(tvm) branch nightly updated (c8deb7fa36 -> a320b63198)

2024-04-30 Thread github-bot
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 c8deb7fa36 Overriding the StructuralEqual() for easy usage (#16908) add bc8742b4c9 [Misc] Add script for

Re: [PR] [Relax] Implement relax.op.view [tvm]

2024-04-30 Thread via GitHub
tqchen commented on PR #16955: URL: https://github.com/apache/tvm/pull/16955#issuecomment-2087792055 `R.memory.ensure_compact` will actually return a new NDArray whose `elem_offset=0`, and for devices that allows explicit ptr moving, this is something that can accelerate backends --

Re: [PR] chore: remove repetitive words [tvm]

2024-04-30 Thread via GitHub
tianzedavid commented on PR #16957: URL: https://github.com/apache/tvm/pull/16957#issuecomment-2087246950 > Is this spam? Perhaps the @tianzedavid is farming trust? Github profile looks suspicious Just a beginner, wanting to submit some code to open source projects. You think and do

Re: [PR] [Relax] Implement relax.op.view [tvm]

2024-04-30 Thread via GitHub
Lunderberg commented on PR #16955: URL: https://github.com/apache/tvm/pull/16955#issuecomment-2087130666 Changes have been made as requested, ready for re-review. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the

Re: [PR] [Relax] Implement relax.op.view [tvm]

2024-04-30 Thread via GitHub
Lunderberg commented on code in PR #16955: URL: https://github.com/apache/tvm/pull/16955#discussion_r1585490595 ## src/relax/op/tensor/view.h: ## @@ -0,0 +1,38 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See

Re: [PR] [Relax] Implement relax.op.view [tvm]

2024-04-30 Thread via GitHub
Lunderberg commented on PR #16955: URL: https://github.com/apache/tvm/pull/16955#issuecomment-2087070331 > Having view operation can in general cause problems mainly because most ops(including generated and external ones) assumes `elem_offset = 0` So long as the operations assert

Re: [PR] [Frontend][PyTorch] torch2.0 ExportedProgram PoC [tvm]

2024-04-30 Thread via GitHub
tqchen commented on PR #16531: URL: https://github.com/apache/tvm/pull/16531#issuecomment-2086557479 @chunit-quic just want to check in and see where we are on this -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the

Re: [PR] [Relax][UnitTest] Validate IRModule with multiple targets [tvm]

2024-04-30 Thread via GitHub
Lunderberg commented on PR #16960: URL: https://github.com/apache/tvm/pull/16960#issuecomment-2086551418 This PR adds a unit test validating an end-to-end compilation. On `main`, this end-to-end test encounters the errors resolved in PRs https://github.com/apache/tvm/pull/16958 and

(tvm) branch main updated (6252fa5802 -> a320b63198)

2024-04-30 Thread wuwei
This is an automated email from the ASF dual-hosted git repository. wuwei pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git from 6252fa5802 [TIR] Enhance CLZ intrinsic support (#16952) add a320b63198 [Unity][Cutlass] Fix C source generation of

[PR] [Relax][UnitTest] Validate IRModule with multiple targets [tvm]

2024-04-30 Thread via GitHub
Lunderberg opened a new pull request, #16960: URL: https://github.com/apache/tvm/pull/16960 This commit adds a unit test to verify that a single `IRModule` can contain functions that will be used on multiple distinct targets. Previously, this test case caused errors when running the

Re: [PR] [Unity][Cutlass] Fix C source generation of dense operation [tvm]

2024-04-30 Thread via GitHub
vinx13 merged PR #16476: URL: https://github.com/apache/tvm/pull/16476 -- 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:

[PR] [DLight] Check for target in function attributes [tvm]

2024-04-30 Thread via GitHub
Lunderberg opened a new pull request, #16958: URL: https://github.com/apache/tvm/pull/16958 Prior to this commit, the `dlight` scheduling rules were applied solely based on the global `tvm.target.Target.current()`. However, a TIR PrimFunc may be annotated with the target, rather than

Re: [PR] [Relax] Implement relax.op.view [tvm]

2024-04-30 Thread via GitHub
tqchen commented on code in PR #16955: URL: https://github.com/apache/tvm/pull/16955#discussion_r1584898774 ## src/relax/op/tensor/view.h: ## @@ -0,0 +1,38 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the

Re: [PR] [Relax] Implement relax.op.view [tvm]

2024-04-30 Thread via GitHub
tqchen commented on PR #16955: URL: https://github.com/apache/tvm/pull/16955#issuecomment-2085436158 Just want to note here. Having view operation can in general cause problems mainly because most ops(including generated and external ones) assumes `elem_offset = 0` for performance

Re: [I] [CI Problem] [tvm]

2024-04-30 Thread via GitHub
ysh329 commented on issue #16953: URL: https://github.com/apache/tvm/issues/16953#issuecomment-2085356059 Thanks for participating in the TVM community! We use https://discuss.tvm.ai for any general usage questions and discussions. The issue tracker is used for actionable items such as

Re: [I] [CI Problem] how i load a tvm model from memory rather than from so file? [tvm]

2024-04-30 Thread via GitHub
ysh329 commented on issue #16944: URL: https://github.com/apache/tvm/issues/16944#issuecomment-2085354112 Thanks for participating in the TVM community! We use https://discuss.tvm.ai for any general usage questions and discussions. The issue tracker is used for actionable items such as

Re: [PR] [Codegen, CUDA] Enable emitting SyncWarp [tvm]

2024-04-30 Thread via GitHub
tqchen commented on PR #16918: URL: https://github.com/apache/tvm/pull/16918#issuecomment-2085335805 @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

Re: [PR] chore: remove repetitive words [tvm]

2024-04-30 Thread via GitHub
darrenvechain commented on PR #16957: URL: https://github.com/apache/tvm/pull/16957#issuecomment-2084809179 Is this spam? Perhaps the @tianzedavid is farming trust? Github profile looks suspicious -- This is an automated message from the Apache Git Service. To respond to the message,

[PR] chore: remove repetitive words [tvm]

2024-04-30 Thread via GitHub
tianzedavid opened a new pull request, #16957: URL: https://github.com/apache/tvm/pull/16957 (no comment) -- 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,

(tvm) branch main updated: [TIR] Enhance CLZ intrinsic support (#16952)

2024-04-30 Thread syfeng
This is an automated email from the ASF dual-hosted git repository. syfeng 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 6252fa5802 [TIR] Enhance CLZ intrinsic support

Re: [PR] [TIR] Enhance CLZ intrinsic support [tvm]

2024-04-30 Thread via GitHub
Hzfengsy merged PR #16952: URL: https://github.com/apache/tvm/pull/16952 -- 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: