document (need figure update)

Samsung · Sep 2, 2024 · 6fd4c47 · 6fd4c47
1 parent 355151b
commit 6fd4c47
Show file tree

Hide file tree

Showing 4 changed files with 3 additions and 54 deletions.
diff --git a/docs/howto/how-to-introduce-a-new-operation-into-runtime.md b/docs/howto/how-to-introduce-a-new-operation-into-runtime.md
@@ -9,12 +9,10 @@ onert support the operation.
 - [How To Introduce a New Operation Into Runtime](#how-to-introduce-a-new-operation-into-runtime)
   - [Index](#index)
   - [Core](#core)
-  - [Frontend](#frontend)
     - [Loaders](#loaders)
       - [Base Loader](#base-loader)
       - [TFLite Loader](#tflite-loader)
       - [Circle Loader](#circle-loader)
-    - [NNAPI](#nnapi)
   - [Backend](#backend)
     - [ShapeFixer](#shapefixer)
       - [acl_cl](#acl_cl)
@@ -232,12 +230,10 @@ void DynamicShapeInferer::visit(const ir::operation::Select &op)
 }
 ```
 
-## Frontend
-
-This module generates IR from a model. There are two kinds of frontend: Loader and NNAPI. First, Loader loads a model file and generates IR from it. Second, NNAPI generates IR from a model set via [Neural Networks API of android](https://developer.android.com/ndk/guides/neuralnetworks)
-
 ### Loaders
 
+This module generates IR from a model.
+
 #### Base Loader
 
 This is where the common parts of loaders are implemented.
@@ -275,41 +271,6 @@ If you want new operation to be loaded on only TFLite Loader, you only need to i
 This loads a circle file generated by the compiler.
 If you want new operation to be loaded on only Circle Loader, you only need to implement loading the operation here.
 
-### NNAPI
-
-1. Add to the OperationFactory to generate IR of new operation
-- [OperationFactory](/runtime/onert/api/nnapi/wrapper/OperationFactory.cc)
-
-```cpp
-  _map[ANEURALNETWORKS_SELECT] = [](const OperationFactory::Param &init_param, Operands &) {
-    assert(init_param.input_count == 3 && init_param.output_count == 1);
-
-    OperandIndexSequence outputs{init_param.outputs[0]};
-
-    // Each input should be interpreted as follows:
-    //
-    //  0 -> Cond Tensor Index
-    //  1 -> Input1 Tensor Index
-    //  2 -> Input2 Tensor Index
-    OperandIndexSequence inputs;
-    for (uint32_t n = 0; n < init_param.input_count; ++n)
-    {
-      inputs.append(OperandIndex{init_param.inputs[n]});
-    }
-
-    return new operation::Select{inputs, outputs};
-  };
-```
-
-2. If you want that NNAPI supports new operation of TFLite's model, you need to update the things related to the operation in [nnapi_delegate](/runtime/libs/tflite/port/1.13.1/src/nnapi_delegate.cpp) like below
-
-```cpp
-      case tflite::BuiltinOperator_SELECT:
-        nnapi_version = 12;  // require NNAPI 1.2
-        nn_op_type = ANEURALNETWORKS_SELECT;
-        break;
-```
-
 ## Backend
 
 This module generates kernels and tensors of backend such as [ComputeLibrary](https://github.com/ARM-software/ComputeLibrary/) from generated graph-based IR. For this, the runtime fairly works on it internally. But this is not enough because of dependence on backend. So, there are several components that require additional implementation on each backend.

diff --git a/docs/howto/how-to-use-nnapi-binding.md b/docs/howto/how-to-use-nnapi-binding.md
diff --git a/docs/howto/index.rst b/docs/howto/index.rst
@@ -21,7 +21,6 @@ How To
   ./how-to-run-package.md
   ./how-to-make-an-application-with-runtime.md
   ./how-to-use-api.md
-  ./how-to-use-nnapi-binding.md
   ./how-to-use-nnfw-api.md
   ./how-to-use-specific-backend.md
   ./how-to-contribute.md

diff --git a/docs/runtime/api.md b/docs/runtime/api.md
@@ -12,13 +12,7 @@ There are three parts - Frontend, Core and Backend. Core works with Frontend and
 
 Frontend API is about from creation/loading the model and
 
-Runtime supports two (frontend) APIs - NN API and NNFW API.
-
-### NN API
-
-NN API stands for Android Neural Networks API. It is part of Android Open Source Project and we provide a binding between NN API and One Runtime.
-
-For usage, refer to [Howto : NN API](../howto/how-to-use-nnapi-binding.md).
+Runtime supports NNFW API.
 
 ### NNFW API