请注意，本文内容源自机器翻译，可能存在语法或其它翻译错误，仅供参考。如需获取准确内容，请参阅链接中的英语原文或自行翻译。
器件型号：TDA4AL-Q1 工具与软件：
你(们)好
我正在努力将焊炬模型导入 TIDL。 该模型是一个简单的3x3卷积内核、我已经使用 Pytorch API 生成了一个 onnx 模型文件。 执行此操作后、我已根据此 链接中给出的说明创建了一个配置文件。 下面显示了相同的情况。
modelType = 2 numParamBits = 8 numFeatureBits = 8 inputNetFile = /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/custom_model.onnx outputNetFile = /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/custom_model.bin outputParamsFile = /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/custom_model_io_ inData = /home/user/rtos_ti_sdk/psdk_rtos_ti_data_set_09_00_00/calib_data.txt outData = /home/user/rtos_ti_sdk/psdk_rtos_ti_data_set_09_00_00/output_data.bin inWidth = 64 inHeight = 64 resizeWidth = 64 resizeHeight = 64 inNumChannels = 3 tensorType = 1 quantizationStyle = 2 perfSimTool = /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/c7x-mma-tidl/tidl_tools/ti_cnnperfsim.out tidlStatsTool = /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/c7x-mma-tidl/ti_dl/test/PC_dsp_test_dl_algo.out graphVizTool = /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/c7x-mma-tidl/tidl_tools/tidl_graphVisualiser.out numFrames = 1
现在、当我运行可执行文件"tidl_model_import.out"时、通过将上述 cfg 文件作为输入、我将面临非法指令错误(已转储内核)。 日志如下所示
ONNX Model (Proto) File : /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/custom_model.onnx TIDL Network File : /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/custom_model.bin TIDL IO Info File : /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/custom_model_io_ Current ONNX OpSet Version : 17 Empty prototxt path, running calibration ~~~~~Running TIDL in PC emulation mode to collect Activations range for each layer~~~~~ Processing config file #0 : /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/tidl_custom_net.txt.qunat_stats_config.txt Illegal instruction (core dumped) ***************** Calibration iteration number 0 started ************************ ~~~~~Running TIDL in PC emulation mode to collect Activations range for each layer~~~~~ Processing config file #0 : /home/user/host_emulation_mode/ti-processor-sdk-rtos-j721s2-evm-09_00_00_02/pytorch_custom_models/tidl_custom_net.txt.qunat_stats_config.txt Illegal instruction (core dumped) ***************** Calibration iteration number 0 completed ************************ ------------------ Network Compiler Traces ----------------------------- NC running for device: 1 Running with OTF buffer optimizations successful Memory allocation TIDL ALLOWLISTING LAYER CHECK: TIDL_E_QUANT_STATS_NOT_AVAILABLE] tidl_quant_stats_tool.out fails to collect dynamic range. Please look into quant stats log. This model will get fault on target. **************************************************** ** 0 WARNINGS 1 ERRORS ** ****************************************************
包含 pytorch 模型定义和用于将其转换为 onnx 模型的命令的 python 脚本、输出 onnx 文件全部包含在以下链接中。
请帮助我了解此过程中缺少什么会导致显示错误。
提前感谢
