Fixes for magic_wand training scripts

PiperOrigin-RevId: 284086588
Change-Id: Ie571601516848f4c4fc9f3658ef7aea9d9842232

tensorflow/lite/experimental/micro/examples/magic_wand/README.md

@@ -15,6 +15,7 @@ then outputs the gesture to the serial port.
 -   [Deploy to SparkFun Edge](#deploy-to-sparkfun-edge)
 -   [Deploy to Adafruit devices](#deploy-to-adafruit)
 -   [Run the tests on a development machine](#run-the-tests-on-a-development-machine)
+-   [Train your own model](#train-your-own-model)
 
 ## Deploy to Arduino
 
@@ -360,3 +361,8 @@ To understand how TensorFlow Lite does this, you can look at the source in
 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
+
+To train your own model, or create a new model for a new set of gestures,
+follow the instructions in [magic_wand/train/README.md](https://github.com/tensorflow/tensorflow/tree/master/tensorflow/lite/experimental/micro/examples/magic_wand/train/README.md).

tensorflow/lite/experimental/micro/examples/magic_wand/train/train_magic_wand_model.ipynb

@@ -63,9 +63,9 @@
         "colab_type": "text"
       },
       "source": [
-        "## Install dependencies\n",
+        "## Configure dependencies\n",
         "\n",
-        "Run the following cell to ensure the required dependencies are installed."
+        "Run the following cell to ensure the correct version of TensorFlow is used."
       ]
     },
     {
@@ -76,8 +76,7 @@
         "colab": {}
       },
       "source": [
-        "!pip uninstall -y tensorflow\n",
-        "!pip install -q tensorflow-gpu==2.0.0-beta1"
+        "%tensorflow_version 2.x\n"
       ],
       "execution_count": 0,
       "outputs": []
@@ -103,7 +102,7 @@
         "# Clone the repository from GitHub\n",
         "!git clone --depth 1 -q https://github.com/tensorflow/tensorflow\n",
         "# Copy the training scripts into our workspace\n",
-        "!cp -r tensorflow/tensorflow/lite/experimental/micro/magic_wand/train train"
+        "!cp -r tensorflow/tensorflow/lite/experimental/micro/examples/magic_wand/train train"
       ],
       "execution_count": 0,
       "outputs": []