Implement a native client using TensorFlow C API

DeepSpeech.py

@@ -17,6 +17,7 @@
 from tensorflow.contrib.session_bundle import exporter
 from tensorflow.python.ops import ctc_ops
 from tensorflow.contrib.rnn.python.ops import core_rnn_cell
+from tensorflow.python.tools import freeze_graph
 from util.gpu import get_available_gpus
 from util.log import merge_logs
 from util.spell import correction
@@ -1151,7 +1152,7 @@ def train():
             # Run inference
 
             # Input tensor will be of shape [batch_size, n_steps, n_input + 2*n_input*n_context]
-            input_tensor = tf.placeholder(tf.float32, [None, None, n_input + 2*n_input*n_context])
+            input_tensor = tf.placeholder(tf.float32, [None, None, n_input + 2*n_input*n_context], name='input_node')
 
             # Calculate input sequence length. This is done by tiling n_steps, batch_size times.
             # If there are multiple sequences, it is assumed they are padded with zeros to be of
@@ -1166,7 +1167,7 @@ def train():
             # Beam search decode the batch
             decoded, _ = ctc_ops.ctc_beam_search_decoder(logits, seq_length, merge_repeated=False)
             decoded = tf.convert_to_tensor(
-                [tf.sparse_tensor_to_dense(sparse_tensor) for sparse_tensor in decoded])
+                [tf.sparse_tensor_to_dense(sparse_tensor) for sparse_tensor in decoded], name='output_node')
 
             # TODO: Transform the decoded output to a string
 
@@ -1178,8 +1179,9 @@ def train():
             # TODO: This restores the most recent checkpoint, but if we use validation to counterract
             #       over-fitting, we may want to restore an earlier checkpoint.
             checkpoint = tf.train.get_checkpoint_state(checkpoint_dir)
-            saver.restore(session, checkpoint.model_checkpoint_path)
-            print 'Restored checkpoint at training epoch %d' % (int(checkpoint.model_checkpoint_path.split('-')[-1]) + 1)
+            checkpoint_path = checkpoint.model_checkpoint_path
+            saver.restore(session, checkpoint_path)
+            print 'Restored checkpoint at training epoch %d' % (int(checkpoint_path.split('-')[-1]) + 1)
 
             # Initialise the model exporter and export the model
             model_exporter.init(session.graph.as_graph_def(),
@@ -1194,8 +1196,28 @@ def train():
                     print 'Removing old export'
                     shutil.rmtree(actual_export_dir)
             try:
+                # Export serving model
                 model_exporter.export(export_dir, tf.constant(export_version), session)
-                print 'Model exported at %s' % (export_dir)
+
+                # Export graph
+                input_graph_name = 'input_graph.pb'
+                tf.train.write_graph(session.graph, export_dir, input_graph_name, as_text=False)
+
+                # Freeze graph
+                input_graph_path = os.path.join(export_dir, input_graph_name)
+                input_saver_def_path = ''
+                input_binary = True
+                output_node_names = 'output_node'
+                restore_op_name = 'save/restore_all'
+                filename_tensor_name = 'save/Const:0'
+                output_graph_path = os.path.join(export_dir, 'output_graph.pb')
+                clear_devices = False
+                freeze_graph.freeze_graph(input_graph_path, input_saver_def_path,
+                                          input_binary, checkpoint_path, output_node_names,
+                                          restore_op_name, filename_tensor_name,
+                                          output_graph_path, clear_devices, '')
+
+                print 'Models exported at %s' % (export_dir)
             except RuntimeError:
                 print  sys.exc_info()[1]
 

native_client/BUILD

@@ -0,0 +1,19 @@
+# Description: Deepspeech native client library.
+
+cc_library(
+    name = "deepspeech",
+    srcs = ["deepspeech.cc",
+            "c_speech_features/c_speech_features.c",
+            "kiss_fft130/kiss_fft.c",
+            "kiss_fft130/tools/kiss_fftr.c"],
+    hdrs = ["deepspeech.h",
+            "c_speech_features/c_speech_features.h",
+            "kiss_fft130/kiss_fft.h",
+            "kiss_fft130/_kiss_fft_guts.h",
+            "kiss_fft130/tools/kiss_fftr.h"],
+    includes = ["c_speech_features",
+                "kiss_fft130"],
+    deps = [
+        "//tensorflow/core:tensorflow"
+    ]
+)

native_client/Makefile

@@ -0,0 +1,14 @@
+
+TFDIR ?= ../../tensorflow
+CFLAGS ?= -O2 -Wall
+
+default: deepspeech
+
+clean:
+	rm -f deepspeech
+
+deepspeech: client.cc
+	c++ -o deepspeech ${CFLAGS} client.cc `pkg-config --cflags --libs sox` -L${TFDIR}/bazel-bin/tensorflow -L${TFDIR}/bazel-bin/native_client -ldeepspeech -ltensorflow
+
+run: deepspeech
+	LD_LIBRARY_PATH=${TFDIR}/bazel-bin/tensorflow:${TFDIR}/bazel-bin/native_client:${LD_LIBRARY_PATH} ./deepspeech ${ARGS}

native_client/README.md

@@ -0,0 +1,48 @@
+# DeepSpeech native client
+
+A native client for running queries on an exported DeepSpeech model.
+
+## Requirements
+
+* [TensorFlow source](https://www.tensorflow.org/install/install_sources)
+* [libsox](https://sourceforge.net/projects/sox/)
+
+## Preparation
+
+Create a symbolic link in the TensorFlow checkout to the deepspeech `native_client` directory.
+
+```
+cd tensorflow
+ln -s ../DeepSpeech/native_client ./
+```
+
+## Building
+
+Before building the TensorFlow stand-alone library, you will need to prepare your environment to configure and build TensorFlow. Follow the [instructions](https://www.tensorflow.org/install/install_sources) on the TensorFlow site for your platform, up to the end of 'Configure the installation'.
+
+To build the TensorFlow library, execute the following command:
+
+```
+bazel build -c opt //tensorflow:libtensorflow.so
+```
+
+Then you can build the DeepSpeech native library.
+
+```
+bazel build -c opt //native_client:deepspeech
+```
+
+Finally, you can change to the `native_client` directory and use the `Makefile`. By default, the `Makefile` will assume there is a TensorFlow checkout in a directory above the DeepSpeech checkout. If that is not the case, set the environment variable `TFDIR` to point to the right directory.
+
+```
+cd ../DeepSpeech/native_client
+make deepspeech
+```
+
+## Running
+
+The client can be run via the `Makefile`. The client will accept audio of any format your installation of SoX supports.
+
+```
+ARGS="/path/to/output_graph.pb /path/to/audio/file.ogg" make run
+```