This project contains a code generator that produces static C NN inference deployment code targeting tiny micro-controllers (TinyML) as replacement for other µTVM runtimes. This tools generates a runtime, which statically executes the compiled model. This reduces the overhead in terms of code size and execution time compared to having a dynamic on-device runtime.
The "solution" for the __nop issue discussed before is not great but I was not able to remove it from the graph easily because inputs and outputs of the __nop node have to be preserved.
If you come up with a nice solution, feel free to propose something else.
Test fails because an issue with the tvm_patches. I will try to fix them but be could also use the https://github.com/tum-ei-eda/tvm form in the CI job in the future...
The "solution" for the
__nop
issue discussed before is not great but I was not able to remove it from the graph easily because inputs and outputs of the __nop node have to be preserved.If you come up with a nice solution, feel free to propose something else.