Update for upstream tflm #63
Conversation
…ls in TfLiteTensor.
TODO: modeify examples return pass/fail stauts
…se 64-bit for compiler for now.
Major refactor switching to fully tracking allocations / identifying head/tail gap
TODO: use instrumented MicroInterpreter (patch) rather than evil-half-brother of AllocateTensors
…erpreter::AllocateTensors()
ScratchBuffer's offsets in arena are captured and dumped in table in compiled application
…ulus data for footprint comparison
Set TF_LITE_MICRO_FOOTPRINT_ONLY
…g id meant it never te4sted streaming version
Ref examples relied on variant code selcted by the "portable_optimized" buildtag
code in tflite_u which is now rename "ifx_fast".
strange that CONV_2D, QUANTIZE and FULLY_CONNECTED are gone?
…iml_deployment/ide_projfiles)
…flite_u package bin dir.
While most of the tflite operators/kernels have a Register_* function returning the type TfLiteRegistration, this is nor the the case for custom TFLite Micro Operators. The AddCustom() function of the MicroOpsResolver still expects a pointer to a static struct. My proposal to fix this issue: - Update external function prototype for registering custom ops to use a pointer - Dereference pointer when building the registrations_ array
Fix registration of custom operators
# Conflicts: # examples/Makefile # src/Compiler.cc # src/RecordAllocations.cc
…t actually multiple subgraphs yet
|
Hi Yair, Thanks for the PR! FYI: the ifx branches (pushing bugfix was actually sloppiness on my part - it should really have been pushed as ifx/bugifx/....) reference a (currently) Infineon-internal version of tflite(u) with some extended APIs and kernel suite plus associated in-house-CI scripts etc. I'm currently wrapping updating these for tflite(u) from tensorflow v2.5.0 when that's done switching to support tflite(u) from the "new" standalone repo is "next on the list". Your PR is obvisouly very helpful in that regard. What would be super-helpful would to fix up the PR so that checks pass again for stock tflite(u). Pressure on "other fronts" has meant this has repeatedly postpone for ifx/ branches which is blocking a proper merge back into Christoph's main code branch. If you feel it would help your efforts I can push the current state of the v.2.5.0 update |
…cess-control compiler flag
|
I'd be glad to see the current state of what you have for v2.5.0 In general - I would be very happy to collaborate on this project. In addition to code size savings, I think it can also serve as a platform for model modification/optimization. |
Hi,
Note: Used branch bugfix/failing_fallback_opdata_parsing as a starting point as it seemed to be the most up to date code.
Made some updates in order to be able to use (sort of) the latest TFLM build.
Rewrote the node/registration scanning loops to support the new subgraph API requirements.
Had some issues with data structures being defined as private in TFLM and the #define private public trick didn't work when used on compiler.cc, so for this to work I had to add a getter to the micro_interpreter for the new MicroGraph type:
MicroGraph *getGraph() {return &graph_;}
Pretty sure this can be worked around, but I couldn't find a decent solution.
Other than that, seems to work on the tests I tried.