Apache TVM v0.10.0 Release

#1

The Apache TVM community is happy to announce the release of TVM v0.10.0.

The TVM community has worked since the v0.10.0 release to deliver many exciting features and improvements.

  • Metaschedule
    • Software pipelining and padding for irregular shapes for auto tensorization
    • Stabilized and polished user-interfaces (e.g. database changes, tune_relay)
    • A new MLP-based cost model
  • TIR
    • New schedule primitive for PadEinsum
    • A new TIR node: DeclBuffer
    • INT8 Intrinsics for TensorCores for CUDA!
  • microTVM
    • Improved schedule primitives for ARM v8-m ISA

Check out the release notes on GitHub and the source download here.

1 Like
#2

Hi AndrewZhaoLuo, why is the v0.10.0 release version same as v0.10.0 rc0 version? https://github.com/apache/tvm/releases/download/v0.10.0/apache-tvm-src-v0.10.0.rc0.tar.gz

#3

rc0 is a “Release Candidate”, that we then vote on whether to release (https://lists.apache.org/thread/5qpcz06mw0xnnj6hdrm906xc71nfrw7t). If I recall correctly, the actual version is derived using git (https://github.com/apache/tvm/blob/main/version.py#L58) so no code change is required.

#4

Thanks for reply. Cheers!