[tvm] branch nightly updated (5cca18bb07 -> 7831a79f7f)

This is an automated email from the ASF dual-hosted git repository. github-bot pushed a change to branch nightly in repository https://gitbox.apache.org/repos/asf/tvm.git from 5cca18bb07 [Frontend] Add ONNX importer for QLinearSoftmax (#14425) add 4011280b16 [OpenCL][Textures] Always

[GitHub] [tvm] tvm-bot commented on pull request #14441: [Unity][Op][Docs] Update comment for `call_tir_dyn`

tvm-bot commented on PR #14441: URL: https://github.com/apache/tvm/pull/14441#issuecomment-1491233010 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] slyubomirsky opened a new pull request, #14441: [Unity][Op][Docs] Update comment for `call_tir_dyn`

slyubomirsky opened a new pull request, #14441: URL: https://github.com/apache/tvm/pull/14441 The comment for the `vm.call_tir_dyn` operator was out of date. This PR corrects it with a more applicable distinction. Note: I was still unable to figure out what it actually does. Only a

[GitHub] [tvm] masahi commented on pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

masahi commented on PR #14440: URL: https://github.com/apache/tvm/pull/14440#issuecomment-1491209909 Hey I'm happy to report that I have pattern-based rewriter for dataflow block working. Compared to the existing call node based matching & rewriting that requires a common post-dominator in

[GitHub] [tvm] zhaoyang-star commented on issue #14437: [Bug] TVMError: CodeGenVM cannot handle this intrinsic now: Op(relax.nn.conv2d)

zhaoyang-star commented on issue #14437: URL: https://github.com/apache/tvm/issues/14437#issuecomment-1491186311 > You need to run `Legalize()` pass before build. Thanks for your quick reply @masahi After add `relax.transform.LegalizeOps()(mod)` before `relax.build`. Another

[GitHub] [tvm] masahi commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

masahi commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153887847 ## src/relax/ir/dataflow_matcher.cc: ## @@ -540,33 +541,48 @@ struct RNode { /** * \brief This method try to match a real node and a pattern node along with its

[GitHub] [tvm] masahi commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

masahi commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153886330 ## src/relax/ir/dataflow_matcher.cc: ## @@ -540,33 +541,48 @@ struct RNode { /** * \brief This method try to match a real node and a pattern node along with its

[GitHub] [tvm] ganler commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

ganler commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153886037 ## src/relax/ir/dataflow_matcher.cc: ## @@ -540,33 +541,48 @@ struct RNode { /** * \brief This method try to match a real node and a pattern node along with its

[GitHub] [tvm] ganler commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

ganler commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153884888 ## src/relax/analysis/udchain.cc: ## @@ -65,36 +65,36 @@ class UDChain : public relax::ExprVisitor { std::pair>, runtime::Array> FunctionUseDef( const Function&

[GitHub] [tvm] masahi commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

masahi commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153884915 ## src/relax/ir/dataflow_matcher.cc: ## @@ -540,33 +541,48 @@ struct RNode { /** * \brief This method try to match a real node and a pattern node along with its

[GitHub] [tvm] ganler commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

ganler commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153883716 ## src/relax/ir/dataflow_matcher.cc: ## @@ -540,33 +541,48 @@ struct RNode { /** * \brief This method try to match a real node and a pattern node along with its

[GitHub] [tvm] masahi commented on a diff in pull request #14291: [Unity][BYOC] Add cuBLAS backend

masahi commented on code in PR #14291: URL: https://github.com/apache/tvm/pull/14291#discussion_r1153882120 ## src/runtime/contrib/cublas/cublas.cc: ## @@ -133,6 +134,120 @@ bool CheckMixPrecisionType(DLDataType in_dtype, DLDataType out_dtype, bool int_s int roundoff(int v,

[GitHub] [tvm] masahi commented on a diff in pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

masahi commented on code in PR #14440: URL: https://github.com/apache/tvm/pull/14440#discussion_r1153876511 ## src/relax/analysis/udchain.cc: ## @@ -65,36 +65,36 @@ class UDChain : public relax::ExprVisitor { std::pair>, runtime::Array> FunctionUseDef( const Function&

[GitHub] [tvm] tvm-bot commented on pull request #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

tvm-bot commented on PR #14440: URL: https://github.com/apache/tvm/pull/14440#issuecomment-1491078284 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] masahi opened a new pull request, #14440: [Unity][Graph matching] Clean up undo stack for parent and child nodes properly

masahi opened a new pull request, #14440: URL: https://github.com/apache/tvm/pull/14440 This is a fix for the issue discribed in https://github.com/apache/tvm/pull/14417#issuecomment-1490049917 -- This is an automated message from the Apache Git Service. To respond to the message, please

[tvm] branch unity updated: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern (#14439)

This is an automated email from the ASF dual-hosted git repository. yaxingcai pushed a commit to branch unity in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/unity by this push: new adf256df79 [Unity][Graph matching] Automatically

[GitHub] [tvm] cyx-6 merged pull request #14439: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern

cyx-6 merged PR #14439: URL: https://github.com/apache/tvm/pull/14439 -- 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] vinx13 commented on a diff in pull request #14291: [Unity][BYOC] Add cuBLAS backend

vinx13 commented on code in PR #14291: URL: https://github.com/apache/tvm/pull/14291#discussion_r1153789203 ## src/runtime/contrib/cublas/cublas.cc: ## @@ -133,6 +134,120 @@ bool CheckMixPrecisionType(DLDataType in_dtype, DLDataType out_dtype, bool int_s int roundoff(int v,

[GitHub] [tvm] masahi commented on pull request #14291: [Unity][BYOC] Add cuBLAS backend

masahi commented on PR #14291: URL: https://github.com/apache/tvm/pull/14291#issuecomment-1490949326 waiting for the next rebase -- 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] ganler commented on a diff in pull request #14439: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern

ganler commented on code in PR #14439: URL: https://github.com/apache/tvm/pull/14439#discussion_r1153779139 ## tests/python/relax/test_dataflow_pattern.py: ## @@ -1034,15 +1034,6 @@ def main( matmul2 = is_op("relax.matmul")(inp_pat, K_weight_pat) matmul3 =

[GitHub] [tvm] ganler commented on a diff in pull request #14439: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern

ganler commented on code in PR #14439: URL: https://github.com/apache/tvm/pull/14439#discussion_r1153776712 ## tests/python/relax/test_dataflow_pattern.py: ## @@ -1034,15 +1034,6 @@ def main( matmul2 = is_op("relax.matmul")(inp_pat, K_weight_pat) matmul3 =

[GitHub] [tvm] Hzfengsy commented on pull request #14291: [Unity][BYOC] Add cuBLAS backend

Hzfengsy commented on PR #14291: URL: https://github.com/apache/tvm/pull/14291#issuecomment-1490937198 any updates? -- 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

[tvm] branch main updated: [Hexagon] Fix deprecated call for data layout size in bits (#14438)

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 7831a79f7f [Hexagon] Fix deprecated call for data

[GitHub] [tvm] masahi merged pull request #14438: [Hexagon] Fix deprecated call for data layout size in bits

masahi merged PR #14438: URL: https://github.com/apache/tvm/pull/14438 -- 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] Hzfengsy commented on a diff in pull request #14398: [TIR] Add merge primitive for TIR schedule

Hzfengsy commented on code in PR #14398: URL: https://github.com/apache/tvm/pull/14398#discussion_r1153731668 ## include/tvm/tir/schedule/schedule.h: ## @@ -292,6 +292,16 @@ class ScheduleNode : public runtime::Object { */ virtual Array GetConsumers(const BlockRV&

[tvm] branch main updated: [TFLite] Support for BATCH_MATMUL tflite operator (#14423)

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 70399da0a2 [TFLite] Support for BATCH_MATMUL tflite

[GitHub] [tvm] masahi merged pull request #14423: [TFLite] Support for BATCH_MATMUL tflite operator

masahi merged PR #14423: URL: https://github.com/apache/tvm/pull/14423 -- 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 a diff in pull request #14439: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern

masahi commented on code in PR #14439: URL: https://github.com/apache/tvm/pull/14439#discussion_r1153724926 ## include/tvm/relax/dataflow_matcher.h: ## @@ -65,15 +65,6 @@ TVM_DLL tvm::runtime::Map MatchGraph(const PatternContext& ctx,

[GitHub] [tvm] tvm-bot commented on pull request #14439: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern

tvm-bot commented on PR #14439: URL: https://github.com/apache/tvm/pull/14439#issuecomment-1490857511 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] masahi opened a new pull request, #14439: [Unity][Graph matching] Automatically add `used-by` constraints for `is_op` pattern

masahi opened a new pull request, #14439: URL: https://github.com/apache/tvm/pull/14439 Compared to the typical, structural matching (e.g. matching `is_op` based pattern against `CallNode`), graph pattern matching requires additionally specifying `used_by` constraints. This leads to an

[GitHub] [tvm] ganler commented on a diff in pull request #14417: [Unity] Remove non-deterministic behavior from graph pattern matching

ganler commented on code in PR #14417: URL: https://github.com/apache/tvm/pull/14417#discussion_r1153615575 ## src/relax/ir/dataflow_matcher.cc: ## @@ -541,15 +541,21 @@ struct RNode { * \brief This method try to match a real node and a pattern node along with its neighbors.

[GitHub] [tvm] ganler commented on a diff in pull request #14417: [Unity] Remove non-deterministic behavior from graph pattern matching

ganler commented on code in PR #14417: URL: https://github.com/apache/tvm/pull/14417#discussion_r1153613456 ## src/relax/ir/dataflow_matcher.cc: ## @@ -541,15 +541,21 @@ struct RNode { * \brief This method try to match a real node and a pattern node along with its neighbors.

[tvm] branch unity updated: [Unity] Remove non-deterministic behavior from graph pattern matching (#14417)

This is an automated email from the ASF dual-hosted git repository. wuwei pushed a commit to branch unity in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/unity by this push: new 0695cfe4e2 [Unity] Remove non-deterministic behavior

[GitHub] [tvm] vinx13 merged pull request #14417: [Unity] Remove non-deterministic behavior from graph pattern matching

vinx13 merged PR #14417: URL: https://github.com/apache/tvm/pull/14417 -- 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] lhutton1 commented on pull request #14426: [CL] Update Compute Library from v22.11 to v23.02.1

lhutton1 commented on PR #14426: URL: https://github.com/apache/tvm/pull/14426#issuecomment-1490577715 Thanks @leandron! -- 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

[tvm] branch main updated: [CL] Update Compute Library from v22.11 to v23.02.1 (#14426)

This is an automated email from the ASF dual-hosted git repository. lukhut 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 41fb9f41d4 [CL] Update Compute Library from v22.11 to

[GitHub] [tvm] lhutton1 merged pull request #14426: [CL] Update Compute Library from v22.11 to v23.02.1

lhutton1 merged PR #14426: URL: https://github.com/apache/tvm/pull/14426 -- 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 (683e7a4555 -> 221215bf60)

This is an automated email from the ASF dual-hosted git repository. leandron pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git from 683e7a4555 [TOPI] Add instance_norm operator (#14410) add 221215bf60 [ETHOSN] Remove requantize dependency on

[GitHub] [tvm] leandron merged pull request #14422: [ETHOSN] Remove requantize dependency on resize

leandron merged PR #14422: URL: https://github.com/apache/tvm/pull/14422 -- 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] mehrdadh commented on pull request #14414: [skip ci] Replace magic_wand model with micro_speech

mehrdadh commented on PR #14414: URL: https://github.com/apache/tvm/pull/14414#issuecomment-1490554259 @Aleksei-grovety yeah I added it here: https://github.com/apache/tvm/actions/runs/4555724050 -- This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [tvm] mehrdadh commented on pull request #14434: [microTVM]Fix more security issues with pyproject

mehrdadh commented on PR #14434: URL: https://github.com/apache/tvm/pull/14434#issuecomment-1490551843 https://github.com/apache/tvm/security/dependabot/245 is still open. I wonder why since I followed what it was recommended in its fix here: https://github.com/apache/tvm/pull/14432/files

[GitHub] [tvm] janetsc closed pull request #14378: [Unity] [Hexagon] Fix deprecated call for data layout size in bits

janetsc closed pull request #14378: [Unity] [Hexagon] Fix deprecated call for data layout size in bits URL: https://github.com/apache/tvm/pull/14378 -- 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

[GitHub] [tvm] janetsc commented on pull request #14378: [Unity] [Hexagon] Fix deprecated call for data layout size in bits

janetsc commented on PR #14378: URL: https://github.com/apache/tvm/pull/14378#issuecomment-1490546991 Done, created a PR for main, closing this one. -- 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

[GitHub] [tvm] tvm-bot commented on pull request #14438: [Hexagon] Fix deprecated call for data layout size in bits

tvm-bot commented on PR #14438: URL: https://github.com/apache/tvm/pull/14438#issuecomment-1490545596 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] janetsc opened a new pull request, #14438: [Hexagon] Fix deprecated call for data layout size in bits

janetsc opened a new pull request, #14438: URL: https://github.com/apache/tvm/pull/14438 Ported from Unity PR - https://github.com/apache/tvm/pull/14378/ Updating the codegen for Hexagon call from llvm::DataLayout::getTypeSizeInBits.getFixedSize() to

[GitHub] [tvm] tqchen commented on pull request #14303: [CI] Add llvm-15 and mlir-15 to Docker setup

tqchen commented on PR #14303: URL: https://github.com/apache/tvm/pull/14303#issuecomment-1490427530 happy to go with what @multiverstack-intellif you think is convenient -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and

[tvm] branch main updated: [TOPI] Add instance_norm operator (#14410)

This is an automated email from the ASF dual-hosted git repository. tqchen 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 683e7a4555 [TOPI] Add instance_norm operator (#14410)

[GitHub] [tvm] tqchen merged pull request #14410: [TOPI] Add instance_norm operator

tqchen merged PR #14410: URL: https://github.com/apache/tvm/pull/14410 -- 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] tqchen commented on pull request #14427: [Unity][Hexagon] Allow scalar tensors to have null shape during allocation

tqchen commented on PR #14427: URL: https://github.com/apache/tvm/pull/14427#issuecomment-1490424009 yes, we plan to do it sometime this weekend -- 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

[tvm] branch main updated: [Typo] Fix name of iter var type 4 (#14436)

This is an automated email from the ASF dual-hosted git repository. tqchen 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 776cf5b3b1 [Typo] Fix name of iter var type 4 (#14436)

[GitHub] [tvm] tqchen merged pull request #14436: [Typo] Fix name of iter var type 4

tqchen merged PR #14436: URL: https://github.com/apache/tvm/pull/14436 -- 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 unity updated: [Unity] Minor updates to DataFlowBlockRewrite (#14431)

This is an automated email from the ASF dual-hosted git repository. tqchen pushed a commit to branch unity in repository https://gitbox.apache.org/repos/asf/tvm.git The following commit(s) were added to refs/heads/unity by this push: new 5d145fdbe2 [Unity] Minor updates to

[GitHub] [tvm] tqchen merged pull request #14431: [Unity] Minor updates to DataFlowBlockRewrite

tqchen merged PR #14431: URL: https://github.com/apache/tvm/pull/14431 -- 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: [Runtime] Introduce runtime module property (#14406)

This is an automated email from the ASF dual-hosted git repository. tqchen 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 7b34a6e0c6 [Runtime] Introduce runtime module property

[GitHub] [tvm] tqchen merged pull request #14406: [Runtime] Introduce runtime module property

tqchen merged PR #14406: URL: https://github.com/apache/tvm/pull/14406 -- 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] lhutton1 commented on pull request #12525: [TVMC] Allow selecting a subset of tasks to be used in `tvmc tune`

lhutton1 commented on PR #12525: URL: https://github.com/apache/tvm/pull/12525#issuecomment-1490283613 Thanks @PhilippvK @areusch @leandron! -- 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 (8e2382eea5 -> ffc1fc0116)

This is an automated email from the ASF dual-hosted git repository. lukhut pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git from 8e2382eea5 [Bugfix] Conv3Dtranspose default kernel layout should be IODHW (#14340) add ffc1fc0116 [TVMC] Allow

[GitHub] [tvm] lhutton1 merged pull request #12525: [TVMC] Allow selecting a subset of tasks to be used in `tvmc tune`

lhutton1 merged PR #12525: URL: https://github.com/apache/tvm/pull/12525 -- 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 closed issue #14437: [Bug] TVMError: CodeGenVM cannot handle this intrinsic now: Op(relax.nn.conv2d)

masahi closed issue #14437: [Bug] TVMError: CodeGenVM cannot handle this intrinsic now: Op(relax.nn.conv2d) URL: https://github.com/apache/tvm/issues/14437 -- 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] masahi commented on issue #14437: [Bug] TVMError: CodeGenVM cannot handle this intrinsic now: Op(relax.nn.conv2d)

masahi commented on issue #14437: URL: https://github.com/apache/tvm/issues/14437#issuecomment-1490180368 You need to run `Legalize()` pass before build. -- 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

[GitHub] [tvm] masahi closed issue #14326: [Bug] Conv3Dtranspose default kernel layout should be IODHW

masahi closed issue #14326: [Bug] Conv3Dtranspose default kernel layout should be IODHW URL: https://github.com/apache/tvm/issues/14326 -- 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] masahi commented on pull request #14340: [Bugfix] Conv3Dtranspose default kernel layout should be IODHW

masahi commented on PR #14340: URL: https://github.com/apache/tvm/pull/14340#issuecomment-1490178917 @rebel-jangys Thanks for your hard work, it's merged finally! -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the

[tvm] branch main updated: [Bugfix] Conv3Dtranspose default kernel layout should be IODHW (#14340)

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 8e2382eea5 [Bugfix] Conv3Dtranspose default kernel

[GitHub] [tvm] rebel-jangys commented on pull request #14340: [Bugfix] Conv3Dtranspose default kernel layout should be IODHW

rebel-jangys commented on PR #14340: URL: https://github.com/apache/tvm/pull/14340#issuecomment-1490169169 cc @apeskov , this PR is the one you commented as TODO https://github.com/apache/tvm/blob/cbe068cfacbd82d75bfc4914455bbd27652608ac/src/runtime/contrib/dnnl/dnnl_json_runtime.cc#L427

[GitHub] [tvm] zhaoyang-star opened a new issue, #14437: [Bug] TVMError: CodeGenVM cannot handle this intrinsic now: Op(relax.nn.conv2d)

zhaoyang-star opened a new issue, #14437: URL: https://github.com/apache/tvm/issues/14437 I try to import ResNet18 PyTorch model to Relax. The error occurs when executing `relax.build` and the error is: ``` File "/WORK/Dev/tvm/src/relax/backend/vm/codegen_vm.cc", line 169

[GitHub] [tvm] nwy2010 commented on issue #13387: [Bug] some bugs in tvm.

nwy2010 commented on issue #13387: URL: https://github.com/apache/tvm/issues/13387#issuecomment-1490103561 (2)cudnn.py in /python/tvm/contribute: idx = -1 if algo_type == "fwd": idx = _FWD_ALGOS.index(algo_name) elif algo_type == "bwd_filter": idx =

[GitHub] [tvm] nwy2010 commented on issue #13387: [Bug] some bugs in tvm.

nwy2010 commented on issue #13387: URL: https://github.com/apache/tvm/issues/13387#issuecomment-1490101231 (2)cudnn.py in /python/tvm/contribute: idx = -1 if algo_type == "fwd": idx = _FWD_ALGOS.index(algo_name) elif algo_type == "bwd_filter": idx =

[GitHub] [tvm] masahi commented on pull request #14417: [Unity] Remove non-deterministic behavior from graph pattern matching

masahi commented on PR #14417: URL: https://github.com/apache/tvm/pull/14417#issuecomment-1490049917 Just hit an interesting bug in the graph matcher. The following fake QKV projection mod and pattern should not match. Without this PR I get segfault. With this PR, it incorrectly

[tvm] branch llvm_update created (now a58645352a)

This is an automated email from the ASF dual-hosted git repository. lukhut pushed a change to branch llvm_update in repository https://gitbox.apache.org/repos/asf/tvm.git at a58645352a [CI] Add llvm-15 and enable mlir build for ci_cpu docker setup. No new revisions were added by this

[GitHub] [tvm] masahi commented on pull request #14378: [Unity] [Hexagon] Fix deprecated call for data layout size in bits

masahi commented on PR #14378: URL: https://github.com/apache/tvm/pull/14378#issuecomment-1489929407 @janetsc Can you send this to `main`? The change to `unity` should be specific to relax. We get the same change in `unity` after weekly rebase. -- This is an automated message from the

[GitHub] [tvm] masahi merged pull request #14400: [TIR] Update LowerTVMBuiltin to use Optional

masahi merged PR #14400: URL: https://github.com/apache/tvm/pull/14400 -- 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 dependabot/pip/apps/android_camera/models/tensorflow-2.11.1 updated (0707494e4e -> 62159866fb)

This is an automated email from the ASF dual-hosted git repository. github-bot pushed a change to branch dependabot/pip/apps/android_camera/models/tensorflow-2.11.1 in repository https://gitbox.apache.org/repos/asf/tvm.git omit 0707494e4e Bump tensorflow from 2.9.3 to 2.11.1 in

[tvm] branch main updated (1d1dbebc73 -> cbe068cfac)

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 1d1dbebc73 [microTVM]Fix more security issues with pyproject (#14434) add cbe068cfac [TIR] Update LowerTVMBuiltin to

[GitHub] [tvm] masahi merged pull request #14434: [microTVM]Fix more security issues with pyproject

masahi merged PR #14434: URL: https://github.com/apache/tvm/pull/14434 -- 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 (fafe39ddab -> 1d1dbebc73)

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 fafe39ddab [Analysis] Improve error message in VerifyWellFormed (#14389) add 1d1dbebc73 [microTVM]Fix more security

[tvm] branch main updated (79027f92ac -> fafe39ddab)

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 79027f92ac [TIR] Remove special-casing of T.address_of in the storage rewrite pass (#14430) add fafe39ddab

[GitHub] [tvm] masahi merged pull request #14389: [Analysis] Improve error message in VerifyWellFormed

masahi merged PR #14389: URL: https://github.com/apache/tvm/pull/14389 -- 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 (4011280b16 -> 79027f92ac)

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 4011280b16 [OpenCL][Textures] Always use SSA for texture loading (#14397) add 79027f92ac [TIR] Remove

[GitHub] [tvm] masahi merged pull request #14430: [TIR] Remove special-casing of T.address_of in the storage rewrite pass

masahi merged PR #14430: URL: https://github.com/apache/tvm/pull/14430 -- 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 closed issue #13976: [Bug][TIR][StorageRewrite] bufferload unvisited when used in builtin::address_of()

masahi closed issue #13976: [Bug][TIR][StorageRewrite] bufferload unvisited when used in builtin::address_of() URL: https://github.com/apache/tvm/issues/13976 -- 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

[GitHub] [tvm] masahi commented on pull request #14431: [Unity] Minor updates to DataFlowBlockRewrite

masahi commented on PR #14431: URL: https://github.com/apache/tvm/pull/14431#issuecomment-1489922852 I just realized that I didn't need to use `DataflowBlockRewrite` for my purpose. But keeping this PR for now. -- This is an automated message from the Apache Git Service. To respond to

[GitHub] [tvm] Aleksei-grovety commented on pull request #14414: [skip ci] Replace magic_wand model with micro_speech

Aleksei-grovety commented on PR #14414: URL: https://github.com/apache/tvm/pull/14414#issuecomment-1489908896 > LGTM, thanks! Just to double check, did you already upload the model to S3? I didn't, someone uploaded [the

[GitHub] [tvm] tvm-bot commented on pull request #14436: [Typo] Fix name of iter var type 4

tvm-bot commented on PR #14436: URL: https://github.com/apache/tvm/pull/14436#issuecomment-1489823824 Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews

[GitHub] [tvm] wrongtest-intellif opened a new pull request, #14436: [Typo] Fix name of iter var type 4

wrongtest-intellif opened a new pull request, #14436: URL: https://github.com/apache/tvm/pull/14436 IterVarType enum 4 should be opaque. -- 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

[GitHub] [tvm] leandron commented on pull request #14426: [CL] Update Compute Library from v22.11 to v23.02.1

leandron commented on PR #14426: URL: https://github.com/apache/tvm/pull/14426#issuecomment-1489819546 @tvm-bot rerun -- 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] multiverstack-intellif commented on pull request #14303: [CI] Add llvm-15 and mlir-15 to Docker setup

multiverstack-intellif commented on PR #14303: URL: https://github.com/apache/tvm/pull/14303#issuecomment-1489807163 > Hi. Friendly nudge so that we can take a decision on how to proceed here. I'd suggest that we build LLVM 15 from source, similar to what is being done in `ci_arm` for now.

[GitHub] [tvm] PhilippvK commented on a diff in pull request #12525: [TVMC] Allow selecting a subset of tasks to be used in `tvmc tune`

PhilippvK commented on code in PR #12525: URL: https://github.com/apache/tvm/pull/12525#discussion_r1152798088 ## tests/python/driver/tvmc/test_autotuner.py: ## @@ -207,3 +209,27 @@ def test_autotune_pass_context(mock_pc, onnx_mnist, tmpdir_factory): # AutoTVM overrides

[GitHub] [tvm] PhilippvK commented on pull request #12525: [TVMC] Allow selecting a subset of tasks to be used in `tvmc tune`

PhilippvK commented on PR #12525: URL: https://github.com/apache/tvm/pull/12525#issuecomment-1489774227 @lhutton1 Hey, I implemented the proposed changes and the CI passed. Feel free to merge after quickly having a look at the refactored unit tests. -- This is an automated message from