GlobalOptimization
-iree-global-opt-cleanup-numeric-narrowing
link
Cleans up any numeric narrowing ops inserted by iree-global-opt-infer-numeric-narrowing.
-iree-global-opt-convert-1x1-filter-conv2d-to-matmul
link
Convert linalg convolution ops with 1x1 kernels into linalg matrix multiplication ops.
-iree-global-opt-convert-strided-contraction-to-contraction
link
Factors out an extract_slice from contraction-like ops with strided inputs.
-iree-global-opt-data-layout-propagation
link
Propagate pack/unpack ops across other ops to improve fusion
-iree-global-opt-decompose-concat
link
Decomposes concatenations into a destination and a sequence of slice inserts.
Optionslink
-enable-concat-transposition : Allows transposing concatenations such that they occur on the inner most dims.
-iree-global-opt-demote-contraction-inputs-to-bf16
link
Demotes inputs (LHS, RHS) of linalg matmul-like ops from f32 to bf16.
Optionslink
-demote-only : Select the type of contraction ops to demote.
-iree-global-opt-detach-elementwise-from-named-ops
link
Detaches elementwise ops from named Linalg ops.
-iree-global-opt-erase-unused-linalg-operands
link
Erases unused linalg operand and remove dead code.
-iree-global-opt-expand-tensor-shapes
link
Expands tensor shape dimensions into SSA values across the program.
-iree-global-opt-fuse-dequantization-matmul
link
Fuses dequantization and matmul linalg.generic ops.
-iree-global-opt-fuse-silu-horizontal-matmul
link
Fuses matmul ops and silu linalg.generic op.
-iree-global-opt-generalize-linalg-named-ops
link
Convert some Linalg named ops into linalg.generics.
Optionslink
-enable-generalize-matmul : Convert linalg named opt to generic ops.
-iree-global-opt-infer-numeric-narrowing
link
Infers and inserts util.numeric.optional_narrow ops at points that may be beneficial.
-iree-global-opt-loop-invariant-code-motion
link
Hoist loop invariants out of loops with zero-trip-check.
-iree-global-opt-materialize-homogeneous-encodings
link
Materializes logical encodings to physical encodings if there is a single device target.
-iree-global-opt-optimize-numerics
link
Optimizes numerics given annotations added via iree-global-opt-infer-numeric-narrowing.
-iree-global-opt-propagate-linalg-transpose
link
Propagates linalg.transpose through a restricted set of ops.
Optionslink
-enable-aggressive-propagation : Enable aggressive propagation to named ops.
-test-sinking-only : Flag used for lit-testing sinking patterns only. Not for general usage
-test-bubbling-only : Flag used for lit-testing bubbling patterns only. Not for general usage
-enable-attention-v-transpose : Enable transposition of attention v operand
-enable-aggressive-propagation-through-conv : enable propagation through convolutions
-iree-global-opt-quantized-conv-to-conv
link
Lower quantized_conv to conv
-iree-global-opt-quantized-matmul-to-matmul
link
Lower quantized_matmul to matmul
-iree-global-opt-raise-special-ops
link
Raises special ops like softmax to the high level linalg.ext representation.
-iree-global-opt-remove-zero-extent-tensors
link
Removes tensors that have 0-extents.
-iree-global-opt-simplify-pack-unpack
link
Simplifies tensor pack and unpack ops.
-iree-global-opt-warn-on-uninitialized-values
link
Catch some uses of uninitialized values