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

Lie about manylinux1 so we can get our packages on PyPI

Browse Source
This commit is contained in:
Reuben Morais 2017-11-15 10:20:12 +01:00
parent 44b76bd71f
commit f1d89c02be
6 changed files with 14 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
README.md
View File

@ -65,7 +65,7 @@ See the help output with `./deepspeech -h` and the [native client README](native
Pre-built binaries that can be used for performing inference with a trained model can be found on TaskCluster. You'll need to download the appropriate Python wheel package.
[deepspeech-0.0.1-cp27-cp27mu-linux_x86_64.whl (Python 2.7, Linux / amd64)](https://index.taskcluster.net/v1/task/project.deepspeech.deepspeech.native_client.master.cpu/artifacts/public/deepspeech-0.0.1-cp27-cp27mu-linux_x86_64.whl)
[deepspeech-0.0.1-cp27-cp27mu-manylinux1_x86_64.whl (Python 2.7, Linux / amd64)](https://index.taskcluster.net/v1/task/project.deepspeech.deepspeech.native_client.master.cpu/artifacts/public/deepspeech-0.0.1-cp27-cp27mu-manylinux1_x86_64.whl)
[Other configurations](https://tools.taskcluster.net/index/artifacts/#project.deepspeech.deepspeech.native_client.master/project.deepspeech.deepspeech.native_client.master)
@ -99,8 +99,8 @@ Install the required dendencies using pip:
```bash
cd DeepSpeech
python util/taskcluster.py --target /tmp --source tensorflow --artifact tensorflow_warpctc-1.3.0rc0-cp27-cp27mu-linux_x86_64.whl
pip install /tmp/tensorflow_warpctc-1.3.0rc0-cp27-cp27mu-linux_x86_64.whl
python util/taskcluster.py --target /tmp --source tensorflow --artifact tensorflow_warpctc-1.3.0rc0-cp27-cp27mu-manylinux1_x86_64.whl
pip install /tmp/tensorflow_warpctc-1.3.0rc0-cp27-cp27mu-manylinux1_x86_64.whl
pip install -r requirements.txt
```
@ -118,8 +118,8 @@ If you have a capable (Nvidia, at least 8GB of VRAM) GPU, it is highly recommend
```bash
pip uninstall tensorflow
python util/taskcluster.py --target /tmp --source tensorflow --arch gpu --artifact tensorflow_gpu_warpctc-1.3.0rc0-cp27-cp27mu-linux_x86_64.whl
pip install /tmp/tensorflow_gpu_warpctc-1.3.0rc0-cp27-cp27mu-linux_x86_64.whl
python util/taskcluster.py --target /tmp --source tensorflow --arch gpu --artifact tensorflow_gpu_warpctc-1.3.0rc0-cp27-cp27mu-manylinux1_x86_64.whl
pip install /tmp/tensorflow_gpu_warpctc-1.3.0rc0-cp27-cp27mu-manylinux1_x86_64.whl
```
### Training a model

2
native_client/README.md
View File

@ -22,7 +22,7 @@ If you're looking to train a model, you now have a `libctc_decoder_with_kenlm.so
For Python bindings, use `--artifact file_name`, where `file_name` is the appropriate file for your Python version and platform. The names of the available artifacts can be found on the listing page: [Linux](https://tools.taskcluster.net/index/artifacts/project.deepspeech.deepspeech.native_client.master/cpu) or [macOS](https://tools.taskcluster.net/index/artifacts/project.deepspeech.deepspeech.native_client.master/osx).
For example, for Python 2.7 bindings on Linux, you can do `python util/taskcluster.py --target /destination --artifact deepspeech-0.0.1-cp27-cp27mu-linux_x86_64.whl`.
For example, for Python 2.7 bindings on Linux, you can do `python util/taskcluster.py --target /destination --artifact deepspeech-0.0.1-cp27-cp27mu-manylinux1_x86_64.whl`.
For Node.JS bindings, use `--artifact deepspeech-0.0.1.tgz`.

3
native_client/definitions.mk
View File

@ -12,6 +12,9 @@ LDFLAGS :=
SOX_CFLAGS := `pkg-config --cflags sox`
SOX_LDFLAGS := `pkg-config --libs sox`
PYTHON_PACKAGES := numpy
ifeq ($(OS),Linux)
PYTHON_PLATFORM_NAME := --plat-name manylinux1_x86_64
endif
endif
ifeq ($(TARGET),rpi3)

3
native_client/setup.cfg
View File

@ -12,3 +12,6 @@ bdist-dir=temp_build
[install_lib]
build-dir=temp_build
[metadata]
description-file = README.md

2
tc-python-tests-prod.sh
View File

@ -41,7 +41,7 @@ pyenv install ${pyver}
pyenv virtualenv ${pyver} ${PYENV_NAME}
source ${PYENV_ROOT}/versions/${pyver}/envs/${PYENV_NAME}/bin/activate
platform=$(python -c 'import sys; import platform; sys.stdout.write("%s_%s" % (platform.system().lower(), platform.machine()));')
platform=$(python -c 'import sys; import platform; plat = platform.system().lower(); plat = "manylinux1" if plat == "linux" else plat; sys.stdout.write("%s_%s" % (plat, platform.machine()));')
deepspeech_pkg="deepspeech-0.0.1-cp${pyver_pkg}-cp${pyver_pkg}${py_unicode_type}-${platform}.whl"
pip install --upgrade scipy==0.19.1 ${DEEPSPEECH_ARTIFACTS_ROOT}/${deepspeech_pkg}

2
tc-python-tests.sh
View File

@ -39,7 +39,7 @@ pyenv install ${pyver}
pyenv virtualenv ${pyver} ${PYENV_NAME}
source ${PYENV_ROOT}/versions/${pyver}/envs/${PYENV_NAME}/bin/activate
platform=$(python -c 'import sys; import platform; sys.stdout.write("%s_%s" % (platform.system().lower(), platform.machine()));')
platform=$(python -c 'import sys; import platform; plat = platform.system().lower(); plat = "manylinux1" if plat == "linux" else plat; sys.stdout.write("%s_%s" % (plat, platform.machine()));')
deepspeech_pkg="deepspeech-0.0.1-cp${pyver_pkg}-cp${pyver_pkg}${py_unicode_type}-${platform}.whl"
if [ "${aot_model}" = "--aot" ]; then