[tvm] branch main updated (36ea17a -> 7f237dd)

This is an automated email from the ASF dual-hosted git repository. comaniac pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from 36ea17a [Docker][Vulkan] Allow Vulkan GPU access in docker container. (#8784) add 7f237dd Extend tune_relay_x86

[GitHub] [tvm] comaniac commented on pull request #8794: Extend tune_relay_x86 tutorial to measure default and kernel level tune

comaniac commented on pull request #8794: URL: https://github.com/apache/tvm/pull/8794#issuecomment-902452000 Thanks @elvin-n -- 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] comaniac merged pull request #8794: Extend tune_relay_x86 tutorial to measure default and kernel level tune

comaniac merged pull request #8794: URL: https://github.com/apache/tvm/pull/8794 -- 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] elvin-n commented on a change in pull request #8794: Extend tune_relay_x86 tutorial to measure default and kernel level tune

elvin-n commented on a change in pull request #8794: URL: https://github.com/apache/tvm/pull/8794#discussion_r692684405 ## File path: tutorials/autotvm/tune_relay_x86.py ## @@ -256,6 +274,29 @@ def tune_and_evaluate(tuning_opt): #[Task 10/12] Current/Best:

[GitHub] [tvm] junrushao1994 commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

junrushao1994 commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692664180 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap

[GitHub] [tvm] junrushao1994 commented on pull request #8803: [TIR] Support fold constants in specialize process

junrushao1994 commented on pull request #8803: URL: https://github.com/apache/tvm/pull/8803#issuecomment-902434198 I wonder if we could just use TIR's `Simplify` pass? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use

[GitHub] [tvm] Hzfengsy commented on a change in pull request #8803: [TIR] Support fold constants in specialize process

Hzfengsy commented on a change in pull request #8803: URL: https://github.com/apache/tvm/pull/8803#discussion_r692652425 ## File path: src/tir/ir/specialize.cc ## @@ -157,12 +179,31 @@ class PrimFuncSpecializer : public StmtExprMutator { } } +

[GitHub] [tvm] wrongtest opened a new pull request #8803: [TIR] Support fold constants in specialize process

wrongtest opened a new pull request #8803: URL: https://github.com/apache/tvm/pull/8803 Often `specialize` utility rewrite vars in tir scripts into constants. If the vars are within expressions, it seems better to do some simplification. For example, the input and output buffer shape can

[GitHub] [tvm-vta] aasorokiin commented on a change in pull request #32: VTA Chisel Wide memory interface.

aasorokiin commented on a change in pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#discussion_r692642018 ## File path: hardware/chisel/src/main/resources/verilog/VTAMemDPI.v ## @@ -18,89 +18,153 @@ */ module VTAMemDPI # -( parameter LEN_BITS = 8, -

[GitHub] [tvm-vta] aasorokiin commented on a change in pull request #32: VTA Chisel Wide memory interface.

aasorokiin commented on a change in pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#discussion_r692641717 ## File path: src/dpi/module.cc ## @@ -180,36 +188,84 @@ void HostDevice::WaitPopResponse(HostResponse* r) { resp_.WaitPop(r); } -void

[GitHub] [tvm-vta] aasorokiin commented on a change in pull request #32: VTA Chisel Wide memory interface.

aasorokiin commented on a change in pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#discussion_r692641676 ## File path: src/dpi/module.cc ## @@ -118,16 +120,22 @@ class HostDevice { class MemDevice { public: - void SetRequest(uint8_t opcode, uint64_t

[GitHub] [tvm-vta] aasorokiin commented on a change in pull request #32: VTA Chisel Wide memory interface.

aasorokiin commented on a change in pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#discussion_r692641297 ## File path: hardware/chisel/src/main/scala/util/SyncQueue.scala ## @@ -0,0 +1,517 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one

[GitHub] [tvm] electriclilies commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

electriclilies commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692639189 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap

[GitHub] [tvm] electriclilies commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

electriclilies commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692615158 ## File path: include/tvm/relay/function.h ## @@ -144,6 +144,8 @@ constexpr const char* kComposite = "Composite"; constexpr const char* kInline =

[GitHub] [tvm] junrushao1994 commented on a change in pull request #8768: [TIR] Fix buffer scope in structural equal

junrushao1994 commented on a change in pull request #8768: URL: https://github.com/apache/tvm/pull/8768#discussion_r692624358 ## File path: include/tvm/ir/type.h ## @@ -164,10 +164,17 @@ class PointerTypeNode : public TypeNode { } bool SEqualReduce(const

[GitHub] [tvm] masahi closed issue #8295: [AMP] Vulkan Support for Mixed Precision Pass

masahi closed issue #8295: URL: https://github.com/apache/tvm/issues/8295 -- 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 commented on issue #8295: [AMP] Vulkan Support for Mixed Precision Pass

masahi commented on issue #8295: URL: https://github.com/apache/tvm/issues/8295#issuecomment-902388931 Vulkan support for fp16 is fully functional, thanks @Lunderberg -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use

[GitHub] [tvm-vta] vegaluisjose commented on a change in pull request #32: VTA Chisel Wide memory interface.

vegaluisjose commented on a change in pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#discussion_r692609985 ## File path: hardware/chisel/src/main/scala/util/SyncQueue.scala ## @@ -0,0 +1,517 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under

[GitHub] [tvm] vinx13 commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

vinx13 commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692605491 ## File path: python/tvm/contrib/popen_pool.py ## @@ -269,21 +285,32 @@ class PopenPoolExecutor: timeout : float Timeout value for each

[GitHub] [tvm] Hzfengsy commented on a change in pull request #8768: [TIR] Fix buffer scope in structural equal

Hzfengsy commented on a change in pull request #8768: URL: https://github.com/apache/tvm/pull/8768#discussion_r692606120 ## File path: include/tvm/ir/type.h ## @@ -164,10 +164,17 @@ class PointerTypeNode : public TypeNode { } bool SEqualReduce(const PointerTypeNode*

[GitHub] [tvm] aasorokiin edited a comment on pull request #8797: VTA cmake change to include Verilator header for building tsim library

aasorokiin edited a comment on pull request #8797: URL: https://github.com/apache/tvm/pull/8797#issuecomment-902371180 It should fix VTA PR There is a more general problem with defines. That g++ argument define -D__STDC_FORMAT_MACROS is redefined in verilator includes and causes

[GitHub] [tvm] aasorokiin edited a comment on pull request #8797: VTA cmake change to include Verilator header for building tsim library

aasorokiin edited a comment on pull request #8797: URL: https://github.com/apache/tvm/pull/8797#issuecomment-902371180 It should fix VTA PR There is a more general problem with defines. That g++ argument define -D__STDC_FORMAT_MACROS is redefined in verilator includes and cause

[GitHub] [tvm] aasorokiin commented on pull request #8797: VTA cmake change to include Verilator header for building tsim library

aasorokiin commented on pull request #8797: URL: https://github.com/apache/tvm/pull/8797#issuecomment-902371180 There is a more general problem with defines. That g++ argument define -D__STDC_FORMAT_MACROS is redefined in verilator includes and cause compilation error with -Werror --

[GitHub] [tvm] khionu commented on a change in pull request #8725: [Rust] Fix memory leak #2

khionu commented on a change in pull request #8725: URL: https://github.com/apache/tvm/pull/8725#discussion_r692601816 ## File path: rust/tvm-rt/src/function.rs ## @@ -291,12 +299,12 @@ where } pub fn register_untyped>( -f: fn(Vec>) -> Result, +f: for<'a> fn(Vec>)

[GitHub] [tvm] khionu commented on a change in pull request #8725: [Rust] Fix memory leak #2

khionu commented on a change in pull request #8725: URL: https://github.com/apache/tvm/pull/8725#discussion_r692601685 ## File path: rust/tvm-rt/src/function.rs ## @@ -113,8 +114,11 @@ impl Function { /// Calls the function that created from `Builder`. pub fn

[GitHub] [tvm] masahi commented on issue #8233: [Bug] PyTorch MaskRCNN GPU OOM error

masahi commented on issue #8233: URL: https://github.com/apache/tvm/issues/8233#issuecomment-902368556 I've just hit this problem when evaluating PT MaskRCNN on coco dataset. I want to take a look at this issue. -- This is an automated message from the Apache Git Service. To respond to

[GitHub] [tvm-vta] tmoreau89 merged pull request #33: Port to the latest stable Chisel release

tmoreau89 merged pull request #33: URL: https://github.com/apache/tvm-vta/pull/33 -- 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-vta] branch main updated (4f7d249 -> 3ec5f21)

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-vta.git. from 4f7d249 Renamed graph_runtime references to graph_executor. (#31) add 3ec5f21 Port to the latest stable Chisel

[GitHub] [tvm] vegaluisjose commented on pull request #8797: VTA cmake change to include Verilator header for building tsim library

vegaluisjose commented on pull request #8797: URL: https://github.com/apache/tvm/pull/8797#issuecomment-902364049 If I understand correctly, this PR fixes the following [VTA PR](https://github.com/apache/tvm-vta/pull/32) right? -- This is an automated message from the Apache Git

[GitHub] [tvm] junrushao1994 commented on a change in pull request #8768: [TIR] Fix buffer scope in structural equal

junrushao1994 commented on a change in pull request #8768: URL: https://github.com/apache/tvm/pull/8768#discussion_r692596108 ## File path: include/tvm/ir/type.h ## @@ -164,10 +164,17 @@ class PointerTypeNode : public TypeNode { } bool SEqualReduce(const

[GitHub] [tvm] junrushao1994 commented on pull request #8784: [Docker][Vulkan] Allow Vulkan GPU access in docker container.

junrushao1994 commented on pull request #8784: URL: https://github.com/apache/tvm/pull/8784#issuecomment-902363627 Thanks @Lunderberg! -- 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

[tvm] branch main updated: [Docker][Vulkan] Allow Vulkan GPU access in docker container. (#8784)

This is an automated email from the ASF dual-hosted git repository. junrushao 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 36ea17a [Docker][Vulkan] Allow Vulkan GPU access

[GitHub] [tvm] junrushao1994 merged pull request #8784: [Docker][Vulkan] Allow Vulkan GPU access in docker container.

junrushao1994 merged pull request #8784: URL: https://github.com/apache/tvm/pull/8784 -- 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 (9697bfd -> 0d3e233)

This is an automated email from the ASF dual-hosted git repository. junrushao pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from 9697bfd Add synr==0.3.0 dependency for Docker images and Python dependency. (#8801) add 0d3e233

[tvm] branch main updated: Add synr==0.3.0 dependency for Docker images and Python dependency. (#8801)

This is an automated email from the ASF dual-hosted git repository. junrushao 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 9697bfd Add synr==0.3.0 dependency for Docker

[GitHub] [tvm] junrushao1994 merged pull request #8798: [UnitTest][Flaky] Increased tolerance on onnx test_forward::test_aten

junrushao1994 merged pull request #8798: URL: https://github.com/apache/tvm/pull/8798 -- 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 merged pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

junrushao1994 merged pull request #8801: URL: https://github.com/apache/tvm/pull/8801 -- 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] huanmei9 commented on pull request #8770: [contrib] Fix "custom_path" already exist bug

huanmei9 commented on pull request #8770: URL: https://github.com/apache/tvm/pull/8770#issuecomment-902361519 Hi, @leandron . Sorry for the late reply. My use case is to check the build fils generated by zephyr in MicroTVM, such as .map and .hex files. In addition, regardless of

[GitHub] [tvm] comaniac commented on pull request #8775: Remove compile_engine.h for real this time.

comaniac commented on pull request #8775: URL: https://github.com/apache/tvm/pull/8775#issuecomment-902349542 Thanks for the response :) I completely understand the concern of unifying the lowering mechanism and feel this is the right track as well. Meanwhile, I have two cases:

[GitHub] [tvm] jroesch commented on a change in pull request #8725: [Rust] Fix memory leak #2

jroesch commented on a change in pull request #8725: URL: https://github.com/apache/tvm/pull/8725#discussion_r692573768 ## File path: rust/tvm-rt/src/map.rs ## @@ -121,6 +121,43 @@ where } } +// pub struct Iter<'a, K, V> { Review comment: Remove this, not

[GitHub] [tvm] jroesch commented on a change in pull request #8725: [Rust] Fix memory leak #2

jroesch commented on a change in pull request #8725: URL: https://github.com/apache/tvm/pull/8725#discussion_r692573613 ## File path: rust/tvm-rt/src/function.rs ## @@ -130,6 +134,10 @@ impl Function { ) }; +if ret_type_code ==

[GitHub] [tvm] tqchen commented on pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

tqchen commented on pull request #8789: URL: https://github.com/apache/tvm/pull/8789#issuecomment-902331067 you can dismiss the reviews as long as the comments are addressed -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub

[GitHub] [tvm] shingjan commented on pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on pull request #8789: URL: https://github.com/apache/tvm/pull/8789#issuecomment-902330521 I think Github used to let contributors revoke `request for changes`. But maybe that is not the case for public repository like TVM. -- This is an automated message from the

[GitHub] [tvm] junrushao1994 commented on pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

junrushao1994 commented on pull request #8789: URL: https://github.com/apache/tvm/pull/8789#issuecomment-902329460 Oh looks like I need @vinx13's approval to merge this PR, but he is now on vacation  -- This is an automated message from the Apache Git Service. To respond to the

[GitHub] [tvm] junrushao1994 commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

junrushao1994 commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692565459 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap

[GitHub] [tvm] shingjan commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692565036 ## File path: tutorials/autotvm/tune_conv2d_cuda.py ## @@ -204,7 +204,7 @@ def conv2d_no_batching(N, H, W, CO, CI, KH, KW, stride, padding): # see many

[GitHub] [tvm] junrushao1994 commented on pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

junrushao1994 commented on pull request #8789: URL: https://github.com/apache/tvm/pull/8789#issuecomment-902327025 Overall it looks pretty good now :-) Please fix the lint and let's get it merged ASAP -- This is an automated message from the Apache Git Service. To respond to the

[GitHub] [tvm] junrushao1994 commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

junrushao1994 commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692563697 ## File path: tutorials/autotvm/tune_conv2d_cuda.py ## @@ -204,7 +204,7 @@ def conv2d_no_batching(N, H, W, CO, CI, KH, KW, stride, padding): # see

[GitHub] [tvm] shingjan commented on pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on pull request #8789: URL: https://github.com/apache/tvm/pull/8789#issuecomment-902325825 @junrushao1994 I added three new test cases for `ffi`, including calling python function via ffi, calling cpp function via ffi and calling python function from cpp function via

[GitHub] [tvm] tqchen commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

tqchen commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692559893 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap devic

[GitHub] [tvm] tqchen commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

tqchen commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692559893 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap devic

[GitHub] [tvm] tqchen commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

tqchen commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692557835 ## File path: include/tvm/relay/function.h ## @@ -144,6 +144,8 @@ constexpr const char* kComposite = "Composite"; constexpr const char* kInline = "Inline";

[GitHub] [tvm] jroesch edited a comment on pull request #8775: Remove compile_engine.h for real this time.

jroesch edited a comment on pull request #8775: URL: https://github.com/apache/tvm/pull/8775#issuecomment-902319579 @comaniac Hey Cody, Sorry for slow response, been having a busy week. Let me clarify all my assumptions and models and see if we can clarify from there: 1. I

[GitHub] [tvm] jroesch commented on pull request #8775: Remove compile_engine.h for real this time.

jroesch commented on pull request #8775: URL: https://github.com/apache/tvm/pull/8775#issuecomment-902319579 @comaniac Hey Cody, Sorry for slow response, been having a busy week. Let me clarify all my assumptions and models and see if we can clarify from there: 1. I think

[GitHub] [tvm] electriclilies commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

electriclilies commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692553994 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap

[GitHub] [tvm] mbs-octoml commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

mbs-octoml commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692530723 ## File path: src/relay/backend/aot_executor_codegen.cc ## @@ -48,6 +48,7 @@ namespace backend { using IntegerArray = Array; using StorageMap =

[GitHub] [tvm] tkonolige commented on a change in pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

tkonolige commented on a change in pull request #8802: URL: https://github.com/apache/tvm/pull/8802#discussion_r692529259 ## File path: src/relay/backend/te_compiler.cc ## @@ -875,6 +878,121 @@ LoweredModule LowerTE(const IRModule& module, TargetMap targets, DeviceMap devic

[GitHub] [tvm] tkonolige commented on a change in pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

tkonolige commented on a change in pull request #8801: URL: https://github.com/apache/tvm/pull/8801#discussion_r692536663 ## File path: python/gen_requirements.py ## @@ -231,7 +231,7 @@ ("sphinx_autodoc_annotation", None), ("sphinx_gallery", None),

[GitHub] [tvm] jroesch commented on a change in pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

jroesch commented on a change in pull request #8801: URL: https://github.com/apache/tvm/pull/8801#discussion_r692534259 ## File path: python/gen_requirements.py ## @@ -231,7 +231,7 @@ ("sphinx_autodoc_annotation", None), ("sphinx_gallery", None),

[GitHub] [tvm] jroesch commented on a change in pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

jroesch commented on a change in pull request #8801: URL: https://github.com/apache/tvm/pull/8801#discussion_r692534259 ## File path: python/gen_requirements.py ## @@ -231,7 +231,7 @@ ("sphinx_autodoc_annotation", None), ("sphinx_gallery", None),

[GitHub] [tvm] electriclilies opened a new pull request #8802: Add LowerTEPass, and convert calls to LowerTE to application of LowerTEPass

electriclilies opened a new pull request #8802: URL: https://github.com/apache/tvm/pull/8802 In this PR, I introduce a pass-version of LowerTE called LowerTEPass. I also introduce two utility functions , IRModuleToLoweredModule and LoweredModuleToIRModule. These functions convert between

[GitHub] [tvm] csullivan edited a comment on pull request #8756: [CODEGEN][OpenCL]: fix tir.erf codegen to opencl directly

csullivan edited a comment on pull request #8756: URL: https://github.com/apache/tvm/pull/8756#issuecomment-902277519 Also please lint your commits to push this commit past the CI sanity check. For applying black you can try `./tests/lint/git-black.sh -i HEAD~2`. You can check the CI

[GitHub] [tvm] csullivan commented on pull request #8756: [CODEGEN][OpenCL]: fix tir.erf codegen to opencl directly

csullivan commented on pull request #8756: URL: https://github.com/apache/tvm/pull/8756#issuecomment-902277519 Also please lint your commits to push this commit past the CI sanity check. For applying black you can try `./tests/lint/git-black.sh -i HEAD~2`. You can check the CI status by

[GitHub] [tvm] csullivan commented on a change in pull request #8756: [CODEGEN][OpenCL]: fix tir.erf codegen to opencl directly

csullivan commented on a change in pull request #8756: URL: https://github.com/apache/tvm/pull/8756#discussion_r692507919 ## File path: tests/python/unittest/test_target_codegen_opencl.py ## @@ -119,7 +119,27 @@ def check_max(dev, n, dtype): check_max(dev, 1, "float32")

[GitHub] [tvm] tkonolige commented on a change in pull request #8787: Fix typos

tkonolige commented on a change in pull request #8787: URL: https://github.com/apache/tvm/pull/8787#discussion_r692500108 ## File path: include/tvm/node/reflection.h ## @@ -43,7 +43,7 @@ using runtime::ObjectPtr; using runtime::ObjectRef; /*! - * \brief Visitor class for

[GitHub] [tvm-vta] tmoreau89 commented on pull request #33: Port to the latest stable Chisel release

tmoreau89 commented on pull request #33: URL: https://github.com/apache/tvm-vta/pull/33#issuecomment-902230052 @vegaluisjose @liangfu can you take a look? thanks! -- 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-vta] tmoreau89 commented on pull request #33: Port to the latest stable Chisel release

tmoreau89 commented on pull request #33: URL: https://github.com/apache/tvm-vta/pull/33#issuecomment-902229418 Thank you @ekiwi - glad to see your changes use the latest stable release. -- This is an automated message from the Apache Git Service. To respond to the message, please log on

[GitHub] [tvm-vta] tmoreau89 merged pull request #31: Renamed graph_runtime references to graph_executor.

tmoreau89 merged pull request #31: URL: https://github.com/apache/tvm-vta/pull/31 -- 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-vta] branch main updated: Renamed graph_runtime references to graph_executor. (#31)

This is an automated email from the ASF dual-hosted git repository. moreau pushed a commit to branch main in repository https://gitbox.apache.org/repos/asf/tvm-vta.git The following commit(s) were added to refs/heads/main by this push: new 4f7d249 Renamed graph_runtime references to

[GitHub] [tvm-vta] tmoreau89 commented on pull request #31: Renamed graph_runtime references to graph_executor.

tmoreau89 commented on pull request #31: URL: https://github.com/apache/tvm-vta/pull/31#issuecomment-902228785 Thanks Eric, LGTM -- 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] grant-arm commented on a change in pull request #8720: Sanitize names of input tensors in interface header

grant-arm commented on a change in pull request #8720: URL: https://github.com/apache/tvm/pull/8720#discussion_r692470290 ## File path: src/target/source/source_module.cc ## @@ -252,7 +252,9 @@ class CSourceCrtMetadataModuleNode : public runtime::ModuleNode { << ")

[GitHub] [tvm] areusch commented on a change in pull request #8468: [TIR][USMP] Added buffer info extraction pass

areusch commented on a change in pull request #8468: URL: https://github.com/apache/tvm/pull/8468#discussion_r692438124 ## File path: include/tvm/tir/stmt.h ## @@ -521,6 +521,11 @@ class AllocateNode : public StmtNode { PrimExpr condition; /*! \brief The body to be

[GitHub] [tvm] areusch commented on a change in pull request #8423: Implementation of relay_to_tir target hook

areusch commented on a change in pull request #8423: URL: https://github.com/apache/tvm/pull/8423#discussion_r692429689 ## File path: src/target/target_kind.cc ## @@ -70,6 +70,15 @@ Optional TargetKind::Get(const String& target_kind_name) { return reg->kind_; } +const

[GitHub] [tvm-rfcs] AndrewZhaoLuo commented on pull request #6: [RFC] [Relay] Automatic Mixed Precision Pass

AndrewZhaoLuo commented on pull request #6: URL: https://github.com/apache/tvm-rfcs/pull/6#issuecomment-902192178 If there is not other objections, this will be merged on monday. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [tvm] gromero commented on a change in pull request #8787: Fix typos

gromero commented on a change in pull request #8787: URL: https://github.com/apache/tvm/pull/8787#discussion_r692426278 ## File path: include/tvm/node/reflection.h ## @@ -43,7 +43,7 @@ using runtime::ObjectPtr; using runtime::ObjectRef; /*! - * \brief Visitor class for to

[GitHub] [tvm-rfcs] areusch commented on a change in pull request #10: Additional Target Hooks RFC

areusch commented on a change in pull request #10: URL: https://github.com/apache/tvm-rfcs/pull/10#discussion_r692403824 ## File path: rfcs/0010-additional-target-hooks.md ## @@ -0,0 +1,155 @@ +Feature Name: additional-target-hooks +Start Date: 2021-07-14 +RFC PR:

[GitHub] [tvm] leandron commented on a change in pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

leandron commented on a change in pull request #8801: URL: https://github.com/apache/tvm/pull/8801#discussion_r692416582 ## File path: python/gen_requirements.py ## @@ -231,7 +231,7 @@ ("sphinx_autodoc_annotation", None), ("sphinx_gallery", None),

[GitHub] [tvm] areusch commented on a change in pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

areusch commented on a change in pull request #8801: URL: https://github.com/apache/tvm/pull/8801#discussion_r692415327 ## File path: python/gen_requirements.py ## @@ -231,7 +231,7 @@ ("sphinx_autodoc_annotation", None), ("sphinx_gallery", None),

[GitHub] [tvm] leandron opened a new pull request #8801: Add synr==0.3.0 dependency for Docker images and Python dependency.

leandron opened a new pull request #8801: URL: https://github.com/apache/tvm/pull/8801 Add synr==0.3.0 dependency for Docker images and Python dependency: - PR #8776 removed `synr` as a dependency to be installed in the Docker images, making the images to need manual

[GitHub] [tvm] areusch commented on a change in pull request #8720: Sanitize names of input tensors in interface header

areusch commented on a change in pull request #8720: URL: https://github.com/apache/tvm/pull/8720#discussion_r692401117 ## File path: src/target/source/source_module.cc ## @@ -252,7 +252,9 @@ class CSourceCrtMetadataModuleNode : public runtime::ModuleNode { << ")

[GitHub] [tvm-vta] tmoreau89 commented on pull request #32: VTA Chisel Wide memory interface.

tmoreau89 commented on pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#issuecomment-902160451 Thank you @aasorokiin - CC-ing @vegaluisjose -- 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] areusch commented on a change in pull request #8720: Sanitize names of input tensors in interface header

areusch commented on a change in pull request #8720: URL: https://github.com/apache/tvm/pull/8720#discussion_r692400277 ## File path: python/tvm/micro/interface_api.py ## @@ -59,7 +60,8 @@ def generate_c_interface_header(module_name, inputs, outputs, output_path):

[GitHub] [tvm-rfcs] areusch commented on a change in pull request #22: [RFC][TIR] TIR Non-scalar Constants

areusch commented on a change in pull request #22: URL: https://github.com/apache/tvm-rfcs/pull/22#discussion_r692378556 ## File path: rfcs/0022-tir-non-scalar-constants.md ## @@ -0,0 +1,107 @@ + +- Feature Name: tir_non_scalar_constants +- Start Date: 2021-06-01 +- RFC PR:

[GitHub] [tvm] aasorokiin commented on pull request #8797: VTA cmake change to include Verilator header for building tsim library

aasorokiin commented on pull request #8797: URL: https://github.com/apache/tvm/pull/8797#issuecomment-902151476 VTA-hardware build flow already requires VERILATOR_INC_DIR or it defines Verilator includes from those paths. Can it be accepted to have the same requirements for TVM tsim lib

[GitHub] [tvm-vta] aasorokiin commented on pull request #32: VTA Chisel Wide memory interface.

aasorokiin commented on pull request #32: URL: https://github.com/apache/tvm-vta/pull/32#issuecomment-902150552 Adding @tmoreau89 . I need help on deciding how to proceed further. Checks flow fails because VTA DPI modification require changes to TVM. A Verilator include directory is

[GitHub] [tvm] shingjan commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692379797 ## File path: python/tvm/contrib/popen_pool.py ## @@ -86,8 +86,12 @@ class PopenWorker: API to interact with a separate process via Popen. """ -

[GitHub] [tvm] shingjan commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692370647 ## File path: python/tvm/contrib/popen_pool.py ## @@ -128,6 +132,9 @@ def _start(self): if self._proc is not None: return +

[GitHub] [tvm] shingjan commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692370486 ## File path: tests/python/contrib/test_popen_pool.py ## @@ -66,6 +66,19 @@ def test_popen_pool_executor(): assert val.value == idx +def

[GitHub] [tvm] shingjan commented on a change in pull request #8789: [CONTRIB] Allow customized initializer in PopenPool

shingjan commented on a change in pull request #8789: URL: https://github.com/apache/tvm/pull/8789#discussion_r692369734 ## File path: python/tvm/contrib/popen_pool.py ## @@ -153,6 +160,11 @@ def _start(self): self._reader = os.fdopen(main_read, "rb")

[GitHub] [tvm-vta] ekiwi opened a new pull request #33: Port to the latest stable Chisel release

ekiwi opened a new pull request #33: URL: https://github.com/apache/tvm-vta/pull/33 This PR ports the chisel hardware to use the latest stable release. It also cleans up any deprecation warnings. -- This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [tvm] junrushao1994 commented on pull request #8775: Remove compile_engine.h for real this time.

junrushao1994 commented on pull request #8775: URL: https://github.com/apache/tvm/pull/8775#issuecomment-902117888 @comaniac This proposal sounds good to me, in terms of minimal API change and being consistent with the TE compiler’s design philosophy @jroesch Would you like to take

[GitHub] [tvm] jroesch opened a new issue #8800: [Rust] Fix ByteArray conversion to TVMRetValue from Rust.

jroesch opened a new issue #8800: URL: https://github.com/apache/tvm/issues/8800 Currently we have an ownership issue when returning `ByteArray` from Rust to C++. We do not know where the `*TVMByteArray` pointer has been allocated from (i.e Rust's global allocator, or malloc) leading to

[GitHub] [tvm] Lyken17 opened a new pull request #8799: Add group_conv2d_transpose_nchw to support `groups` argument

Lyken17 opened a new pull request #8799: URL: https://github.com/apache/tvm/pull/8799 As mentioned in https://github.com/apache/tvm/issues/8182, `TranposedConv2d` is an important operator in GAN related applications and `groups` should also be supported for this operator. This PR was set

[GitHub] [tvm] comaniac commented on a change in pull request #8794: Extend tune_relay_x86 tutorial to measure default and kernel level tune

comaniac commented on a change in pull request #8794: URL: https://github.com/apache/tvm/pull/8794#discussion_r692337709 ## File path: tutorials/autotvm/tune_relay_x86.py ## @@ -256,6 +274,29 @@ def tune_and_evaluate(tuning_opt): #[Task 10/12] Current/Best:

[GitHub] [tvm] comaniac commented on pull request #8775: Remove compile_engine.h for real this time.

comaniac commented on pull request #8775: URL: https://github.com/apache/tvm/pull/8775#issuecomment-902096517 In TE compiler, the input tensors are internally created by `ScheduleBuilder`. Thus, it could be a common case for developers who want to lower a Relay function to TE but don't

[GitHub] [tvm] Lunderberg commented on pull request #8784: [Docker][Vulkan] Allow Vulkan GPU access in docker container.

Lunderberg commented on pull request #8784: URL: https://github.com/apache/tvm/pull/8784#issuecomment-902077945 Bump to restart CI. Previous failure was on flaky test, PR to resolve in #8798. -- This is an automated message from the Apache Git Service. To respond to the message, please

[GitHub] [tvm] Lunderberg opened a new pull request #8798: [UnitTest][Flaky] Increased tolerance on onnx test_forward::test_aten

Lunderberg opened a new pull request #8798: URL: https://github.com/apache/tvm/pull/8798 Default tolerance had about 2-3% failure rate (8/300 iterations), and caused failures on unrelated PRs (e.g.

[GitHub] [tvm] aasorokiin opened a new pull request #8797: VTA cmake change to include Verilator header for building tsim library

aasorokiin opened a new pull request #8797: URL: https://github.com/apache/tvm/pull/8797 Proposed changes to VTA require verilator include directory to build module.cc. Background: VTA modification is proposed to support 64/128/256/512 bit AXI data bus. (vta pull request VTA Chisel

[GitHub] [tvm] Lunderberg commented on pull request #8784: [Docker][Vulkan] Allow Vulkan GPU access in docker container.

Lunderberg commented on pull request #8784: URL: https://github.com/apache/tvm/pull/8784#issuecomment-902073010 Thank you, and definitely agree that we don't want to slow down the CI significantly as part of increasing coverage. I intend to run some local runs on the full test suite, as

[tvm] branch main updated (e1a0ea1 -> 1936609)

This is an automated email from the ASF dual-hosted git repository. jroesch pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from e1a0ea1 [microTVM][RVM] Fix base-box-tool command in README.md (#8613) add 1936609 [Android][RPC] Fix Vulkan

  1   2   >