search

xmos / ai_tools

AI applications and tools
Other
25 stars 10 forks source link

Remove custom flexbuffer Python bindings #326

Closed keithm-xmos closed 1 year ago

keithm-xmos commented 3 years ago

We implemented Python bindings to support the flexbuffer format for custom TensorFlow Lite operator options. In the next release of flatbuffers (2.0), Python bindings will be supported. So, we no longer need to maintain our own code for them. This code can be removed and the code that creates the custom options from a dictionary can be modified to use the bindings now provided by the flexbuffer library. See https://github.com/google/flatbuffers/blob/master/python/flatbuffers/flexbuffers.py

NOTE: This enhancement will not be necessary if implementing https://github.com/xmos/ai_tools/issues/352

keithm-xmos commented 3 years ago

Flatbuffers release 2 adds Python support for Flexbuffers. See: https://github.com/google/flatbuffers/releases/tag/v2.0.0