RTOS的tidl_model_import.out支持双输入么？
目前的配置文件
tidl_import_yolox.txt
modelType = 2
numParamBits = 16
numFeatureBits = 16
quantizationStyle = 3
quantizationLevel = 4
memPoolSizeL3 = 1073741824
memPoolSize = 134217728
inputNetFile = "../../test/testvecs/25025test/250525_op11_forced.onnx"
outputNetFile = "../../test/testvecs/25025test/output/25025_tidl_net_SDK9_1.bin"
outputParamsFile = "../../test/testvecs/25025test/output/25025_tidl_io_SDK9_1.bin"
inDataNorm = 1
inMean = 8 8 8 8
inScale = 0.001988 0.001988 0.001988 0.001988
inDataFormat = 1
inWidth = 1080
inHeight = 1920
inNumChannels = 4
numFrames = 20
inData = "../../test/testvecs/25025test/detection_list.txt"
perfSimConfig = ../../test/testvecs/25025test/device_configs/j721e_config.cfg
debugTraceLevel = 2
模型*.onnx信息
(tidl_9.1) wp@wp:~/jm/tda4vm/download/ti-processor-sdk-rtos-j721e-evm-09_01_00_06/c7x-mma-tidl/ti_dl/test/testvecs/25025test$ python3 onnx_check.py
Opset: 17
Operators: ['Add', 'Clip', 'Constant', 'Conv', 'ConvTranspose', 'LeakyRelu']
Input: current_frame [1, 4, 1080, 1920]
Input: pre_fusion [1, 4, 1080, 1920]
Output: denoise_img [1, 4, 1080, 1920]
Output: fusion_img [1, 4, 1080, 1920]
执行和还原现场信息：
使用./out/tidl_model_import.out ${TIDL_INSTALL_PATH}/ti_dl/test/testvecs/25025test/tidl_import_yolox.txt
无法正常执行，输出结果。
执行现象如下：
(tidl_9.1) wp@wp:~/jm/tda4vm/download/ti-processor-sdk-rtos-j721e-evm-09_01_00_06/c7x-mma-tidl/ti_dl/utils/tidlModelImport$ ./out/tidl_model_import.out ${TIDL_INSTALL_PATH}/ti_dl/test/testvecs/25025test/tidl_import_yolox.txt
ONNX Model (Proto) File : ../../test/testvecs/25025test/250525_op11_forced.onnx
TIDL Network File : ../../test/testvecs/25025test/output/25025_tidl_net_SDK9_1.bin
TIDL IO Info File : ../../test/testvecs/25025test/output/25025_tidl_io_SDK9_1.bin
Current ONNX OpSet Version : 11
Running tidl_optimizeNet
tidl_model_import.out: tidl_import_common.cpp:201: void* my_malloc(int): Assertion `ptr != NULL' failed.
已中止 (核心已转储)
(tidl_9.1) wp@wp:~/jm/tda4vm/download/ti-processor-sdk-rtos-j721e-evm-09_01_00_06/c7x-mma-tidl/ti_dl/utils/tidlModelImport$
问题：
tidl_model_import.out不支持多输入的模型么？