From afc18650dcbe18b76d2ff44bcec31061da9abb60 Mon Sep 17 00:00:00 2001 From: Teresa Charlin Date: Fri, 20 Nov 2020 13:08:42 +0000 Subject: IVGCVSW-5347 Update Readme for 20.11 * Adding delegate readme.md and TensorFlowLiteDelegateSupport.md Change-Id: I1b8012440cf4cd6120902ad69c5b3a2a5e410d71 Signed-off-by: Teresa Charlin --- delegate/README.md | 7 +++ delegate/TensorFlowLiteDelegateSupport.md | 85 ++++++++++++++++++++++++++ src/armnnDeserializer/DeserializerSupport.md | 1 + src/armnnSerializer/SerializerSupport.md | 1 + src/armnnTfLiteParser/TensorFlowLiteSupport.md | 2 + 5 files changed, 96 insertions(+) create mode 100644 delegate/README.md create mode 100644 delegate/TensorFlowLiteDelegateSupport.md diff --git a/delegate/README.md b/delegate/README.md new file mode 100644 index 0000000000..7430f19c00 --- /dev/null +++ b/delegate/README.md @@ -0,0 +1,7 @@ +# The Arm NN TensorFlow Lite delegate + +'armnnDelegate' is a library for accelerating certain TensorFlow Lite operators on Arm hardware by providing +the TensorFlow Lite interpreter with an alternative implementation of the operators via its delegation mechanism. + +For more information about the TensorFlow Lite operators that are supported, +see [TensorFlowLiteDelegateSupport.md](./TensorFlowLiteDelegateSupport.md). diff --git a/delegate/TensorFlowLiteDelegateSupport.md b/delegate/TensorFlowLiteDelegateSupport.md new file mode 100644 index 0000000000..b1b39f616b --- /dev/null +++ b/delegate/TensorFlowLiteDelegateSupport.md @@ -0,0 +1,85 @@ +# TensorFlow Lite operators that the Arm NN TensorFlow Lite Delegate supports + +This reference guide provides a list of TensorFlow Lite operators the Arm NN SDK currently supports. + +## Fully supported + +The Arm NN SDK TensorFlow Lite delegate currently supports the following operators: + +* ABS + +* ADD + +* AVERAGE_POOL_2D, Supported Fused Activation: RELU , RELU6 , TANH, NONE + +* CONCATENATION, Supported Fused Activation: RELU , RELU6 , TANH, NONE + +* CONV_2D, Supported Fused Activation: RELU , RELU6 , TANH, NONE + +* DEPTHWISE_CONV_2D, Supported Fused Activation: RELU , RELU6 , TANH, NONE + +* DEQUANTIZE + +* DIV + +* EQUAL + +* EXP + +* FULLY_CONNECTED, Supported Fused Activation: RELU , RELU6 , TANH, NONE + +* GREATER + +* GREATER_OR_EQUAL + +* LESS + +* LESS_OR_EQUAL + +* LOGISTIC + +* LOG_SOFTMAX + +* L2_POOL_2D + +* MAXIMUM + +* MAX_POOL_2D, Supported Fused Activation: RELU , RELU6 , TANH, NONE + +* MEAN + +* MINIMUM + +* MUL + +* NEG + +* NOT_EQUAL + +* QUANTIZE + +* RESHAPE + +* RESIZE_BILINEAR + +* RESIZE_NEAREST_NEIGHBOR + +* RELU + +* RELU6 + +* RSQRT + +* SOFTMAX + +* SQRT + +* SUB + +* TANH + +* TRANSPOSE + +* TRANSPOSE_CONV + +More machine learning operators will be supported in future releases. diff --git a/src/armnnDeserializer/DeserializerSupport.md b/src/armnnDeserializer/DeserializerSupport.md index 4e2ead41bc..2ff3a24503 100644 --- a/src/armnnDeserializer/DeserializerSupport.md +++ b/src/armnnDeserializer/DeserializerSupport.md @@ -29,6 +29,7 @@ The Arm NN SDK Deserialize parser currently supports the following layers: * Input * InstanceNormalization * L2Normalization +* Logical * LogSoftmax * Lstm * Maximum diff --git a/src/armnnSerializer/SerializerSupport.md b/src/armnnSerializer/SerializerSupport.md index 438335341a..67dc5d1596 100644 --- a/src/armnnSerializer/SerializerSupport.md +++ b/src/armnnSerializer/SerializerSupport.md @@ -28,6 +28,7 @@ The Arm NN SDK Serializer currently supports the following layers: * Input * InstanceNormalization * L2Normalization +* Logical * LogSoftmax * Lstm * Maximum diff --git a/src/armnnTfLiteParser/TensorFlowLiteSupport.md b/src/armnnTfLiteParser/TensorFlowLiteSupport.md index 9718b22798..faad3d0fa9 100644 --- a/src/armnnTfLiteParser/TensorFlowLiteSupport.md +++ b/src/armnnTfLiteParser/TensorFlowLiteSupport.md @@ -104,6 +104,8 @@ Arm tested these operators with the following TensorFlow Lite neural network: * FSRCNN +* EfficientNet-lite + * RDN converted from [TensorFlow model](https://github.com/hengchuan/RDN-TensorFlow) * Quantized RDN (CpuRef) -- cgit v1.2.1