目前在做https://aistudio.baidu.com/aistudio/projectdetail/1059610项目，先在aistudio上训练好模型，然后paddle_lite_opt \ --model_file=inference/ssd_mobilenet_v1_voc/__model__ \ --param_file=inference/ssd_mobilenet_v1_voc/__params__ \ --optimize_out=./inference/ssd_mobilenet_v1_voc \ --optimize_out_type=naive_buffer \ --valid_targets=arm转格式，再部署到树莓派（树莓派3b,64G内存，64位系统），Paddle-Lite-Demo自带的项目预测小狗照片那个能够成功运行，用我的训练的nb文件进行视频流预测时，运行./run.sh报错了:

[100%] Built target object_detection_demo
[I 4/ 5 21:55:39. 6 ...ang/Paddle-Lite/lite/core/device_info.cc:1065 Setup] ARM multiprocessors name: MODEL NAME : ARMV7 PROCESSOR REV 4 (V7L)
HARDWARE : BCM2835

[I 4/ 5 21:55:39. 7 ...ang/Paddle-Lite/lite/core/device_info.cc:1066 Setup] ARM multiprocessors number: 4
[I 4/ 5 21:55:39. 7 ...ang/Paddle-Lite/lite/core/device_info.cc:1068 Setup] ARM multiprocessors ID: 0, max freq: 1200, min freq: 1200, cluster ID: 0, CPU ARCH: A53

[I 4/ 5 21:55:39. 8 ...ang/Paddle-Lite/lite/core/device_info.cc:434 check_cpu_online] Failed to query the online statue of CPU id:3
[I 4/ 5 21:55:39. 8 ...ang/Paddle-Lite/lite/core/device_info.cc:439 check_cpu_online] CPU id:3 is offine

[F 4/ 5 21:55:39. 8 ...e-Lite/lite/model_parser/model_parser.cc:862 LoadModelNaiveFromFile] Paddle-Lite v2.7 has upgraded the naive-buffer model format. Please use the OPT to generate a new model. Thanks!

[F 4/ 5 21:55:39. 8 ...engyang/Paddle-Lite/lite/core/program.cc:173 RuntimeProgram] Check failed: block_size: No block found!

[F 4/ 5 21:55:39. 8 ...engyang/Paddle-Lite/lite/core/program.cc:174 RuntimeProgram] Check failed: block_idx >= 0 && block_idx < block_size: Invalid block index, expected [0,4294967295] but got 0

./run.sh：行 19: 18357 段错误 LD_LIBRARY_PATH=$LD_LIBRARY_PATH:${PADDLE_LITE_DIR}/libs/${TARGET_ARCH_ABI} ./object_detection_demo ../models/ssd_mobilenet_v1_pascalvoc_for_cpu/ssd_mobilenet_v1_voc.nb ../labels/helmet_label