(tvm) branch nightly updated (6a3fadc065 -> 268d15c987)

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 6a3fadc065 [Unity][Transform] Handle `call_tir_inplace` in `FuseTIR` and `FuseOps` (#16487) add 2dcf9ec5a6

Re: [PR] [Unity][TVMScript] Optionally hide StructInfo that can be inferred [tvm]

Lunderberg commented on PR #16356: URL: https://github.com/apache/tvm/pull/16356#issuecomment-1933248740 Sounds good, and thank you for the review! Before merging, I'm going to rebase onto main, since the CI results look a bit out of date. -- This is an automated message from the Apache

Re: [PR] [Unity][TVMScript] Optionally hide StructInfo that can be inferred [tvm]

Lunderberg commented on code in PR #16356: URL: https://github.com/apache/tvm/pull/16356#discussion_r1482323270 ## src/script/printer/relax/utils.h: ## @@ -82,10 +84,47 @@ inline Optional StructInfoAsAnn(const relax::Var& v, const ObjectPath& if (!v->struct_info_.defined())

Re: [PR] [Runtime] Add "TVM_DLL" to NDArray cache load func [tvm]

MasterJH5574 commented on PR #16541: URL: https://github.com/apache/tvm/pull/16541#issuecomment-1933223239 @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.

Re: [PR] [TIR] Expand debug symbol output for CodeGenLLVM [tvm]

Lunderberg commented on PR #16544: URL: https://github.com/apache/tvm/pull/16544#issuecomment-1933217085 This came about while debugging the implementation of https://github.com/apache/tvm/pull/16542, but is otherwise unrelated. -- This is an automated message from the Apache Git

Re: [PR] [TIR] Fix segfaults from ordering of Let/Assert in MakePackedAPI [tvm]

Lunderberg commented on PR #16543: URL: https://github.com/apache/tvm/pull/16543#issuecomment-1933217017 This came about while debugging the implementation of https://github.com/apache/tvm/pull/16542, but is otherwise unrelated. -- This is an automated message from the Apache Git

[PR] [TIR] Fix segfaults from ordering of Let/Assert in MakePackedAPI [tvm]

Lunderberg opened a new pull request, #16543: URL: https://github.com/apache/tvm/pull/16543 Prior to this commit, the `MakePackedAPI` pass would output steps in the following order: 1. Check the number of arguments. 2. All `LetStmt` produced by the `ArgBinder` 3. `AssertStmt`

[PR] [Relax] Support callback as argument [tvm]

Lunderberg opened a new pull request, #16542: URL: https://github.com/apache/tvm/pull/16542 Prior to this commit, calls from Relax to external PackedFuncs could only be done through the TVM global registry. While Relax functions accepting a callback could be written as `callback_arg:

(tvm) branch main updated (2dcf9ec5a6 -> 268d15c987)

This is an automated email from the ASF dual-hosted git repository. tqchen pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git from 2dcf9ec5a6 [Keras] Enable Dense operator for any input dims (#16526) add 268d15c987 [CI] Fix CI Script and Broken

Re: [PR] [CI] Fix CI Script and Broken Tests [tvm]

tqchen merged PR #16521: URL: https://github.com/apache/tvm/pull/16521 -- 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] [Upd] Enable lld search to include /opt/rocm/llvm/bin for rocm [tvm]

shreygupta2809 opened a new pull request, #16540: URL: https://github.com/apache/tvm/pull/16540 Closes [#1216](https://github.com/mlc-ai/mlc-llm/issues/1216) and [#1614](https://github.com/mlc-ai/mlc-llm/issues/1614) @tqchen -- This is an automated message from the Apache Git Service.

Re: [PR] Improve error message in NDArray::CopyFromTo [tvm]

ekalda commented on PR #16539: URL: https://github.com/apache/tvm/pull/16539#issuecomment-1932248856 cc @lhutton1 @eirenevp -- 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.

[PR] Improve error message in NDArray::CopyFromTo [tvm]

ekalda opened a new pull request, #16539: URL: https://github.com/apache/tvm/pull/16539 Make it explicit that the quoted numbers are bytes. -- 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

Re: [PR] [SVE] Change the dtype of Ramp and Broadcast lanes to PrimExpr [tvm]

tqchen commented on code in PR #16523: URL: https://github.com/apache/tvm/pull/16523#discussion_r1481512533 ## include/tvm/runtime/data_type.h: ## @@ -114,17 +118,28 @@ class DataType { /*! \return whether type is a handle type. */ bool is_handle() const { return code()

[I] [BUG] List index out of range compiling `test_tflite_large_irregular` on `arm_cpu` target [tvm]

lhutton1 opened a new issue, #16538: URL: https://github.com/apache/tvm/issues/16538 ### Expected behaviour: The test `tests/python/relay/test_op_qnn_conv2d.py:test_tflite_large_irregular` runs successfully when the target is `arm_cpu`. ### Actual behaviour: The test fails to

Re: [PR] [SVE] Change the dtype of Ramp and Broadcast lanes to PrimExpr [tvm]

ekalda commented on code in PR #16523: URL: https://github.com/apache/tvm/pull/16523#discussion_r1481340370 ## src/arith/scalable_expression.cc: ## @@ -0,0 +1,62 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements.

Re: [PR] [SVE] Change the dtype of Ramp and Broadcast lanes to PrimExpr [tvm]

ekalda commented on code in PR #16523: URL: https://github.com/apache/tvm/pull/16523#discussion_r1481333743 ## src/arith/rewrite_simplify.h: ## @@ -221,6 +221,8 @@ class RewriteSimplifier::Impl : public IRMutatorWithAnalyzer { bool CanProveGreaterEqual(const PrimExpr& x,

Re: [PR] [SVE] Change the dtype of Ramp and Broadcast lanes to PrimExpr [tvm]

ekalda commented on code in PR #16523: URL: https://github.com/apache/tvm/pull/16523#discussion_r1481333106 ## src/arith/int_set.cc: ## @@ -466,14 +466,21 @@ class IntervalSetEvaluator : public ExprFunctor { if (stride.Match(op->stride)) { DataType t =

Re: [PR] [SVE] Change the dtype of Ramp and Broadcast lanes to PrimExpr [tvm]

ekalda commented on code in PR #16523: URL: https://github.com/apache/tvm/pull/16523#discussion_r1481332421 ## include/tvm/runtime/data_type.h: ## @@ -114,17 +118,28 @@ class DataType { /*! \return whether type is a handle type. */ bool is_handle() const { return code()

[I] [BUG] NCHW4c is an unsupported convolution layout for `arm_cpu` [tvm]

lhutton1 opened a new issue, #16537: URL: https://github.com/apache/tvm/issues/16537 ### Expected behaviour: When compiled with target `arm_cpu` the model should compile successfully. ### Actual behaviour: The test fails to run and gives the following error: ``` def

Re: [PR] [Keras] Enable Dense operator for any input dims [tvm]

masahi merged PR #16526: URL: https://github.com/apache/tvm/pull/16526 -- 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: [Keras] Enable Dense operator for any input dims (#16526)

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 2dcf9ec5a6 [Keras] Enable Dense operator for any input

[I] [BUG] Schedule that doesn't support dynamic height/width is selected when compiling convolution for `arm_cpu` [tvm]

lhutton1 opened a new issue, #16536: URL: https://github.com/apache/tvm/issues/16536 ### Expected behaviour: When an `arm_cpu` target is used, the model should compile successfully without an error. ### Actual behaviour: When compiled on an `arm_cpu` target, the models results