TexasInstruments / edgeai-tidl-tools

Edgeai TIDL Tools and Examples - This repository contains Tools and example developed for Deep learning runtime (DLRT) offering provided by TI’s edge AI solutions.
Other
121 stars 27 forks source link

Compiler's configuration parameters #56

Closed yurkovak closed 1 year ago

yurkovak commented 1 year ago

I found this docs page, Configuration Parameters, and I'm interested in the MSMCSIZE_KB parameter for TDA4VM. Could you please elaborate on how to use this config? The docs are very scarce, it seems that it should be used with one of the _edgeai-tidl-tools/tidltools/*.so files, but it's not explained which one and at which stage.

  1. Can MSMCSIZE_KB be configured during the python pipeline described here?
  2. Should it be used in the compilation or inference stage?
  3. Should it be used on a host device or on the chip itself?

Thanks in advance!

kumardesappan commented 1 year ago

MSMCSIZE_KB can be set in the "edgeai-tidl-tools/tidl_tools/device_config.cfg". It is used compilation steps and models artefacts are generated based on this compilation parameters It is applicable for both host device or on the chip