Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update requirements.txt and install script #16

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Update requirements.txt and install script #16

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 8 additions & 20 deletions PyTorch/SpeechSynthesis/FastPitch/install.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ while getopts "cu:" opt; do
esac
done
: ${CONDA-"false"} # default value
: ${USER-`woami`} # default value
: ${USER-`whoami`} # default value

if [ "$CONDA" = "true" ]
then
Expand All @@ -26,23 +26,16 @@ source /disk/scratch1/${USER}/miniconda3/bin/activate

SERVERNAME=`hostname -s`
conda create -n fastpitch_${SERVERNAME} python=3.8
source activate fastpitch_${SERVERNAME}

## Get a version of gcc > 5.0. The current anaconda default (June 2021) is 9.3 which seems to work (so far!)
conda install gcc_linux-64 gxx_linux-64
## Make these the default C and C++ compilers (you could also set CC and CXX environment variabes)
## But aliasing seems to work well enough
alias gcc=x86_64-conda_cos6-linux-gnu-cc
alias g++=x86_64-conda_cos6-linux-gnu-c++
## Get a version of gcc > 5.0 which works to compile apex
conda install -n fastpitch_${SERVERNAME} gcc_linux-64=8.4 gxx_linux-64=8.4
## activating the environment now sets CC and CXX environment variables
## to point to our gcc/g++
source activate fastpitch_${SERVERNAME}

export CUDA_HOME=/opt/cuda-10.2.89_440_33
conda install pytorch torchvision cudatoolkit=10.2 -c pytorch

conda uninstall pytorch

## Then we reinstall and this for some reason downgrades the gcc to 7 and then installing apex works/
conda install pytorch torchvision cudatoolkit=10.2 -c pytorch

## Apex
cd /disk/scratch1/${USER}/FastPitches/PyTorch/SpeechSynthesis/FastPitch/
git clone https://github.com/NVIDIA/apex
Expand All @@ -51,13 +44,8 @@ pip install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp
cd ../

## Python requirements
## ignore warnings around numba installation
pip install -r requirements.txt
pip install tqdm tensorboard
pip install librosa
pip install wandb
pip install llvmlite==0.35.0
## Ignore warning around here
pip install numba==0.49.1

## for logging
## if needed, create a free account here: https://app.wandb.ai/login?signup=true
Expand All @@ -70,7 +58,7 @@ export CUDA_VISIBLE_DEVICES=1
./scripts/download_fastpitch.sh
./scripts/download_waveglow.sh
mkdir output
python inference.py --cuda --fastpitch pretrained_models/fastpitch/nvidia_fastpitch_210824.pt --waveglow pretrained_models/waveglow/nvidia_waveglow256pyt_fp16.pt --wn-channels 256 -i phrases/devset10.tsv -o output/wavs_devset10
python inference.py --cuda --fastpitch pretrained_models/fastpitch/nvidia_fastpitch_210824.pt --waveglow pretrained_models/waveglow/nvidia_waveglow256pyt_fp16.pt --wn-channels 256 -i phrases/devset10.tsv -o output/wavs_devset10


## Get set up with LJ
Expand Down
11 changes: 8 additions & 3 deletions PyTorch/SpeechSynthesis/FastPitch/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,12 @@
matplotlib
numpy
git+https://github.com/NVIDIA/dllogger#egg=dllogger
inflect
librosa==0.8.0
llvmlite==0.35.0
matplotlib
numba==0.49.1
numpy
scipy
tensorboard
tensorboardX==2.0
git+git://github.com/NVIDIA/dllogger.git@26a0f8f1958de2c0c460925ff6102a4d2486d6cc#egg=dllogger
tqdm
wandb