[GitHub] [tvm] zhanghaohit commented on pull request #8274: [TIR][VM] Revert a change to lower_tvm_builtin.cc from #6126

2021-09-09 Thread GitBox
zhanghaohit commented on pull request #8274: URL: https://github.com/apache/tvm/pull/8274#issuecomment-916646147 > @zhanghaohit would you be able to create an issue that tracks the two hidden bugs that require deeper investigation? Sure. I create two issues: #8977 #8978 -- This

[GitHub] [tvm] tmoreau89 commented on issue #8838: [Tracking Issue][ONNX] Quantized operator support in ONNX importer

2021-09-09 Thread GitBox
tmoreau89 commented on issue #8838: URL: https://github.com/apache/tvm/issues/8838#issuecomment-916645695 @avquicinc excellent point and many thanks for digging into the registry. I agree that the focus here is to shortlist these ops to the set defined in the registry. To use the example

[GitHub] [tvm] zhanghaohit opened a new issue #8978: [Bug][VTA][OpenCL]

2021-09-09 Thread GitBox
zhanghaohit opened a new issue #8978: URL: https://github.com/apache/tvm/issues/8978 In [`src/tir/transforms/lower_tvm_builtin.cc`](https://github.com/apache/tvm/blob/f6a404447406f91ddf6195ac044dc1f0b0e867e7/src/tir/transforms/lower_tvm_builtin.cc#L116): ```c++ if

[GitHub] [tvm] zhanghaohit opened a new issue #8977: [Bug][VM] If not allocate on stack, VM runtime cannot work?

2021-09-09 Thread GitBox
zhanghaohit opened a new issue #8977: URL: https://github.com/apache/tvm/issues/8977 In [`src/tir/transforms/lower_tvm_builtin.cc`](https://github.com/apache/tvm/blob/f6a404447406f91ddf6195ac044dc1f0b0e867e7/src/tir/transforms/lower_tvm_builtin.cc#L116) ```c++ if

[GitHub] [tvm] chunit-quic commented on pull request #8893: [Doc] Current passes

2021-09-09 Thread GitBox
chunit-quic commented on pull request #8893: URL: https://github.com/apache/tvm/pull/8893#issuecomment-916607347 Thank you @comaniac, it is a good suggestion! Automatically generate these results are much more decent. Perhaps it will take longer time but I would like to try the way of

[GitHub] [tvm] tmoreau89 commented on pull request #8952: [TVM] Add importer for ONNX QLinearMatMul op

2021-09-09 Thread GitBox
tmoreau89 commented on pull request #8952: URL: https://github.com/apache/tvm/pull/8952#issuecomment-916586108 @cconvey may need to rebase and resolve conflicts given that test_forward has changed. @anwang2009 and @mbrookhart if you get cycles on Friday could you help review? thanks!

[GitHub] [tvm-rfcs] BBuf commented on a change in pull request #24: [RFC][Frontend] Add a OneFlow Frontend

2021-09-09 Thread GitBox
BBuf commented on a change in pull request #24: URL: https://github.com/apache/tvm-rfcs/pull/24#discussion_r705831608 ## File path: rfcs/0024-add-oneflow-frontend.md ## @@ -0,0 +1,134 @@ +- Feature Name: (`add oneflow frontend`) +- Start Date: (2021-8-20) +- RFC PR:

[GitHub] [tvm-rfcs] BBuf commented on a change in pull request #24: [RFC][Frontend] Add a OneFlow Frontend

2021-09-09 Thread GitBox
BBuf commented on a change in pull request #24: URL: https://github.com/apache/tvm-rfcs/pull/24#discussion_r705829042 ## File path: rfcs/0024-add-oneflow-frontend.md ## @@ -0,0 +1,134 @@ +- Feature Name: (`add oneflow frontend`) +- Start Date: (2021-8-20) +- RFC PR:

[GitHub] [tvm-rfcs] tqchen edited a comment on pull request #18: [RFC] Adding initial SVE implementation

2021-09-09 Thread GitBox
tqchen edited a comment on pull request #18: URL: https://github.com/apache/tvm-rfcs/pull/18#issuecomment-916553296 Thanks @MeeraN7 @giuseros, to make the discussion more concrete, right now the IR after legalization looks like ```c++ for (i: int32, 0, 17;i+=VL) {

[GitHub] [tvm-rfcs] tqchen commented on pull request #18: [RFC] Adding initial SVE implementation

2021-09-09 Thread GitBox
tqchen commented on pull request #18: URL: https://github.com/apache/tvm-rfcs/pull/18#issuecomment-916553296 Thanks @MeeraN7 @giuseros, to make the discussion more concrete, right now the IR after legalization looks like ```c++ for (i: int32, 0, 17;i+=VL) { C_2[ramp(i,

[GitHub] [tvm] junrushao1994 opened a new issue #8976: [Release] v0.8 Release Planning

2021-09-09 Thread GitBox
junrushao1994 opened a new issue #8976: URL: https://github.com/apache/tvm/issues/8976 Thanks to everyone who contributed in the past cycle! I would love to use this thread to plan for the v0.8 release. The community has always been focusing on high quality releases. One of the

[GitHub] [tvm-rfcs] tqchen commented on a change in pull request #27: Formalize TVM Documentation Organization

2021-09-09 Thread GitBox
tqchen commented on a change in pull request #27: URL: https://github.com/apache/tvm-rfcs/pull/27#discussion_r705820621 ## File path: rfcs/0027-formalize-documentation-organization.md ## @@ -0,0 +1,442 @@ +- Feature Name: `Formalize TVM Documentation Organization` +- Start

[GitHub] [tvm] csullivan commented on issue #8928: [VOTE] Adopt New Code Review Guideline

2021-09-09 Thread GitBox
csullivan commented on issue #8928: URL: https://github.com/apache/tvm/issues/8928#issuecomment-916519549 +1 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To

[GitHub] [tvm] csullivan commented on pull request #8960: [Hexagon] Add contrib tests for blocked conv2d and maxpool2d

2021-09-09 Thread GitBox
csullivan commented on pull request #8960: URL: https://github.com/apache/tvm/pull/8960#issuecomment-916516137 Thanks for the feedback @tmoreau89 @adstraw @Lunderberg, I've updated the commit. Please take a look when you have a chance. -- This is an automated message from the Apache Git

[GitHub] [tvm] csullivan commented on a change in pull request #8960: [Hexagon] Add contrib tests for blocked conv2d and maxpool2d

2021-09-09 Thread GitBox
csullivan commented on a change in pull request #8960: URL: https://github.com/apache/tvm/pull/8960#discussion_r705795133 ## File path: tests/python/contrib/test_hexagon/test_conv2d_blocked.py ## @@ -0,0 +1,474 @@ +# Licensed to the Apache Software Foundation (ASF) under one

[GitHub] [tvm] csullivan commented on a change in pull request #8960: [Hexagon] Add contrib tests for blocked conv2d and maxpool2d

2021-09-09 Thread GitBox
csullivan commented on a change in pull request #8960: URL: https://github.com/apache/tvm/pull/8960#discussion_r705793968 ## File path: tests/python/contrib/test_hexagon/infrastructure.py ## @@ -0,0 +1,57 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or

[GitHub] [tvm] csullivan commented on a change in pull request #8960: [Hexagon] Add contrib tests for blocked conv2d and maxpool2d

2021-09-09 Thread GitBox
csullivan commented on a change in pull request #8960: URL: https://github.com/apache/tvm/pull/8960#discussion_r705793778 ## File path: tests/python/contrib/test_hexagon/test_conv2d_blocked.py ## @@ -0,0 +1,474 @@ +# Licensed to the Apache Software Foundation (ASF) under one

[GitHub] [tvm-rfcs] hogepodge commented on a change in pull request #18: [RFC] Adding initial SVE implementation

2021-09-09 Thread GitBox
hogepodge commented on a change in pull request #18: URL: https://github.com/apache/tvm-rfcs/pull/18#discussion_r705787815 ## File path: rfcs/0018-initial-sve-addition.md ## @@ -0,0 +1,221 @@ +- Feature Name: Adding Initial SVE Support to TVM +- Start Date: 2021-07-30 +- RFC

[GitHub] [tvm-rfcs] tkonolige commented on a change in pull request #27: Formalize TVM Documentation Organization

2021-09-09 Thread GitBox
tkonolige commented on a change in pull request #27: URL: https://github.com/apache/tvm-rfcs/pull/27#discussion_r705737942 ## File path: rfcs/0027-formalize-documentation-organization.md ## @@ -0,0 +1,442 @@ +- Feature Name: `Formalize TVM Documentation Organization` +- Start

[GitHub] [tvm-rfcs] hogepodge commented on a change in pull request #24: [RFC][Frontend] Add a OneFlow Frontend

2021-09-09 Thread GitBox
hogepodge commented on a change in pull request #24: URL: https://github.com/apache/tvm-rfcs/pull/24#discussion_r705741352 ## File path: rfcs/0024-add-oneflow-frontend.md ## @@ -0,0 +1,134 @@ +- Feature Name: (`add oneflow frontend`) +- Start Date: (2021-8-20) +- RFC PR:

[GitHub] [tvm-rfcs] hogepodge commented on a change in pull request #25: [RFC]PyTorchTVM

2021-09-09 Thread GitBox
hogepodge commented on a change in pull request #25: URL: https://github.com/apache/tvm-rfcs/pull/25#discussion_r705738135 ## File path: rfcs/0025-add-pytorch-tvm.md ## @@ -0,0 +1,265 @@ +- Feature Name: PyTorchTVM +- Start Date: 2021-08-24 +- RFC PR:

[GitHub] [tvm-rfcs] hogepodge commented on pull request #27: Formalize TVM Documentation Organization

2021-09-09 Thread GitBox
hogepodge commented on pull request #27: URL: https://github.com/apache/tvm-rfcs/pull/27#issuecomment-916474862 Looking for some more reviews to move this forward: @tqchen @masahi @comaniac @icemelon @leandron -- This is an automated message from the Apache Git Service. To respond to

[GitHub] [tvm-rfcs] hogepodge commented on a change in pull request #27: Formalize TVM Documentation Organization

2021-09-09 Thread GitBox
hogepodge commented on a change in pull request #27: URL: https://github.com/apache/tvm-rfcs/pull/27#discussion_r705733165 ## File path: rfcs/0027-formalize-documentation-organization.md ## @@ -0,0 +1,430 @@ +- Feature Name: `Formalize TVM Documentation Organization` +- Start

[GitHub] [tvm-rfcs] hogepodge commented on a change in pull request #27: Formalize TVM Documentation Organization

2021-09-09 Thread GitBox
hogepodge commented on a change in pull request #27: URL: https://github.com/apache/tvm-rfcs/pull/27#discussion_r705732937 ## File path: rfcs/0027-formalize-documentation-organization.md ## @@ -0,0 +1,430 @@ +- Feature Name: `Formalize TVM Documentation Organization` +- Start

[GitHub] [tvm-rfcs] hogepodge commented on pull request #30: Command Line Configuration Files

2021-09-09 Thread GitBox
hogepodge commented on pull request #30: URL: https://github.com/apache/tvm-rfcs/pull/30#issuecomment-916472594 Can you address the common usage of environment variables for configuration, why that wasn't considered, and the potential for future work? -- This is an automated message

[GitHub] [tvm] comaniac commented on pull request #8893: [Doc] Current passes

2021-09-09 Thread GitBox
comaniac commented on pull request #8893: URL: https://github.com/apache/tvm/pull/8893#issuecomment-916464350 I agree with the idea that it would be great to have examples for each pass. I'm just thinking whether we could have a more systematic way to do so instead of manually adding

[GitHub] [tvm] Laurawly commented on pull request #8808: [BYOC][TensorRT] Add TensorRT own int8 calibration support to TensorRT BYOC integration

2021-09-09 Thread GitBox
Laurawly commented on pull request #8808: URL: https://github.com/apache/tvm/pull/8808#issuecomment-916425015 @FrozenGene @trevor-m Could you update your reviews? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the

[GitHub] [tvm] huajsj commented on pull request #8702: [Runtime] Pipeline Executor Initial patch.

2021-09-09 Thread GitBox
huajsj commented on pull request #8702: URL: https://github.com/apache/tvm/pull/8702#issuecomment-916417269 thanks @masahi, all review comments get addressed, please review again. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [tvm] huajsj commented on a change in pull request #8702: [Runtime] Pipeline Executor Initial patch.

2021-09-09 Thread GitBox
huajsj commented on a change in pull request #8702: URL: https://github.com/apache/tvm/pull/8702#discussion_r70542 ## File path: python/tvm/contrib/pipeline_executor.py ## @@ -0,0 +1,533 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more

[tvm] branch main updated: [microTVM] Refactor `platform` used as board name in microTVM (#8940)

2021-09-09 Thread areusch
This is an automated email from the ASF dual-hosted git repository. areusch pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new f6a4044 [microTVM] Refactor `platform` used as board

[GitHub] [tvm] areusch merged pull request #8940: [microTVM] Refactor `platform` used as board name in microTVM

2021-09-09 Thread GitBox
areusch merged pull request #8940: URL: https://github.com/apache/tvm/pull/8940 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm-rfcs] junrushao1994 commented on pull request #25: [RFC]PyTorchTVM

2021-09-09 Thread GitBox
junrushao1994 commented on pull request #25: URL: https://github.com/apache/tvm-rfcs/pull/25#issuecomment-916379086 @masahi @t-vi Please take another look :-) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL

[GitHub] [tvm] electriclilies commented on a change in pull request #8974: Remove memory planing from LowerTEPass

2021-09-09 Thread GitBox
electriclilies commented on a change in pull request #8974: URL: https://github.com/apache/tvm/pull/8974#discussion_r705609087 ## File path: src/relay/backend/aot_executor_codegen.cc ## @@ -45,6 +45,13 @@ namespace tvm { namespace relay { namespace backend { +struct

[GitHub] [tvm] gromero commented on a change in pull request #8940: [microTVM] Refactor `platform` used as board name in microTVM

2021-09-09 Thread GitBox
gromero commented on a change in pull request #8940: URL: https://github.com/apache/tvm/pull/8940#discussion_r705585495 ## File path: tests/micro/zephyr/conftest.py ## @@ -20,33 +20,44 @@ import pytest +from tvm.micro import project import tvm.contrib.utils import

[GitHub] [tvm] guberti commented on a change in pull request #8940: [microTVM] Refactor `platform` used as board name in microTVM

2021-09-09 Thread GitBox
guberti commented on a change in pull request #8940: URL: https://github.com/apache/tvm/pull/8940#discussion_r705577741 ## File path: apps/microtvm/arduino/template_project/microtvm_api_server.py ## @@ -43,6 +43,19 @@ IS_TEMPLATE = not (API_SERVER_DIR /

[GitHub] [tvm] junrushao1994 merged pull request #8943: [TensorIR][M2a] Compute-At

2021-09-09 Thread GitBox
junrushao1994 merged pull request #8943: URL: https://github.com/apache/tvm/pull/8943 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] junrushao1994 commented on a change in pull request #8943: [TensorIR][M2a] Compute-At

2021-09-09 Thread GitBox
junrushao1994 commented on a change in pull request #8943: URL: https://github.com/apache/tvm/pull/8943#discussion_r705561926 ## File path: src/tir/schedule/primitive/compute_at.cc ## @@ -0,0 +1,584 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or

[GitHub] [tvm] vvchernov commented on a change in pull request #8975: [Debug] iters_num and repeat were added to debug executor init args

2021-09-09 Thread GitBox
vvchernov commented on a change in pull request #8975: URL: https://github.com/apache/tvm/pull/8975#discussion_r705519614 ## File path: python/tvm/contrib/debugger/debug_executor.py ## @@ -68,7 +73,7 @@ def create(graph_json_str, libmod, device, dump_root=None):

[tvm] branch main updated (90676c4 -> 2232399)

2021-09-09 Thread jwfromm
This is an automated email from the ASF dual-hosted git repository. jwfromm pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from 90676c4 [TIR][VM] Revert a change to lower_tvm_builtin.cc from #6126 (#8274) add 2232399 [RPC][IOS] Add random

[GitHub] [tvm] jwfromm merged pull request #8935: [RPC][IOS] Add random to ios_rpc

2021-09-09 Thread GitBox
jwfromm merged pull request #8935: URL: https://github.com/apache/tvm/pull/8935 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] jwfromm commented on pull request #8935: [RPC][IOS] Add random to ios_rpc

2021-09-09 Thread GitBox
jwfromm commented on pull request #8935: URL: https://github.com/apache/tvm/pull/8935#issuecomment-916262842 Thanks for the fix. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific

[GitHub] [tvm] tmoreau89 commented on a change in pull request #8960: [Hexagon] Add contrib tests for blocked conv2d and maxpool2d

2021-09-09 Thread GitBox
tmoreau89 commented on a change in pull request #8960: URL: https://github.com/apache/tvm/pull/8960#discussion_r705516793 ## File path: tests/python/contrib/test_hexagon/test_conv2d_blocked.py ## @@ -0,0 +1,474 @@ +# Licensed to the Apache Software Foundation (ASF) under one

[GitHub] [tvm] tkonolige commented on a change in pull request #8975: [Debug] iters_num and repeat were added to debug executor init args

2021-09-09 Thread GitBox
tkonolige commented on a change in pull request #8975: URL: https://github.com/apache/tvm/pull/8975#discussion_r705512883 ## File path: python/tvm/contrib/debugger/debug_executor.py ## @@ -68,7 +73,7 @@ def create(graph_json_str, libmod, device, dump_root=None):

[GitHub] [tvm] jwfromm removed a comment on pull request #8636: [OpenCL] Add vectorization to cuda conv2d_nhwc schedule

2021-09-09 Thread GitBox
jwfromm removed a comment on pull request #8636: URL: https://github.com/apache/tvm/pull/8636#issuecomment-916260974 @jcf94 can you take another look at this PR and see if @echuraev changes look good? -- This is an automated message from the Apache Git Service. To respond to the

[GitHub] [tvm] jwfromm commented on pull request #8636: [OpenCL] Add vectorization to cuda conv2d_nhwc schedule

2021-09-09 Thread GitBox
jwfromm commented on pull request #8636: URL: https://github.com/apache/tvm/pull/8636#issuecomment-916260974 @jcf94 can you take another look at this PR and see if @echuraev changes look good? -- This is an automated message from the Apache Git Service. To respond to the message, please

[GitHub] [tvm] areusch commented on a change in pull request #8940: [microTVM] Refactor `platform` used as board name in microTVM

2021-09-09 Thread GitBox
areusch commented on a change in pull request #8940: URL: https://github.com/apache/tvm/pull/8940#discussion_r705502904 ## File path: tests/micro/zephyr/conftest.py ## @@ -20,33 +20,44 @@ import pytest +from tvm.micro import project import tvm.contrib.utils import

[GitHub] [tvm] areusch commented on issue #8673: [CI] skip non-docs phases when a PR only modifies `docs/`

2021-09-09 Thread GitBox
areusch commented on issue #8673: URL: https://github.com/apache/tvm/issues/8673#issuecomment-916233235 the tricky thing is that we still do want to run some CI tasks--we just want to skip most of them. for example, you could checkin broken ReStructuredText--we should still fail the CI if

[GitHub] [tvm] tmoreau89 commented on pull request #8274: [TIR][VM] Revert a change to lower_tvm_builtin.cc from #6126

2021-09-09 Thread GitBox
tmoreau89 commented on pull request #8274: URL: https://github.com/apache/tvm/pull/8274#issuecomment-916223413 @zhanghaohit would you be able to create an issue that tracks the two hidden bugs that require deeper investigation? -- This is an automated message from the Apache Git

[tvm] branch main updated (aa2b37d -> 90676c4)

2021-09-09 Thread moreau
This is an automated email from the ASF dual-hosted git repository. moreau pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from aa2b37d [microTVM] Add support for AutoTVM (#8715) add 90676c4 [TIR][VM] Revert a change to lower_tvm_builtin.cc

[GitHub] [tvm] tmoreau89 commented on pull request #8274: [TIR][VM] Revert a change to lower_tvm_builtin.cc from #6126

2021-09-09 Thread GitBox
tmoreau89 commented on pull request #8274: URL: https://github.com/apache/tvm/pull/8274#issuecomment-916222726 Thanks @mbrookhart and @zhanghaohit the fix has been merged. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and

[GitHub] [tvm] tmoreau89 merged pull request #8274: [TIR][VM] Revert a change to lower_tvm_builtin.cc from #6126

2021-09-09 Thread GitBox
tmoreau89 merged pull request #8274: URL: https://github.com/apache/tvm/pull/8274 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] vegaluisjose commented on pull request #8973: Update TVM VTA (VTA Chisel Wide memory interface)

2021-09-09 Thread GitBox
vegaluisjose commented on pull request #8973: URL: https://github.com/apache/tvm/pull/8973#issuecomment-916208424 oh yeah, that is right. @tmoreau89 is this all needed right? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [tvm] aasorokiin commented on pull request #8973: Update TVM VTA (VTA Chisel Wide memory interface)

2021-09-09 Thread GitBox
aasorokiin commented on pull request #8973: URL: https://github.com/apache/tvm/pull/8973#issuecomment-916198215 Should this PR be done to update TVM to use the latest VTA? @vegaluisjose -- This is an automated message from the Apache Git Service. To respond to the message, please log on

[GitHub] [tvm] mikepapadim commented on issue #8673: [CI] skip non-docs phases when a PR only modifies `docs/`

2021-09-09 Thread GitBox
mikepapadim commented on issue #8673: URL: https://github.com/apache/tvm/issues/8673#issuecomment-916169753 There is a straightforward way to do it. We could add the `\doc` directory, as well as specific file extensions (i.e., txt, md etc) with regex in the exclude paths of the Jenkins

[GitHub] [tvm] kparzysz-quic commented on issue #8928: [VOTE] Adopt New Code Review Guideline

2021-09-09 Thread GitBox
kparzysz-quic commented on issue #8928: URL: https://github.com/apache/tvm/issues/8928#issuecomment-916143196 +1 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To

[GitHub] [tvm] Lunderberg commented on pull request #8960: [Hexagon] Add contrib tests for blocked conv2d and maxpool2d

2021-09-09 Thread GitBox
Lunderberg commented on pull request #8960: URL: https://github.com/apache/tvm/pull/8960#issuecomment-916091088 The changes look good to me. I like the hoisting of the shape definitions up into the `conftest.py`, to avoid redefinition across multiple files. -- This is an automated

[GitHub] [tvm] vvchernov opened a new pull request #8975: WIP: [Debug] iters_num and repeat were added to debug executor

2021-09-09 Thread GitBox
vvchernov opened a new pull request #8975: URL: https://github.com/apache/tvm/pull/8975 iters_num and repeat were added to debug executor to control them on high level of API. Before it they were hard coded -- This is an automated message from the Apache Git Service. To respond to

[GitHub] [tvm] echuraev commented on pull request #8636: [OpenCL] Add vectorization to cuda conv2d_nhwc schedule

2021-09-09 Thread GitBox
echuraev commented on pull request #8636: URL: https://github.com/apache/tvm/pull/8636#issuecomment-915951069 > I think it is good to go as long as there would be no concern for regression. Does the new schedule subsume the old schedule in its search space? Yes, it does. When

[GitHub] [tvm] masahi commented on pull request #8636: [OpenCL] Add vectorization to cuda conv2d_nhwc schedule

2021-09-09 Thread GitBox
masahi commented on pull request #8636: URL: https://github.com/apache/tvm/pull/8636#issuecomment-915938295 I think it is good to go as long as there would be no concern for regression. Does the new schedule subsume the old schedule in its search space? -- This is an automated message

[tvm] branch main updated: [microTVM] Add support for AutoTVM (#8715)

2021-09-09 Thread masahi
This is an automated email from the ASF dual-hosted git repository. masahi pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new aa2b37d [microTVM] Add support for AutoTVM (#8715)

[GitHub] [tvm] masahi merged pull request #8715: [microTVM] Add support for AutoTVM

2021-09-09 Thread GitBox
masahi merged pull request #8715: URL: https://github.com/apache/tvm/pull/8715 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[tvm] branch main updated (e47fc6a -> e4478aa)

2021-09-09 Thread masahi
This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from e47fc6a [Bugfix] Fix div zero error in rewrite_simplify (#8961) add e4478aa flaky off (#8972) No new revisions

[tvm] branch main updated (b8fcad8 -> e47fc6a)

2021-09-09 Thread masahi
This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from b8fcad8 [LLVM/CG] Sort PrimFuncs when creating LLVM module (#8958) add e47fc6a [Bugfix] Fix div zero error in

[GitHub] [tvm] masahi merged pull request #8972: [Onnx] Turn off one more flaky test

2021-09-09 Thread GitBox
masahi merged pull request #8972: URL: https://github.com/apache/tvm/pull/8972 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] masahi merged pull request #8961: [Bugfix] Fix div zero error in rewrite_simplify

2021-09-09 Thread GitBox
masahi merged pull request #8961: URL: https://github.com/apache/tvm/pull/8961 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] Alex-grovety commented on issue #8717: Fusion of operations and cast in mobilenet v1 conv2d causing large feature maps

2021-09-09 Thread GitBox
Alex-grovety commented on issue #8717: URL: https://github.com/apache/tvm/issues/8717#issuecomment-915902237 output tensor is 8-bit for case when hardware has support for fast int8 arithmetic operations, if hardware hasn't support for fast int8 arithmetic operations we get: ```

[GitHub] [tvm] manupa-arm commented on issue #8717: Fusion of operations and cast in mobilenet v1 conv2d causing large feature maps

2021-09-09 Thread GitBox
manupa-arm commented on issue #8717: URL: https://github.com/apache/tvm/issues/8717#issuecomment-915884010 2.) is impressive :) -- 2.4MB to 4 bytes. Also the tensor going out is a 8-bit tensor. Any catch ? cc : @u99127 @Mousius -- This is an automated message from the

[tvm] branch main updated: [LLVM/CG] Sort PrimFuncs when creating LLVM module (#8958)

2021-09-09 Thread masahi
This is an automated email from the ASF dual-hosted git repository. masahi pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new b8fcad8 [LLVM/CG] Sort PrimFuncs when creating LLVM

[GitHub] [tvm] masahi merged pull request #8958: [LLVM/CG] Sort PrimFuncs when creating LLVM module

2021-09-09 Thread GitBox
masahi merged pull request #8958: URL: https://github.com/apache/tvm/pull/8958 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[tvm] branch main updated: change the doc to reflect previous code change (#8970)

2021-09-09 Thread masahi
This is an automated email from the ASF dual-hosted git repository. masahi pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/main by this push: new 7f986d7 change the doc to reflect previous code

[GitHub] [tvm] masahi merged pull request #8970: [AutoScheduler] Change the doc to reflect previous code change

2021-09-09 Thread GitBox
masahi merged pull request #8970: URL: https://github.com/apache/tvm/pull/8970 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] Alex-grovety commented on issue #8717: Fusion of operations and cast in mobilenet v1 conv2d causing large feature maps

2021-09-09 Thread GitBox
Alex-grovety commented on issue #8717: URL: https://github.com/apache/tvm/issues/8717#issuecomment-915870297 2.) We can use `compute_at` for DepthwiseConv2d and get: ``` PrimFunc([placeholder, placeholder, placeholder, placeholder, T_cast]) attrs={"from_legacy_te_schedule": (bool)1,

[GitHub] [tvm] NoobCheese commented on issue #8925: [Bug]Got "ValueError: not enough values to unpack (expected 2, got 1)" when BatchMatMul using two same tensors

2021-09-09 Thread GitBox
NoobCheese commented on issue #8925: URL: https://github.com/apache/tvm/issues/8925#issuecomment-915869593 > I checked the below example, seems Relay currently doesn't support this. > > ``` > import tvm > from tvm import relay > a = relay.Var('a', relay.TensorType((1,8,

[GitHub] [tvm] NoobCheese closed issue #8925: [Bug]Got "ValueError: not enough values to unpack (expected 2, got 1)" when BatchMatMul using two same tensors

2021-09-09 Thread GitBox
NoobCheese closed issue #8925: URL: https://github.com/apache/tvm/issues/8925 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail:

[GitHub] [tvm] manupa-arm edited a comment on pull request #8914: Run InferType in LowerTEPass and remove per_target_modules_ from the Interpreter

2021-09-09 Thread GitBox
manupa-arm edited a comment on pull request #8914: URL: https://github.com/apache/tvm/pull/8914#issuecomment-915862469 @Mousius PTAL when you have some time? and explicitly approve, possibly -- This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [tvm] MasterJH5574 edited a comment on pull request #8943: [TensorIR][M2a] Compute-At

2021-09-09 Thread GitBox
MasterJH5574 edited a comment on pull request #8943: URL: https://github.com/apache/tvm/pull/8943#issuecomment-915860412 There's another issue that I want @junrushao1994 to take a look at. I've informed him privately. So we may wait for his response first. -- This is an automated

[GitHub] [tvm] manupa-arm commented on pull request #8914: Run InferType in LowerTEPass and remove per_target_modules_ from the Interpreter

2021-09-09 Thread GitBox
manupa-arm commented on pull request #8914: URL: https://github.com/apache/tvm/pull/8914#issuecomment-915862469 @Mousius PTAL when you have some time? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to

[GitHub] [tvm] MasterJH5574 commented on pull request #8943: [TensorIR][M2a] Compute-At

2021-09-09 Thread GitBox
MasterJH5574 commented on pull request #8943: URL: https://github.com/apache/tvm/pull/8943#issuecomment-915860412 There's another issue that I want @junrushao1994 to take a look. I've informed him privately. So we may wait for his response first. -- This is an automated message from the

[GitHub] [tvm] mikepapadim opened a new pull request #8974: Remove memory planing from LowerTEPass

2021-09-09 Thread GitBox
mikepapadim opened a new pull request #8974: URL: https://github.com/apache/tvm/pull/8974 This PR is on top #8914 and removes the memory planing from `LowerTEPass`. @jroesch @electriclilies @mbs-octoml -- This is an automated message from the Apache Git Service. To respond to

[GitHub] [tvm-rfcs] suluner commented on pull request #25: [RFC]PyTorchTVM

2021-09-09 Thread GitBox
suluner commented on pull request #25: URL: https://github.com/apache/tvm-rfcs/pull/25#issuecomment-915849025 Can this method split the graph automaticly given a torchscript, and convert the subgraph that tvm Relay IR can support or tvm outperforming pytorch? -- This is an automated

[GitHub] [tvm] MasterJH5574 commented on a change in pull request #8943: [TensorIR][M2a] Compute-At

2021-09-09 Thread GitBox
MasterJH5574 commented on a change in pull request #8943: URL: https://github.com/apache/tvm/pull/8943#discussion_r705040467 ## File path: src/tir/schedule/concrete_schedule.cc ## @@ -439,6 +439,44 @@ BlockRV ConcreteScheduleNode::CacheWrite(const BlockRV& block_rv, int

[GitHub] [tvm] Hzfengsy commented on a change in pull request #8943: [TensorIR][M2a] Compute-At

2021-09-09 Thread GitBox
Hzfengsy commented on a change in pull request #8943: URL: https://github.com/apache/tvm/pull/8943#discussion_r705019396 ## File path: src/tir/schedule/primitive/cache_read_write.cc ## @@ -160,6 +160,21 @@ Block MakeCacheStage(const BufferRegion& cache_region, CacheStageInfo*