You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@tvm.apache.org by gi...@apache.org on 2025/09/22 05:03:28 UTC

(tvm) branch nightly updated (874de945ac -> 4c82c71933)

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 874de945ac Fixing datatype error for gpt-2 (#18328)
     add eb45a46d70 [CMake][Web] Install `web/` directory in cmake for Python package (#18327)
     add a54af64872 [Relax][Backend] Implement R.call_py_func operator for calling Python functions from compiled TVM (#18326)
     add 4c82c71933 [flashinfer] Support directing JIT to FlashInfer GroupedGemm kernels (#18325)

No new revisions were added by this update.

Summary of changes:
 CMakeLists.txt                                    |   3 +
 python/tvm/relax/backend/cuda/flashinfer.py       |  96 ++++-
 python/tvm/relax/base_py_module.py                |  38 ++
 src/relax/backend/vm/codegen_vm.cc                |   1 -
 src/relax/backend/vm/lower_runtime_builtin.cc     |  20 +
 src/runtime/vm/builtin.cc                         |  74 ++++
 tests/python/relax/test_base_py_module_printer.py |  96 ++---
 tests/python/relax/test_group_gemm_flashinfer.py  | 496 ++++++++++++++++++++++
 tests/python/relax/test_relax_operators.py        |  76 ++++
 9 files changed, 842 insertions(+), 58 deletions(-)
 create mode 100644 tests/python/relax/test_group_gemm_flashinfer.py