NN non-square and 1-D pooling #586
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Conflicts: # README.md
# Conflicts: # ARM.CMSIS.pdsc
# Conflicts: # ARM.CMSIS.pdsc # CMSIS/DoxyGen/General/general.dxy # CMSIS/DoxyGen/General/src/introduction.txt # CMSIS/DoxyGen/RTOS2/src/cmsis_os2.txt # CMSIS/RTOS2/RTX/Include/rtx_os.h # CMSIS/RTOS2/RTX/Library/ARM/MDK/RTX_CM.uvprojx # CMSIS/RTOS2/RTX/Library/ARM/RTX_CM0.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_CM3.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_CM4F.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_V8MB.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_V8MBN.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_V8MM.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_V8MMF.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_V8MMFN.lib # CMSIS/RTOS2/RTX/Library/ARM/RTX_V8MMN.lib # CMSIS/RTOS2/RTX/Library/GCC/MDK/RTX_CM.uvprojx # CMSIS/RTOS2/RTX/Library/GCC/libRTX_CM0.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_CM3.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_CM4F.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_V8MB.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_V8MBN.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_V8MM.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_V8MMF.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_V8MMFN.a # CMSIS/RTOS2/RTX/Library/GCC/libRTX_V8MMN.a # CMSIS/RTOS2/RTX/Library/IAR/RTX_CM0.a # CMSIS/RTOS2/RTX/Library/IAR/RTX_CM3.a # CMSIS/RTOS2/RTX/Library/IAR/RTX_CM4F.a # README.md
…l.com. This redirect is going to be overwritten with the full documentation at next full release.
…one documentation.
Change-Id: I3de07af0117bfcdcbb4cab85b6de281c6c4ac773
Change-Id: I9a99d0c814f3c74eb4414eb633faf5e2a08ef575
Rename the define EVR_RTX_KERNEL_INITIALIZE_COMPLETED_DISABLE to EVR_RTX_KERNEL_INITIALIZED_DISABLE so it matches the rest of the codebase. The commit 858f429 - "RTX5: minor change in events (name consistency)" initially changed this define but missed this one spot.
Fix define around EvrRtxKernelInitialized
…define Revert "Fix define around EvrRtxKernelInitialized"
Didn't find any reference to 'Serial Viewer Output' on any ARM page. Therefore I assume that the correct expression is 'Serial Wire Output'.
This reverts commit 85f9713.
Change-Id: I01e3cdd00cfca44e15987062c53f1c131ef0c5fe
Change-Id: I74ec801603d2de3eba34767da49ba0f30f87eb2e
Change-Id: I2d6dff97fac04ba88dde42361178e82ab82f62fd
Change-Id: I89b24f5f6fc8877570b967588d93d1904259d39f
Change-Id: Ide55b6cffb84ef61b4a5b2c7b0513407fbe77b52
Change-Id: Icd8983e8e2f94cb4cf02b3c78b7fa179b654b4f0
Change-Id: I78ad450657733426cff4bf22fba0d3c181378a86
…erview. Change-Id: I2eae8b40bb71b0475c945a0e2c5c3d5c242bf282
- Added general version history. Change-Id: Iae74ff8bd138d25e290cff7c331db995bf2be911
…ic test for 1d pooling
|
Sorry I broke this on accident, I'm going to reopen it |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed four reference functions for non-square and 1D pooling (Max and Ave). It doesn't make a bunch of sense to have non-square CNNS but not non-square pooling as referenced in issue 455 ( #455 ) Also added the 1D convolution because it can be done slightly faster by separating it into its own function.
I also added 1D wrappers around the non-square CNN functions to allow simplified code for 1D CNNs to work with the 1D pooling. This would make issue 541 simpler. ( #541 ) I didn't seperate the CNN's into their own functions because there was not an obvious potential to save OPs
Note I did add test to the test code but I don't have Keil so I haven't tested my implementations so there are probably bugs in the shapes of the Non-Square functions