liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383696605
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383707907
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383708697
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383693836
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383709402
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383710409
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383706780
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu removed a comment on issue #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#issuecomment-590702012
Hi @huajsj , thanks for the great work! I have tried your work on PYNQ
board, it worked correctly as expected.
--
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383710142
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
liangfu commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r383695691
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Fou
lhutton1 commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383812005
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Ap
lhutton1 commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383812005
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Ap
lhutton1 commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383812005
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Ap
siju-samuel opened a new pull request #4937: [CI Upgrade]Update existing
testcases to run on TF/Lite 2.0 version
URL: https://github.com/apache/incubator-tvm/pull/4937
The existing CI upgrade part of tf2.0 upgrade as per [the
discussion](https://discuss.tvm.ai/t/moving-up-to-tensorflow-1-1
FrozenGene merged pull request #4923: [LLVM] Fix build breaks with StringRef
changes
URL: https://github.com/apache/incubator-tvm/pull/4923
This is an automated message from the Apache Git Service.
To respond to the message,
This is an automated email from the ASF dual-hosted git repository.
zhaowu pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from a2429c1 [Relay][External Codegen] Support data types for
CSourceModuleCodegen args and output (#4934)
a
FrozenGene commented on issue #4923: [LLVM] Fix build breaks with StringRef
changes
URL: https://github.com/apache/incubator-tvm/pull/4923#issuecomment-590822514
Thanks @wpan11nv It is merged.
This is an automated message fro
lhutton1 commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383820228
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to th
lhutton1 commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383815227
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Ap
lhutton1 commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383821438
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to th
wweic commented on issue #4628: [Object] Add String container
URL: https://github.com/apache/incubator-tvm/pull/4628#issuecomment-590905104
@tqchen @FrozenGene Just to give a quick update, I'm testing with clang
9.0.1 in Mac OS X. These test macros(`__cpp_lib_experimental_string_view`, and
mbaret commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383922454
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Apac
mbaret commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383923497
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Apac
mbaret commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383918746
##
File path: python/tvm/relay/op/contrib/dnnl.py
##
@@ -0,0 +1,72 @@
+# Licensed to the Apache S
mbaret commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r383924345
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to the
notoraptor opened a new pull request #4938: [topi][relay] add operation tan to
TVM
URL: https://github.com/apache/incubator-tvm/pull/4938
Hi! This pull request adds operation tan to TVM. The main goal is to make it
available in relay module.
---
cchung100m commented on issue #4925: [DOCS] Fix Sphinx Warning: the target
found for cross-reference
URL: https://github.com/apache/incubator-tvm/pull/4925#issuecomment-590960984
Hi @comaniac
Thanks for the review.
I solve the undefined label warnings and the output file as s
comaniac commented on a change in pull request #4936: [Tutorial] Add a tutorial
for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r383999070
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,166 @@
+# Licensed to the Apache Software
comaniac commented on a change in pull request #4936: [Tutorial] Add a tutorial
for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r383998281
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,166 @@
+# Licensed to the Apache Software
comaniac commented on a change in pull request #4936: [Tutorial] Add a tutorial
for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r38490
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,166 @@
+# Licensed to the Apache Software
yongfeng-nv commented on a change in pull request #4931: WIP: Alternative
solution to Split node min range is not stringent (4885). Don't Merge.
URL: https://github.com/apache/incubator-tvm/pull/4931#discussion_r384007120
##
File path: src/te/schedule/message_passing.cc
##
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384014229
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to th
zhiics commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384019498
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Apac
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384015814
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to th
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384017579
##
File path: python/tvm/relay/op/contrib/dnnl.py
##
@@ -0,0 +1,72 @@
+# Licensed to the Apache
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384019199
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Ap
tqchen commented on issue #4628: [Object] Add String container
URL: https://github.com/apache/incubator-tvm/pull/4628#issuecomment-590977222
Interesting, i think it is fine given that we will fallback to the
std::string impl. The standard only requires it to available in C++17, so they
mig
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384021494
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Ap
tqchen commented on issue #4628: [Object] Add String container
URL: https://github.com/apache/incubator-tvm/pull/4628#issuecomment-590978688
please look into the compilation error.
This is an automated message from the Apache
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384022363
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to th
comaniac commented on issue #4933: [Relay] Target annotation for external
codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-590980676
> I've mostly got a few questions about the direction of the PR. The thing
that concerns me is the lack of support for composite o
zhiics commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384026695
##
File path: src/relay/pass/annotate_target.cc
##
@@ -0,0 +1,102 @@
+/*
+ * Licensed to the Apac
tqchen commented on issue #4937: [CI Upgrade]Update existing testcases to run
on TF/Lite 2.0 version
URL: https://github.com/apache/incubator-tvm/pull/4937#issuecomment-590983662
NOTE: the CI image won't be updated automatically as per script. Please send
testcase update to make sure thing
zhiics commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384029307
##
File path: tests/python/relay/test_annotate_target.py
##
@@ -0,0 +1,152 @@
+# Licensed to the
kevinthesun commented on a change in pull request #4936: [Tutorial] Add a
tutorial for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r384035500
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,166 @@
+# Licensed to the Apache Softw
kevinthesun commented on a change in pull request #4936: [Tutorial] Add a
tutorial for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r384035500
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,166 @@
+# Licensed to the Apache Softw
soiferj commented on issue #4919: [Relay][Pass] Don't consider constants as
free vars in MergeComposite
URL: https://github.com/apache/incubator-tvm/pull/4919#issuecomment-590991944
Sounds good. I'll close this PR for now.
Th
soiferj closed pull request #4919: [Relay][Pass] Don't consider constants as
free vars in MergeComposite
URL: https://github.com/apache/incubator-tvm/pull/4919
This is an automated message from the Apache Git Service.
To res
soiferj opened a new pull request #4939: [Relay][Pass] Make MergeComposite
generated function params ordered alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939
This change makes the order of the params to `MergeComposite`'s generated
function alphabetical. This is to mak
kevinthesun commented on a change in pull request #4930: [Frontend] [MXNet]
make_loss operator support
URL: https://github.com/apache/incubator-tvm/pull/4930#discussion_r384037726
##
File path: python/tvm/relay/frontend/mxnet.py
##
@@ -646,6 +646,11 @@ def _mx_arange(input
zhiics commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591005365
@mbaret We can check composite in the op register only if it is an `op`.
Otherwise, we can check the CallNode whose op is a function
mbaret commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591006773
What I'd like to understand is what the composite equivalent of this line
will be:
```
static auto fannotate = Op::GetAttr("tar
soiferj commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591008923
Sorry, I don't quite understand - where do you think the change should be
made? The main issue here is
huajsj commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r384056975
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Foun
zhiics commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591009487
@mbaret We shouldn't even get to that line because the op in the call is not
a primitive op, but a FunctionNode. We need to check the
soiferj edited a comment on issue #4939: [Relay][Pass] Sort MergeComposite
generated function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591008923
Sorry, I don't quite understand - where do you think the change should be
made? The main issue h
huajsj commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r384061674
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Foun
huajsj commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r384061553
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Foun
comaniac commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591013575
The example I just made may not be the best. If your issue is only caused by
`FreeVars()`, then maybe
comaniac edited a comment on issue #4939: [Relay][Pass] Sort MergeComposite
generated function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591013575
The example I just made may not be the best. If your issue is only caused by
`FreeVars()`, then
soiferj commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591014458
Got it, I see what you're saying. I'm still not convinced that we want to
generalize sorting free vars
huajsj commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r384064866
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Foun
huajsj commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r384065248
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Foun
soiferj edited a comment on issue #4939: [Relay][Pass] Sort MergeComposite
generated function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591014458
Got it, I see what you're saying. I'm still not convinced that we want to
generalize sorting fre
comaniac commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591017348
In my opinion, we should try to simplify the task of Relay passes. If we let
the merge composite pass
huajsj commented on issue #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#issuecomment-591017634
Hi @liangfu, nice to know the tutorial worked in your environment, and
thanks for the kindly follow up , all review comments get addressed, please
check.
---
huajsj commented on a change in pull request #4887: [VTA] YoloV3 Support
URL: https://github.com/apache/incubator-tvm/pull/4887#discussion_r384070679
##
File path: vta/tutorials/frontend/deploy_detection_on_vta.py
##
@@ -0,0 +1,334 @@
+# Licensed to the Apache Software Foun
MarisaKirisame commented on issue #4881: [WIP] Fixing an Infinite Loop case in
UnmatchedChecker.
URL: https://github.com/apache/incubator-tvm/pull/4881#issuecomment-591021694
@slyubomirsky @zhiics @junrushao1994 can you guys review?
-
alexwong commented on a change in pull request #4936: [Tutorial] Add a tutorial
for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r384073516
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,166 @@
+# Licensed to the Apache Software
zhiics commented on a change in pull request #4881: [WIP] Fixing an Infinite
Loop case in UnmatchedChecker.
URL: https://github.com/apache/incubator-tvm/pull/4881#discussion_r384077999
##
File path: tests/python/relay/test_pass_unmatched_cases.py
##
@@ -273,3 +274,27 @@ de
maheshambule commented on a change in pull request #4930: [Frontend] [MXNet]
make_loss operator support
URL: https://github.com/apache/incubator-tvm/pull/4930#discussion_r384082327
##
File path: python/tvm/relay/frontend/mxnet.py
##
@@ -646,6 +646,11 @@ def _mx_arange(inpu
zhiics commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591032603
IMHO, this might seem like a workaround, but I think it doesn't really solve
the ordering problem of pat
zhiics edited a comment on issue #4939: [Relay][Pass] Sort MergeComposite
generated function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591032603
IMHO, this might seem like a workaround, but I think it doesn't really solve
the ordering problem
soiferj commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591040506
Thanks, @zhiics. This fix is not intended to solve the ordering of pattern
matching, but rather to solv
mbaret commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591047600
Where will we register the composite equivalent of FTVMAnnotateTarget?
--
zhiics commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591051281
no need to register if we don't have an op for it. They are annotated
already.
-
mbaret commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591051986
But if we want to register a check that depends on the attrs/args, then we
will need to provide a function here.
---
zhiics commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591052375
Yeah, my point is that we probably want to think a bit to make patter
matcher a little stronger or have
anijain2305 commented on issue #4873: [Relay][FastMath] Relay pass to use fast
exp/tanh
URL: https://github.com/apache/incubator-tvm/pull/4873#issuecomment-591052536
I investigated more. I am also little uncomfortable about adding another
Relay build config option. It changes API of PassCo
zhiics commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591062207
I probably missed something, but we already have a function because the
composite pass has been executed, right?
---
comaniac commented on issue #4933: [Relay] Target annotation for external
codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591063851
More specifically,
```python
auto op = call->op;
if (op.as()) {
# Check function attribute "composite" which type
soiferj commented on issue #4939: [Relay][Pass] Sort MergeComposite generated
function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591064120
I definitely agree that we'll want that eventually. However, improved
pattern matching won't change the
soiferj edited a comment on issue #4939: [Relay][Pass] Sort MergeComposite
generated function params alphabetically
URL: https://github.com/apache/incubator-tvm/pull/4939#issuecomment-591064120
I definitely agree that we'll want that eventually. However, improved
pattern matching won't cha
mbaret commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591065765
To give an example, we match qnn.conv_2d -> bias_add -> requantize to a
composite convolution (which represents a tflite quantized co
yzhliu merged pull request #4935: [Fix] remove unnecessary spliting in the
cached chunk
URL: https://github.com/apache/incubator-tvm/pull/4935
This is an automated message from the Apache Git Service.
To respond to the messa
This is an automated email from the ASF dual-hosted git repository.
liuyizhi pushed a change to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git.
from 588523d [LLVM] Fix build breaks from StringRef changes (#4923)
add 545f6ea [Fix] remove unnecessary s
comaniac commented on issue #4933: [Relay] Target annotation for external
codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591073220
IMHO, if you want to define a pattern by yourself, you should be clear about
what can be supported and specify it in the pattern.
icemelon9 commented on a change in pull request #4933: [Relay] Target
annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384133372
##
File path: python/tvm/relay/op/contrib/__init__.py
##
@@ -15,7 +15,8 @@
# specific languag
comaniac commented on a change in pull request #4933: [Relay] Target annotation
for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#discussion_r384136754
##
File path: python/tvm/relay/op/contrib/__init__.py
##
@@ -15,7 +15,8 @@
# specific language
kamathhrishi opened a new issue #4940: Build issues
URL: https://github.com/apache/incubator-tvm/issues/4940
I tried to install TVM with SGX support. When building using CMake the
versions seem fine.
This was on Ubuntu 18.04.
```(tf2) hrishikesh~/tvm/build$ cmake ..
-- Th
kamathhrishi commented on issue #4940: Build issues
URL: https://github.com/apache/incubator-tvm/issues/4940#issuecomment-591081176
Further
```[ 87%] Building CXX object
CMakeFiles/tvm.dir/src/runtime/rpc/rpc_device_api.cc.o
[ 87%] Building CXX object
CMakeFiles/tvm.dir/src/run
mbaret commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591083262
The motivation behind merge composite was to make it so that this annotation
mechanism could handle the case where there's a many-to-
yzhliu commented on a change in pull request #4922: [Relay][pass] call graph
for relay
URL: https://github.com/apache/incubator-tvm/pull/4922#discussion_r384153634
##
File path: src/relay/pass/call_graph.cc
##
@@ -0,0 +1,339 @@
+/*
+ * Licensed to the Apache Software Found
zhiics commented on a change in pull request #4922: [Relay][pass] call graph
for relay
URL: https://github.com/apache/incubator-tvm/pull/4922#discussion_r384160228
##
File path: src/relay/pass/call_graph.cc
##
@@ -0,0 +1,339 @@
+/*
+ * Licensed to the Apache Software Found
zhiics commented on a change in pull request #4922: [Relay][pass] call graph
for relay
URL: https://github.com/apache/incubator-tvm/pull/4922#discussion_r384160228
##
File path: src/relay/pass/call_graph.cc
##
@@ -0,0 +1,339 @@
+/*
+ * Licensed to the Apache Software Found
yzhliu commented on a change in pull request #4936: [Tutorial] Add a tutorial
for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r384162338
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,171 @@
+# Licensed to the Apache Software F
zhiics commented on issue #4933: [Relay] Target annotation for external codegen
URL: https://github.com/apache/incubator-tvm/pull/4933#issuecomment-591105578
@mbaret Thanks for all discussions. I think this is actually a separate
issue to this PR. We probably want to have an RFC for it. We
alexwong commented on a change in pull request #4936: [Tutorial] Add a tutorial
for PyTorch
URL: https://github.com/apache/incubator-tvm/pull/4936#discussion_r384167453
##
File path: tutorials/frontend/from_pytorch.py
##
@@ -0,0 +1,171 @@
+# Licensed to the Apache Software
yongfeng-nv commented on issue #4931: Tighten split's extent
URL: https://github.com/apache/incubator-tvm/pull/4931#issuecomment-591127743
@merrymercy, I have updated this PR description and refactor the code. All
tests are clear. Please review it. Thank you.
---
icemelon9 opened a new pull request #4941: Bump up dev version
URL: https://github.com/apache/incubator-tvm/pull/4941
Not for release purpose. This is only for convenience purposes to
differentiate TVM with and w/o op strategy.
cc @tqchen @comaniac
-
1 - 100 of 164 matches
Mail list logo