnn --models models --det dbnet_op --cls angle_op --rec crnn_lite_op --keys keys.txt --image $TARGET_IMG --numThread $NUM_THREADS --padding 50 --maxSideLen 0 --boxScoreThresh 0.6 --boxThresh 0.3 --unClipRatio 2.0 --doAngle 1 --mostAngle 1 --GPU $GPU_INDEX

[benjaminwan]

param is too old, please regenerate! 模型的转换路线：onnx->ncnn ncnn库有一次大的改版，大概是202207，ncnn库自带了onnx转ncnn的工具。 如果用老版本转换的模型，但是配上最新版的ncnn引擎，就会出现这个param is too old错误。

[wode2016501]

具体怎么做呢

[benjaminwan]

下载，解压，models文件夹 https://github.com/benjaminwan/OcrLiteNcnn/releases/download/1.8.1/Project_OcrLiteNcnn-1.8.1.7z