diff options
Diffstat (limited to 'src/armnnSerializer/SerializerSupport.md')
| -rw-r--r-- | src/armnnSerializer/SerializerSupport.md | 14 |
1 files changed, 8 insertions, 6 deletions
diff --git a/src/armnnSerializer/SerializerSupport.md b/src/armnnSerializer/SerializerSupport.md index 2def918f21..4b12f9ca23 100644 --- a/src/armnnSerializer/SerializerSupport.md +++ b/src/armnnSerializer/SerializerSupport.md @@ -20,11 +20,9 @@ The Arm NN SDK Serializer currently supports the following layers: * Dequantize * DetectionPostProcess * Division -* Equal * Floor * FullyConnected * Gather -* Greater * Input * InstanceNormalization * L2Normalization @@ -45,7 +43,6 @@ The Arm NN SDK Serializer currently supports the following layers: * QuantizedLstm * Reshape * Resize -* ResizeBilinear * Rsqrt * Slice * Softmax @@ -60,6 +57,11 @@ The Arm NN SDK Serializer currently supports the following layers: More machine learning layers will be supported in future releases. -Note: Merger layer has been renamed Concat. Serializations of the old - format with Merger layers will deserialize to Concat layers to - maintain backward compatibility. +## Deprecated layers + +Some layers have been deprecated and replaced by others layers. In order to maintain backward compatibility, serializations of these deprecated layers will deserialize to the layers that have replaced them, as follows: + +* Equal will deserialize as Comparison +* Merger will deserialize as Concat +* Greater will deserialize as Comparison +* ResizeBilinear will deserialize as Resize |