[GitHub] [incubator-tvm] MarisaKirisame commented on issue #5427: [Relay] Lack complex tests for parser

MarisaKirisame commented on issue #5427: URL: https://github.com/apache/incubator-tvm/issues/5427#issuecomment-618890273 Hi, in the parser test file, you can see that we use roundtrip to do test - in particular, we test that parsing/pretty printing or prettyprinting/parsing give the same

[GitHub] [incubator-tvm] mbaret commented on pull request #5409: [BYOC] Don't annotate constants

mbaret commented on pull request #5409: URL: https://github.com/apache/incubator-tvm/pull/5409#issuecomment-618895831 I'm concerned that if we enforce annotation of individual constant nodes, MergeCompilerRegions will become a necessary pass to run. Otherwise we'll generate partitions

[GitHub] [incubator-tvm] libaihong opened a new pull request #5428: [CODEGEN][CUDA] Fix a bug when vectorized load was involved for…

libaihong opened a new pull request #5428: URL: https://github.com/apache/incubator-tvm/pull/5428 When I test with vectorized load for char2, there is a cuda compilation error: _RuntimeError: Compilation error: /tmp/tmpa87wlle2/my_kernel.cu(3093): error: no operator "=" matches

[incubator-tvm] branch master updated: [PYTORCH]where, addcdiv, addcmul op support (#5383)

This is an automated email from the ASF dual-hosted git repository. masahi pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git The following commit(s) were added to refs/heads/master by this push: new ba38222 [PYTORCH]where, addcdiv,

[GitHub] [incubator-tvm] lixiaoquan opened a new pull request #5429: [RELAY][TF] Support symbolic newshape for Reshape

lixiaoquan opened a new pull request #5429: URL: https://github.com/apache/incubator-tvm/pull/5429 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the

[GitHub] [incubator-tvm] mbaret commented on pull request #5345: [RELAY] Move frontend utils

mbaret commented on pull request #5345: URL: https://github.com/apache/incubator-tvm/pull/5345#issuecomment-618893040 I couldn't find an appropriate folder to move this to, so I've instead inlined this function as it's only used in two places.

[incubator-tvm] branch master updated: [FRONTEND][TFLITE]Gather, StridedSlice op support added (#4788)

This is an automated email from the ASF dual-hosted git repository. zhaowu pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git The following commit(s) were added to refs/heads/master by this push: new 8efd546 [FRONTEND][TFLITE]Gather,

[GitHub] [incubator-tvm] FrozenGene commented on pull request #4788: [FRONTEND][TFLITE]Gather, StridedSlice op support added

FrozenGene commented on pull request #4788: URL: https://github.com/apache/incubator-tvm/pull/4788#issuecomment-618942742 Thanks @siju-samuel @wyc-ruiker @u99127 This is an automated message from the Apache Git Service. To

[GitHub] [incubator-tvm] siju-samuel commented on pull request #5383: [PYTORCH]where, addcdiv, addcmul op support

siju-samuel commented on pull request #5383: URL: https://github.com/apache/incubator-tvm/pull/5383#issuecomment-618937603 @masahi Could you please review and merge. Thanks. This is an automated message from the Apache Git

[GitHub] [incubator-tvm] masahi commented on pull request #5383: [PYTORCH]where, addcdiv, addcmul op support

masahi commented on pull request #5383: URL: https://github.com/apache/incubator-tvm/pull/5383#issuecomment-618941150 Thanks @siju-samuel This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [incubator-tvm] antinucleon opened a new issue #5427: [Relay] Lack complex tests for fromtext

antinucleon opened a new issue #5427: URL: https://github.com/apache/incubator-tvm/issues/5427 Complex tests such as full models should be added to relay parser tests. As today, I tested MobileNet / Bert and a few more models, none is able to be successfully loaded.

[GitHub] [incubator-tvm] JishinMaster opened a new pull request #5432: Corrected TVM autotuning on GPU

JishinMaster opened a new pull request #5432: URL: https://github.com/apache/incubator-tvm/pull/5432 Added missing "tir" in tvm.tir.analysis.verify_gpu_code(f, kwargs) , which made the tuning on GPU (tested with RTX2080) fail. Requesting a review from @eqy or @merrymercy

[GitHub] [incubator-tvm] dhruvaray commented on pull request #5430: [RELAY,TOPI][TFLITE] Sparse to dense operator - WIP

dhruvaray commented on pull request #5430: URL: https://github.com/apache/incubator-tvm/pull/5430#issuecomment-619016587 Closing this as I messed up on the rebase. Please ignore This is an automated message from the Apache

[GitHub] [incubator-tvm] dhruvaray opened a new pull request #5430: [RELAY,TOPI][TFLITE] Sparse to dense operator

dhruvaray opened a new pull request #5430: URL: https://github.com/apache/incubator-tvm/pull/5430 Added support for sparse_to_dense operator @u99127, @FrozenGene, @siju-samuel @jwfromm @kazum - Kindly review This is

[GitHub] [incubator-tvm] t-vi opened a new pull request #5431: misc fixes for ROCm

t-vi opened a new pull request #5431: URL: https://github.com/apache/incubator-tvm/pull/5431 This fixes two things in the AMDGPU LLVM codegen. - pointer lifetime of the LLVM context passed to codegen as a pointer - the runtime::String refactor caused typing issues for one of the nodes

[GitHub] [incubator-tvm] hzfan commented on a change in pull request #5367: Improve IntervalSet's floormod

hzfan commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414552082 ## File path: src/te/operation/compute_op.cc ## @@ -231,20 +231,18 @@ void ComputeOpNode::PropBoundToInputs( // undefined behaviour), so

[GitHub] [incubator-tvm] hzfan commented on a change in pull request #5367: Improve IntervalSet's floormod

hzfan commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414548737 ## File path: src/te/operation/compute_op.cc ## @@ -231,20 +231,18 @@ void ComputeOpNode::PropBoundToInputs( // undefined behaviour), so

[GitHub] [incubator-tvm] hzfan commented on a change in pull request #5367: Improve IntervalSet's floormod

hzfan commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414548737 ## File path: src/te/operation/compute_op.cc ## @@ -231,20 +231,18 @@ void ComputeOpNode::PropBoundToInputs( // undefined behaviour), so

[GitHub] [incubator-tvm] zhiics commented on a change in pull request #5345: [RELAY] Move frontend utils

zhiics commented on a change in pull request #5345: URL: https://github.com/apache/incubator-tvm/pull/5345#discussion_r414572622 ## File path: python/tvm/relay/frontend/tflite.py ## @@ -2219,6 +2218,20 @@ def get_expr(self, input_tensor_idx): def has_expr(self,

[GitHub] [incubator-tvm] zhiics commented on a change in pull request #5345: [RELAY] Move frontend utils

zhiics commented on a change in pull request #5345: URL: https://github.com/apache/incubator-tvm/pull/5345#discussion_r414572622 ## File path: python/tvm/relay/frontend/tflite.py ## @@ -2219,6 +2218,20 @@ def get_expr(self, input_tensor_idx): def has_expr(self,

[GitHub] [incubator-tvm] t-vi opened a new pull request #5433: fix miopen padding

t-vi opened a new pull request #5433: URL: https://github.com/apache/incubator-tvm/pull/5433 The MIOpen convolution accidentally passed the pad left + pad right (and top + bottom) instead of just the value applied to both sides. This patch fixes this, and also conditions the use of MIOpen

[GitHub] [incubator-tvm] tqchen opened a new issue #5435: [FFI] Improve FFI Type Error Message

tqchen opened a new issue #5435: URL: https://github.com/apache/incubator-tvm/issues/5435 Currently, the PrintType for Array/Map does not try to nest into the content, but the FFI type checking do. This can create confusion when there is an error message. We should improve the TypeName

[incubator-tvm] branch master updated: misc fixes for ROCm (pointer lifetime, runtime::String refactor) (#5431)

This is an automated email from the ASF dual-hosted git repository. tqchen pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git The following commit(s) were added to refs/heads/master by this push: new 9a989e5 misc fixes for ROCm (pointer

[GitHub] [incubator-tvm] tqchen commented on pull request #5433: fix miopen padding

tqchen commented on pull request #5433: URL: https://github.com/apache/incubator-tvm/pull/5433#issuecomment-619108217 thanks @t-vi please rebase This is an automated message from the Apache Git Service. To respond to the

[GitHub] [incubator-tvm] zhiics commented on issue #5427: [Relay] Lack complex tests for parser

zhiics commented on issue #5427: URL: https://github.com/apache/incubator-tvm/issues/5427#issuecomment-619123723 I think I met some other issues before with the roundtrip parser. If I remember correctly, they don't support functions in prelude

[GitHub] [incubator-tvm] zhiics edited a comment on issue #5427: [Relay] Lack complex tests for parser

zhiics edited a comment on issue #5427: URL: https://github.com/apache/incubator-tvm/issues/5427#issuecomment-619123723 I think I met some other issues before with the roundtrip parser. If I remember correctly, they don't support all functions in prelude

[GitHub] [incubator-tvm] maheshambule commented on a change in pull request #5330: [Frontend][TFLite] support for FILL and SPLIT_V operators

maheshambule commented on a change in pull request #5330: URL: https://github.com/apache/incubator-tvm/pull/5330#discussion_r414740666 ## File path: python/tvm/relay/frontend/tflite.py ## @@ -1067,6 +1069,27 @@ def convert_zeros_like(self, op): return out +def

[GitHub] [incubator-tvm] maheshambule commented on pull request #5330: [Frontend][TFLite] support for FILL and SPLIT_V operators

maheshambule commented on pull request #5330: URL: https://github.com/apache/incubator-tvm/pull/5330#issuecomment-619145838 @u99127 I have merged the master and removed the duplicate code. This is an automated message from

[GitHub] [incubator-tvm] michalpiszczek opened a new pull request #5436: [TFLite Runtime] Re-enable test for remote execution via RPC

michalpiszczek opened a new pull request #5436: URL: https://github.com/apache/incubator-tvm/pull/5436 Follow up to: https://github.com/apache/incubator-tvm/pull/5425 Changes: - Re-enables a test for remote execution with the TFLite runtime via RPC - The test for local

[GitHub] [incubator-tvm] tmoreau89 commented on pull request #5436: [TFLite Runtime] Re-enable test for remote execution via RPC

tmoreau89 commented on pull request #5436: URL: https://github.com/apache/incubator-tvm/pull/5436#issuecomment-619150720 @michalpiszczek one suggestion is just to leave the CI commented out, and scope this PR to just changing the dockerfile

[GitHub] [incubator-tvm] jroesch commented on pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

jroesch commented on pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#issuecomment-619150640 My one worry is that this is a bit fragile, what happens if the number of children changes but we forget to update the bound?

[GitHub] [incubator-tvm] mbaret commented on pull request #5409: [BYOC] Don't annotate constants

mbaret commented on pull request #5409: URL: https://github.com/apache/incubator-tvm/pull/5409#issuecomment-619080360 It looks like the refactor to AnnotatedRegionSet has broken this 'fix'. I think the important property to maintain is that 'every node should belong to a region' rather

[GitHub] [incubator-tvm] siju-samuel opened a new pull request #5434: [PYTORCH]Rsub, Embedded, OneHot ops support

siju-samuel opened a new pull request #5434: URL: https://github.com/apache/incubator-tvm/pull/5434 - rsub - embedded - one_hot #5133 @masahi please help me to review this pytorch frontend ops. Thanks. This

[incubator-tvm] branch master updated (9a989e5 -> 9687307)

This is an automated email from the ASF dual-hosted git repository. tqchen pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git. from 9a989e5 misc fixes for ROCm (pointer lifetime, runtime::String refactor) (#5431) add 9687307

[GitHub] [incubator-tvm] tqchen commented on pull request #5432: Corrected TVM autotuning on GPU

tqchen commented on pull request #5432: URL: https://github.com/apache/incubator-tvm/pull/5432#issuecomment-619126547 Thanks @JishinMaster ! This is an automated message from the Apache Git Service. To respond to the

[GitHub] [incubator-tvm] tqchen commented on pull request #5436: [TFLite Runtime] Re-enable test for remote execution via RPC

tqchen commented on pull request #5436: URL: https://github.com/apache/incubator-tvm/pull/5436#issuecomment-619125877 @michalpiszczek please PR the docker changes separately as we need to first build the docker binary and update them to the CI env.

[GitHub] [incubator-tvm] tqchen commented on a change in pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

tqchen commented on a change in pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#discussion_r414735177 ## File path: include/tvm/ir/expr.h ## @@ -161,7 +163,8 @@ class RelayExprNode : public BaseExprNode { template inline const TTypeNode*

[GitHub] [incubator-tvm] yongfeng-nv commented on a change in pull request #5367: Improve IntervalSet's floormod

yongfeng-nv commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414674995 ## File path: src/te/operation/compute_op.cc ## @@ -231,20 +231,18 @@ void ComputeOpNode::PropBoundToInputs( // undefined

[GitHub] [incubator-tvm] mbrookhart commented on a change in pull request #5231: [POC] Pattern Language, Matcher, Rewriter, and Function Paritioner

mbrookhart commented on a change in pull request #5231: URL: https://github.com/apache/incubator-tvm/pull/5231#discussion_r414721399 ## File path: python/tvm/relay/df_pattern/__init__.py ## @@ -0,0 +1,488 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or

[GitHub] [incubator-tvm] jroesch commented on a change in pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

jroesch commented on a change in pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#discussion_r414733387 ## File path: include/tvm/ir/expr.h ## @@ -161,7 +163,8 @@ class RelayExprNode : public BaseExprNode { template inline const TTypeNode*

[GitHub] [incubator-tvm] tqchen opened a new pull request #5438: [IR] Initial stab at std::string->String upgrade

tqchen opened a new pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438 This PR is an example of String Update with backward compact(see json_compact.py) This is an automated message from the Apache

[GitHub] [incubator-tvm] michalpiszczek opened a new pull request #5437: [TFLite Runtime] Add TFLite Runtime dependencies to CI CPU docker build

michalpiszczek opened a new pull request #5437: URL: https://github.com/apache/incubator-tvm/pull/5437 This PR contains only the docker build changes needed to re-enable the tests in: https://github.com/apache/incubator-tvm/pull/5436.

[GitHub] [incubator-tvm] tqchen commented on pull request #5438: [IR] Initial stab at std::string->String upgrade

tqchen commented on pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438#issuecomment-619151418 cc @jroesch @icemelon9 This is an automated message from the Apache Git Service. To respond to the

[GitHub] [incubator-tvm] tqchen commented on pull request #5431: misc fixes for ROCm

tqchen commented on pull request #5431: URL: https://github.com/apache/incubator-tvm/pull/5431#issuecomment-619098492 Thansk @t-vi ! This is an automated message from the Apache Git Service. To respond to the message, please

[GitHub] [incubator-tvm] tqchen commented on issue #5427: [Relay] Lack complex tests for parser

tqchen commented on issue #5427: URL: https://github.com/apache/incubator-tvm/issues/5427#issuecomment-619099214 It would be great if we can add some of these models(or found the root cause get a fragement) and add them to the round trip test

[GitHub] [incubator-tvm] tqchen edited a comment on pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

tqchen edited a comment on pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#issuecomment-619153621 The code still works, the overflow children will need to go through the overflow path, and call

[GitHub] [incubator-tvm] tqchen commented on pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

tqchen commented on pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#issuecomment-619153621 The code still works, the overflow children will need to go through the overflow path, and call

[GitHub] [incubator-tvm] anijain2305 opened a new pull request #5439: TFlite e2e FP32 Object detection model

anijain2305 opened a new pull request #5439: URL: https://github.com/apache/incubator-tvm/pull/5439 TBD This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use

[GitHub] [incubator-tvm] wpan11nv edited a comment on pull request #5428: [CODEGEN][CUDA] Fix a bug when vectorized load was involved for…

wpan11nv edited a comment on pull request #5428: URL: https://github.com/apache/incubator-tvm/pull/5428#issuecomment-619214302 Could you please add a few tests.Do you mean loading uchar2x4? We could load/store them as uint16_t x 4. I do not see similar code in PrintType below:

[GitHub] [incubator-tvm] wpan11nv commented on pull request #5428: [CODEGEN][CUDA] Fix a bug when vectorized load was involved for…

wpan11nv commented on pull request #5428: URL: https://github.com/apache/incubator-tvm/pull/5428#issuecomment-619214302 Could you please add a few tests.Do you mean supporting loading uchar2x4? We could load/store them as uint16_t x 4. I do not see similar code in PrintType below:

[GitHub] [incubator-tvm] michalpiszczek commented on pull request #5436: [TFLite Runtime] Re-enable test for remote execution via RPC

michalpiszczek commented on pull request #5436: URL: https://github.com/apache/incubator-tvm/pull/5436#issuecomment-619160168 @tmoreau89 @tqchen I've created a separate PR with just the docker changes here: https://github.com/apache/incubator-tvm/pull/5437

[GitHub] [incubator-tvm] tqchen commented on a change in pull request #5438: [POC][IR] Initial stab at std::string->String upgrade

tqchen commented on a change in pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438#discussion_r414821178 ## File path: python/tvm/ir/json_compact.py ## @@ -84,12 +91,26 @@ def _update_global_key(item, _): del item["global_key"] return

[GitHub] [incubator-tvm] u99127 commented on a change in pull request #5437: [TFLite Runtime] Add TFLite Runtime dependencies to CI CPU docker build

u99127 commented on a change in pull request #5437: URL: https://github.com/apache/incubator-tvm/pull/5437#discussion_r414830870 ## File path: docker/install/ubuntu_install_tflite.sh ## @@ -26,12 +26,18 @@ cd flatbuffers cmake -G "Unix Makefiles" -DCMAKE_BUILD_TYPE=Release

[GitHub] [incubator-tvm] tqchen commented on pull request #5438: [POC][IR] Initial stab at std::string->String upgrade

tqchen commented on pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438#issuecomment-619211235 @jroesch @zhiics wrt to the String serialization. Right now String is serialized as an normal object. So if the content of String is printable, it will show up as

[GitHub] [incubator-tvm] wpan11nv edited a comment on pull request #5382: [TE] Fix MakeLoopNest for warp memory

wpan11nv edited a comment on pull request #5382: URL: https://github.com/apache/incubator-tvm/pull/5382#issuecomment-619201349 Can we make it explicit, e.g by tagging the binding iter_var with "warp" scope. For CUDA, when the block configuration is (1, 32), threadIdx.y may "define" the

[GitHub] [incubator-tvm] tqchen edited a comment on pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

tqchen edited a comment on pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#issuecomment-619153621 The code still works, the overflow children will need to go through the overflow path, and call

[GitHub] [incubator-tvm] yongfeng-nv commented on a change in pull request #5367: Improve IntervalSet's floormod

yongfeng-nv commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414674995 ## File path: src/te/operation/compute_op.cc ## @@ -231,20 +231,18 @@ void ComputeOpNode::PropBoundToInputs( // undefined

[GitHub] [incubator-tvm] tqchen commented on pull request #5423: [RUNTIME][OBJECT] Introduce static slots for common objects.

tqchen commented on pull request #5423: URL: https://github.com/apache/incubator-tvm/pull/5423#issuecomment-619192968 ping @jroesch This is an automated message from the Apache Git Service. To respond to the message, please

[GitHub] [incubator-tvm] zhiics commented on a change in pull request #5438: [POC][IR] Initial stab at std::string->String upgrade

zhiics commented on a change in pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438#discussion_r414804149 ## File path: python/tvm/ir/json_compact.py ## @@ -84,12 +91,26 @@ def _update_global_key(item, _): del item["global_key"] return

[GitHub] [incubator-tvm] comaniac commented on pull request #5409: [BYOC] Don't annotate constants

comaniac commented on pull request #5409: URL: https://github.com/apache/incubator-tvm/pull/5409#issuecomment-619198333 > I'm concerned that if we enforce annotation of individual constant nodes, MergeCompilerRegions will become a necessary pass to run. Otherwise we'll generate partitions

[GitHub] [incubator-tvm] wpan11nv commented on pull request #5382: [TE] Fix MakeLoopNest for warp memory

wpan11nv commented on pull request #5382: URL: https://github.com/apache/incubator-tvm/pull/5382#issuecomment-619201349 Can we make it explicit, e.g by tagging the binding iter_var with "warp" scope. The existing "warp" scope attribute on stage looks odd to me. Should it be an attribute

[incubator-tvm] branch master updated (9687307 -> 3cc4971)

This is an automated email from the ASF dual-hosted git repository. jroesch pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git. from 9687307 Corrected TVM autotuning on GPU (#5432) add 3cc4971 [RUNTIME][OBJECT] Introduce static slots

[GitHub] [incubator-tvm] michalpiszczek commented on a change in pull request #5437: [TFLite Runtime] Add TFLite Runtime dependencies to CI CPU docker build

michalpiszczek commented on a change in pull request #5437: URL: https://github.com/apache/incubator-tvm/pull/5437#discussion_r414833737 ## File path: docker/install/ubuntu_install_tflite.sh ## @@ -26,12 +26,18 @@ cd flatbuffers cmake -G "Unix Makefiles"

[GitHub] [incubator-tvm] zhiics commented on pull request #5438: [POC][IR] Initial stab at std::string->String upgrade

zhiics commented on pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438#issuecomment-619216822 @tqchen ahh, I see. I was thinking if it is necessary to have similar stuff like Load/Save for tvm::String, so that we don't need to put them under serialization.cc. Looks

[GitHub] [incubator-tvm] zhiics edited a comment on pull request #5438: [POC][IR] Initial stab at std::string->String upgrade

zhiics edited a comment on pull request #5438: URL: https://github.com/apache/incubator-tvm/pull/5438#issuecomment-619216822 @tqchen ahh, I see. I was thinking if it is necessary to have some stuff like Load/Save for tvm::String separately, so that we don't need to put them under

[GitHub] [incubator-tvm] dreamqin68 opened a new issue #5440: import vta: Cannot find config in /3rdparty/vta-hw/config/vta_config.json

dreamqin68 opened a new issue #5440: URL: https://github.com/apache/incubator-tvm/issues/5440 I installed VTA according to the VTA Simulator Installation guide. Then I set the environment. But when I imported vta, there was a bug. `Traceback (most recent call last): File

[GitHub] [incubator-tvm] jwfromm commented on a change in pull request #5441: Add TopK to ONNX Frontend

jwfromm commented on a change in pull request #5441: URL: https://github.com/apache/incubator-tvm/pull/5441#discussion_r414874263 ## File path: python/tvm/relay/frontend/onnx.py ## @@ -1470,6 +1470,23 @@ def _impl_v9(cls, inputs, attr, params): output =

[GitHub] [incubator-tvm] yzhliu commented on a change in pull request #5367: Improve IntervalSet's floormod

yzhliu commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414870644 ## File path: include/tvm/arith/analyzer.h ## @@ -411,8 +412,9 @@ class TVM_DLL Analyzer { * * \param var The variable. * \param expr

[GitHub] [incubator-tvm] weberlo commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

weberlo commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414765031 ## File path: python/tvm/autotvm/tuner/tuner.py ## @@ -150,7 +150,15 @@ def tune(self, n_trial, measure_option, early_stopping=None, callbacks=(),

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414918275 ## File path: src/runtime/micro/micro_session.cc ## @@ -489,6 +629,16 @@ PackedFunc MicroSession::GetFunction( return

[GitHub] [incubator-tvm] mbrookhart commented on issue #3670: [RFC] AlterOpLayout Pass Refactoring

mbrookhart commented on issue #3670: URL: https://github.com/apache/incubator-tvm/issues/3670#issuecomment-619253852 @yzhliu I've hit a couple of issue that would be completely solved by this change :). Any chance you've made progress? If not, would you be offended if I gave it a shot

[GitHub] [incubator-tvm] mbrookhart commented on a change in pull request #5441: Add TopK to ONNX Frontend

mbrookhart commented on a change in pull request #5441: URL: https://github.com/apache/incubator-tvm/pull/5441#discussion_r414914928 ## File path: python/tvm/relay/frontend/onnx.py ## @@ -1470,6 +1470,23 @@ def _impl_v9(cls, inputs, attr, params): output =

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414919196 ## File path: tests/python/unittest/test_runtime_micro.py ## @@ -25,8 +25,25 @@ from tvm.micro import create_micro_mod from tvm.relay.testing

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414919333 ## File path: topi/python/topi/arm_cpu/conv2d_spatial_pack.py ## @@ -173,7 +173,7 @@ def schedule_conv2d_spatial_pack_nchw(cfg, s, data_vec,

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414918682 ## File path: src/runtime/micro/openocd_low_level_device.cc ## @@ -210,9 +210,9 @@ class OpenOCDLowLevelDevice final : public LowLevelDevice {

[GitHub] [incubator-tvm] wpan11nv edited a comment on pull request #5382: [TE] Fix MakeLoopNest for warp memory

wpan11nv edited a comment on pull request #5382: URL: https://github.com/apache/incubator-tvm/pull/5382#issuecomment-619201349 Can we make it explicit, e.g by tagging the binding iter_var with "warp" scope. For CUDA, when the block configuration is (1, 32), threadIdx.y may be the thread

[GitHub] [incubator-tvm] hcho3 commented on issue #4953: [CI][Docker] xgboost 1.0.1 causes segfault on test_autotvm_xgboost_model.py

hcho3 commented on issue #4953: URL: https://github.com/apache/incubator-tvm/issues/4953#issuecomment-619255802 @leandron @areusch @tqchen I've put up RC1 for the upcoming XGBoost 1.1.0 release. Feel free to try it: ``` python3 -m pip install xgboost==1.1.0rc1 ``` The unit

[GitHub] [incubator-tvm] yzhliu commented on issue #3670: [RFC] AlterOpLayout Pass Refactoring

yzhliu commented on issue #3670: URL: https://github.com/apache/incubator-tvm/issues/3670#issuecomment-619263443 @mbrookhart not at all, please feel free to proceed :) This is an automated message from the Apache Git

[GitHub] [incubator-tvm] yzhliu commented on pull request #5367: Improve IntervalSet's floormod

yzhliu commented on pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#issuecomment-619238149 @icemelon9 is this what you're looking for? This is an automated message from the Apache Git Service. To

[GitHub] [incubator-tvm] mbrookhart opened a new pull request #5441: Add TopK to ONNX Frontend

mbrookhart opened a new pull request #5441: URL: https://github.com/apache/incubator-tvm/pull/5441 @masahi @jwfromm Thanks! This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414919556 ## File path: topi/python/topi/arm_cpu/cortex_m7/micro_kernel/gemm.py ## @@ -0,0 +1,221 @@ +# Licensed to the Apache Software Foundation (ASF)

[GitHub] [incubator-tvm] roastduck commented on pull request #5382: [TE] Fix MakeLoopNest for warp memory

roastduck commented on pull request #5382: URL: https://github.com/apache/incubator-tvm/pull/5382#issuecomment-619291622 Do you mean requiring the users to tag the iter_var or we do it in InferBound? If the former, maybe we can merge this PR first and then start an RFC for the API change.

[incubator-tvm] branch master updated (3cc4971 -> 52bf1b3)

This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git. from 3cc4971 [RUNTIME][OBJECT] Introduce static slots for common objects. (#5423) add 52bf1b3

[GitHub] [incubator-tvm] yongfeng-nv commented on a change in pull request #5367: Improve IntervalSet's floormod

yongfeng-nv commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414944483 ## File path: include/tvm/arith/analyzer.h ## @@ -411,8 +412,9 @@ class TVM_DLL Analyzer { * * \param var The variable. * \param

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414956089 ## File path: python/tvm/contrib/binutil.py ## @@ -220,17 +228,32 @@ def tvm_callback_relocate_binary(

[GitHub] [incubator-tvm] kevinthesun commented on a change in pull request #4312: [TOPI][Relay][OP] Dynamic NMS and strided_slice

kevinthesun commented on a change in pull request #4312: URL: https://github.com/apache/incubator-tvm/pull/4312#discussion_r414959052 ## File path: src/relay/op/tensor/transform.cc ## @@ -1891,81 +1952,163 @@ Array > StridedSliceInferCorrectLayout( }

[GitHub] [incubator-tvm] tmoreau89 opened a new pull request #5442: [Docs] VTA install doc migration from md to rst

tmoreau89 opened a new pull request #5442: URL: https://github.com/apache/incubator-tvm/pull/5442 Addresses issue #5396 This is an automated message from the Apache Git Service. To respond to the message, please log

[GitHub] [incubator-tvm] yongfeng-nv commented on a change in pull request #5367: Improve IntervalSet's floormod

yongfeng-nv commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414940958 ## File path: src/arith/int_set.cc ## @@ -311,6 +311,16 @@ inline IntervalSet Combine(Analyzer* analyzer, LOG(FATAL) << "Modular by

[GitHub] [incubator-tvm] masahi commented on pull request #5434: [PYTORCH]Rsub, Embedded, OneHot ops support

masahi commented on pull request #5434: URL: https://github.com/apache/incubator-tvm/pull/5434#issuecomment-619298153 Thanks @siju-samuel This is an automated message from the Apache Git Service. To respond to the message,

[incubator-tvm] branch master updated (52bf1b3 -> 83930a3)

This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git. from 52bf1b3 [RELAY][PYTORCH]cosh,sinh,log2,log10,log1p op support (#5395) add 83930a3 [PYTORCH]Rsub,

[GitHub] [incubator-tvm] yongfeng-nv commented on a change in pull request #5367: Improve IntervalSet's floormod

yongfeng-nv commented on a change in pull request #5367: URL: https://github.com/apache/incubator-tvm/pull/5367#discussion_r414946295 ## File path: src/te/operation/compute_op.cc ## @@ -231,20 +231,18 @@ void ComputeOpNode::PropBoundToInputs( // undefined

[incubator-tvm] branch master updated (83930a3 -> 2dbe626)

This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-tvm.git. from 83930a3 [PYTORCH]Rsub, Embedded, OneHot ops support (#5434) add 2dbe626 fix miopen pad (#5433) No new

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414955335 ## File path: python/tvm/contrib/debugger/debug_runtime.py ## @@ -181,6 +181,7 @@ def _run_debug(self): """

[GitHub] [incubator-tvm] masahi commented on pull request #5433: fix miopen padding

masahi commented on pull request #5433: URL: https://github.com/apache/incubator-tvm/pull/5433#issuecomment-619299101 Thanks @t-vi This is an automated message from the Apache Git Service. To respond to the message, please

[GitHub] [incubator-tvm] masahi commented on pull request #5431: misc fixes for ROCm

masahi commented on pull request #5431: URL: https://github.com/apache/incubator-tvm/pull/5431#issuecomment-619299391 Thanks @t-vi for fixing this. I recently did a clean install of rocm 3.3 and hit segfault from CodeGenLLVM destructor because of this. Now TVM + rocm is working again.

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414955744 ## File path: python/tvm/micro/base.py ## @@ -133,44 +152,91 @@ def __exit__(self, exc_type, exc_value, exc_traceback): self._exit()

[GitHub] [incubator-tvm] areusch commented on a change in pull request #5417: [RUNTIME][uTVM] AutoTVM + uTVM for Cortex-M7

areusch commented on a change in pull request #5417: URL: https://github.com/apache/incubator-tvm/pull/5417#discussion_r414955794 ## File path: python/tvm/micro/device/host.py ## @@ -38,59 +52,65 @@ def create_micro_lib(obj_path, src_path, lib_type, options=None):