lzlwakeup

tflite和libonnx的框架，可以xxd固化模型参数，Tengine是否也有类似方式，不适用模型读取，直接加载hex数据进行推理？

example use other input

1

In cONNXr/examples/example1/example.c Onnx__TensorProto *inp0set0 = openTensorProtoFile("../test/mnist/test_data_set_0/input_0.pb"); Onnx__TensorProto *out0set0 = openTensorProtoFile("../test/mnist/test_data_set_0/output_0.pb"); For test, this can be save image with .pb for read. I want to port connx on MCU, chip only...

Unsupported opset => Gather-11 (ai.onnx)

2

IR Version: v6 Producer: pytorch 1.11.0 Domain: Imports: ai.onnx v11 Conv_0: Conv-11 (ai.onnx) Inputs: input.1: float32[1 x 3 x 352 x 352] = [...] onnx::Conv_760: float32[24 x 3 x 3...

h5

2

想请教下，在模型转h文件时，必须要加入x_test做转换么？ 如果我采用训练完备的h5模型，是无法直接转换使用。 关于直接采用训练完备的模型，如何快速转weight.h，而不需要重新训练。

CEVA x2动态空间问题。

2

PC端VS2019可以运行成功KWS。 移植Ceva dsp的过程中，出现内存申请无法运行问题。 例如计算mfcc中，若在create中申请，则编译烧录Ceva后，出现mfcc成员数据打印错误，即数值飞掉。将mfcc中成员指针改为具体大小的数组，则可以正确输出。 model = nnom_model_create(); 输出打印m->head null的错误，查找了一下问题，指向nnom_input.c中的input_s()函数，其他需要申请空间的位置应该都有该问题。 layer = nnom_mem(sizeof(nnom_io_layer_t) + sizeof(nnom_layer_io_t) * 2); if (layer == NULL) { printf("input err \n"); } 编译烧录打印，串口查看input err。其他情况相同。 我想请教，这种问题是否有解决方法或者解决思路？

Hi, guys there is a embedded porting issues. I want to use MCU load darknet for image classification, but MCU cannot load a weight file directly, the .weight file should...

For the conversion tool: https://convertmodel.com/#outputFormat=tengine err: ReferenceError: SharedArrayBuffer is not defined When I want to convert the tflite model, or yolo model(weight and cfg) to tengine mode, build err. The...

Platform-independent operation

1

Hi team， Some questions are bothering me. When I use code generation, arm dependent code is automatically generated, "depthwise_kernel3x3_stride1_inplace_CHW_fpreq.c", "#include "arm_nnsupportfunctions.h" //TODO: remove this in the future for self-contained" This...

Hi gays, thanks so much for this great project. The CMSIS NN library is required in the generated code, but my device doesn't support it(rsic-v). Is there any other way...

[Performance] Python infer and C++ are different for audio process

6

### Describe the issue Hi, I want to run an onnx model in C++ environment， code in attach, performance like this   C++ result has more background noise, python...