[VTA][Relay] Extending Vision model coverage compilation for VTA #3740
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
vegaluisjose
approved these changes
Aug 9, 2019
tmoreau89
commented
Aug 9, 2019
| @@ -61,6 +62,8 @@ | |||
| # Make sure that TVM was compiled with RPC=1 | |||
| assert tvm.module.enabled("rpc") | |||
|
|
|||
| # Increase python recursion limit to traverse Relay program | |||
| sys.setrecursionlimit(10000) | |||
There was a problem hiding this comment.
@jroesch is there a better place to put this line?
weberlo
approved these changes
Aug 9, 2019
|
@tmoreau89 maybe we should set limit in tvm init.py |
|
@MarisaKirisame If we set it in init.py, then do we not revert it back to its original value?? |
|
@tmoreau89 why revert at all? tvm is supposed to be used with a high recursion limit. if user wish otherwise she should pick a better constant herself. |
|
Ok, which init.py should we include the lines in? the top level? |
|
@tmoreau89 I think relay is fine for now. If tvm also need it, we can move it up lateron. |
tmoreau89
changed the title
liangfu
reviewed
Aug 18, 2019
liangfu
approved these changes
Aug 18, 2019
jroesch
approved these changes
Sep 5, 2019
MarisaKirisame
pushed a commit
to MarisaKirisame/tvm
that referenced
this pull request
Sep 7, 2019
…che#3740) * adding support for graphpack over multiply op * increasing resnet model coverage * fix indentation * lint * moving recursion limit fix into graphpack pass * moving recursionlimit to relay init * pooling on NCHWnc format * adding more models * deploy_resnet_on_vta.py * trailing line * generalizing to vision models * merge conflicts * fix, apply quantization to VTA only * improving comments * trimming models that have runtime issues for the moment * lint * lint * lint
wweic
pushed a commit
to wweic/tvm
that referenced
this pull request
Sep 16, 2019
…che#3740) * adding support for graphpack over multiply op * increasing resnet model coverage * fix indentation * lint * moving recursion limit fix into graphpack pass * moving recursionlimit to relay init * pooling on NCHWnc format * adding more models * deploy_resnet_on_vta.py * trailing line * generalizing to vision models * merge conflicts * fix, apply quantization to VTA only * improving comments * trimming models that have runtime issues for the moment * lint * lint * lint
wweic
pushed a commit
to wweic/tvm
that referenced
this pull request
Sep 16, 2019
…che#3740) * adding support for graphpack over multiply op * increasing resnet model coverage * fix indentation * lint * moving recursion limit fix into graphpack pass * moving recursionlimit to relay init * pooling on NCHWnc format * adding more models * deploy_resnet_on_vta.py * trailing line * generalizing to vision models * merge conflicts * fix, apply quantization to VTA only * improving comments * trimming models that have runtime issues for the moment * lint * lint * lint
wweic
pushed a commit
to neo-ai/tvm
that referenced
this pull request
Sep 16, 2019
…che#3740) * adding support for graphpack over multiply op * increasing resnet model coverage * fix indentation * lint * moving recursion limit fix into graphpack pass * moving recursionlimit to relay init * pooling on NCHWnc format * adding more models * deploy_resnet_on_vta.py * trailing line * generalizing to vision models * merge conflicts * fix, apply quantization to VTA only * improving comments * trimming models that have runtime issues for the moment * lint * lint * lint
tqchen
pushed a commit
to tqchen/tvm
that referenced
this pull request
Mar 29, 2020
…che#3740) * adding support for graphpack over multiply op * increasing resnet model coverage * fix indentation * lint * moving recursion limit fix into graphpack pass * moving recursionlimit to relay init * pooling on NCHWnc format * adding more models * deploy_resnet_on_vta.py * trailing line * generalizing to vision models * merge conflicts * fix, apply quantization to VTA only * improving comments * trimming models that have runtime issues for the moment * lint * lint * lint
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.
This fixes Relay compilation for ResNet18_v2, ResNet34_v2, ResNet50_v2, ResNet101_v2, ResNet152_v2, Alexnet, VGG11, VGG13, VGG16, VGG19.
Currently the following models will work in HW/SIM (white-listed in tutorial until more models are supported): ResNet18_v2, ResNet34_v2, ResNet50_v2, ResNet101_v2
Performance on
pynqIoT platform: 371.68ms, 470.86ms, 662.13ms and 916.26ms for ResNet-18, 34, 50, and 101 respectively.Bug fixes:
1<<24to avoid out of memory errors on the pynq[resolved]: getting quantization to pass on these models is pending on #3543 being merged. I've addressed this dependence with a TODO.