Hello guys,
I’m curious if the uTVM can support audio models and especially LSTM or GRU operations in there for ARM-based bare-metal devices such CortexM7 alike? Is there any example for that …
BR/Peter
Hello guys,
I’m curious if the uTVM can support audio models and especially LSTM or GRU operations in there for ARM-based bare-metal devices such CortexM7 alike? Is there any example for that …
BR/Peter
hi @peter73421,
we haven’t currently demonstrated LSTM or GRU models on microTVM. However, as it uses a similar GraphRuntime to the TVM C++ runtime, the LSTM and GRU models that run with TVM should theoretically also run. you could try to prototype this on a traditional OS by adapting apps/bundle_deploy
to use the model you’d like to run.
Andrew