commits  

Messages by Thread

• (tvm) branch main updated: [TOPI][CUDA] Fix topk/sort gridDim overflow by remapping grid axes in sort_ir (#19900) tlopex
• (tvm) branch main updated: [Relax][cuDNN] Do not offload causal / non-fp16 attention, and fix the default softmax scale (#20078) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][Torch] Emit int64 indices for sort and argsort (#20254) tlopex
• (tvm-ffi) branch main updated: [Fix][Rust] Map children of inline pre-order replacements (#784) tqchen
• (tvm-ffi) branch main updated: [FEAT] Add optional in-place permission to default structural descent (#781) tqchen
• (tvm) branch dependabot/npm_and_yarn/web/js-yaml-3.15.2 created (now 3f66ecf4de) github-bot
• (tvm) branch main updated: [FIX][TIRx][CUDA] Support SM100 weight-stationary B collectors (#20329) tqchen
• (tvm-ffi) branch main updated: [FIX] Forward implicitly convertible values into UnchangedOr (#782) tqchen
• (tvm) branch main updated: [REFACTOR][IR] New Functor/Visitor/Mutator Infra at Base Layer (#20327) bohan
• (tvm-ffi) branch main updated: [PERF] Reduce structural traversal result overhead (#780) tqchen
• (tvm-ffi) branch main updated: [IR] Expose protected remap state and skip empty lookups (#779) tqchen
• (tvm) branch main updated: [Relax][PyTorch] Support aten.scatter.src in ExportedProgram importer (#19935) tlopex
• (tvm-ffi) branch main updated: [EXTRA] Enable typed mutation result conversions and subsumption (#777) tqchen
• (tvm-ffi) branch main updated: [EXTRA] Allow callers to disable unique in-place mutation (#776) tqchen
• svn commit: r87452 - dev/tvm/tvm-ffi-v0.1.14-rc5 yaxingcai
• (tvm-ffi) tag v0.1.14-rc5 created (now 70e5af34) yaxingcai
• (tvm) branch main updated: [Relax][TensorRT] Build and embed engines during code generation (#20301) tlopex
• (tvm) branch main updated: [Relax][ONNX] Add CastLike support and dynamic-k Trilu to expand backend coverage (#19898) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@9fbdd94c3b67a81a7edc88182aa4d28412e8b4e2) tqchen
• (tvm) branch main updated: [Fix][Relax][Frontend][TFLite] Correct quantized SSD inference (#20291) tlopex
• (tvm) branch dependabot/npm_and_yarn/web/baseline-browser-mapping-2.11.22 created (now e99c6c4282) github-bot
• (tvm-ffi) branch main updated: [PERF] Forward single StructuralMutate callbacks directly (#774) tqchen
• (tvm) branch main updated: [Relax][VM] Improve diagnostics for unlowered Relax operators (#19899) tlopex
• (tvm) branch main updated: [TIRx][Schedule] Support rfactor for arg reducers selecting last index (#19909) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@e4e489b788d08bf96fada95d221aa7c8a5fb566e) tqchen
• (tvm) branch main updated: [TOPI] Use branchless boundary index for reflect/replicate pad (#19928) tlopex
• (tvm) branch main updated: [REFACTOR][TIR] Remove IRTransform in favour of tvm_ffi.structural_map (#20304) tqchen
• (tvm) branch main updated: [Fix][Relax][Frontend][Torch] Fix `torch.round(x, decimals)` via `from_exported_program` and negative-decimals rounding (#20239) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][Torch] Support `aten.diagonal` from decomposed repeated-subscript einsum (#20237) tlopex
• (tvm) branch main updated: [Relax] Exclude R.null_value()-bound vars from KillAfterLastUse (#20267) tlopex
• (tvm) branch main updated: [Fix][Codegen] Avoid extraneous parentheses in if_then_else generated code (#20285) tlopex
• (tvm) branch main updated: [Relax] Preserve out_dtype in AdjustMatmulOrder (#20296) tlopex
• (tvm) branch main updated: [Fix][Relax] Canonicalize strided slice begin indices (#20284) tlopex
• (tvm) branch main updated: [REFACTOR][TIR] Inline StructuralWalk at variable-use checks (#20306) tqchen
• (tvm) branch main updated: [REFACTOR][TIR] Use StructuralWalk in place of PostOrderVisit (#20308) tqchen
• (tvm-ffi) tag v0.1.14-rc3 deleted (was 786a9b22) yaxingcai
• (tvm-ffi) branch main updated: [REFACTOR] Type default structural mutation results as UnchangedOr<Any> (#773) tqchen
• (tvm) branch main updated: [REFACTOR][IR] Add structural hooks to remaining Type and Relax Expr nodes (#20302) tqchen
• (tvm-ffi) tag v0.1.14-rc2 deleted (was bdd4989f) yaxingcai
• (tvm-ffi) tag v0.1.14-rc4 created (now c5e636bb) yaxingcai
• (tvm) branch main updated: [REFACTOR][Arith] Delete IntGroupBounds::Substitute (#20305) tqchen
• (tvm) branch main updated: [REFACTOR][ARITH] Use StructuralWalk and Map in arith (#20300) tqchen
• (tvm) branch main updated: [Fix][TIRx] Fix buffer lifetime in LowerWarpMemory (#20295) tlopex
• (tvm) branch main updated: [FFI] Bump tvm-ffi for StructuralMap policy (#20299) tqchen
• (tvm-ffi) branch main updated: [EXTRA] Simplify StructuralMap engine policy (#771) tqchen
• (tvm-ffi) branch main updated: Revert "[Rust][Fix] Preserve object lifetime and thread-safety contracts" (#772) tqchen
• (tvm-ffi) branch revert-766-fix/rust-stubgen-object-contracts created (now 494386f8) tlopex
  • (tvm-ffi) 01/01: Revert "[Rust][Fix] Preserve object lifetime and thread-safety contracts (#766)" tlopex
• (tvm) branch main updated: [TIR] Ignore None-valued pragma annotations (#20265) tlopex
• (tvm-ffi) branch main updated: [FFI][REFACTOR] Introduce UnchangedOr for StructuralMutate (#768) tqchen
• (tvm-ffi) branch main updated: [Rust][Fix] Preserve object lifetime and thread-safety contracts (#766) tlopex
• (tvm-ffi) branch main updated: [FFI][REFACTOR] Clarify DefRegionKind (#769) tqchen
• (tvm) branch main updated: [Fix][Relax] Preserve identity permute_dims in AdjustMatmulOrder (#20287) tlopex
• (tvm) branch main updated: [Fix][LLVM] Preserve 64-bit AllocBuffer extents (#20141) tlopex
• (tvm-ffi) tag v0.1.14-rc3 created (now 786a9b22) yaxingcai
• (tvm) branch main updated: [Relax][ONNX] Import Min/Max/Sum/Mean when an input has no static shape (#20288) tlopex
• (tvm) branch main updated: [Relax][ONNX] Preserve bool dtype when folding constant comparisons (#20286) tlopex
• (tvm) branch main updated: [S-TIR][Test] Use tvm.testing.main() so schedule tests can run standalone (#20283) tlopex
• (tvm) branch main updated: [Test] Pin CUDA round ties-to-even on exact midpoints (#20274) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][Torch] Validate flatten dims in `from_fx` (#20245) tlopex
• (tvm-ffi) branch main updated: [FFI][PERF] AnyUnsafe::CheckAnyViewStrict for a borrowed AnyView (#767) tlopex
• (tvm-ffi) branch main updated: [FFI][FIX] Use int64_t for the parsed TVM_TRACEBACK_LIMIT (#765) tlopex
• (tvm-ffi) branch main updated: [FFI][PERF] Force-inline ObjectPtr::reset and WeakObjectPtr::reset (#764) tqchen
• (tvm-ffi) branch main updated: [FFI][FIX] Backtrace: refresh DbgHelp modules, tolerate bad TVM_TRACEBACK_LIMIT, match object_is_not_callable exactly (#763) tqchen
• (tvm-ffi) branch main updated: [FFI][REFACTOR] Remove TVM_FFI_S_VISIT_RETURN_NONE (#762) tqchen
• (tvm-ffi) branch main updated: [FFI][PERF] Force-inline ObjectPtr wrapper special members (#761) tqchen
• (tvm-ffi) branch main updated: [FFI][PERF] Strip cv in IsObjectInstance so as<Object>() folds to true (#760) tqchen
• (tvm-ffi) branch main updated: [FFI][PERF] Mark non-trivial special members TVM_FFI_INLINE (#759) tlopex
• (tvm-ffi) branch main updated: [PERF][EXTRA] Split lazy structural mutation loops (#756) tqchen
• (tvm-ffi) branch main updated: [STUBGEN][RUST] Skip crate-provided types and add a custom constructor example (#755) tlopex
• (tvm-ffi) branch main updated: [WIN] Enhance threadsafety of DbgHelp (#754) tqchen
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Add TVM_FFI_S_VISIT_RETURN_NONE for visit hook tails (#753) tqchen
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Add StructuralMutate, a callback-driven mutation primitive (#751) tqchen
• (tvm-ffi) branch main updated: [STUBGEN] Let a file declare a namespace with a `prefix` directive (#752) tqchen
• (tvm) branch main updated: [TIRx][Test] Gate collector-A-only block-scale MMA test on CUDA 13.4 (#20276) bohan
• (tvm-ffi) branch main updated: [STUBGEN][RUST] Error on dependencies a partial run does not ask for (#748) tlopex
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Add StructuralVisit, a callback-driven traversal primitive (#750) tqchen
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Allow StructuralMapMutator to carry state via Parent (#749) tqchen
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Allow StructuralWalkVisitor to carry state via Parent (#747) tqchen
• (tvm-ffi) branch main updated: [STUBGEN] Add a `--check` mode that reports stale stub files (#746) tqchen
• (tvm-ffi) branch main updated: [Test][Rust] Use existing C++ registrations for structural walk and map (#745) tqchen
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Move structural-mutate error construction out of the hot path (#744) tqchen
• (tvm-ffi) branch main updated: [STUBGEN][RUST] Rename the wrapper slot to `base` and tighten `rust_ident` (#742) tqchen
• (tvm-ffi) branch main updated: [FIX][RUST] Wait for fixture registration in structural visit tests (#743) tqchen
• (tvm) branch main updated: [TIRx][Test] Gate tcgen05.mma collector certification on nvcc >= 13.4 (#20270) bohan
• (tvm) branch main updated: [CUDA][TIRx] Allow collector-A-only SM107 block-scale MMA and open cp.async wait_group immediates (#20271) bohan
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Optimize StructuralMap engine (#741) tqchen
• (tvm-ffi) branch main updated: [STUBGEN][RUST] Generate complete object mirrors and allocators (#740) tlopex
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Streamline container mutation hooks (#739) tqchen
• (tvm-ffi) branch main updated: [STUBGEN][RUST] Add the Rust backend with opaque object bindings (#738) tlopex
• (tvm) branch main updated: [CUDA][TIRx] Add collector-qualified tcgen05.mma block_scale forms and bind tcgen05.ld.red redval as output (#20266) bohan
• (tvm-ffi) branch main updated: [STUBGEN] Classify native layouts from reflected byte facts (#730) tlopex
• (tvm-ffi) branch main updated: [RUST] Add borrowed object node cast (#734) tqchen
• (tvm) branch main updated: [Relax][Frontend][ONNX] Support dynamic Range bounds (#20109) tlopex
• (tvm) branch main updated: [Fix][Metal] Bound symbolic stack allocations (#20236) tlopex
• (tvm-ffi) branch main updated: [STUBGEN] Generalize one-line directive handling (#736) tlopex
• svn commit: r87157 - dev/tvm/tvm-ffi-v0.1.14-rc2 yaxingcai
• (tvm-ffi) branch main updated: [REFACTOR][EXTRA] Optimize StructuralMap (#733) yaxingcai
• (tvm-ffi) tag v0.1.14-rc1 deleted (was 79697974) yaxingcai
• (tvm-ffi) tag v0.1.14-rc0 deleted (was b1936099) yaxingcai
• (tvm-ffi) tag v0.1.14-rc2 created (now bdd4989f) yaxingcai
• (tvm-ffi) branch rust-borrowed-object-cast created (now 0adb5a54) tlopex
  • (tvm-ffi) 01/01: [RUST] Add borrowed object node cast tlopex
• (tvm-ffi) branch main updated: [FFI] Make Object::unique() require no weak references (#731) tqchen
• (tvm) branch dependabot/npm_and_yarn/web/browserslist-4.28.8 created (now 2a049df922) github-bot
• (tvm-ffi) branch main updated: [FIX] Support free-threaded Python cross-builds (#706) tqchen
• (tvm-ffi) branch main updated: [CORE] Publish native layout traits from ObjectDef (#721) tlopex
• (tvm-ffi) branch main updated: [REFACTOR][RUST] Keep mutate dispatch state on the dispatcher (#728) tqchen
• (tvm) branch main updated: [CUDA] Add NVIDIA Jetson AGX Thor target tag (#20259) mshr
• (tvm-ffi) branch main updated: [STUBGEN] Carry native layout facts through the language-neutral data. (#720) tqchen
• svn commit: r87137 - dev/tvm/tvm-ffi-v0.1.14-rc1 yaxingcai
• (tvm-ffi) tag v0.1.14-rc1 created (now 79697974) yaxingcai
• (tvm) branch main updated: [Fix][Relax][Metal] Constrain wide-head prefill tiling (#20235) tlopex
• (tvm) branch main updated: [Fix][S-TIR] Preserve general reduction predicates (#20242) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][Torch] Honor the `dtype` argument of `aten.mean` (`torch.Tensor.mean` / `torch.mean`) (#20241) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@c46018d37018d547078fc1dcdfda5eebfa5916a8) tqchen
  • (tvm-site) branch asf-site updated: deploying docs (apache/tvm@c46018d37018d547078fc1dcdfda5eebfa5916a8) tqchen
• (tvm-ffi) branch rust-mutator-owned-state updated (b8af1b12 -> 7b85be09) tlopex
• (tvm-ffi) branch rust-mutator-owned-state created (now b8af1b12) tlopex
  • (tvm-ffi) 01/01: [REFACTOR][RUST] Keep mutate dispatch state on the dispatcher tlopex
• (tvm) branch main updated: [BugFix][Arith] Reject padded IterMapSimplify fallback (#20169) tlopex
• (tvm) branch main updated: [REFACTOR][TIRx] Make BufferRegion a typed expression (#20256) tqchen
• (tvm) branch dependabot/pip/docker/python/pip-26.2 created (now 72151c8f96) github-bot
• (tvm-ffi) branch main updated: [FEAT][FFI] Accept flexible callback return types in C++ StructuralMap (#725) tqchen
• (tvm) branch main updated: [Fix][Relax][Frontend][PyTorch] Fix `x.split(int)` with a non-divisible `split_size` (#20240) tlopex
• (tvm-ffi) tag v0.1.14-rc0 created (now b1936099) yaxingcai
• svn commit: r87119 - dev/tvm/tvm-ffi-v0.1.14-rc0 yaxingcai
• (tvm) branch main updated: [REFACTOR][IR] Make expression subscription eager and remove SubscriptProxy (#20257) tqchen
• (tvm) branch main updated: [Test] Run round ties-to-even test on every backend that implements it (#20252) tlopex
• (tvm) branch main updated: [TIRx][CUDA] Preserve local mbarrier predicate and count (#20250) tqchen
• (tvm-ffi) branch main updated: [Rust] Decouple structural mutation dispatch from recursion control (#723) tqchen
• (tvm) branch main updated: [FIX][IR] Complete lazy subscript realization (#20251) tqchen
• (tvm) branch main updated: [REFACTOR][IR] Consolidate expression operator overloading into the base layer (#20248) tqchen
• (tvm) branch main updated: [BugFix] Align default (C) tirx.round lowering to ties-to-even (#20131) tlopex
• (tvm-ffi) branch main updated: [FIX][RUST] Seal structural callback result traits and update return-value docs (#724) tlopex
• (tvm-ffi) branch rust-mutate-context-dispatch updated (6f5debf8 -> 9609fb4a) tlopex
• (tvm-ffi) branch rust-mutate-context-dispatch updated (e45f0264 -> 6f5debf8) tlopex
• (tvm-ffi) branch rust-mutate-context-dispatch updated (ba9883ec -> e45f0264) tlopex
• (tvm-ffi) branch rust-mutate-context-dispatch created (now ba9883ec) tlopex
  • (tvm-ffi) 01/01: [FEAT][RUST] Pass mutation context to typed dispatch callbacks tlopex
• (tvm) branch main updated: [REFACTOR][IR] Unify expression subscription realization (#20246) tqchen
• (tvm-ffi) branch main updated: [FEAT][RUST] Accept any Into<Any> return value in structural map/mutate callbacks (#722) tqchen
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@81bb3dcfe6bcacee62a9639558fea5bd6eb4cc90) tqchen
  • (tvm-site) branch asf-site updated: deploying docs (apache/tvm@81bb3dcfe6bcacee62a9639558fea5bd6eb4cc90) tqchen
• (tvm) branch main updated: [REFACTOR][TIR] Split masked buffer access into special calls (#20244) tqchen
• (tvm-ffi) branch main updated: [CORE] Propagate static object type bounds in Any conversions (#719) tqchen
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Validate Flatten axis range in `from_onnx` (#20145) tlopex
• (tvm) branch main updated: [Fix][Relax][Torch] Materialize runtime scalar shape values (#20138) tlopex
• (tvm) branch main updated: [Fix][TIRx] Preserve index semantics when narrowing to int32 (#20129) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Support Pad mode="wrap" and axes input for op… (#20152) tlopex
• (tvm) branch main updated: [Fix][Relax] Preserve match-cast storage liveness (#20220) tlopex
• (tvm) branch main updated: [Relax][Frontend][ONNX] Support symbolic shapes in Min/Max broadcast (#20218) tlopex
• (tvm-ffi) branch main updated: [FEAT][RUST] Add object binding runtime support (#716) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@66efb01242610bfdfc7338d10afd0f739e8e445e) tqchen
• (tvm) branch main updated: [CUDA][TIRx] Preserve device state during cleanup and skip invalid Top-K references (#20233) bohan
• (tvm-ffi) branch main updated: [REFACTOR][RUST] Clarify structural mutator dispatch and recursion APIs (#714) tqchen
• (tvm-ffi) branch main updated: [FIX] Stabilize Python annotation metadata (#718) tqchen
• (tvm-ffi) branch main updated: [FEAT][RUST] Support Any elements in FFI containers (#715) tqchen
• (tvm) branch main updated: [REFACTOR][TE] Represent tensor loads with opaque callees (#20225) tqchen
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@46bc8c6a38e41b55fbc41f796e77773cb7b4f887) tqchen
• (tvm) branch main updated: [TIRx][CUDA] Allow launch bounds with required block size (#20226) bohan
• (tvm) branch main updated: [Fix][Arith] Give each materialized Z3 solver a private context (#20221) tqchen
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@3aa0eb16f70e42a7ed716b2e5a0642eda342b3fe) tqchen
• (tvm) branch main updated: [Fix][Relax] Normalize negative indices in Gather/Scatter/OneHot Ops (#20219) tlopex
• (tvm) branch main updated: [Fix][TIRx] Restore IterVar span reflection (#20214) tlopex
• (tvm) branch main updated: [Fix][Relax] Skip ReorderPermuteDimsAfterConcat for unknown-rank inputs (#20216) tlopex
• (tvm) branch main updated: [Fix][Relax][ONNX] Where: broadcast size-1 shape expressions, materialize ShapeExpr inputs (#20210) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@0d9c9384b1473e2a2e7d84f077f151bb3b9130bc) tqchen
• (tvm) branch main updated: [TIRx][CUDA] Support exact required block dimensions (#20223) hongyij
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@fb00a9bdb4e1b439eaf1781ec487276bbdc9cf64) tqchen
• (tvm) branch main updated: [Fix][Support] Use sbsa-linux CUDA include dir on ARM64 Linux (#20222) hongyij
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@ddea9c3a609d274700f44fccede90f8c9254ee37) tqchen
• (tvm) branch main updated: [Perf][Arith] Materialize Z3 solvers lazily on first query (#20215) tqchen
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@9e5e709662050f68eccf27e7bb0e3f8f2c19b0ab) tqchen
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Fix Softplus accuracy loss from hardcoded threshold (#20212) tlopex
• (tvm) branch main updated: [Fix][DLight] Localize private scalar reduction buffers (#20160) tlopex
• (tvm) branch main updated: [BugFix][Relax] Skip parallel matmul fusion for mixed output dtypes (#20208) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Support Shape outputs as Gather indices (#20179) tlopex
• (tvm) branch main updated: [Fix][Relax][ONNX] Correct fmod mapping in Mod constant folding (#20170) tlopex
• (tvm) branch main updated: [BugFix][Relax] Preserve take mode in ReorderTakeAfterMatmul (#20206) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@46c37fc523a2ae751272cc131376633cb502908b) tqchen
• (tvm) branch main updated: [Fix][Relax] Preserve tensor-derived symbols during fusion (#20139) tlopex
• (tvm) branch main updated: [Fix][S-TIR][DLight] Guard non-affine reduction write-back (#20057) tlopex
• (tvm-site) branch asf-site updated: deploying docs (apache/tvm@b4e62497dc76461f1365e71df13bd6da2e90fb66) tqchen
• (tvm) branch main updated: fix(tirx): stabilize multi-GPU correctness tests (#20213) hongyij
• (tvm) branch main updated: [Fix][Relax] Raise error on non-unit dim ONNX Squeeze axis (#20188) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Fix Scatter with indices smaller than data (#20187) tlopex
• (tvm) branch main updated: [Web] Decode packed BF16 tensor records in place (#20167) akaashrp
• (tvm-ffi) branch main updated: [FEAT][RUST] Add AnyCompatible::to_any. (#712) tqchen
• (tvm) branch main updated: [Codegen][LLVM] Add LLVM 23 compatibility (#20189) tqchen
• (tvm) branch main updated: [Feat][Web] Support per-parameter tensor cache encoding (#20136) akaashrp
• (tvm) branch main updated: [Fix][Arith] Preserve nested floormod semantics (#20181) ruihangl
• (tvm) branch main updated: [Fix][Relax][Torch] Align retained expand dimensions by trailing rank (#20137) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Support broadcastable multi-axis PRelu slopes (#20149) tlopex
• (tvm) branch main updated: [Runtime] Add PagedAttentionKVCache checkpoint primitives (#20035) tlopex
• (tvm) branch main updated: [Fix][Relax][Frontend][ONNX] Fix Mean/Sum/Min/Max with all-constant inputs (#20147) tlopex
• (tvm) branch main updated: [TIRx][CUDA] Preserve explicit single-CTA cluster launches (#20180) bohan
• (tvm) branch main updated: [Fix][Relax] Lower non-contiguous WebGPU cumsum (#20133) tlopex
• (tvm) branch main updated: [Fix][Relax] Run destructors for non-trivially-destructible types in Arena (#20163) tlopex
• (tvm-ffi) branch main updated: [FIX] Make ffi.Error picklable (#709) tqchen

• Earlier messages
• Later messages