This is an automated email from the ASF dual-hosted git repository.
github-actions[bot] pushed a change to branch nightly
in repository https://gitbox.apache.org/repos/asf/tvm.git
from a8f1aced2f [FIX] Skip metal target tag registration for unsupported
LLVM CPUs (#19427)
add 7eea6df1b6 [Relax][FRONTEND][ONNX] Support Softmax, LogSoftmax and
Hardmax when opset version ≤12 (#19428)
add a6625cbb94 [Fix][CUDA] Version compatibility of CUDA symbols (#19432)
add 2b87313c98 [ARITH] Expose allow_override parameter in Python
Analyzer.bind() (#19417)
add 2c72b902d7 [Relax][NN] Use int64 for RoPE apply flag (#19430)
No new revisions were added by this update.
Summary of changes:
python/tvm/arith/analyzer.py | 12 +-
.../relax/frontend/nn/llm/position_embedding.py | 2 +-
python/tvm/relax/frontend/onnx/onnx_frontend.py | 165 ++++++++++++++++++++-
src/arith/analyzer.cc | 5 +-
src/runtime/contrib/cublas/cublas.cc | 10 ++
src/runtime/vm/paged_kv_cache.cc | 2 +-
tests/python/arith/test_arith_simplify.py | 14 ++
tests/python/relax/test_frontend_onnx.py | 51 +++++++
8 files changed, 247 insertions(+), 14 deletions(-)
