diff options
| author | Charles Xu <charles.xu@arm.com> | 2020-05-13 10:15:26 +0200 |
|---|---|---|
| committer | Tim Hall <tim.hall@arm.com> | 2020-06-18 17:53:52 +0100 |
| commit | 78792223369fa34dacd0e69e189af035283da2ae (patch) | |
| tree | ac3826df5528866319fd65d7a99eef8e87cd4084 /ethosu/vela/graph_optimiser.py | |
| parent | 620d88c60482bad4d96da4d32cc4cca5561cca9e (diff) | |
| download | ethos-u-vela-78792223369fa34dacd0e69e189af035283da2ae.tar.gz | |
Add elementwise vector scalars support
Write the constant scalars into flash. In case it's Dram
or OffChipFlash, DMA the scalars from flash to sram.
Signed-off-by: Charles Xu <charles.xu@arm.com>
Change-Id: I42300a05dfe968d623b8aec8549644549e0f54b5
Diffstat (limited to 'ethosu/vela/graph_optimiser.py')
| -rw-r--r-- | ethosu/vela/graph_optimiser.py | 13 |
1 files changed, 13 insertions, 0 deletions
diff --git a/ethosu/vela/graph_optimiser.py b/ethosu/vela/graph_optimiser.py index 913b9a6a..351716e0 100644 --- a/ethosu/vela/graph_optimiser.py +++ b/ethosu/vela/graph_optimiser.py @@ -25,6 +25,7 @@ from .data_type import DataType from .operation import NpuBlockType from .operation import Operation from .tensor import Tensor +from .numeric_util import full_shape passthrough_nodes = set(("Identity",)) @@ -313,6 +314,7 @@ fc_op = set( depthwise_op = set(("DepthwiseConv2dNative", "DepthwiseConv2dBiasAct",)) pool_op = set(("AvgPool", "MaxPool", "QuantizedAvgPool", "QuantizedMaxPool", "AvgPoolAct", "MaxPoolAct", "ResizeBilinear",)) elementwise_op = set(("AddAct", "MulAct", "SubAct", "Maximum", "Minimum", "LeakyRelu", "Abs")) +binary_elementwise_op = set(("AddAct", "MulAct", "SubAct", "Maximum", "Minimum")) activation_ops = set(("Relu", "Relu6", "ReluN1To1", "Sigmoid", "Tanh")) memory_only_ops = set(("Reshape",)) @@ -399,6 +401,16 @@ def fixup_act_reorder(op, arch): op.type = "Identity" return op +def fixup_elementwise_with_scalars(op, arch): + if op.type in binary_elementwise_op: + ifm_tensor, ifm2_tensor, _, ofm_tensor = op.get_ifm_ifm2_weights_ofm() + if ifm2_tensor.shape != [] and ifm_tensor.shape != []: + diff = len(ifm_tensor.shape) - len(ifm2_tensor.shape) + if diff > 0: + ifm2_tensor.shape = full_shape(len(ifm_tensor.shape), ifm2_tensor.shape, 1) + elif diff < 0: + ifm_tensor.shape = full_shape(len(ifm2_tensor.shape), ifm_tensor.shape, 1) + return op # Set input/output tensor equivalence to the same id for memory operations def set_tensor_equivalence(op, arch): @@ -492,6 +504,7 @@ def optimise_graph_a(nng, arch, verbose_graph=False): fixup_act_reorder, add_padding_fields, mark_npu_block_type, + fixup_elementwise_with_scalars, # convert_mul_max_to_abs_or_lrelu # TODO: enable optimisation once quantisation issues are resolved ] |