experiments
/
STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Update Hello World example. 

PiperOrigin-RevId: 310263402
Change-Id: I921176f6a8dc4c76bd45e6a508548d3b1936f89d
This commit is contained in:
Meghna Natraj 2020-05-06 17:44:37 -07:00 committed by TensorFlower Gardener
parent fac30b7a87
commit 48e57d17bd
19 changed files with 4008 additions and 1738 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tensorflow/lite/g3doc/microcontrollers/build_convert.md
View File

@ -71,7 +71,7 @@ important to change the array declaration to `const` for better memory
efficiency on embedded platforms.
For an example of how to include and use a model in your program, see
[`sine_model_data.cc`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/sine_model_data.cc)
[`model.cc`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/model.cc)
in the *Hello World* example.
## Model architecture and training

61
tensorflow/lite/g3doc/microcontrollers/get_started.md
View File

@ -86,12 +86,10 @@ World README.md</a>
The following section walks through the *Hello World* example's
[`hello_world_test.cc`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/hello_world_test.cc),
which demonstrates how to run inference using TensorFlow Lite for
Microcontrollers.
unit test which demonstrates how to run inference using TensorFlow Lite for
Microcontrollers. It loads the model and runs inference several times.
The test loads the model and then uses it to run inference several times.
### Include the library headers
### 1. Include the library headers
To use the TensorFlow Lite for Microcontrollers library, we must include the
following header files:
@ -116,22 +114,20 @@ following header files:
- [`version.h`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/version.h)
provides versioning information for the TensorFlow Lite schema.
### Include the model
### 2. Include the model header
The TensorFlow Lite for Microcontrollers interpreter expects the model to be
provided as a C++ array. In the *Hello World* example, the model is defined in
`sine_model_data.h` and `sine_model_data.cc`. The header is included with the
following line:
provided as a C++ array. The model is defined in `model.h` and `model.cc` files.
The header is included with the following line:
```C++
#include "tensorflow/lite/micro/examples/hello_world/sine_model_data.h"
#include "tensorflow/lite/micro/examples/hello_world/model.h"
```
### Set up the unit test
### 3. Include the unit test framework header
The code we are walking through is a unit test that uses the TensorFlow Lite for
Microcontrollers unit test framework. To load the framework, we include the
following file:
In order to create a unit test, we include the TensorFlow Lite for
Microcontrollers unit test framework by including the following line:
```C++
#include "tensorflow/lite/micro/testing/micro_test.h"
@ -143,11 +139,16 @@ The test is defined using the following macros:
TF_LITE_MICRO_TESTS_BEGIN
TF_LITE_MICRO_TEST(LoadModelAndPerformInference) {
. // add code here
.
}
TF_LITE_MICRO_TESTS_END
```
The remainder of the code demonstrates how to load the model and run inference.
We now discuss the code included in the macro above.
### Set up logging
### 4. Set up logging
To set up logging, a `tflite::ErrorReporter` pointer is created using a pointer
to a `tflite::MicroErrorReporter` instance:
@ -162,14 +163,14 @@ logs. Since microcontrollers often have a variety of mechanisms for logging, the
implementation of `tflite::MicroErrorReporter` is designed to be customized for
your particular device.
### Load a model
### 5. Load a model
In the following code, the model is instantiated using data from a `char` array,
`g_sine_model_data`, which is declared in `sine_model_data.h`. We then check the
model to ensure its schema version is compatible with the version we are using:
`g_model`, which is declared in `model.h`. We then check the model to ensure its
schema version is compatible with the version we are using:
```C++
const tflite::Model* model = ::tflite::GetModel(g_sine_model_data);
const tflite::Model* model = ::tflite::GetModel(g_model);
if (model->version() != TFLITE_SCHEMA_VERSION) {
TF_LITE_REPORT_ERROR(error_reporter,
"Model provided is schema version %d not equal "
@ -178,7 +179,7 @@ if (model->version() != TFLITE_SCHEMA_VERSION) {
}
```
### Instantiate operations resolver
### 6. Instantiate operations resolver
An
[`AllOpsResolver`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/kernels/all_ops_resolver.h)
@ -198,7 +199,7 @@ This is done using a different class, `MicroMutableOpResolver`. You can see how
to use it in the *Micro speech* example's
[`micro_speech_test.cc`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/micro_speech/micro_speech_test.cc).
### Allocate memory
### 7. Allocate memory
We need to preallocate a certain amount of memory for input, output, and
intermediate arrays. This is provided as a `uint8_t` array of size
@ -212,7 +213,7 @@ uint8_t tensor_arena[tensor_arena_size];
The size required will depend on the model you are using, and may need to be
determined by experimentation.
### Instantiate interpreter
### 8. Instantiate interpreter
We create a `tflite::MicroInterpreter` instance, passing in the variables
created earlier:
@ -222,7 +223,7 @@ tflite::MicroInterpreter interpreter(model, resolver, tensor_arena,
tensor_arena_size, error_reporter);
```
### Allocate tensors
### 9. Allocate tensors
We tell the interpreter to allocate memory from the `tensor_arena` for the
model's tensors:
@ -231,7 +232,7 @@ model's tensors:
interpreter.AllocateTensors();
```
### Validate input shape
### 10. Validate input shape
The `MicroInterpreter` instance can provide us with a pointer to the model's
input tensor by calling `.input(0)`, where `0` represents the first (and only)
@ -265,7 +266,7 @@ The enum value `kTfLiteFloat32` is a reference to one of the TensorFlow Lite
data types, and is defined in
[`common.h`](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/c/common.h).
### Provide an input value
### 11. Provide an input value
To provide an input to the model, we set the contents of the input tensor, as
follows:
@ -276,7 +277,7 @@ input->data.f[0] = 0.;
In this case, we input a floating point value representing `0`.
### Run the model
### 12. Run the model
To run the model, we can call `Invoke()` on our `tflite::MicroInterpreter`
instance:
@ -300,7 +301,7 @@ successfully run.
TF_LITE_MICRO_EXPECT_EQ(kTfLiteOk, invoke_status);
```
### Obtain the output
### 12. Obtain the output
The model's output tensor can be obtained by calling `output(0)` on the
`tflite::MicroInterpreter`, where `0` represents the first (and only) output
@ -327,7 +328,7 @@ float value = output->data.f[0];
TF_LITE_MICRO_EXPECT_NEAR(0., value, 0.05);
```
### Run inference again
### 13. Run inference again
The remainder of the code runs inference several more times. In each instance,
we assign a value to the input tensor, invoke the interpreter, and read the
@ -350,7 +351,7 @@ value = output->data.f[0];
TF_LITE_MICRO_EXPECT_NEAR(-0.959, value, 0.05);
```
### Read the application code
### 14. Read the application code
Once you have walked through this unit test, you should be able to understand
the example's application code, located in

10
tensorflow/lite/micro/examples/hello_world/BUILD
View File

@ -16,12 +16,12 @@ package(default_visibility = ["//visibility:public"])
licenses(["notice"]) # Apache 2.0
cc_library(
name = "sine_model_data",
name = "model",
srcs = [
"sine_model_data.cc",
"model.cc",
],
hdrs = [
"sine_model_data.h",
"model.h",
],
build_for_embedded = True,
copts = micro_copts(),
@ -33,9 +33,9 @@ tflite_micro_cc_test(
"hello_world_test.cc",
],
deps = [
":model",
"//tensorflow/lite:schema_fbs_version",
"//tensorflow/lite/micro:micro_framework",
"//tensorflow/lite/micro/examples/hello_world:sine_model_data",
"//tensorflow/lite/micro/kernels:all_ops_resolver",
"//tensorflow/lite/micro/kernels:micro_ops",
"//tensorflow/lite/micro/testing:micro_test",
@ -83,10 +83,10 @@ cc_binary(
],
deps = [
":constants",
":model",
":output_handler",
"//tensorflow/lite:schema_fbs_version",
"//tensorflow/lite/micro:micro_framework",
"//tensorflow/lite/micro/examples/hello_world:sine_model_data",
"//tensorflow/lite/micro/kernels:all_ops_resolver",
"//tensorflow/lite/schema:schema_fbs",
],

8
tensorflow/lite/micro/examples/hello_world/Makefile.inc
View File

@ -1,9 +1,9 @@
HELLO_WORLD_TEST_SRCS := \
tensorflow/lite/micro/examples/hello_world/hello_world_test.cc \
tensorflow/lite/micro/examples/hello_world/sine_model_data.cc
tensorflow/lite/micro/examples/hello_world/model.cc
HELLO_WORLD_TEST_HDRS := \
tensorflow/lite/micro/examples/hello_world/sine_model_data.h
tensorflow/lite/micro/examples/hello_world/model.h
OUTPUT_HANDLER_TEST_SRCS := \
tensorflow/lite/micro/examples/hello_world/output_handler_test.cc \
@ -16,12 +16,12 @@ tensorflow/lite/micro/examples/hello_world/constants.h
HELLO_WORLD_SRCS := \
tensorflow/lite/micro/examples/hello_world/main.cc \
tensorflow/lite/micro/examples/hello_world/main_functions.cc \
tensorflow/lite/micro/examples/hello_world/sine_model_data.cc \
tensorflow/lite/micro/examples/hello_world/model.cc \
tensorflow/lite/micro/examples/hello_world/output_handler.cc \
tensorflow/lite/micro/examples/hello_world/constants.cc
HELLO_WORLD_HDRS := \
tensorflow/lite/micro/examples/hello_world/sine_model_data.h \
tensorflow/lite/micro/examples/hello_world/model.h \
tensorflow/lite/micro/examples/hello_world/output_handler.h \
tensorflow/lite/micro/examples/hello_world/constants.h \
tensorflow/lite/micro/examples/hello_world/main_functions.h

40
tensorflow/lite/micro/examples/hello_world/README.md
View File

@ -1,41 +1,32 @@
# Hello World example
# Hello World Example
This example is designed to demonstrate the absolute basics of using [TensorFlow
Lite for Microcontrollers](https://www.tensorflow.org/lite/microcontrollers).
It includes the full end-to-end workflow of training a model, converting it for
use with TensorFlow Lite, and running inference on a microcontroller.
use with TensorFlow Lite for Microcontrollers for running inference on a
microcontroller.
The sample is built around a model trained to replicate a `sine` function. It
contains implementations for several platforms. In each case, the model is used
to generate a pattern of data that is used to either blink LEDs or control an
animation.
The model is trained to replicate a `sine` function and generates a pattern of
data to either blink LEDs or control an animation, depending on the capabilities
of the device.
![Animation of example running on STM32F746](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/images/STM32F746.gif)
![Animation on STM32F746](images/animation_on_STM32F746.gif)
## Table of contents
- [Understand the model](#understand-the-model)
- [Deploy to Arduino](#deploy-to-arduino)
- [Deploy to ESP32](#deploy-to-esp32)
- [Deploy to SparkFun Edge](#deploy-to-sparkfun-edge)
- [Deploy to STM32F746](#deploy-to-STM32F746)
- [Run the tests on a development machine](#run-the-tests-on-a-development-machine)
## Understand the model
The sample comes with a pre-trained model. The code used to train and convert
the model is available as a tutorial in [create_sine_model.ipynb](https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/create_sine_model.ipynb).
Walk through this tutorial to understand what the model does,
how it works, and how it was converted for use with TensorFlow Lite for
Microcontrollers.
- [Train your own model](#train-your-own-model)
## Deploy to Arduino
The following instructions will help you build and deploy this sample
to [Arduino](https://www.arduino.cc/) devices.
![Animation of example running on Arduino MKRZERO](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/images/arduino_mkrzero.gif)
![Animation on Arduino MKRZERO](images/animation_on_arduino_mkrzero.gif)
The sample has been tested with the following devices:
@ -132,7 +123,7 @@ idf.py --port /dev/ttyUSB0 flash monitor
The following instructions will help you build and deploy this sample on the
[SparkFun Edge development board](https://sparkfun.com/products/15170).
![Animation of example running on SparkFun Edge](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/images/sparkfun_edge.gif)
![Animation on SparkFun Edge](images/animation_on_sparkfun_edge.gif)
If you're new to using this board, we recommend walking through the
[AI on a microcontroller with TensorFlow Lite and SparkFun Edge](https://codelabs.developers.google.com/codelabs/sparkfun-tensorflow)
@ -272,7 +263,7 @@ The following instructions will help you build and deploy the sample to the
[STM32F7 discovery kit](https://os.mbed.com/platforms/ST-Discovery-F746NG/)
using [ARM Mbed](https://github.com/ARMmbed/mbed-cli).
![Animation of example running on STM32F746](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/images/STM32F746.gif)
![Animation on STM32F746](images/animation_on_STM32F746.gif)
Before we begin, you'll need the following:
@ -400,7 +391,14 @@ the trained TensorFlow model, runs some example inputs through it, and got the
expected outputs.
To understand how TensorFlow Lite does this, you can look at the source in
[hello_world_test.cc](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/lite/micro/examples/hello_world/hello_world_test.cc).
[hello_world_test.cc](hello_world_test.cc).
It's a fairly small amount of code that creates an interpreter, gets a handle to
a model that's been compiled into the program, and then invokes the interpreter
with the model and sample inputs.
### Train your own model
So far you have used an existing trained model to run inference on
microcontrollers. If you wish to train your own model, follow the instructions
given in the [train/](train/) directory.

1333
tensorflow/lite/micro/examples/hello_world/create_sine_model.ipynb
View File

File diff suppressed because one or more lines are too long

20
tensorflow/lite/micro/examples/hello_world/hello_world_test.cc
View File

@ -14,7 +14,7 @@ limitations under the License.
==============================================================================*/
// #include "tensorflow/lite/c/common.h"
#include "tensorflow/lite/micro/examples/hello_world/sine_model_data.h"
#include "tensorflow/lite/micro/examples/hello_world/model.h"
#include "tensorflow/lite/micro/kernels/all_ops_resolver.h"
#include "tensorflow/lite/micro/micro_error_reporter.h"
#include "tensorflow/lite/micro/micro_interpreter.h"
@ -31,7 +31,7 @@ TF_LITE_MICRO_TEST(LoadModelAndPerformInference) {
// Map the model into a usable data structure. This doesn't involve any
// copying or parsing, it's a very lightweight operation.
const tflite::Model* model = ::tflite::GetModel(g_sine_model_data);
const tflite::Model* model = ::tflite::GetModel(g_model);
if (model->version() != TFLITE_SCHEMA_VERSION) {
TF_LITE_REPORT_ERROR(error_reporter,
"Model provided is schema version %d not equal "
@ -43,8 +43,13 @@ TF_LITE_MICRO_TEST(LoadModelAndPerformInference) {
tflite::ops::micro::AllOpsResolver resolver;
// Create an area of memory to use for input, output, and intermediate arrays.
// `arena_used_bytes` can be used to retrieve the optimal size.
const int tensor_arena_size = 2208 + 16 + 100 /* some reserved space */;
// Minimum arena size, at the time of writing. After allocating tensors
// you can retrieve this value by invoking interpreter.arena_used_bytes().
const int model_arena_size = 2352;
/* Extra headroom for model + alignment + future interpreter changes */
const int extra_arena_size = 560 + 16 + 100;
const int tensor_arena_size = model_arena_size + extra_arena_size;
uint8_t tensor_arena[tensor_arena_size];
// Build an interpreter to run the model with
@ -53,11 +58,10 @@ TF_LITE_MICRO_TEST(LoadModelAndPerformInference) {
// Allocate memory from the tensor_arena for the model's tensors
TF_LITE_MICRO_EXPECT_EQ(interpreter.AllocateTensors(), kTfLiteOk);
// At the time of writing, the hello world model uses 2208 bytes, we leave
// 100 bytes head room here to make the test less fragile and in the same
// time, alert for substantial increase.
TF_LITE_MICRO_EXPECT_LE(interpreter.arena_used_bytes(), 2208 + 100);
// Alert for substantial increase in arena size usage.
TF_LITE_MICRO_EXPECT_LE(interpreter.arena_used_bytes(),
model_arena_size + 100);
// Obtain a pointer to the model's input tensor
TfLiteTensor* input = interpreter.input(0);

0
tensorflow/lite/micro/examples/hello_world/images/STM32F746.gif → tensorflow/lite/micro/examples/hello_world/images/animation_on_STM32F746.gif
View File

Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 292 KiB

After

Width:  |  Height:  |  Size: 292 KiB

0
tensorflow/lite/micro/examples/hello_world/images/arduino_mkrzero.gif → tensorflow/lite/micro/examples/hello_world/images/animation_on_arduino_mkrzero.gif
View File

Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 529 KiB

After

Width:  |  Height:  |  Size: 529 KiB

0
tensorflow/lite/micro/examples/hello_world/images/sparkfun_edge.gif → tensorflow/lite/micro/examples/hello_world/images/animation_on_sparkfun_edge.gif
View File

Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 625 KiB

After

Width:  |  Height:  |  Size: 625 KiB

BIN
tensorflow/lite/micro/examples/hello_world/images/model_architecture.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 89 KiB

6
tensorflow/lite/micro/examples/hello_world/main_functions.cc
View File

@ -1,4 +1,4 @@
/* Copyright 2019 The TensorFlow Authors. All Rights Reserved.
/* Copyright 2020 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
@ -16,8 +16,8 @@ limitations under the License.
#include "tensorflow/lite/micro/examples/hello_world/main_functions.h"
#include "tensorflow/lite/micro/examples/hello_world/constants.h"
#include "tensorflow/lite/micro/examples/hello_world/model.h"
#include "tensorflow/lite/micro/examples/hello_world/output_handler.h"
#include "tensorflow/lite/micro/examples/hello_world/sine_model_data.h"
#include "tensorflow/lite/micro/kernels/all_ops_resolver.h"
#include "tensorflow/lite/micro/micro_error_reporter.h"
#include "tensorflow/lite/micro/micro_interpreter.h"
@ -49,7 +49,7 @@ void setup() {
// Map the model into a usable data structure. This doesn't involve any
// copying or parsing, it's a very lightweight operation.
model = tflite::GetModel(g_sine_model_data);
model = tflite::GetModel(g_model);
if (model->version() != TFLITE_SCHEMA_VERSION) {
TF_LITE_REPORT_ERROR(error_reporter,
"Model provided is schema version %d not equal "

250
tensorflow/lite/micro/examples/hello_world/model.cc Normal file
View File

@ -0,0 +1,250 @@
/* Copyright 2020 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
// Automatically created from a TensorFlow Lite flatbuffer using the command:
// xxd -i model.tflite > model.cc
// This is a standard TensorFlow Lite model file that has been converted into a
// C data array, so it can be easily compiled into a binary for devices that
// don't have a file system.
// See train/README.md for a full description of the creation process.
#include "tensorflow/lite/micro/examples/hello_world/model.h"
// We need to keep the data array aligned on some architectures.
#ifdef __has_attribute
#define HAVE_ATTRIBUTE(x) __has_attribute(x)
#else
#define HAVE_ATTRIBUTE(x) 0
#endif
#if HAVE_ATTRIBUTE(aligned) || (defined(__GNUC__) && !defined(__clang__))
#define DATA_ALIGN_ATTRIBUTE __attribute__((aligned(4)))
#else
#define DATA_ALIGN_ATTRIBUTE
#endif
const unsigned char g_model[] DATA_ALIGN_ATTRIBUTE = {
0x1c, 0x00, 0x00, 0x00, 0x54, 0x46, 0x4c, 0x33, 0x00, 0x00, 0x12, 0x00,
0x1c, 0x00, 0x04, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x14, 0x00,
0x00, 0x00, 0x18, 0x00, 0x12, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00,
0x60, 0x09, 0x00, 0x00, 0xa8, 0x02, 0x00, 0x00, 0x90, 0x02, 0x00, 0x00,
0x3c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x04, 0x00, 0x08, 0x00,
0x08, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00, 0x0b, 0x00, 0x00, 0x00,
0x13, 0x00, 0x00, 0x00, 0x6d, 0x69, 0x6e, 0x5f, 0x72, 0x75, 0x6e, 0x74,
0x69, 0x6d, 0x65, 0x5f, 0x76, 0x65, 0x72, 0x73, 0x69, 0x6f, 0x6e, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x48, 0x02, 0x00, 0x00, 0x34, 0x02, 0x00, 0x00,
0x0c, 0x02, 0x00, 0x00, 0xfc, 0x00, 0x00, 0x00, 0xac, 0x00, 0x00, 0x00,
0x8c, 0x00, 0x00, 0x00, 0x3c, 0x00, 0x00, 0x00, 0x34, 0x00, 0x00, 0x00,
0x2c, 0x00, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00, 0x1c, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0xfe, 0xfd, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x05, 0x00, 0x00, 0x00, 0x31, 0x2e, 0x35, 0x2e, 0x30, 0x00, 0x00, 0x00,
0x7c, 0xfd, 0xff, 0xff, 0x80, 0xfd, 0xff, 0xff, 0x84, 0xfd, 0xff, 0xff,
0x88, 0xfd, 0xff, 0xff, 0x22, 0xfe, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xfc, 0x04, 0x00, 0x00,
0x9f, 0x0a, 0x00, 0x00, 0x65, 0x06, 0x00, 0x00, 0x3d, 0xf8, 0xff, 0xff,
0x00, 0x00, 0x00, 0x00, 0xeb, 0x0a, 0x00, 0x00, 0x2f, 0xf8, 0xff, 0xff,
0xe8, 0x04, 0x00, 0x00, 0x21, 0x0a, 0x00, 0x00, 0x46, 0xfe, 0xff, 0xff,
0xc8, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xa3, 0xf7, 0xff, 0xff,
0x28, 0xf9, 0xff, 0xff, 0x9a, 0x05, 0x00, 0x00, 0x6e, 0xfe, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x73, 0x1c, 0x11, 0xe1,
0x0c, 0x81, 0xa5, 0x43, 0xfe, 0xd5, 0xd5, 0xb2, 0x60, 0x77, 0x19, 0xdf,
0x8a, 0xfe, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x51, 0x0b, 0x00, 0x00, 0x47, 0xf6, 0xff, 0xff,
0x00, 0x00, 0x00, 0x00, 0x1c, 0x0c, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x9b, 0x05, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0xe7, 0x20, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x92, 0x07, 0x00, 0x00, 0xf4, 0xf4, 0xff, 0xff, 0x55, 0xf0, 0xff, 0xff,
0x00, 0x00, 0x00, 0x00, 0xd6, 0xfe, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x00, 0x01, 0x00, 0x00, 0xee, 0xfc, 0x00, 0xec, 0x05, 0x16, 0xef, 0xec,
0xe6, 0xf8, 0x03, 0x01, 0x00, 0xfa, 0xf8, 0xf5, 0xda, 0xeb, 0x27, 0x14,
0xef, 0xde, 0xe2, 0xda, 0xf0, 0xdf, 0x32, 0x06, 0x01, 0xe6, 0xee, 0xf9,
0x00, 0x16, 0x07, 0xe0, 0xfe, 0xff, 0xe9, 0x05, 0xe7, 0xef, 0x81, 0x1b,
0x18, 0xea, 0xca, 0x01, 0x0f, 0x00, 0xdb, 0xf7, 0x0e, 0xec, 0x12, 0x1e,
0x04, 0x13, 0xb2, 0xe7, 0xfd, 0x06, 0xbb, 0xe0, 0x0c, 0xec, 0xf0, 0xdf,
0xeb, 0xf7, 0x05, 0x26, 0x19, 0xe4, 0x70, 0x1a, 0xea, 0x1e, 0x34, 0xdf,
0x19, 0xf3, 0xf1, 0x19, 0x0e, 0x03, 0x1b, 0xe1, 0xde, 0x13, 0xf6, 0x19,
0xff, 0xf6, 0x1a, 0x17, 0xf1, 0x1c, 0xdb, 0x1a, 0x1a, 0x20, 0xe6, 0x19,
0xf5, 0xff, 0x97, 0x0b, 0x00, 0x00, 0xce, 0xdf, 0x0d, 0xf7, 0x15, 0xe4,
0xed, 0xfc, 0x0d, 0xe9, 0xfb, 0xec, 0x5c, 0xfc, 0x1d, 0x02, 0x58, 0xe3,
0xe0, 0xf4, 0x15, 0xec, 0xf9, 0x00, 0x13, 0x05, 0xec, 0x0c, 0x1c, 0x14,
0x0c, 0xe9, 0x0a, 0xf4, 0x18, 0x00, 0xd7, 0x05, 0x27, 0x02, 0x15, 0xea,
0xea, 0x02, 0x9b, 0x00, 0x0c, 0xfa, 0xe9, 0xea, 0xfe, 0x01, 0x14, 0xfd,
0x0b, 0x02, 0xf0, 0xef, 0x06, 0xee, 0x01, 0x0d, 0x06, 0xe7, 0xf7, 0x11,
0xf5, 0x0a, 0xf9, 0xf1, 0x23, 0xff, 0x0d, 0xf2, 0xec, 0x11, 0x26, 0x1d,
0xf2, 0xea, 0x28, 0x18, 0xe0, 0xfb, 0xf3, 0xf4, 0x05, 0x1c, 0x1d, 0xfb,
0xfd, 0x1e, 0xfc, 0x11, 0xe8, 0x06, 0x09, 0x03, 0x12, 0xf2, 0x35, 0xfb,
0xdd, 0x1b, 0xf9, 0xef, 0xf3, 0xe7, 0x6f, 0x0c, 0x1d, 0x00, 0x43, 0xfd,
0x0d, 0xf1, 0x0a, 0x19, 0x1a, 0xfa, 0xe0, 0x18, 0x1e, 0x13, 0x37, 0x1c,
0x12, 0xec, 0x3a, 0x0c, 0xb6, 0xcb, 0xe6, 0x13, 0xf7, 0xeb, 0xf1, 0x05,
0x1b, 0xfa, 0x19, 0xe5, 0xec, 0xcf, 0x0c, 0xf4, 0xe2, 0xff, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x21, 0xa2, 0x8c, 0xc9,
0x5f, 0x1d, 0xce, 0x41, 0x9f, 0xcd, 0x20, 0xb1, 0xdf, 0x53, 0x2f, 0x81,
0x00, 0x00, 0x06, 0x00, 0x08, 0x00, 0x04, 0x00, 0x06, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xe2, 0xee, 0xff, 0xff,
0x80, 0xff, 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00, 0x54, 0x4f, 0x43, 0x4f,
0x20, 0x43, 0x6f, 0x6e, 0x76, 0x65, 0x72, 0x74, 0x65, 0x64, 0x2e, 0x00,
0x01, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xbc, 0xf9, 0xff, 0xff,
0x48, 0x01, 0x00, 0x00, 0x3c, 0x01, 0x00, 0x00, 0x30, 0x01, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00, 0x04, 0x01, 0x00, 0x00,
0xb8, 0x00, 0x00, 0x00, 0x70, 0x00, 0x00, 0x00, 0x28, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x1a, 0xff, 0xff, 0xff, 0x02, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x0b, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0xca, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x08, 0x1c, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00, 0x04, 0x00, 0x04, 0x00,
0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00,
0x09, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0e, 0x00, 0x14, 0x00, 0x00, 0x00,
0x08, 0x00, 0x0c, 0x00, 0x07, 0x00, 0x10, 0x00, 0x0e, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x08, 0x1c, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0xba, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x01,
0x01, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00,
0x00, 0x00, 0x0e, 0x00, 0x16, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00,
0x07, 0x00, 0x10, 0x00, 0x0e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08,
0x24, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x00, 0x00, 0x06, 0x00, 0x08, 0x00, 0x07, 0x00, 0x06, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x01, 0x01, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x00, 0x10, 0x00, 0x04, 0x00,
0x08, 0x00, 0x0c, 0x00, 0x0a, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x0a, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x0b, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x0a, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0xdc, 0x04, 0x00, 0x00,
0x54, 0x04, 0x00, 0x00, 0xc4, 0x03, 0x00, 0x00, 0x54, 0x03, 0x00, 0x00,
0xd0, 0x02, 0x00, 0x00, 0x4c, 0x02, 0x00, 0x00, 0xe0, 0x01, 0x00, 0x00,
0x5c, 0x01, 0x00, 0x00, 0xd8, 0x00, 0x00, 0x00, 0x6c, 0x00, 0x00, 0x00,
0x3c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xd8, 0xff, 0xff, 0xff,
0x18, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00,
0x49, 0x64, 0x65, 0x6e, 0x74, 0x69, 0x74, 0x79, 0x00, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x0c, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x1c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x0d, 0x00, 0x00, 0x00, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32, 0x5f,
0x69, 0x6e, 0x70, 0x75, 0x74, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0xc2, 0xfb, 0xff, 0xff,
0x00, 0x00, 0x00, 0x02, 0x58, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x28, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xc4, 0xfc, 0xff, 0xff,
0x14, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0xba, 0x2b, 0x4f, 0x38, 0x20, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x34, 0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x5f,
0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x2a, 0xfc, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09,
0x6c, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x28, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x2c, 0xfd, 0xff, 0xff, 0x14, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0xb9, 0x36, 0x0b, 0x3c,
0x34, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x34,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x2f, 0x52, 0x65, 0x61, 0x64,
0x56, 0x61, 0x72, 0x69, 0x61, 0x62, 0x6c, 0x65, 0x4f, 0x70, 0x2f, 0x74,
0x72, 0x61, 0x6e, 0x73, 0x70, 0x6f, 0x73, 0x65, 0x00, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0xaa, 0xfc, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09, 0x6c, 0x00, 0x00, 0x00,
0x09, 0x00, 0x00, 0x00, 0x44, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x9c, 0xfc, 0xff, 0xff, 0x30, 0x00, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00,
0x18, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x80, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0xaa, 0x7b, 0xbe, 0x3b, 0x01, 0x00, 0x00, 0x00,
0x2e, 0xbd, 0xbd, 0x3f, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x19, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x33,
0x2f, 0x52, 0x65, 0x6c, 0x75, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x2a, 0xfd, 0xff, 0xff,
0x00, 0x00, 0x00, 0x02, 0x58, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00,
0x28, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x2c, 0xfe, 0xff, 0xff,
0x14, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0xe3, 0x04, 0x20, 0x39, 0x20, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x33, 0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x5f,
0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x92, 0xfd, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09,
0x6c, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x28, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x94, 0xfe, 0xff, 0xff, 0x14, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0xe8, 0x76, 0x51, 0x3c,
0x34, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x33,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x2f, 0x52, 0x65, 0x61, 0x64,
0x56, 0x61, 0x72, 0x69, 0x61, 0x62, 0x6c, 0x65, 0x4f, 0x70, 0x2f, 0x74,
0x72, 0x61, 0x6e, 0x73, 0x70, 0x6f, 0x73, 0x65, 0x00, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x12, 0xfe, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09, 0x6c, 0x00, 0x00, 0x00,
0x07, 0x00, 0x00, 0x00, 0x44, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x04, 0xfe, 0xff, 0xff, 0x30, 0x00, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00,
0x18, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x80, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0xd2, 0x91, 0x43, 0x3c, 0x01, 0x00, 0x00, 0x00,
0x40, 0xce, 0x42, 0x40, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x19, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32,
0x2f, 0x52, 0x65, 0x6c, 0x75, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x92, 0xfe, 0xff, 0xff,
0x00, 0x00, 0x00, 0x02, 0x5c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x2c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x94, 0xff, 0xff, 0xff,
0x18, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x28, 0xb3, 0xd9, 0x38, 0x20, 0x00, 0x00, 0x00,
0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31,
0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32, 0x2f, 0x4d, 0x61, 0x74,
0x4d, 0x75, 0x6c, 0x5f, 0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0xfe, 0xfe, 0xff, 0xff,
0x00, 0x00, 0x00, 0x09, 0x78, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00,
0x34, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x0c, 0x00,
0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x14, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0xd5, 0x6b, 0x8a, 0x3b, 0x34, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x32, 0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x2f,
0x52, 0x65, 0x61, 0x64, 0x56, 0x61, 0x72, 0x69, 0x61, 0x62, 0x6c, 0x65,
0x4f, 0x70, 0x2f, 0x74, 0x72, 0x61, 0x6e, 0x73, 0x70, 0x6f, 0x73, 0x65,
0x00, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x8a, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09,
0x60, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x7c, 0xff, 0xff, 0xff, 0x2c, 0x00, 0x00, 0x00,
0x20, 0x00, 0x00, 0x00, 0x14, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x80, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0x01, 0x00, 0x00, 0x00, 0x5d, 0x4f, 0xc9, 0x3c, 0x01, 0x00, 0x00, 0x00,
0x0e, 0x86, 0xc8, 0x40, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x12, 0x00, 0x00, 0x00, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32, 0x5f,
0x69, 0x6e, 0x70, 0x75, 0x74, 0x5f, 0x69, 0x6e, 0x74, 0x38, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x0e, 0x00, 0x18, 0x00, 0x08, 0x00, 0x07, 0x00, 0x0c, 0x00,
0x10, 0x00, 0x14, 0x00, 0x0e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09,
0x6c, 0x00, 0x00, 0x00, 0x0a, 0x00, 0x00, 0x00, 0x50, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x14, 0x00, 0x04, 0x00, 0x08, 0x00,
0x0c, 0x00, 0x10, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x30, 0x00, 0x00, 0x00,
0x24, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x1a, 0xde, 0x0a, 0x3c,
0x01, 0x00, 0x00, 0x00, 0x66, 0x64, 0x87, 0x3f, 0x01, 0x00, 0x00, 0x00,
0x13, 0x42, 0x8d, 0xbf, 0x0d, 0x00, 0x00, 0x00, 0x49, 0x64, 0x65, 0x6e,
0x74, 0x69, 0x74, 0x79, 0x5f, 0x69, 0x6e, 0x74, 0x38, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x3c, 0x00, 0x00, 0x00, 0x28, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x00, 0x0e, 0x00, 0x07, 0x00,
0x00, 0x00, 0x08, 0x00, 0x0a, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x06,
0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x06, 0x00, 0x06, 0x00, 0x05, 0x00,
0x06, 0x00, 0x00, 0x00, 0x00, 0x72, 0x0a, 0x00, 0x0c, 0x00, 0x07, 0x00,
0x00, 0x00, 0x08, 0x00, 0x0a, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09,
0x04, 0x00, 0x00, 0x00};
const int g_model_len = 2512;

20
tensorflow/lite/micro/examples/hello_world/sine_model_data.h → tensorflow/lite/micro/examples/hello_world/model.h
View File

@ -1,4 +1,4 @@
/* Copyright 2019 The TensorFlow Authors. All Rights Reserved.
/* Copyright 2020 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
@ -13,15 +13,19 @@ See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
// Automatically created from a TensorFlow Lite flatbuffer using the command:
// xxd -i model.tflite > model.cc
// This is a standard TensorFlow Lite model file that has been converted into a
// C data array, so it can be easily compiled into a binary for devices that
// don't have a file system. It was created using the command:
// xxd -i sine_model.tflite > sine_model_data.cc
// don't have a file system.
#ifndef TENSORFLOW_LITE_MICRO_EXAMPLES_HELLO_WORLD_SINE_MODEL_DATA_H_
#define TENSORFLOW_LITE_MICRO_EXAMPLES_HELLO_WORLD_SINE_MODEL_DATA_H_
// See train/README.md for a full description of the creation process.
extern const unsigned char g_sine_model_data[];
extern const int g_sine_model_data_len;
#ifndef TENSORFLOW_LITE_MICRO_EXAMPLES_HELLO_WORLD_MODEL_H_
#define TENSORFLOW_LITE_MICRO_EXAMPLES_HELLO_WORLD_MODEL_H_
#endif // TENSORFLOW_LITE_MICRO_EXAMPLES_HELLO_WORLD_SINE_MODEL_DATA_H_
extern const unsigned char g_model[];
extern const int g_model_len;
#endif // TENSORFLOW_LITE_MICRO_EXAMPLES_HELLO_WORLD_MODEL_H_

255
tensorflow/lite/micro/examples/hello_world/sine_model_data.cc
View File

@ -1,255 +0,0 @@
/* Copyright 2019 The TensorFlow Authors. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
// Automatically created from a TensorFlow Lite flatbuffer using the command:
// xxd -i sine_model.tflite > sine_model_data.cc
// See the README for a full description of the creation process.
#include "tensorflow/lite/micro/examples/hello_world/sine_model_data.h"
// We need to keep the data array aligned on some architectures.
#ifdef __has_attribute
#define HAVE_ATTRIBUTE(x) __has_attribute(x)
#else
#define HAVE_ATTRIBUTE(x) 0
#endif
#if HAVE_ATTRIBUTE(aligned) || (defined(__GNUC__) && !defined(__clang__))
#define DATA_ALIGN_ATTRIBUTE __attribute__((aligned(4)))
#else
#define DATA_ALIGN_ATTRIBUTE
#endif
const unsigned char g_sine_model_data[] DATA_ALIGN_ATTRIBUTE = {
0x18, 0x00, 0x00, 0x00, 0x54, 0x46, 0x4c, 0x33, 0x00, 0x00, 0x0e, 0x00,
0x18, 0x00, 0x04, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x14, 0x00,
0x0e, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x10, 0x0a, 0x00, 0x00,
0xb8, 0x05, 0x00, 0x00, 0xa0, 0x05, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x0b, 0x00, 0x00, 0x00, 0x90, 0x05, 0x00, 0x00, 0x7c, 0x05, 0x00, 0x00,
0x24, 0x05, 0x00, 0x00, 0xd4, 0x04, 0x00, 0x00, 0xc4, 0x00, 0x00, 0x00,
0x74, 0x00, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00, 0x1c, 0x00, 0x00, 0x00,
0x14, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x54, 0xf6, 0xff, 0xff, 0x58, 0xf6, 0xff, 0xff, 0x5c, 0xf6, 0xff, 0xff,
0x60, 0xf6, 0xff, 0xff, 0xc2, 0xfa, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x40, 0x00, 0x00, 0x00, 0x7c, 0x19, 0xa7, 0x3e, 0x99, 0x81, 0xb9, 0x3e,
0x56, 0x8b, 0x9f, 0x3e, 0x88, 0xd8, 0x12, 0xbf, 0x74, 0x10, 0x56, 0x3e,
0xfe, 0xc6, 0xdf, 0xbe, 0xf2, 0x10, 0x5a, 0xbe, 0xf0, 0xe2, 0x0a, 0xbe,
0x10, 0x5a, 0x98, 0xbe, 0xb9, 0x36, 0xce, 0x3d, 0x8f, 0x7f, 0x87, 0x3e,
0x2c, 0xb1, 0xfd, 0xbd, 0xe6, 0xa6, 0x8a, 0xbe, 0xa5, 0x3e, 0xda, 0x3e,
0x50, 0x34, 0xed, 0xbd, 0x90, 0x91, 0x69, 0xbe, 0x0e, 0xfb, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0x67, 0x41, 0x48, 0xbf,
0x24, 0xcd, 0xa0, 0xbe, 0xb7, 0x92, 0x0c, 0xbf, 0x00, 0x00, 0x00, 0x00,
0x98, 0xfe, 0x3c, 0x3f, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x4a, 0x17, 0x9a, 0xbe,
0x41, 0xcb, 0xb6, 0xbe, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x13, 0xd6, 0x1e, 0x3e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x5a, 0xfb, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00,
0x4b, 0x98, 0xdd, 0xbd, 0x40, 0x6b, 0xcb, 0xbe, 0x36, 0x0c, 0xd4, 0x3c,
0xbd, 0x44, 0xb5, 0x3e, 0x95, 0x70, 0xe3, 0x3e, 0xe7, 0xac, 0x86, 0x3e,
0x00, 0xc4, 0x4e, 0x3d, 0x7e, 0xa6, 0x1d, 0x3e, 0xbd, 0x87, 0xbb, 0x3e,
0xb4, 0xb8, 0x09, 0xbf, 0xa1, 0x1f, 0xf8, 0xbe, 0x8d, 0x90, 0xdd, 0x3e,
0xde, 0xfa, 0x6f, 0xbe, 0xb2, 0x75, 0xe4, 0x3d, 0x6e, 0xfe, 0x36, 0x3e,
0x20, 0x18, 0xc2, 0xbe, 0x39, 0xc7, 0xfb, 0xbe, 0xfe, 0xa4, 0x30, 0xbe,
0xf7, 0x91, 0xde, 0xbe, 0xde, 0xab, 0x24, 0x3e, 0xfb, 0xbb, 0xce, 0x3e,
0xeb, 0x23, 0x80, 0xbe, 0x7b, 0x58, 0x73, 0xbe, 0x9a, 0x2e, 0x03, 0x3e,
0x10, 0x42, 0xa9, 0xbc, 0x10, 0x12, 0x64, 0xbd, 0xe3, 0x8d, 0x0c, 0x3d,
0x9e, 0x48, 0x97, 0xbe, 0x34, 0x51, 0xd4, 0xbe, 0x02, 0x3b, 0x0d, 0x3e,
0x62, 0x67, 0x89, 0xbe, 0x74, 0xdf, 0xa2, 0x3d, 0xf3, 0x25, 0xb3, 0xbe,
0xef, 0x34, 0x7b, 0x3d, 0x61, 0x70, 0xe3, 0x3d, 0xba, 0x76, 0xc0, 0xbe,
0x7d, 0xe9, 0xa7, 0x3e, 0xc3, 0xab, 0xd0, 0xbe, 0xcf, 0x7c, 0xdb, 0xbe,
0x70, 0x27, 0x9a, 0xbe, 0x98, 0xf5, 0x3c, 0xbd, 0xff, 0x4b, 0x4b, 0x3e,
0x7e, 0xa0, 0xf8, 0xbd, 0xd4, 0x6e, 0x86, 0x3d, 0x00, 0x4a, 0x07, 0x3a,
0x4c, 0x24, 0x61, 0xbe, 0x54, 0x68, 0xf7, 0xbd, 0x02, 0x3f, 0x77, 0xbe,
0x23, 0x79, 0xb3, 0x3e, 0x1c, 0x83, 0xad, 0xbd, 0xc8, 0x92, 0x8d, 0x3e,
0xa8, 0xf3, 0x15, 0xbd, 0xe6, 0x4d, 0x6c, 0x3d, 0xac, 0xe7, 0x98, 0xbe,
0x81, 0xec, 0xbd, 0x3e, 0xe2, 0x55, 0x73, 0x3e, 0xc1, 0x77, 0xc7, 0x3e,
0x6e, 0x1b, 0x5e, 0x3d, 0x27, 0x78, 0x02, 0x3f, 0xd4, 0x21, 0x90, 0x3d,
0x52, 0xdc, 0x1f, 0x3e, 0xbf, 0xda, 0x88, 0x3e, 0x80, 0x79, 0xe3, 0xbd,
0x40, 0x6f, 0x10, 0xbe, 0x20, 0x43, 0x2e, 0xbd, 0xf0, 0x76, 0xc5, 0xbd,
0xcc, 0xa0, 0x04, 0xbe, 0xf0, 0x69, 0xd7, 0xbe, 0xb1, 0xfe, 0x64, 0xbe,
0x20, 0x41, 0x84, 0xbe, 0xb2, 0xc3, 0x26, 0xbe, 0xd8, 0xf4, 0x09, 0xbe,
0x64, 0x44, 0xd1, 0x3d, 0xd5, 0xe1, 0xc8, 0xbe, 0x35, 0xbc, 0x3f, 0xbe,
0xc0, 0x94, 0x82, 0x3d, 0xdc, 0x2b, 0xb1, 0xbd, 0x02, 0xdb, 0xbf, 0xbe,
0xa5, 0x7f, 0x8a, 0x3e, 0x21, 0xb4, 0xa2, 0x3e, 0xcd, 0x86, 0x56, 0xbf,
0x9c, 0x3b, 0x76, 0xbc, 0x85, 0x6d, 0x60, 0xbf, 0x86, 0x00, 0x3c, 0xbe,
0xc1, 0x23, 0x7e, 0x3e, 0x96, 0xcd, 0x3f, 0x3e, 0x86, 0x91, 0x2d, 0x3e,
0x55, 0xef, 0x87, 0x3e, 0x7e, 0x97, 0x03, 0xbe, 0x2a, 0xcd, 0x01, 0x3e,
0x32, 0xc9, 0x8e, 0xbe, 0x72, 0x77, 0x3b, 0xbe, 0xe0, 0xa1, 0xbc, 0xbe,
0x8d, 0xb7, 0xa7, 0x3e, 0x1c, 0x05, 0x95, 0xbe, 0xf7, 0x1f, 0xbb, 0x3e,
0xc9, 0x3e, 0xd6, 0x3e, 0x80, 0x42, 0xe9, 0xbd, 0x27, 0x0c, 0xd2, 0xbe,
0x5c, 0x32, 0x34, 0xbe, 0x14, 0xcb, 0xca, 0xbd, 0xdd, 0x3a, 0x67, 0xbe,
0x1c, 0xbb, 0x8d, 0xbe, 0x91, 0xac, 0x5c, 0xbe, 0x52, 0x40, 0x6f, 0xbe,
0xd7, 0x71, 0x94, 0x3e, 0x18, 0x71, 0x09, 0xbe, 0x9b, 0x29, 0xd9, 0xbe,
0x7d, 0x66, 0xd2, 0xbe, 0x98, 0xd6, 0xb2, 0xbe, 0x00, 0xc9, 0x84, 0x3a,
0xbc, 0xda, 0xc2, 0xbd, 0x1d, 0xc2, 0x1b, 0xbf, 0xd4, 0xdd, 0x92, 0x3e,
0x07, 0x87, 0x6c, 0xbe, 0x40, 0xc2, 0x3b, 0xbe, 0xbd, 0xe2, 0x9c, 0x3e,
0x0a, 0xb5, 0xa0, 0xbe, 0xe2, 0xd5, 0x9c, 0xbe, 0x3e, 0xbb, 0x7c, 0x3e,
0x17, 0xb4, 0xcf, 0x3e, 0xd5, 0x8e, 0xc8, 0xbe, 0x7c, 0xf9, 0x5c, 0x3e,
0x80, 0xfc, 0x0d, 0x3d, 0xc5, 0xd5, 0x8b, 0x3e, 0xf5, 0x17, 0xa2, 0x3e,
0xc7, 0x60, 0x89, 0xbe, 0xec, 0x95, 0x87, 0x3d, 0x7a, 0xc2, 0x5d, 0xbf,
0x77, 0x94, 0x98, 0x3e, 0x77, 0x39, 0x07, 0xbc, 0x42, 0x29, 0x00, 0x3e,
0xaf, 0xd0, 0xa9, 0x3e, 0x31, 0x23, 0xc4, 0xbe, 0x95, 0x36, 0x5b, 0xbe,
0xc7, 0xdc, 0x83, 0xbe, 0x1e, 0x6b, 0x47, 0x3e, 0x5b, 0x24, 0x99, 0x3e,
0x99, 0x27, 0x54, 0x3e, 0xc8, 0x20, 0xdd, 0xbd, 0x5a, 0x86, 0x2f, 0x3e,
0x80, 0xf0, 0x69, 0xbe, 0x44, 0xfc, 0x84, 0xbd, 0x82, 0xa0, 0x2a, 0xbe,
0x87, 0xe6, 0x2a, 0x3e, 0xd8, 0x34, 0xae, 0x3d, 0x50, 0xbd, 0xb5, 0x3e,
0xc4, 0x8c, 0x88, 0xbe, 0xe3, 0xbc, 0xa5, 0x3e, 0xa9, 0xda, 0x9e, 0x3e,
0x3e, 0xb8, 0x23, 0xbe, 0x80, 0x90, 0x15, 0x3d, 0x97, 0x3f, 0xc3, 0x3e,
0xca, 0x5c, 0x9d, 0x3e, 0x21, 0xe8, 0xe1, 0x3e, 0xc0, 0x49, 0x01, 0xbc,
0x00, 0x0b, 0x88, 0xbd, 0x3f, 0xf7, 0xca, 0x3c, 0xfb, 0x5a, 0xb1, 0x3e,
0x60, 0xd2, 0x0d, 0x3c, 0xce, 0x23, 0x78, 0xbf, 0x8f, 0x4f, 0xb9, 0xbe,
0x69, 0x6a, 0x34, 0xbf, 0x4b, 0x5e, 0xa9, 0x3e, 0x64, 0x8c, 0xd9, 0x3e,
0x52, 0x77, 0x36, 0x3e, 0xeb, 0xaf, 0xbe, 0x3e, 0x40, 0xbe, 0x36, 0x3c,
0x08, 0x65, 0x3b, 0xbd, 0x55, 0xe0, 0x66, 0xbd, 0xd2, 0xe8, 0x9b, 0xbe,
0x86, 0xe3, 0x09, 0xbe, 0x93, 0x3d, 0xdd, 0x3e, 0x0f, 0x66, 0x18, 0x3f,
0x18, 0x05, 0x33, 0xbd, 0xde, 0x15, 0xd7, 0xbe, 0xaa, 0xcf, 0x49, 0xbe,
0xa2, 0xa5, 0x64, 0x3e, 0xe6, 0x9c, 0x42, 0xbe, 0x54, 0x42, 0xcc, 0x3d,
0xa0, 0xbd, 0x9d, 0xbe, 0xc2, 0x69, 0x48, 0x3e, 0x5b, 0x8b, 0xa2, 0xbe,
0xc0, 0x13, 0x87, 0x3d, 0x36, 0xfd, 0x69, 0x3e, 0x05, 0x86, 0x40, 0xbe,
0x1e, 0x7a, 0xce, 0xbe, 0x46, 0x13, 0xa7, 0xbe, 0x68, 0x52, 0x86, 0xbe,
0x04, 0x9e, 0x86, 0xbd, 0x8c, 0x54, 0xc1, 0x3d, 0xe0, 0x3b, 0xad, 0x3c,
0x42, 0x67, 0x85, 0xbd, 0xea, 0x97, 0x42, 0x3e, 0x6e, 0x13, 0x3b, 0xbf,
0x56, 0x5b, 0x16, 0x3e, 0xaa, 0xab, 0xdf, 0x3e, 0xc8, 0x41, 0x36, 0x3d,
0x24, 0x2d, 0x47, 0xbe, 0x77, 0xa5, 0xae, 0x3e, 0xc0, 0xc2, 0x5b, 0x3c,
0xac, 0xac, 0x4e, 0x3e, 0x99, 0xec, 0x13, 0xbe, 0xf2, 0xab, 0x73, 0x3e,
0xaa, 0xa1, 0x48, 0xbe, 0xe8, 0xd3, 0x01, 0xbe, 0x60, 0xb7, 0xc7, 0xbd,
0x64, 0x72, 0xd3, 0x3d, 0x83, 0xd3, 0x99, 0x3e, 0x0c, 0x76, 0x34, 0xbe,
0x42, 0xda, 0x0d, 0x3e, 0xfb, 0x47, 0x9a, 0x3e, 0x8b, 0xdc, 0x92, 0xbe,
0x56, 0x7f, 0x6b, 0x3e, 0x04, 0xd4, 0x88, 0xbd, 0x11, 0x9e, 0x80, 0x3e,
0x3c, 0x89, 0xff, 0x3d, 0xb3, 0x3e, 0x88, 0x3e, 0xf7, 0xf0, 0x88, 0x3e,
0x28, 0xfb, 0xc9, 0xbe, 0x53, 0x3e, 0xcf, 0x3e, 0xac, 0x75, 0xdc, 0xbe,
0xdd, 0xca, 0xd7, 0x3e, 0x01, 0x58, 0xa7, 0x3e, 0x29, 0xb8, 0x13, 0xbf,
0x76, 0x81, 0x12, 0xbc, 0x28, 0x8b, 0x16, 0xbf, 0x0e, 0xec, 0x0e, 0x3e,
0x40, 0x0a, 0xdb, 0xbd, 0x98, 0xec, 0xbf, 0xbd, 0x32, 0x55, 0x0c, 0xbe,
0xfb, 0xf9, 0xc9, 0x3e, 0x83, 0x4a, 0x6d, 0xbe, 0x76, 0x59, 0xe2, 0xbe,
0x54, 0x7d, 0x9f, 0xbb, 0x9d, 0xe8, 0x95, 0x3e, 0x5c, 0xd3, 0xd0, 0x3d,
0x19, 0x8a, 0xb0, 0x3e, 0xde, 0x6f, 0x2e, 0xbe, 0xd0, 0x16, 0x83, 0x3d,
0x9c, 0x7d, 0x11, 0xbf, 0x2b, 0xcc, 0x25, 0x3c, 0x2a, 0xa5, 0x27, 0xbe,
0x22, 0x14, 0xc7, 0xbe, 0x5e, 0x7a, 0xac, 0x3e, 0x4e, 0x41, 0x94, 0xbe,
0x5a, 0x68, 0x7b, 0x3e, 0x86, 0xfd, 0x4e, 0x3e, 0xa2, 0x56, 0x6a, 0xbe,
0xca, 0xfe, 0x81, 0xbe, 0x43, 0xc3, 0xb1, 0xbd, 0xc5, 0xb8, 0xa7, 0x3e,
0x55, 0x23, 0xcd, 0x3e, 0xaf, 0x2e, 0x76, 0x3e, 0x69, 0xa8, 0x90, 0xbe,
0x0d, 0xba, 0xb9, 0x3e, 0x66, 0xff, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x40, 0x00, 0x00, 0x00, 0x53, 0xd6, 0xe2, 0x3d, 0x66, 0xb6, 0xcc, 0x3e,
0x03, 0xe7, 0xf6, 0x3e, 0xe0, 0x28, 0x10, 0xbf, 0x00, 0x00, 0x00, 0x00,
0x3e, 0x3d, 0xb0, 0x3e, 0x00, 0x00, 0x00, 0x00, 0x62, 0xf0, 0x77, 0x3e,
0xa6, 0x9d, 0xa4, 0x3e, 0x3a, 0x4b, 0xf3, 0xbe, 0x71, 0x9e, 0xa7, 0x3e,
0x00, 0x00, 0x00, 0x00, 0x34, 0x39, 0xa2, 0x3e, 0x00, 0x00, 0x00, 0x00,
0xcc, 0x9c, 0x4a, 0x3e, 0xab, 0x40, 0xa3, 0x3e, 0xb2, 0xff, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0xb3, 0x71, 0x67, 0x3f,
0x9a, 0x7a, 0x95, 0xbf, 0xe1, 0x48, 0xe8, 0xbe, 0x8a, 0x72, 0x96, 0x3e,
0x00, 0xd2, 0xd3, 0xbb, 0x1a, 0xc5, 0xd7, 0x3f, 0xac, 0x7e, 0xc8, 0xbe,
0x90, 0xa7, 0x95, 0xbe, 0x3b, 0xd7, 0xdc, 0xbe, 0x41, 0xa8, 0x16, 0x3f,
0x50, 0x5b, 0xcb, 0x3f, 0x52, 0xb9, 0xed, 0xbe, 0x2e, 0xa7, 0xc6, 0xbe,
0xaf, 0x0f, 0x14, 0xbf, 0xb3, 0xda, 0x59, 0x3f, 0x02, 0xec, 0xd7, 0xbe,
0x00, 0x00, 0x06, 0x00, 0x08, 0x00, 0x04, 0x00, 0x06, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x66, 0x11, 0x1f, 0xbf,
0xb8, 0xfb, 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00, 0x54, 0x4f, 0x43, 0x4f,
0x20, 0x43, 0x6f, 0x6e, 0x76, 0x65, 0x72, 0x74, 0x65, 0x64, 0x2e, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x14, 0x00,
0x04, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x0c, 0x00, 0x00, 0x00,
0xf0, 0x00, 0x00, 0x00, 0xe4, 0x00, 0x00, 0x00, 0xd8, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x90, 0x00, 0x00, 0x00,
0x48, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xce, 0xff, 0xff, 0xff,
0x00, 0x00, 0x00, 0x08, 0x18, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x1c, 0xfc, 0xff, 0xff, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00,
0x08, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0e, 0x00,
0x14, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x07, 0x00, 0x10, 0x00,
0x0e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08, 0x1c, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xba, 0xff, 0xff, 0xff,
0x00, 0x00, 0x00, 0x01, 0x01, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00,
0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0e, 0x00, 0x16, 0x00, 0x00, 0x00,
0x08, 0x00, 0x0c, 0x00, 0x07, 0x00, 0x10, 0x00, 0x0e, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x08, 0x24, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x00, 0x00, 0x06, 0x00, 0x08, 0x00, 0x07, 0x00,
0x06, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x01, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x0a, 0x00, 0x00, 0x00, 0x10, 0x03, 0x00, 0x00, 0xa4, 0x02, 0x00, 0x00,
0x40, 0x02, 0x00, 0x00, 0xf4, 0x01, 0x00, 0x00, 0xac, 0x01, 0x00, 0x00,
0x48, 0x01, 0x00, 0x00, 0xfc, 0x00, 0x00, 0x00, 0xb4, 0x00, 0x00, 0x00,
0x50, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x26, 0xfd, 0xff, 0xff,
0x3c, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x18, 0xfd, 0xff, 0xff, 0x20, 0x00, 0x00, 0x00,
0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31,
0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x34, 0x2f, 0x4d, 0x61, 0x74,
0x4d, 0x75, 0x6c, 0x5f, 0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x6e, 0xfd, 0xff, 0xff,
0x50, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x60, 0xfd, 0xff, 0xff, 0x34, 0x00, 0x00, 0x00,
0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31,
0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x34, 0x2f, 0x4d, 0x61, 0x74,
0x4d, 0x75, 0x6c, 0x2f, 0x52, 0x65, 0x61, 0x64, 0x56, 0x61, 0x72, 0x69,
0x61, 0x62, 0x6c, 0x65, 0x4f, 0x70, 0x2f, 0x74, 0x72, 0x61, 0x6e, 0x73,
0x70, 0x6f, 0x73, 0x65, 0x00, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0xce, 0xfd, 0xff, 0xff,
0x34, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0xc0, 0xfd, 0xff, 0xff, 0x19, 0x00, 0x00, 0x00,
0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31,
0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x33, 0x2f, 0x52, 0x65, 0x6c,
0x75, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x12, 0xfe, 0xff, 0xff, 0x3c, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x04, 0xfe, 0xff, 0xff, 0x20, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x33, 0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x5f,
0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x5a, 0xfe, 0xff, 0xff, 0x50, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x4c, 0xfe, 0xff, 0xff, 0x34, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x33, 0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x2f,
0x52, 0x65, 0x61, 0x64, 0x56, 0x61, 0x72, 0x69, 0x61, 0x62, 0x6c, 0x65,
0x4f, 0x70, 0x2f, 0x74, 0x72, 0x61, 0x6e, 0x73, 0x70, 0x6f, 0x73, 0x65,
0x00, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0xba, 0xfe, 0xff, 0xff, 0x34, 0x00, 0x00, 0x00,
0x0a, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0xac, 0xfe, 0xff, 0xff, 0x19, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x32, 0x2f, 0x52, 0x65, 0x6c, 0x75, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0xfe, 0xfe, 0xff, 0xff, 0x3c, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xf0, 0xfe, 0xff, 0xff,
0x20, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x5f, 0x62, 0x69, 0x61, 0x73,
0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x46, 0xff, 0xff, 0xff, 0x50, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x38, 0xff, 0xff, 0xff,
0x34, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x2f, 0x52, 0x65, 0x61, 0x64,
0x56, 0x61, 0x72, 0x69, 0x61, 0x62, 0x6c, 0x65, 0x4f, 0x70, 0x2f, 0x74,
0x72, 0x61, 0x6e, 0x73, 0x70, 0x6f, 0x73, 0x65, 0x00, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0xa6, 0xff, 0xff, 0xff, 0x48, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00,
0x2c, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00,
0x04, 0x00, 0x08, 0x00, 0x08, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x7f, 0x43,
0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0d, 0x00, 0x00, 0x00,
0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32, 0x5f, 0x69, 0x6e, 0x70, 0x75,
0x74, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0e, 0x00, 0x14, 0x00, 0x04, 0x00,
0x00, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x0e, 0x00, 0x00, 0x00,
0x28, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x08, 0x00, 0x00, 0x00, 0x04, 0x00, 0x04, 0x00, 0x04, 0x00, 0x00, 0x00,
0x08, 0x00, 0x00, 0x00, 0x49, 0x64, 0x65, 0x6e, 0x74, 0x69, 0x74, 0x79,
0x00, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x00, 0x00, 0x0a, 0x00, 0x0c, 0x00, 0x07, 0x00, 0x00, 0x00, 0x08, 0x00,
0x0a, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09, 0x03, 0x00, 0x00, 0x00};
const int g_sine_model_data_len = 2640;

69
tensorflow/lite/micro/examples/hello_world/train/README.md Normal file
View File

@ -0,0 +1,69 @@
# Hello World Training
This example shows how to train a 2.5 kB model to generate a `sine` wave.
## Table of contents
- [Overview](#overview)
- [Training](#training)
- [Trained Models](#trained-models)
- [Model Architecture](#model-architecture)
## Overview
1. Dataset: Data is generated locally in the Jupyter Notebook.
2. Dataset Type: **Structured Data**
3. Deep Learning Framework: **TensorFlow 2**
4. Language: **Python 3.7**
5. Model Size: **2.5 kB**
6. Model Category: **Regression**
## Training
Train the model in the cloud using Google Colaboratory or locally using a
Jupyter Notebook.
<table class="tfo-notebook-buttons" align="left">
<td>
<a target="_blank" href="https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/train/train_hello_world_model.ipynb"><img src="https://www.tensorflow.org/images/colab_logo_32px.png" />Google Colaboratory</a>
</td>
<td>
<a target="_blank" href="https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/hello_world/train/train_hello_world_model.ipynb"><img src="https://www.tensorflow.org/images/GitHub-Mark-32px.png" />Jupyter Notebook</a>
</td>
</table>
*Estimated Training Time: 10 minutes.*
## Trained Models
| Download Link | [hello_world.zip](https://storage.googleapis.com/download.tensorflow.org/models/tflite/micro/hello_world_2020_04_13.zip) |
| ------------- |-------------|
The `models` directory in the above zip file can be generated by following the
instructions in the [Training](#training) section above. It
includes the following 3 model files:
| Name | Format | Target Framework | Target Device |
| :------------- |:-------------|:-------------|-----|
| `model.pb` | Keras SavedModel | TensorFlow | Large-Scale/Cloud/Servers |
| `model.tflite` *(2.5 kB)* | Fully Quantized* TFLite Model | TensorFlow Lite | Mobile Devices|
| `model.cc` | C Source File | TensorFlow Lite for Microcontrollers | Microcontrollers |
**Fully quantized implies that the model is **strictly int8** quantized
**excluding** the input(s) and output(s).*
<!-- **Fully quantized implies that the model is **strictly int8** quantized
including the input(s)and output(s).* -->
## Model Architecture
The final model used to simulate a sine wave is displayed below. It is a
simple feed forward deep neural network with 2 fully connected layers with
ReLu activations and a final fully connected output layer with as shown below.
![model_architecture.png](../images/model_architecture.png)
*This image was derived from visualizing the 'model.tflite' file in [Netron](https://github.com/lutzroeder/netron)*

3530
tensorflow/lite/micro/examples/hello_world/train/train_hello_world_model.ipynb Normal file
View File

File diff suppressed because one or more lines are too long

2
tensorflow/lite/micro/examples/micro_speech/README.md
View File

@ -545,4 +545,4 @@ with the model and sample inputs.
So far you have used an existing trained model to run inference on
microcontrollers. If you wish to train your own model, follow the instructions
in [train/README.md](train/README.md).
given in the [train/](train/) directory.

140
tensorflow/lite/micro/examples/micro_speech/train/README.md
View File

@ -23,28 +23,48 @@ stop
go
```
The scripts used in training the model have been sourced from the
[Simple Audio Recognition](https://www.tensorflow.org/tutorials/sequences/audio_recognition)
tutorial.
## Table of contents
- [Overview](#overview)
- [Trained Models](#trained-models)
- [Training](#training)
- [Trained Models](#trained-models)
- [Model Architecture](#model-architecture)
- [Dataset](#dataset)
- [Preprocessing Speech Input](#preprocessing-speech-input)
- [Other Training Methods](#other-training-methods)
## Overview
1. Training Jupyter Notebook: [`train_micro_speech_model.ipynb`](train_micro_speech_model.ipynb)
. The training scripts used in this notebook are in the
[Simple Audio Recognition](https://www.tensorflow.org/tutorials/sequences/audio_recognition)
tutorial.
2. Dataset Type: **Speech**
3. Dataset: Speech Commands, Version 2. ([Download Link](https://storage.cloud.google.com/download.tensorflow.org/data/speech_commands_v0.02.tar.gz)
1. Dataset: Speech Commands, Version 2. ([Download Link](https://storage.cloud.google.com/download.tensorflow.org/data/speech_commands_v0.02.tar.gz)
, [Paper](https://arxiv.org/abs/1804.03209))
4. Deep Learning Framework: **TensorFlow 1.5**
5. Language: **Python 3.7**
6. Model Size: **<20 kB**
7. Model Category: **Multiclass Classification**
2. Dataset Type: **Speech**
3. Deep Learning Framework: **TensorFlow 1.5**
4. Language: **Python 3.7**
5. Model Size: **<20 kB**
6. Model Category: **Multiclass Classification**
## Training
Train the model in the cloud using Google Colaboratory or locally using a
Jupyter Notebook.
<table class="tfo-notebook-buttons" align="left">
<td>
<a target="_blank" href="https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/micro_speech/train/train_micro_speech_model.ipynb"><img src="https://www.tensorflow.org/images/colab_logo_32px.png" />Google Colaboratory</a>
</td>
<td>
<a target="_blank" href="https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/micro_speech/train/train_micro_speech_model.ipynb"><img src="https://www.tensorflow.org/images/GitHub-Mark-32px.png" />Jupyter Notebook</a>
</td>
</table>
*Estimated Training Time: ~2 Hours.*
For more options, refer to the [Other Training Methods](#other-training-methods)
section.
## Trained Models
@ -52,7 +72,7 @@ tutorial.
| ------------- |-------------|
The `models` directory in the above zip file can be generated by following the
instructions in the [Training](#training) section below. It
instructions in the [Training](#training) section above. It
includes the following 3 model files:
| Name | Format | Target Framework | Target Device |
@ -61,67 +81,11 @@ includes the following 3 model files:
| `model.tflite` *(<20 kB)* | Fully Quantized* TFLite Model | TensorFlow Lite | Mobile Devices|
| `model.cc` | C Source File | TensorFlow Lite for Microcontrollers | Microcontrollers |
*Fully quantized implies that the model is **strictly int8** quantized
**Fully quantized implies that the model is **strictly int8** quantized
**including** the input(s) and output(s).*
<!-- **Fully quantized implies that the model is **strictly int8** except the
input(s) and output(s) which remain float.* -->
## Training
### 1. Use [Google Colaboratory](https://colab.research.google.com)
*We strongly recommend trying this approach first.*
| Run in Google Colaboratory | <a target="_blank" href="https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/micro/examples/micro_speech/train/train_micro_speech_model.ipynb"><img src="https://www.tensorflow.org/images/colab_logo_32px.png"/>train_micro_speech_model.ipynb</a> |
| ------------- |-------------|
**Estimated Training Time:** ~2 hours.
**Advantage:** It allows the use of a free Tesla K80 GPU for training and avoids
the need to install dependencies.
**Disadvantage:** Your training time is limited as the session can only run
upto 12 hours in a row if you keep the browser open and 90 minutes if you close
the browser.
### 2. Use [Google Cloud](https://cloud.google.com/)
1. Create a Virtual Machine (VM) using a pre-configured Deep Learning VM Image.
```
export IMAGE_FAMILY="tf-latest-cpu"
export ZONE="us-west1-b" # Or any other required region
export INSTANCE_NAME="model-trainer"
export INSTANCE_TYPE="n1-standard-8" # or any other instance type
gcloud compute instances create $INSTANCE_NAME \
--zone=$ZONE \
--image-family=$IMAGE_FAMILY \
--image-project=deeplearning-platform-release \
--machine-type=$INSTANCE_TYPE \
--boot-disk-size=120GB \
--min-cpu-platform=Intel\ Skylake
```
2. As soon as instance has been created you can SSH to it:
```
gcloud compute ssh "jupyter@${INSTANCE_NAME}"
```
3. Train a model by following the instructions in the [`train_micro_speech_model.ipynb`](train_micro_speech_model.ipynb)
jupyter notebook.
4. Finally, don't forget to remove the instance when training is done:
```
gcloud compute instances delete "${INSTANCE_NAME}" --zone="${ZONE}"
```
**Estimated Training Time:** ~2 hours (with GPU) and ~1 day (with CPU).
**Advantage:** There are no time constraints on how long the training process
can take and it avoids the need to install dependencies.
**Disadvantage:** Google Cloud isn't free. You need to pay
depending on how long you use run the VM and what resources you use.
## Model Architecture
This is a simple model comprising of a Convolutional 2D layer, a Fully Connected
@ -197,3 +161,41 @@ python tensorflow/tensorflow/examples/speech_commands/wav_to_features.py \
--window_stride=20 --preprocess=average --quantize=1
```
## Other Training Methods
### Use [Google Cloud](https://cloud.google.com/).
*Note: Google Cloud isn't free. You need to pay depending on how long you use
run the VM and what resources you use.*
1. Create a Virtual Machine (VM) using a pre-configured Deep Learning VM Image.
```
export IMAGE_FAMILY="tf-latest-cpu"
export ZONE="us-west1-b" # Or any other required region
export INSTANCE_NAME="model-trainer"
export INSTANCE_TYPE="n1-standard-8" # or any other instance type
gcloud compute instances create $INSTANCE_NAME \
--zone=$ZONE \
--image-family=$IMAGE_FAMILY \
--image-project=deeplearning-platform-release \
--machine-type=$INSTANCE_TYPE \
--boot-disk-size=120GB \
--min-cpu-platform=Intel\ Skylake
```
2. As soon as instance has been created you can SSH to it:
```
gcloud compute ssh "jupyter@${INSTANCE_NAME}"
```
3. Train a model by following the instructions in the [`train_micro_speech_model.ipynb`](train_micro_speech_model.ipynb)
jupyter notebook.
4. Finally, don't forget to remove the instance when training is done:
```
gcloud compute instances delete "${INSTANCE_NAME}" --zone="${ZONE}"
```