onnxruntime
Remove empty (DQ -> Q -> graph output) sequence in TransposeOptimizer
#22172
Merged

Commits
  • Allow DQ->Q fusion into QNN Convert for same quantization type
    adrianlizarraga committed 1 year ago
  • Update TransposeOptimizer to also remove (DQ -> Q -> graph output) sequence
    adrianlizarraga committed 1 year ago
  • Undo use of double
    adrianlizarraga committed 1 year ago
  • Fix removal order
    adrianlizarraga committed 1 year ago
  • Add unit test for TransposeOptimizer
    adrianlizarraga committed 1 year ago
  • Merge branch 'main' into adrianl/handle-dq-q-at-graph-output
    adrianlizarraga committed 1 year ago
  • Comment-out code to save debug model
    adrianlizarraga committed 1 year ago
  • Run lintrunner
    adrianlizarraga committed 1 year ago
  • Move QNN-EP specific changes to a separate PR
    adrianlizarraga committed 1 year ago
  • Merge branch 'main' into adrianl/handle-dq-q-at-graph-output
    adrianlizarraga committed 1 year ago
  • Update onnxruntime/core/optimizer/transpose_optimization/onnx_transpose_optimization.cc
    adrianlizarraga committed 1 year ago
Loading