## What's Changed * [Relax][PyTorch] Cast non-bool inputs to bool in logical_and converter by @javierdejesusda in https://github.com/apache/tvm/pull/19679 * [CI] Remove PyPI-only tag ref guard from wheel publishing by @tlopex in https://github.com/apache/tvm/pull/19685 * [CI] Target apache-tvm for PyPI wheel publishing by @tlopex in https://github.com/apache/tvm/pull/19686 * [Web] Bump tvmjs version to 0.25.0-dev1 by @akaashrp in https://github.com/apache/tvm/pull/19687 * [Fix] CommReduce could handle 0-dim data by @flashmouse in https://github.com/apache/tvm/pull/19683 * [CI] Pin actions by version tag, trim wheel perms by @MasterJH5574 in https://github.com/apache/tvm/pull/19703 * [Tests] Fix s_tir tests using removed T.block API in TIRx script by @tlopex in https://github.com/apache/tvm/pull/19706 * [CI] Fix release verification script by @MasterJH5574 in https://github.com/apache/tvm/pull/19700 * [Refactor][Meta-schedule] Remove meta-schedule as_string mechanism in favor of default representation by @tlopex in https://github.com/apache/tvm/pull/19709 * [Python] Bump apache-tvm-ffi floor to >=0.1.12 by @MasterJH5574 in https://github.com/apache/tvm/pull/19710 * [Relax][CoreML] Fix CoreML partition pass by @tlopex in https://github.com/apache/tvm/pull/19711 * [Tests] Skip test modules cleanly when optional deps are missing by @tlopex in https://github.com/apache/tvm/pull/19704 * [CI] Merge PR against its target branch instead of main by @MasterJH5574 in https://github.com/apache/tvm/pull/19712 * [CI] Fix CI script test subprocess environment by @tlopex in https://github.com/apache/tvm/pull/19713 * [Codegen][LLVM] Accept splat form in VLA broadcast test by @tlopex in https://github.com/apache/tvm/pull/19716 * [DOCS] Clarify loading serialized artifacts requires a trusted source by @tqchen in https://github.com/apache/tvm/pull/19720 * [REFACTOR][PYTHON] Slim tvm.libinfo to info-only helpers by @tqchen in https://github.com/apache/tvm/pull/19719 * [Codegen][NVPTX] Skip runtime execution in Vulkan codegen tests by @tlopex in https://github.com/apache/tvm/pull/19717 * [REFACTOR][PYTHON] Remove tvm.ffi shim; import tvm_ffi directly by @tqchen in https://github.com/apache/tvm/pull/19721 * [Runtime][Tests] Fix contrib wheel tests by @tlopex in https://github.com/apache/tvm/pull/19714 * [Tests][Disco] Skip CCL tests when runtime support is absent by @tlopex in https://github.com/apache/tvm/pull/19724 * [Tests][Relax] Gate multi-GPU VM test on three devices by @tlopex in https://github.com/apache/tvm/pull/19725 * [REFACTOR][IR] Phase out diagnostic.h for visit-context-aware pass errors by @tqchen in https://github.com/apache/tvm/pull/19722 * [Tests][Hexagon] Lazily import pytest plugin dependencies by @tlopex in https://github.com/apache/tvm/pull/19726 * [Python] Refactor pyproject.toml dependencies by @tlopex in https://github.com/apache/tvm/pull/19723 * [Tests][NNAPI] Skip tests cleanly when remote environment is unavailable by @tlopex in https://github.com/apache/tvm/pull/19730 * [Tests][S-TIR] Fix stale MetaSchedule sketch expectations and migrate let binds to T.let by @tlopex in https://github.com/apache/tvm/pull/19729 * [Tests] Remove test_runtime_ndarray (covered by tvm-ffi) by @tlopex in https://github.com/apache/tvm/pull/19715 * [TIRx] Preserve Triton call_kernel compile options by @tlopex in https://github.com/apache/tvm/pull/19728 * [Relax][PyTorch][DLight] Fix exported-program CUDA test failures by @tlopex in https://github.com/apache/tvm/pull/19732 * [PYTHON] Autoload backends; simplify library loading; remove TVMError for native errors by @tqchen in https://github.com/apache/tvm/pull/19727 * [Script][Tests] Fix dialect redirect module re-execution and stray category-less tirx.intrin_test op by @tlopex in https://github.com/apache/tvm/pull/19731 * [S-TIR][Tests] Fix transform test failures after TIRx bringup by @tlopex in https://github.com/apache/tvm/pull/19735 * [TIRx] Use canonical PTX async script API in s_tir test by @tlopex in https://github.com/apache/tvm/pull/19739 * [Tests] Check WebGPU volatile allreduce annotation structurally by @tlopex in https://github.com/apache/tvm/pull/19740 * [S-TIR] Fix software pipeline offsets for legacy MMA intrinsics by @tlopex in https://github.com/apache/tvm/pull/19742 * [Tests] Fix flaky popen pool executor test by @tlopex in https://github.com/apache/tvm/pull/19746 * [Hexagon][Tests] Clean up stale hexagon tests by @tlopex in https://github.com/apache/tvm/pull/19747 * [Runtime][Disco] Fix session attribute storage, NVSHMEM build, and test gating by @tlopex in https://github.com/apache/tvm/pull/19736 * [CI] Align cuda-python with PyTorch cuda-bindings by @tlopex in https://github.com/apache/tvm/pull/19738 * [Codegen][LLVM][Tests] Gate +v9a vscale_range expectation on LLVM version by @tlopex in https://github.com/apache/tvm/pull/19744 * [Runtime][Tests] Drop int4 from random_fill test, fix dtype error message by @tlopex in https://github.com/apache/tvm/pull/19748 * [Tests][LLVM] Gate stepvector intrinsic rename on LLVM 20 by @tlopex in https://github.com/apache/tvm/pull/19745 * [S-TIR][Tests] Mark test_cp_async_in_if_then_else as xfail by @tlopex in https://github.com/apache/tvm/pull/19751 * [CI] Run s_tir/transform tests in the python-unittest stage by @tlopex in https://github.com/apache/tvm/pull/19737 * [CI] Updated cibw to 4.1.0 by @tlopex in https://github.com/apache/tvm/pull/19754 * [TIRX][Tests] Fix LLVM version gate for vectorized lround by @tlopex in https://github.com/apache/tvm/pull/19753 * [S-TIR][CUDA] Fix legacy predicated cp.async zero fill by @tlopex in https://github.com/apache/tvm/pull/19741 * [Tests][AArch64] Make SVE codegen assertions robust across LLVM versions by @tlopex in https://github.com/apache/tvm/pull/19752 * [Relax][PyTorch] Add logical_or and logical_xor converters by @javierdejesusda in https://github.com/apache/tvm/pull/19756 * [TIRx] Post-bringup follow-ups: op-dispatch, namespaces, launch bounds, gemm-async, backend reorg by @spectrometerHBH in https://github.com/apache/tvm/pull/19757 * [REFACTOR][VM] Move CUDA graph VM builtin back under VM runtime by @tqchen in https://github.com/apache/tvm/pull/19758 * [Runtime][CoreML] Fix FFI casts in CoreML runtime by @tlopex in https://github.com/apache/tvm/pull/19762 * [CI] Drop redundant cmake/ninja install from the Linux wheel CUDA sidecar by @tlopex in https://github.com/apache/tvm/pull/19761 * [REFACTOR][DataType] Phase out target custom datatype support by @tqchen in https://github.com/apache/tvm/pull/19760 * [REFACTOR][TARGET] Cleanup backend target registration by @tqchen in https://github.com/apache/tvm/pull/19759 * [MetaScheduler] Improve print info about builder/runner state by @cbalint13 in https://github.com/apache/tvm/pull/19767 * [REFACTOR][CUDA] Phase out l2 cache flush preproc test by @tqchen in https://github.com/apache/tvm/pull/19768 * [Relax][ONNX] Fix LayerNormalization no-bias zero tensor shape and dtype by @javierdejesusda in https://github.com/apache/tvm/pull/19772 * [Relax][ONNX] Support exclusive option in CumSum by @javierdejesusda in https://github.com/apache/tvm/pull/19773 * [CPP_RPC] Bugfix race conditions and enhance print infos by @cbalint13 in https://github.com/apache/tvm/pull/19778 * [CMAKE] Upgrade TVM build baseline to C++20 by @Ubospica in https://github.com/apache/tvm/pull/19734 * [REFACTOR][CUDA] Phase out cuda_common.h by @tqchen in https://github.com/apache/tvm/pull/19770 * [REFACTOR][PYTHON] Consolidate backend autoload infra by @tqchen in https://github.com/apache/tvm/pull/19769 * [Fix] nn.attention support dynamic batch_size by @flashmouse in https://github.com/apache/tvm/pull/19779 * [Relax][ONNX] Make ReduceMax/ReduceMin NaN propagation order-independent(numpy semantics) by @cchung100m in https://github.com/apache/tvm/pull/19755 * [Docs][CI] Bump tlcpack-sphinx-addon to restore search result summaries by @tlopex in https://github.com/apache/tvm/pull/19782 * [REFACTOR][IR] Cleanup IR naming utilities by @tqchen in https://github.com/apache/tvm/pull/19781 * [CUDA] Narrow the cuda extra from cuda-python to cuda-bindings by @tlopex in https://github.com/apache/tvm/pull/19784 * [AGENT] Migrate agent instructions to vendor-neutral layout by @tqchen in https://github.com/apache/tvm/pull/19783 * [Tests] Modernize test gating by @tlopex in https://github.com/apache/tvm/pull/19777 * [TIRX][CUDA] Framework support for FA4, CLC intrinsics, and nvfp4 tcgen05 GEMM by @spectrometerHBH in https://github.com/apache/tvm/pull/19785 * [Relax][TensorRT] Update TensorRT runtime to 10 by @tlopex in https://github.com/apache/tvm/pull/19789 * [Tests] Make TargetCreation.DeduplicateKeys host-agnostic on AArch64 by @tlopex in https://github.com/apache/tvm/pull/19786 * [Tests] Replace remaining requires_* helpers with standard pytest by @tlopex in https://github.com/apache/tvm/pull/19787 * [TIRx][RISC-V] Use scalable RVV loops for fixed vectorize by @ZephyrLi-pro in https://github.com/apache/tvm/pull/19776 * [Docs] Modernize test-gating documentation by @tlopex in https://github.com/apache/tvm/pull/19788 * [Web] Destroy GPUDevice once on buffer creation error by @guan404ming in https://github.com/apache/tvm/pull/19790 * [REFACTOR] Phase out unused queue and rang license entries by @tqchen in https://github.com/apache/tvm/pull/19794 * [REFACTOR][HEXAGON] Phase out Hexagon app and test wrappers by @tqchen in https://github.com/apache/tvm/pull/19796 * [CI] Pin GitHub Actions to SHA for ASF INFRA compliance by @guan404ming in https://github.com/apache/tvm/pull/19793 * refactor(web): use singular requestFileHandle() instead of requestFileHandles() by @tomayac in https://github.com/apache/tvm/pull/19780 * [REFACTOR][IR] Simplify CallingConv attribute access by @tqchen in https://github.com/apache/tvm/pull/19799 * [CI] Remove Jenkins PR linter step by @tqchen in https://github.com/apache/tvm/pull/19798 * [Relax][Frontend][TFLite] Add support for FFT/complex operators: REAL, IMAG, COMPLEX_ABS by @fnhirwa in https://github.com/apache/tvm/pull/19763 * [Tests][Refactor] Remove unused testing helpers by @tlopex in https://github.com/apache/tvm/pull/19800 * [REFACTOR][TIRX] Add IntImm common scalar ctor and streamline MakeConst by @tqchen in https://github.com/apache/tvm/pull/19797 * [Relax] Fix FuseOpsByPattern dropping bound constants on duplicate-structure functions by @tlopex in https://github.com/apache/tvm/pull/19801 * [Tests] Migrate tvm.testing.parameters() to pytest.mark.parametrize by @tlopex in https://github.com/apache/tvm/pull/19803 * [Tests] Remove the now-unused tvm.testing.parameters() helper by @tlopex in https://github.com/apache/tvm/pull/19807 * [Tests] Drop tautological env-probe implication tests by @tlopex in https://github.com/apache/tvm/pull/19811 * [Relax][TensorRT] Update BYOC operator converters from Relay to Relax by @tlopex in https://github.com/apache/tvm/pull/19810 * [FIX][TIRX] Fix dangling Op static reference warning by @tqchen in https://github.com/apache/tvm/pull/19806 * [Relax][Frontend][TFLite] Add RFFT2D op and supporting TIR kernels by @Aharrypotter in https://github.com/apache/tvm/pull/19812 * [BUILD] Hide static linked library symbols in shared libs by @tqchen in https://github.com/apache/tvm/pull/19817 * Preserve ONNX BatchNormalization inference mode by @yinli-systems in https://github.com/apache/tvm/pull/19818 * [CI] Drop redundant conda LLVM install in GH Actions build by @tlopex in https://github.com/apache/tvm/pull/19823 * [Tests] Inline thin gating helpers in the pytest plugin and tvm.testing.env by @tlopex in https://github.com/apache/tvm/pull/19819 * [Relax][Frontend][NN] Fix SourceModule include resolution for installed wheels by @tlopex in https://github.com/apache/tvm/pull/19822 * [ARITH] Add optional Z3-backed proving to Analyzer by @Ubospica in https://github.com/apache/tvm/pull/19667 * [Relax][Frontend][TFLite] Add missing TFLite operator mappings by @Aharrypotter in https://github.com/apache/tvm/pull/19813 * [Tests] Migrate off tvm.testing.parametrize_targets to native pytest by @tlopex in https://github.com/apache/tvm/pull/19826 * [FFI][ABI] Bump tvm-ffi to latest by @tqchen in https://github.com/apache/tvm/pull/19831 * [Relax][TensorRT] Add partition_for_tensorrt and a pattern table by @tlopex in https://github.com/apache/tvm/pull/19820 * [Arith] Let IRMutatorWithAnalyzer take a const Analyzer& by @tlopex in https://github.com/apache/tvm/pull/19829 * fix: Support 5D volumetric inputs in ONNX GridSample frontend converter by @mvanhorn in https://github.com/apache/tvm/pull/19816 * [Tests] Remove dead helpers and unused probes from tvm.testing by @tlopex in https://github.com/apache/tvm/pull/19821 * [Arith] Restrict floormod coefficient reduction to keep DetectIterMapstable by @tlopex in https://github.com/apache/tvm/pull/19832 * [Web] Avoid redundant OPFS lookup on cache hit by @guan404ming in https://github.com/apache/tvm/pull/19791 * [Docs] Rework Bring Your Own Codegen tutorial and add TensorRT example by @tlopex in https://github.com/apache/tvm/pull/19839 * [Arith] Add Analyzer::Clone for deep-copying analyzer state by @tlopex in https://github.com/apache/tvm/pull/19836 * [BUILD] Bump tvm-ffi after RuntimeTypeIndex optimization by @tqchen in https://github.com/apache/tvm/pull/19834 * [CPP_RPC] Replace legacy OS-specific API with std:: libraries by @cbalint13 in https://github.com/apache/tvm/pull/19840 * [Relax][ONNX] Preserve NaN in Relu to align with ONNX Runtime by @cchung100m in https://github.com/apache/tvm/pull/19750 * [Relax] Fix matmul and reductions with zero-size dimension return uninitialized memory by @cchung100m in https://github.com/apache/tvm/pull/19680 * [TIRx] Phase out flat device-intrinsic op aliases by @tlopex in https://github.com/apache/tvm/pull/19838 * [BUILD] Sync fallback version strings to 0.26 dev cycle by @MasterJH5574 in https://github.com/apache/tvm/pull/19845 * [Docker] Bump CI image deps: sphinx-book-theme + z3-static by @tlopex in https://github.com/apache/tvm/pull/19835 * [Tests] Clean unused tvm.testing helpers by @tlopex in https://github.com/apache/tvm/pull/19846 * [Relax][ONNX] Drop NaN-preservation isnan-where wrappers by @tlopex in https://github.com/apache/tvm/pull/19847 * [Relax][ONNX] Accept 1-D scalar inputs in NonMaxSuppression by @guan404ming in https://github.com/apache/tvm/pull/19843 * [Relax][PyTorch] Add atan2 converter by @javierdejesusda in https://github.com/apache/tvm/pull/19850 * [REFACTOR][IR] Unify StructInfo and Type by @tqchen in https://github.com/apache/tvm/pull/19853 * [Docs] Add TIRx documentation section by @spectrometerHBH in https://github.com/apache/tvm/pull/19855 * [Metal] Enable Metal 4 shader compilation by @oraluben in https://github.com/apache/tvm/pull/19595 * [Relax][Frontend][TFLite] Add explicit operator marker handling by @Aharrypotter in https://github.com/apache/tvm/pull/19824 * [Relax][ONNX] Support align_corners in AffineGrid op by @guan404ming in https://github.com/apache/tvm/pull/19864 * [REFACTOR][RELAX] Phase out Relax PrimType by @tqchen in https://github.com/apache/tvm/pull/19858 * [REFACTOR][IR] Phase out Downcast usages by @tqchen in https://github.com/apache/tvm/pull/19857 * [Relax] Legalize nn.dropout as inference no-op by @guan404ming in https://github.com/apache/tvm/pull/19841 * [Docs][CI] Switch docs theme and bump images to 20260619-214849-4174cdf5 by @tlopex in https://github.com/apache/tvm/pull/19828 * [Relax] Legalize dilated conv_transpose by @guan404ming in https://github.com/apache/tvm/pull/19842 * [Relax] Update dropout call_tir out_ty spelling by @tqchen in https://github.com/apache/tvm/pull/19874 * [Docs] Right-align documentation footer by @tlopex in https://github.com/apache/tvm/pull/19868 * [Relax][TFLite] Add remaining operator tests and reverse_sequence op by @Aharrypotter in https://github.com/apache/tvm/pull/19814 * [CI] Stop building ci_lint Docker image by @tlopex in https://github.com/apache/tvm/pull/19872 * [DOCS] Align footer dropdown menu by @tlopex in https://github.com/apache/tvm/pull/19876 * [Relax][Frontend][TFLite] Support dynamic RANGE scalar bounds by @Aharrypotter in https://github.com/apache/tvm/pull/19867 * [Relax][Frontend][TFLite] Support static hashtable find by @Aharrypotter in https://github.com/apache/tvm/pull/19879 * [REFACTOR][IR] Unify PrimExpr type mechanism to PrimType instead of DataType by @tqchen in https://github.com/apache/tvm/pull/19875 * [DOCS] Add PyPI install guidance and update install command by @tlopex in https://github.com/apache/tvm/pull/19883 * [REFACTOR][IR] Clean up PrimType follow-ups by @tqchen in https://github.com/apache/tvm/pull/19884 * [TFLite] Use structural checks in frontend tests by @tlopex in https://github.com/apache/tvm/pull/19888 * [CI][CMAKE] Install CUDA driver stub for libtvm_runtime_cuda by @MasterJH5574 in https://github.com/apache/tvm/pull/19886 * [ARITH] Use IntImm in canonical scalar hot paths by @tqchen in https://github.com/apache/tvm/pull/19885 * [REFACTOR][RELAX] Rename Relax base type to AnyType by @tqchen in https://github.com/apache/tvm/pull/19889 * [Relax][Frontend][TFLite] Support dynamic DYNAMIC_UPDATE_SLICE starts by @Aharrypotter in https://github.com/apache/tvm/pull/19881 * [CI] Remove retired ci_lint Docker files by @tlopex in https://github.com/apache/tvm/pull/19878 * [TIRx] Replace vars in buffer strides and elem_offset by @guan404ming in https://github.com/apache/tvm/pull/19871 * [TIRx][LLVM] Support scalable Ramp lowering by @ZephyrLi-pro in https://github.com/apache/tvm/pull/19866 * [REFACTOR][Relax] Phase out PrimValue and Relax expression wrappers by @tqchen in https://github.com/apache/tvm/pull/19891 * [Relax] Use optional dtype for absent Relax dtype fields by @tqchen in https://github.com/apache/tvm/pull/19890 * [ONNX] Use structural checks for composite frontend tests by @tlopex in https://github.com/apache/tvm/pull/19880 * [Frontend][ONNX] Fix structural tests for TVMScript checks by @tlopex in https://github.com/apache/tvm/pull/19894 * [DOCS][TIRX] Add in-kernel profiling (CudaProfiler) tutorial by @spectrometerHBH in https://github.com/apache/tvm/pull/19895 * [Relax][PyTorch] Add rnn_tanh.input converter by @cchung100m in https://github.com/apache/tvm/pull/19837 * [DOCS] Refine tvm pypi wheel optional install guidance by @tlopex in https://github.com/apache/tvm/pull/19892 * [TIRx] Bundle CUDA tile primitive and op dispatch updates by @spectrometerHBH in https://github.com/apache/tvm/pull/19896 * [FFI] Bump tvm-ffi to latest June 28 by @tqchen in https://github.com/apache/tvm/pull/19903 * [CI][Docker] Move CI images to Ubuntu 24.04 (noble) by @tlopex in https://github.com/apache/tvm/pull/19893 * [Relax] Fix RemoveUnusedParameters symbolic var promotion by @MasterJH5574 in https://github.com/apache/tvm/pull/19901 * [Relax] Fix int64 row index cast in GPU multinomial sampling by @MasterJH5574 in https://github.com/apache/tvm/pull/19902 * [CI][Docker] Install libpolly-15-dev for noble llvm-15 static linking by @tlopex in https://github.com/apache/tvm/pull/19913 * [DOCS] Refine footer navigation links by @tlopex in https://github.com/apache/tvm/pull/19914 * [Relax] Clean up deprecated void-dtype sentinel usage by @guan404ming in https://github.com/apache/tvm/pull/19908 * [Target][RISC-V] Use riscv_cpu device key for RISC-V target tags by @Ga1axy0 in https://github.com/apache/tvm/pull/19915 * [TVMScript] Render invisible paths in structural diagnostics by @tqchen in https://github.com/apache/tvm/pull/19916 * [Runtime][KVCache] Adapt FlashInfer attention backend to 0.6.3 by @MasterJH5574 in https://github.com/apache/tvm/pull/19904 * [Relax][ONNX] Support 3D AffineGrid by @guan404ming in https://github.com/apache/tvm/pull/19863 * [TIRx] Add a dedicated boolean buffer lowering pass by @guan404ming in https://github.com/apache/tvm/pull/19873 * [ONNX] Fix missing helper in AffineGrid test by @tlopex in https://github.com/apache/tvm/pull/19920 * [REFACTOR][IR] Unify PrimExpr with Expr typed view by @tqchen in https://github.com/apache/tvm/pull/19910 * [TIRX] Remove SizeVar in favor of contextual constraints by @tqchen in https://github.com/apache/tvm/pull/19930 * [TIRx] Generalize expression functor signatures by @tqchen in https://github.com/apache/tvm/pull/19931 * [TIR] Construct proven scalar integer constants directly by @tqchen in https://github.com/apache/tvm/pull/19934 * [IR][Relax] Include expression types in structural identity by @tqchen in https://github.com/apache/tvm/pull/19933 * [FFI] Use thread-safe packed function initialization by @tqchen in https://github.com/apache/tvm/pull/19939 * [CI] Batch Python unittest pytest targets into a single invocation by @tqchen in https://github.com/apache/tvm/pull/19941 * [TEST] Serialize local GPU execution under pytest-xdist by @tqchen in https://github.com/apache/tvm/pull/19942 * [REFACTOR][SCRIPT] Keep dependent shape recursion in docsifier by @tqchen in https://github.com/apache/tvm/pull/19940 * Refactor Tensor arithmetic dispatch away from tirx.generic by @tqchen in https://github.com/apache/tvm/pull/19943 * [TIRx] Phase out duplicate Var type_annotation by @tqchen in https://github.com/apache/tvm/pull/19944 * [CI] Remove stale GitHub automation by @tqchen in https://github.com/apache/tvm/pull/19946 * [CI] Simplify Jenkins pytest execution by @tqchen in https://github.com/apache/tvm/pull/19947 * [CI] Restore dependabot configuration by @tqchen in https://github.com/apache/tvm/pull/19951 * [Cleanup] Remove macOS Clang warnings by @tqchen in https://github.com/apache/tvm/pull/19954 * [ARITH][TIR] Track positive loop extents in analyzer visitors by @tlopex in https://github.com/apache/tvm/pull/19927 * [CI] Repair Python test cleanup regressions by @tqchen in https://github.com/apache/tvm/pull/19955 * [FFI] Bump tvm-ffi for stable Optional layout by @tqchen in https://github.com/apache/tvm/pull/19956 * [ARITH] Scope interval constraints to mapped variables by @tqchen in https://github.com/apache/tvm/pull/19963 * Phase out Relax-specific Id aliases by @tqchen in https://github.com/apache/tvm/pull/19959 * [Relax][PyTorch] Bind symbolic scalar inputs in from_fx by @guan404ming in https://github.com/apache/tvm/pull/19964 * [Relax] Fix divide-by-zero in reshape pattern detection by @guan404ming in https://github.com/apache/tvm/pull/19958 * [Relax][Frontend][ONNX] Add support for Pad mode="wrap" for opset 19 by @napronald in https://github.com/apache/tvm/pull/19827 * [Relax][PyTorch] Fix masked_select VM build by @V-aerus in https://github.com/apache/tvm/pull/19937 * [TIRx] Reuse pass-through input names for inverse index map vars by @guan404ming in https://github.com/apache/tvm/pull/19906 * [Relax] Fix bucketize output dtype during legalization by @V-aerus in https://github.com/apache/tvm/pull/19936 * [Metal] Let compile callback declare payload format via (payload, fmt) by @echuraev in https://github.com/apache/tvm/pull/19924 * [Runtime] Fix CUDA build breaks in fp8 cutlass and thrust by @MasterJH5574 in https://github.com/apache/tvm/pull/19980 * [Fix][Relax][ONNX] Import TopK indices as int64 by @viiccwen in https://github.com/apache/tvm/pull/19973 * [Fix][Relax][ONNX] Cast BatchNorm params to input dtype by @viiccwen in https://github.com/apache/tvm/pull/19979 * [Relax] Legalize shape_to_tensor to device kernel by @guan404ming in https://github.com/apache/tvm/pull/19957 * [CI] Enable parallel GitHub Actions wheel builds by @tlopex in https://github.com/apache/tvm/pull/19983 * [Relax][Frontend][ONNX] Add GroupNormalization support by @napronald in https://github.com/apache/tvm/pull/19907 * [Fix] Add `origins` option to `requestFileHandle` by @tomayac in https://github.com/apache/tvm/pull/19960 * [Relax][PyTorch] Use make_tensor in exported program tests by @mshr-h in https://github.com/apache/tvm/pull/19989 * [Fix][Relax][TFLite] Use astype for frontend casts by @Aharrypotter in https://github.com/apache/tvm/pull/19932 * [Fix][Python] Use standard scikit-build directory by @tlopex in https://github.com/apache/tvm/pull/19990 * [Arith] Fix const-int-bound modular-set tightening for Mod/FloorMod by @sbinabdullah in https://github.com/apache/tvm/pull/19978 * [Fix][Relax][PyTorch] Compare Dynamo output against PyTorch reference by @mshr-h in https://github.com/apache/tvm/pull/19994 * [Fix][Relax][ONNX] Preserve rank-expanding Expand by @viiccwen in https://github.com/apache/tvm/pull/19992 * [DLight][CUDA] Fix undefined TX in GEMV broadcast epilogue by @Nanmur in https://github.com/apache/tvm/pull/19970 * [Relax][Frontend][ONNX] Support Modern QDQ opset attributes by @napronald in https://github.com/apache/tvm/pull/19993 * [Fix][Relax][ONNX] Recover ConstantOfShape initializer shape by @viiccwen in https://github.com/apache/tvm/pull/20002 * [CI] Bump CI at the Ubuntu 24.04 images and re-enable USE_Z3 by @tlopex in https://github.com/apache/tvm/pull/19911 * [Tests][TIRx] Localize hardware test gates by @tlopex in https://github.com/apache/tvm/pull/19985 * [RUNTIME][PYTHON] Add explicit Target device conversion by @tqchen in https://github.com/apache/tvm/pull/20005 * [Fix][Relax][ONNX] Preserve integer Div truncation during import by @viiccwen in https://github.com/apache/tvm/pull/19975 * [IR][Relax][TIRx] Unify Var identity by @tqchen in https://github.com/apache/tvm/pull/20004 * [RELAX] Unify call_tir primitive arguments by @tqchen in https://github.com/apache/tvm/pull/20009 * [Tests] Reduce runtime of slow Python tests by @tlopex in https://github.com/apache/tvm/pull/20006 * [Fix][Relax][ONNX] Relax op normalization for onnx subgraphs by @cbalint13 in https://github.com/apache/tvm/pull/20010 * [REFACTOR][IR] Use CamelCase Var copy helpers by @tqchen in https://github.com/apache/tvm/pull/20008 * [REFACTOR] Remove redundant defensive code guaranteed by IR invariants by @tqchen in https://github.com/apache/tvm/pull/20011 * [Tests][Frontend] Remove redundant ONNX and TFLite tests by @tlopex in https://github.com/apache/tvm/pull/20012 * [Relax][TensorRT] Fix YOLO BYOC offload and partitioning gaps by @tlopex in https://github.com/apache/tvm/pull/19998 * [Web] Link TVMFFIHandleInitOnce into WASM runtime by @akaashrp in https://github.com/apache/tvm/pull/20020 * [REFACTOR][TIR] Phase out redundant TIRx attr names by @tqchen in https://github.com/apache/tvm/pull/20017 * [IR] Rename Var name_hint field to name by @tqchen in https://github.com/apache/tvm/pull/20016 * [TIRx] Introduce first-class Return statement by @tqchen in https://github.com/apache/tvm/pull/20018 * [Tests][Frontend] Remove redundant PyTorch frontend tests by @tlopex in https://github.com/apache/tvm/pull/20021 * [REFACTOR][TIR] Remove buffer type and axis separators by @tqchen in https://github.com/apache/tvm/pull/20019 * [Tests] Update test_adaptive_pooling_window expected IR for const-int-bound fix by @sbinabdullah in https://github.com/apache/tvm/pull/20023 * [S-TIR] Remove unused meta-schedule annotation constants by @tqchen in https://github.com/apache/tvm/pull/20022 * [Fix][Relax][ONNX] Preserve ONNX Squeeze axes attribute for opset < 13 by @OmarAzizi in https://github.com/apache/tvm/pull/19966 * [Relax][Frontend][ONNX] Support dynamic index for Gather on shape by @hamzaqureshi5 in https://github.com/apache/tvm/pull/19968 * [Tests] Reduce redundant ONNX and PyTorch integration tests by @tlopex in https://github.com/apache/tvm/pull/20026 * [REFACTOR][TIRx] Keep AttrStmt node values unboxed by @tqchen in https://github.com/apache/tvm/pull/20030 * [CI] Bump tvm-ffi with compatible Python wrappers by @tqchen in https://github.com/apache/tvm/pull/20032 * [Vulkan] Fix SPIR-V 1.4+ entry-point interfaces by @wilx in https://github.com/apache/tvm/pull/20028 * [Web] Expose RNG state for deterministic restore by @akaashrp in https://github.com/apache/tvm/pull/20034 * [TIRx] Improve BufferStore cast warning context by @tlopex in https://github.com/apache/tvm/pull/20038 * [CI] Verify packed uint1 tvm-ffi revision by @tqchen in https://github.com/apache/tvm/pull/20041 * [Relax] Legalize grouped conv with symbolic channels by @guan404ming in https://github.com/apache/tvm/pull/20039 * [Fix][TIRx] Ignore statement spans in structural identity by @tlopex in https://github.com/apache/tvm/pull/20043 * [Web] Link custom allocator into WASM runtime by @akaashrp in https://github.com/apache/tvm/pull/20046 * [Relax][Frontend][ONNX] Support Shape start and end attributes by @napronald in https://github.com/apache/tvm/pull/20050 * [Fix][LLVM] Keep packed init callbacks local on Mach-O by @akaashrp in https://github.com/apache/tvm/pull/20052 * [Relax][Frontend][ONNX] Fix LpPool conversion by @napronald in https://github.com/apache/tvm/pull/20053 * [Fix][Relax] Return frontend tensor dtype value by @akaashrp in https://github.com/apache/tvm/pull/20051 * [Fix][TIRx] Handle vector access pointer addresses in C codegen by @tlopex in https://github.com/apache/tvm/pull/20058 * [FIX][TIRx] Use cluster arrivals for remote mbarrier views by @jinhongyii in https://github.com/apache/tvm/pull/20074 * [TVMSCRIPT][TIRx] Preserve parser source spans in IR by @jinhongyii in https://github.com/apache/tvm/pull/20073 * [FIX][TIRx] Make TilePrimitiveCall serializable by @jinhongyii in https://github.com/apache/tvm/pull/20071 * [FIX][TIRx] Preserve pointer expression types by @jinhongyii in https://github.com/apache/tvm/pull/20070 * [FIX][TIRx] Remap buffers consistently in ConvertSSA by @jinhongyii in https://github.com/apache/tvm/pull/20069 * [FIX][TIRx][CUDA] Fix tcgen05 register fragment layouts by @jinhongyii in https://github.com/apache/tvm/pull/20068 * [FIX][TIRx] Constant-fold copy slice extents by @jinhongyii in https://github.com/apache/tvm/pull/20067 * [FEATURE][TIRx][CUDA] Support TMEM datapath B by @jinhongyii in https://github.com/apache/tvm/pull/20075 * [TIRX] Represent buffers as typed variables by @tqchen in https://github.com/apache/tvm/pull/20079 * [Web] Batch GPU-to-GPU copies, fix WebGPU synchronization, and add tests for command batching by @akaashrp in https://github.com/apache/tvm/pull/20059 * [TIRx] tcgen05 dispatch paths, buffer dim-surgery views, FlashMLA lowering, and typed-buffer migration fixes by @spectrometerHBH in https://github.com/apache/tvm/pull/20080 * [TIRx] Flatten cuda/trn backend operator folder by @spectrometerHBH in https://github.com/apache/tvm/pull/20081 * [FFI] Bump tvm-ffi to latest Aug 3 by @tqchen in https://github.com/apache/tvm/pull/20083 * [BUILD] Migrate the Z3 dependency to mlc-z3-static by @Ubospica in https://github.com/apache/tvm/pull/20084 * [Relax][Test] Cover default GPU pipeline scheduling for R.power/elementwise kernels by @cchung100m in https://github.com/apache/tvm/pull/19923 * [FIX][TIRx] Use physical order for Buffer.local views by @jinhongyii in https://github.com/apache/tvm/pull/20076 * feat(lower-tirx): align NVIDIA IKET profiling with the official ABI by @spectrometerHBH in https://github.com/apache/tvm/pull/20085 * [FFI] Bump tvm-ffi to 0.1.13.post2 by @tqchen in https://github.com/apache/tvm/pull/20088 * tirx: represent buffer parameters with BufferType by @tqchen in https://github.com/apache/tvm/pull/20086 * [Web] Bump tvmjs version to 0.26.0 and apache-tvm-ffi floor to >=0.1.13.post2 by @MasterJH5574 in https://github.com/apache/tvm/pull/20093 * [Fix][TIRx] Fix MSVC build of IndexDataTypeNormalizer by @MasterJH5574 in https://github.com/apache/tvm/pull/20096
## New Contributors * @flashmouse made their first contribution in https://github.com/apache/tvm/pull/19683 * @ZephyrLi-pro made their first contribution in https://github.com/apache/tvm/pull/19776 * @yinli-systems made their first contribution in https://github.com/apache/tvm/pull/19818 * @mvanhorn made their first contribution in https://github.com/apache/tvm/pull/19816 * @Ga1axy0 made their first contribution in https://github.com/apache/tvm/pull/19915 * @napronald made their first contribution in https://github.com/apache/tvm/pull/19827 * @V-aerus made their first contribution in https://github.com/apache/tvm/pull/19937 * @viiccwen made their first contribution in https://github.com/apache/tvm/pull/19973 * @sbinabdullah made their first contribution in https://github.com/apache/tvm/pull/19978 * @Nanmur made their first contribution in https://github.com/apache/tvm/pull/19970 * @hamzaqureshi5 made their first contribution in https://github.com/apache/tvm/pull/19968 * @wilx made their first contribution in https://github.com/apache/tvm/pull/20028 **Full Changelog**: https://github.com/apache/tvm/compare/v0.25.0...v0.26.0 -- View it on GitHub: https://github.com/apache/tvm/releases/tag/v0.26.0 You are receiving this because you are subscribed to this thread. Message ID: <apache/tvm/releases/[email protected]>
