Age | Commit message (Collapse) | Author |
|
A workaround had been added previously to support building of libmetal
library with ARMCC v6. Since such support has recently been added
upstream in the libmetal repository, the workaround can now be removed.
Change-Id: Idb5e6ba3112e329a0c2193d35c80c3cb45f90486
Signed-off-by: Ledion Daja <ledion.daja@arm.com>
|
|
FlatBuffers version updated to v23.5.26 to match tflite-micro patch 6bb1e3a.
Change-Id: I9c092d0bf461d790a967b137f4dfedc8660496f9
Signed-off-by: Adrian Lundell <adrian.lundell@arm.com>
|
|
The inference parser copies the description from the model using
strncpy, which will only add a null-terminator if the source string is
shorter than the destination. Otherwise, if the bytes copied are not
null-terminated, the destination is left without one.
To ensure that the description is always null-terminated, the last byte
of the description is now always overwritten with one.
Change-Id: I49c23acd12d661e1f5c37088dba2e37935f25fa5
Signed-off-by: Mikael Olsson <mikael.olsson@arm.com>
|
|
AllOpsResolver is no longer available. Create a new similar fallback
solution that is based on a MicroMutableOpResolver instance instead.
Change-Id: I398d0d99c4d0a5230ab1435796e1e1c6dea47f8f
|
|
Change-Id: Ie19178905af511a1117c8501ec1c84d1c996a94c
|
|
Added '-Wno-cast-align' and '-Wno-unused-but-set-variable' options to
suppress openamp related warnings
Change-Id: I67c82f0639a580fc4d8d49ef12e45ead36d4fc82
|
|
Replace deprecated Length() call in Flatbuffers with size(), as
suggested by the deprecation warning.
Change-Id: If28f85547120598860e3606b3983dc4607247bb0
|
|
Private members are protected and run job is virtual to make it more
flexible to customize running jobs.
Change-Id: I2e4a0760000a451eead062ef758b467ea248dffe
|
|
TFLM now mandates to set external context during prepare or invoke
state.
Change-Id: I87d3f386aac5c47b71a7abfb9a8c48ff0255fc67
|
|
Change-Id: I831e4546832aeae22c48b92cd2e8adfac797e704
|
|
Add support for ArmClang including the correct errno.h.
Silence warnings about unsupported pragma.
Change-Id: Ieb88938399679ddc39f7b8c390e41afef2b46606
|
|
Change-Id: Id36d97fa3aa4a60179bedbdc078c4e1e75486a49
|
|
Reinsert INFERENCE_PROCESS_OPS_RESOLVER that was accidentally removed
in previous commit.
Change-Id: I56b84aadc44b2a5f3c9e35e1fce8ce51036048ba
|
|
Interface libraries ethosu_core and ethosu_applications are removed.
Each application shall depend directly on the libraries they need.
Change-Id: I4d37c6c50aa276beaa8e2681a63cea90977477ce
|
|
Update the Ethos-U Monitor to merge performance samples where the
qread or status register have not changed.
Change-Id: I8b4ea728e004bd2ff05ad1c90326a864dc592f4d
|
|
Added compilation flag to allow the inference_process library to
use a MicroMutableOpResolver as an alternative to AllOpsResolver. This
allows to only make use of the needed operators instead of the complete
list of operators, reducing thus the memory footprint of the
application.
Change-Id: If1d6751b12e8aa301bb466e3ffae92406200eab4
|
|
Await upstream fix for cortex-m55 armclang compilation error
Change-Id: I8a25a7aa7f29fab8ca7f0ad4bb63d6e4f89d8a6f
|
|
Change-Id: I01bbafc45ca2a56759d7d5afbbbacefb40fd9a26
|
|
Setting policy CMP0135 to NEW adopts the new behavior without breaking
backwards compatibility.
Change-Id: I2a6cf783a6b2ed00d6c2119cf6d780c3054c25c5
|
|
Change-Id: I449f5c5039801053af380e07b812f2484e209412
|
|
Change-Id: I71637859e96d799482c74a8f55f7f792c2ed606b
|
|
MicroErrorReporter has been removed from TFLM.
Change-Id: I8acb14eff2bc5d4b05d0d50f0d2a44db846e160f
|
|
Add cycle counter, QREAD and STATUS to Event Monitor record.
Change-Id: I9c225558a9d46b05fe322270b4a921ead6bb0233
|
|
To avoid unwanted matches of regexps, example: "cortex-m3" matches
string "cortex-m33", several string MATCHES have been reworked.
Change-Id: Ib923fbed9edbfba7de0af005c67b5527acdf40f9
|
|
This will decrease current stack usage requirement and memory footprint.
Change-Id: I8ac314238efcdf1622cdcafd479c6cc1a8580509
|
|
Change-Id: Ie75eb87cf06cd7bad5f17ce121ea64037c0e4788
|
|
Added base64 encoded printing of ofm data to speed up
logging of large ofms.
Change-Id: Ibebd0a3557396eb68d69789f2c016161e085e64a
|
|
Previous version of CMake did not allow CMAKE_SYSTEM_PROCESSOR
to contain CPU features for Arm Clang toolchain. A workaround
was added to manually strip the CPU features.
Recent versions of CMake have removed this limitation and instead
requires the toolchain to manually add -mcpu=<CPU+features>, which
means that the workaround can be removed.
Change-Id: I3e8647c47df27a96cdd103614ecf2540f150ab36
|
|
Change-Id: Iaf22f1d7ad3e732f9163702cc08c8c2a3bb3a69d
|
|
Change-Id: Id0005b8e68c4a439923941930eddc5f180c8342e
|
|
If the buffer does not point to a well defined flatbuffer the parsing
segfaults.
Change-Id: Icb8dfef37dc28b2b7a22c6d3804851be8198aa9d
|
|
Change-Id: Ia884010492ee9858c8c53d8f547954c7ad6e1a71
|
|
Change-Id: Ie78bdf3330765131ab5aee248ef42f175a652218
|
|
Change 'Inference runtime' to measure CPU cycles for the
Tensorflow Lite Micro interpreter.Invoke() call.
Add 'Operator(s) runtime' print that prints a summary for
cycles spent on all operators during an inference. (This is
equivalent to the old reported 'Inference runtime')
Move prints out of the EndEvent() function in ArmProfiler as
it otherwise interferes with the inference cycle measurement.
Change-Id: Ie11b5abb5b12a3bcf5a67841f04834d05dfd796d
|
|
All drivers have been moved to core_platform.
Change-Id: I260e2ac94c25bc06ea69a20b00218b45278538cb
|
|
Remove old unused cmake toolchain files.
Adapt FreeRTOS to be used with Cortex-M85: fallback to use Cortex-M33
port files in FreeRTOS like for Cortex-M55.
Change-Id: I79af29ed920e1920c970bc6f516e9b711019692a
|
|
Add compilation flags to silence warnings about unused functions
and parameters.
Change-Id: I2c87a39c2adc05a19a59cf3978c5612f568be61a
|
|
Updating tflite_micro.cmake to include the arena allocator.
Change-Id: I793353afeccdf7f8a791b0e7202078eaffe72d7d
|
|
Change-Id: If90ebe17cc33b8fef9e6a4ebae43a2e5749d8a37
|
|
Since tensorflow/lite/c/common.c has changed name to
tensorflow/lite/c/common.cc.
Change-Id: Ib2ad4a02e3c78f9a0dac841f45b430e88d1a0e94
|
|
The printing of the output tensor can be truncated, by adding the
original size of the output in bytes the included data can be
checked for completeness. The dims will tell the number of elements
of the output tensor and is not useful for this purpose as the
element size can vary.
Change-Id: Ied5447e34416a8c6668eeae5c01a5474f103023f
|
|
The branch on the ThreadX fork doesn't build for Cortex-M33 so only
enable it for Cortex-M55.
Change-Id: I2e562e2f872902a9d87584fbacc4dd6b556d0b51
|
|
Change-Id: I4b229dbf750cbbb6306e54ff29db744a2c411ca8
|
|
For larger networks the GetTotalTicks() might overflow if the function
returns int32_t. GetTotalTicks() is a non virtual function that can
be overridden with a different return type in the implementation class.
Change-Id: I8dd2e448fd13e0823fe1e5f9cc335998d2806416
|
|
Change-Id: Iefa21223b9157e25e11f88873541b39292d12872
|
|
This avoids a warning in TFLu
Change-Id: I90f6d6cefbe69c720d3c65c550d7f65e727c48fb
|
|
Remove PMU configuration from the InferenceJob struct and add an
external context parameter intead. The external context is passed
to the TFLu interpreter and will be returned in the
ethosu_inference_begin() and ethosu_inference_end() callbacks.
Change-Id: I6dab04c0ab5088b1325be365d77d65d1182e7441
|
|
Split runJob() into smaller functions to make the code easier to
read and maintain.
Correct return type of Layer By Layer profiler GetTotalTicks().
Change-Id: Ie414522017b3b6c6da9a09182439dbe4b2efdb1f
|
|
Change-Id: Idffa9b21385451883795412076ea48218d6c74ba
|
|
Adapt layer-by-layer profiler to handle PMU event configuration,
PMU setup and retrieving PMU counter values.
Adapt the inference process application to support the lbl-profiler
PMU setup and retrieve/save PMU counter values in the InferenceJob
struct.
Change-Id: I1667a5b11c43c54e7d28232b594dd118bf3f79a8
|