MLBEDSW-5102 Update removal of memory only operators
Memory only operators such as Reshape, Squeeze and ExpandDims are
removed in the graph optimiser step.
- Added semantic check that memory only operators have same
quantisation parameters on ifm/ofm.
- Added support for the ExpandDims operator.
- Addition and cleanup of related unit tests.
- Removed TOSA from the generated SUPPORTED_OPS.md documentation.
Signed-off-by: Jonas Ohlsson <jonas.ohlsson@arm.com>
Change-Id: If848d8afc58c18806e10997ed94e4dae83f30879
diff --git a/ethosu/vela/tflite_supported_operators.py b/ethosu/vela/tflite_supported_operators.py
index 016d44e..933302f 100644
--- a/ethosu/vela/tflite_supported_operators.py
+++ b/ethosu/vela/tflite_supported_operators.py
@@ -86,7 +86,7 @@
)
split_ops = set((Op.Split, Op.SplitV, Op.StridedSlice, Op.Slice, Op.UnpackReshaped, Op.Unpack,))
concat_ops = set((Op.Concat, Op.ConcatTFLite, Op.PackReshaped, Op.Pack,))
- memory_only_ops = set((Op.Reshape, Op.QuantizedReshape, Op.Squeeze,)) | concat_ops | split_ops
+ memory_only_ops = set((Op.Reshape, Op.QuantizedReshape, Op.Squeeze, Op.ExpandDims,)) | concat_ops | split_ops
per_axis_quant_ops = convolution_like_ops # per-axis/channel quantization only currently supported for conv ops
supported_fused_activations = relu_ops | set((Op.Tanh, Op.Sigmoid, Op.LUT,))
supported_operators = npu_pre_ops | mac_main_ops | elem_wise_main_ops | pad_ops | npu_post_ops | memory_only_ops