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

fix install_beamsearch_decoders #7011

Merged
merged 4 commits into from
Jul 13, 2023
Merged

fix install_beamsearch_decoders #7011

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
c4804aa
aliases
karpnv Jul 11, 2023
3b0e941
add NEMO_PATH
karpnv Jul 11, 2023
6e29fad
Merge branch 'main' of https://github.com/NVIDIA/NeMo into karpnv/ins…
karpnv Jul 12, 2023
a7ce1b3
expand_aliases
karpnv Jul 12, 2023
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
13 changes: 9 additions & 4 deletions scripts/asr_language_modeling/ngram_lm/install_beamsearch_decoders.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,8 @@
# limitations under the License.

# Use this script to install KenLM, OpenSeq2Seq decoder, Flashlight decoder
shopt -s expand_aliases

NEMO_PATH=/workspace/nemo # Path to NeMo folder: /workspace/nemo if you use NeMo/Dockerfile
if [ "$#" -eq 1 ]
then
Expand All @@ -24,11 +26,14 @@ KENLM_MAX_ORDER=10 # Maximum order of KenLM model, also specified in the setup_o
cd $NEMO_PATH

if [ $(id -u) -eq 0 ]; then
alias sudo=eval
alias aptupdate='apt-get update'
alias b2install='./b2'
else
alias aptupdate='sudo apt-get update'
alias b2install='sudo ./b2'
fi

sudo apt-get update && apt-get upgrade -y && apt-get install -y liblzma-dev && rm -rf /var/lib/apt/lists/* # liblzma needed for flashlight decoder

aptupdate && apt-get upgrade -y && apt-get install -y liblzma-dev && rm -rf /var/lib/apt/lists/* # liblzma needed for flashlight decoder'

git clone https://github.com/NVIDIA/OpenSeq2Seq
cd OpenSeq2Seq
Expand All @@ -42,7 +47,7 @@ cp $NEMO_PATH/scripts/installers/setup_os2s_decoders.py ./setup.py
./setup.sh

# install Boost package for KenLM
wget https://boostorg.jfrog.io/artifactory/main/release/1.80.0/source/boost_1_80_0.tar.bz2 --no-check-certificate && tar --bzip2 -xf $NEMO_PATH/decoders/boost_1_80_0.tar.bz2 && cd boost_1_80_0 && ./bootstrap.sh && sudo ./b2 --layout=tagged link=static,shared threading=multi,single install -j4 || echo FAILURE
wget https://boostorg.jfrog.io/artifactory/main/release/1.80.0/source/boost_1_80_0.tar.bz2 --no-check-certificate && tar --bzip2 -xf $NEMO_PATH/decoders/boost_1_80_0.tar.bz2 && cd boost_1_80_0 && ./bootstrap.sh && b2install --layout=tagged link=static,shared threading=multi,single install -j4 || echo FAILURE
export BOOST_ROOT=$NEMO_PATH/decoders/boost_1_80_0

# install KenLM
Expand Down
20 changes: 11 additions & 9 deletions tutorials/asr/Offline_ASR.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,9 +3,7 @@
"nbformat_minor": 0,
"metadata": {
"colab": {
"name": "Offline_ASR.ipynb",
"provenance": [],
"collapsed_sections": [],
"toc_visible": true
},
"kernelspec": {
Expand All @@ -31,7 +29,9 @@
"\n",
"You may find more info on how to train and use language models for ASR models here:\n",
"https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/asr_language_modeling.html\n",
"\n\nNOTE: User is responsible for checking the content of datasets and the applicable licenses and determining if suitable for the intended use.\n"
"\n",
"\n",
"NOTE: User is responsible for checking the content of datasets and the applicable licenses and determining if suitable for the intended use.\n"
]
},
{
Expand All @@ -41,7 +41,7 @@
},
"source": [
"## Installation\n",
"NeMo can be installed via simple pip command. \n",
"NeMo can be installed via simple pip command.\n",
"\n",
"Optional CTC beam search decoder might require restart of Colab runtime after installation."
]
Expand Down Expand Up @@ -77,12 +77,14 @@
" import ctc_decoders\n",
"except ModuleNotFoundError:\n",
" # install beam search decoder\n",
" import os\n",
" !apt-get update && apt-get install -y swig\n",
" !git clone https://github.com/NVIDIA/NeMo -b \"$BRANCH\"\n",
" !cd NeMo && bash scripts/asr_language_modeling/ngram_lm/install_beamsearch_decoders.sh\n",
" pwd = !pwd\n",
" NEMO_PATH = os.path.join(pwd[0], \"NeMo\")\n",
" !cd NeMo && bash scripts/asr_language_modeling/ngram_lm/install_beamsearch_decoders.sh $NEMO_PATH\n",
" print('Restarting Colab runtime to successfully import built module.')\n",
" print('Please re-run the notebook.')\n",
" import os\n",
" os.kill(os.getpid(), 9)"
],
"execution_count": null,
Expand Down Expand Up @@ -110,7 +112,7 @@
},
"source": [
"## Instantiate pre-trained NeMo model\n",
"``from_pretrained(...)`` API downloads and initializes model directly from the cloud. \n",
"``from_pretrained(...)`` API downloads and initializes model directly from the cloud.\n",
"\n",
"Alternatively, ``restore_from(...)`` allows loading a model from a disk.\n",
"\n",
Expand Down Expand Up @@ -403,7 +405,7 @@
"source": [
"## Offline inference with beam search decoder and N-gram language model re-scoring\n",
"\n",
"It is possible to use an external [KenLM](https://kheafield.com/code/kenlm/)-based N-gram language model to rescore multiple transcription candidates. \n",
"It is possible to use an external [KenLM](https://kheafield.com/code/kenlm/)-based N-gram language model to rescore multiple transcription candidates.\n",
"\n",
"Let's download and preprocess LibriSpeech 3-gram language model."
]
Expand Down Expand Up @@ -653,4 +655,4 @@
"outputs": []
}
]
}
}