add h5py

udacity · Jun 16, 2017 · 0f57282 · 0f57282
1 parent 0594533
commit 0f57282
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 18 deletions.
diff --git a/requirements.txt b/requirements.txt
@@ -1,3 +1,4 @@
+h5py==2.7.0
 ipykernel==4.6.1
 jupyter==1.0.0
 keras==2.0.5

diff --git a/vui_notebook.ipynb b/vui_notebook.ipynb
@@ -66,7 +66,6 @@
    "cell_type": "code",
    "execution_count": null,
    "metadata": {
-    "collapsed": true,
     "scrolled": false
    },
    "outputs": [],
@@ -200,9 +199,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "collapsed": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "#####################################################################\n",
@@ -638,9 +635,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "collapsed": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "# specify the model\n",
@@ -690,15 +685,14 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "collapsed": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "import numpy as np\n",
     "from data_generator import AudioGenerator\n",
     "from keras import backend as K\n",
     "from utils import int_sequence_to_text\n",
+    "from IPython.display import Audio\n",
     "\n",
     "def get_predictions(index, partition, input_to_softmax, model_path):\n",
     "    \"\"\" Print a model's decoded predictions\n",
@@ -716,10 +710,12 @@
     "    # obtain the true transcription and the audio features \n",
     "    if partition == 'validation':\n",
     "        transcr = data_gen.valid_texts[index]\n",
-    "        data_point = data_gen.normalize(data_gen.featurize(data_gen.valid_audio_paths[index]))\n",
+    "        audio_path = data_gen.valid_audio_paths[index]\n",
+    "        data_point = data_gen.normalize(data_gen.featurize(audio_path))\n",
     "    elif partition == 'train':\n",
     "        transcr = data_gen.train_texts[index]\n",
-    "        data_point = data_gen.normalize(data_gen.featurize(data_gen.train_audio_paths[index]))\n",
+    "        audio_path = data_gen.train_audio_paths[index]\n",
+    "        data_point = data_gen.normalize(data_gen.featurize(audio_path))\n",
     "    else:\n",
     "        raise Exception('Invalid partition!  Must be \"train\" or \"validation\"')\n",
     "        \n",
@@ -730,8 +726,9 @@
     "    pred_ints = (K.eval(K.ctc_decode(\n",
     "                prediction, output_length)[0][0])+1).flatten().tolist()\n",
     "    \n",
-    "    # display the true and predicted transcriptions\n",
+    "    # play the audio file, and display the true and predicted transcriptions\n",
     "    print('-'*80)\n",
+    "    Audio(audio_path)\n",
     "    print('True transcription:\\n' + '\\n' + transcr)\n",
     "    print('-'*80)\n",
     "    print('Predicted transcription:\\n' + '\\n' + ''.join(int_sequence_to_text(pred_ints)))\n",
@@ -748,15 +745,13 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {
-    "collapsed": true
-   },
+   "metadata": {},
    "outputs": [],
    "source": [
     "get_predictions(index=0, \n",
     "                partition='train',\n",
     "                input_to_softmax=final_model(), \n",
-    "                model_path='results/model_end.h5')"
+    "                model_path='model_end.h5')"
    ]
   },
   {
@@ -777,7 +772,7 @@
     "get_predictions(index=0, \n",
     "                partition='validation',\n",
     "                input_to_softmax=final_model(), \n",
-    "                model_path='results/model_end.h5')"
+    "                model_path='model_end.h5')"
    ]
   },
   {