General

This is the 0.1.1 release of Deep Speech, an open speech-to-text engine. This release includes source code

v0.1.1.tar.gz

and a model, not yet optimized for size,

deepspeech-0.1.1-models.tar.gz

trained on American English which achieves a ~~5.6% word error rate~~ (The language model included some test data.) on the LibriSpeech clean test corpus, and example audio

audio-0.1.1.tar.gz

which can be used to test the engine and checkpoint files

deepspeech-0.1.1-checkpoint.tar.gz

which can be used as the basis for further fine-tuning. Unfortunately licensing issues prevent us from releasing the text used to train the language model.

Notable changes from the previous release

Rust bindings were contributed by RustAudio
Lowering dependency on AVX2 to AVX instruction sets (mozilla/tensorflow#46)
Pre-built binaries now work with upstream TensorFlow 1.4 (mozilla/tensorflow#43)
Switching GPU build to CUDA 8.0 / CuDNN v6 (mozilla/tensorflow#43)
Added support for Node.JS 7/8/9 (#1042)
Initializing a training run from a frozen graph (eg. a release model) is now easier (#1149)
The Python package no longer holds the GIL during inference and can be used in multi-threaded Python programs (#1164)
The Python package now works on macOS 10.10 and 10.11 (#1065)

Hyperparameters for fine-tuning

The hyperparameters used to train the model are useful for fine tuning. Thus, we document them here along with the hardware used, a two node cluster where each node has 8 TitanX Pascal GPU's.

train_files Fisher, LibriSpeech, and Switchboard training corpora.
dev_files LibriSpeech clean dev corpus
test_files LibriSpeech clean test corpus
train_batch_size 12
dev_batch_size 8
test_batch_size 8
epoch 13
learning_rate 0.0001
display_step 0
validation_step 1
dropout_rate 0.2367
default_stddev 0.046875
checkpoint_step 1
log_level 0
checkpoint_dir value specific to hardware setup
wer_log_pattern "GLOBAL LOG: logwer('${COMPUTE_ID}', '%s', '%s', %f)"
decoder_library_path value specific to hardware setup
n_hidden 2048

Bindings

This release also includes a Python based command line tool deepspeech, installed through

pip install deepspeech

Alternatively, quicker inference can be performed using a supported NVIDIA GPU on Linux. (See below to find which GPU's are supported.) This is done by instead installing the GPU specific package:

pip install deepspeech-gpu

Also, it exposes bindings for the following languages

Python (Versions 2.7, 3.4, 3.5, and 3.6) installed via
```
pip install deepspeech
```
Alternatively, quicker inference can be performed using a supported NVIDIA GPU on Linux. (See below to find which GPU's are supported.) This is done by instead installing the GPU specific package:
```
pip install deepspeech-gpu
```
NodeJS (Versions 4.x, 5.x, 6.x, 7.x, 8.x and 9.x) installed via
```
npm install deepspeech
```
Alternatively, quicker inference can be performed using a supported NVIDIA GPU on Linux. (See below to find which GPU's are supported.) This is done by instead installing the GPU specific package:
```
npm install deepspeech-gpu
```
C++ which requires the appropriate shared objects are installed from native_client.tar.xz (See the section in the main README which describes native_client.tar.xz installation.)

In addition there are third party bindings that are supported by external developers, for example

Rust which is installed by following the instructions on the external Rust repo.

Supported Platforms

OS X 10.10, 10.11, 10.12 and 10.13
Linux x86 64 bit with a modern CPU (Needs at least AVX/FMA)
Linux x86 64 bit with a modern CPU + NVIDIA GPU (Compute Capability at least 3.0, see NVIDIA docs)
Raspbian Jessie on Raspberry Pi 3

Contact/Getting Help

FAQ - We have a list of common questions, and their answers, in our FAQ. When just getting started, it's best to first check the FAQ to see if your question is addressed.
Discourse Forums - If your question is not addressed in the FAQ, the Discourse Forums is the next place to look. They contain conversations on General Topics, Using Deep Speech, Alternative Platforms, and Deep Speech Development.
IRC - If your question is not addressed by either the FAQ or Discourse Forums, you can contact us on the #machinelearning channel on Mozilla IRC; people there can try to answer/help
Issues - Finally, if all else fails, you can open an issue in our repo if there is a bug with the current code base.

Contributors to 0.1.1 release

mozilla/DeepSpeech v0.1.1 Deep Speech 0.1.1 on GitHub