chore: import upstream snapshot with attribution
Create PR to main with cherry-pick from release / cherry-pick (push) Failing after 0s
CICD NeMo / pre-flight (push) Failing after 0s
CICD NeMo / configure (push) Has been skipped
Build, validate, and release Neural Modules / pre-flight (push) Failing after 1s
CICD NeMo / code-linting (push) Has been skipped
Build, validate, and release Neural Modules / release (push) Has been skipped
Build, validate, and release Neural Modules / release-summary (push) Has been cancelled
CICD NeMo / cicd-test-container-build (push) Has been cancelled
CICD NeMo / cicd-import-tests (push) Has been cancelled
CICD NeMo / L0_Setup_Test_Data_And_Models (push) Has been cancelled
CICD NeMo / cicd-main-unit-tests (push) Has been cancelled
CICD NeMo / cicd-main-speech (push) Has been cancelled
CICD NeMo / Nemo_CICD_Test (push) Has been cancelled
CICD NeMo / Coverage (e2e) (push) Has been cancelled
CICD NeMo / Coverage (unit-test) (push) Has been cancelled
CodeQL / Analyze (python) (push) Has been cancelled
CICD NeMo / cicd-wait-in-queue (push) Has been cancelled

This commit is contained in:
wehub-resource-sync
2026-07-13 13:28:58 +08:00
commit ba4be087d5
2316 changed files with 2668701 additions and 0 deletions
+30
View File
@@ -0,0 +1,30 @@
#!/usr/bin/env bash
# Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Use this script to install KenLM, OpenSeq2Seq decoder, Flashlight decoder, OpenGRM Ngram tool to contaner
# How to use? Build it from NeMo root folder:
# 1. git clone https://github.com/NVIDIA/NeMo.git && cd NeMo
# 2. DOCKER_BUILDKIT=1 docker build -t nemo:23.03.1 -f ./scripts/installers/Dockerfile.ngramtools .
from nvcr.io/nvidia/nemo:23.03
WORKDIR /workspace/nemo
COPY scripts/. /workspace/nemo/scripts/
RUN /bin/bash scripts/asr_language_modeling/ngram_lm/install_beamsearch_decoders.sh
RUN /bin/bash scripts/installers/install_opengrm.sh
@@ -0,0 +1,48 @@
# Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This Dockerfile installs the necessary dependencies for streaming speech translation with vLLM.
ARG BASE_IMAGE=pytorch/pytorch:2.9.0-cuda12.8-cudnn9-runtime
FROM ${BASE_IMAGE}
ARG GIT_URL="https://github.com/NVIDIA-NeMo/NeMo.git"
ARG CHECKOUT="main"
ENV DEBIAN_FRONTEND=noninteractive
# Install dependencies
RUN apt-get update && \
apt-get upgrade -y && \
apt-get install -y \
libtool libltdl-dev \
libsndfile1 sox \
libfreetype6 \
libsox-fmt-all \
swig \
ffmpeg \
libavdevice-dev \
git \
build-essential \
cmake && \
rm -rf /var/lib/apt/lists/*
# Install NeMo
RUN git clone --recursive ${GIT_URL} nemo_dir && cd nemo_dir && \
git checkout $CHECKOUT && \
pip install ".[all]"
# Install vLLM for streaming speech translation
RUN pip install vllm==0.12.0
+30
View File
@@ -0,0 +1,30 @@
#!/bin/bash
# Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
echo "Install latest AIS CLI"
AIS_CLI_URL=https://github.com/NVIDIA/aistore/releases/latest/download/ais-linux-amd64.tar.gz
echo "Download AIS CLI from ${AIS_CLI_URL}"
curl -LO ${AIS_CLI_URL}
echo "Extract"
tar -xzvf ais-linux-amd64.tar.gz
echo "Move to /usr/local/bin/"
mv ./ais /usr/local/bin/.
echo "Cleanup"
rm ais-linux-amd64.tar.gz
+51
View File
@@ -0,0 +1,51 @@
#!/bin/bash
# Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
DOCKER=false
MAYBE_SUDO=""
GRAPHVIZ_REPO=https://gitlab.com/graphviz/graphviz.git
GRAPHVIZ_LATEST_RELEASE=9.0.0
GRAPHVIZ_PY_REPO=https://github.com/xflr6/graphviz
GRAPHVIZ_PY_LATEST_RELEASE=448d1a0 # Temporary fix until 0.20.2
if [[ $* == *--docker* ]]; then
echo "Docker installation"
DOCKER=true
else
echo "Local installation"
if [[ $(sudo -n -v 2) ]]; then
MAYBE_SUDO="sudo"
else
echo "No sudo detected"
fi
fi
{
$MAYBE_SUDO apt-get update
$MAYBE_SUDO apt-get remove -y graphviz
pip uninstall -y graphviz
if [[ $DOCKER == false ]]; then
$MAYBE_SUDO apt-get install -y libtool libltdl-dev automake autoconf bison flex tcl \
ghostscript libgd-dev fontconfig libcairo2-dev libpango1.0-dev libgts-dev
fi
git clone ${GRAPHVIZ_REPO} -b ${GRAPHVIZ_LATEST_RELEASE} && cd graphviz
./autogen.sh && ./configure --disable-python --disable-perl
$MAYBE_SUDO make -j && $MAYBE_SUDO make install
cd .. && $MAYBE_SUDO rm -rf graphviz
pip install -v "git+${GRAPHVIZ_PY_REPO}@${GRAPHVIZ_PY_LATEST_RELEASE}#egg=graphviz"
} || { echo "graphviz installed with errors! Please check installation manually."; exit 1; }
echo "graphviz (re-) installed successfully!"
+29
View File
@@ -0,0 +1,29 @@
#!/bin/bash
# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
K2_REPO=https://github.com/k2-fsa/k2
LATEST_RELEASE=0a09f67 # fix for PyTorch 2.6.0
# uncomment the following line after the next k2 version is released (>1.24.4)
#LATEST_RELEASE=$(git -c 'versionsort.suffix=-' \
# ls-remote --exit-code --refs --sort='version:refname' --tags ${K2_REPO} '*.*' \
# | tail --lines=1 \
# | cut -d '/' -f 3)
# "cut --delimiter '/' --fields 3" doesn't work on macOS, use "-d ... -f ..." instead
pip install wheel setuptools cmake
K2_MAKE_ARGS="-j" pip install -v "git+${K2_REPO}@${LATEST_RELEASE}#egg=k2" || { echo "k2 could not be installed!"; exit 1; }
python3 -m k2.version > /dev/null || { echo "k2 installed with errors! Please check installation manually."; exit 1; }
echo "k2 installed successfully!"
+32
View File
@@ -0,0 +1,32 @@
#!/bin/bash
# Copyright (c) 2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This script install OpenFST and Ngram tools from OpenGRM library
# Optionally, you can specify a path where to install it as a first positional argument: scripts/installers/install_opengrm.sh /path/to/install/openfst .
# Alternatively, in the Linux Debian you can use: sudo apt install libngram-tools
DECODERS_PATH=/workspace/nemo/decoders # Path to decoders folder: /workspace/nemo/decoders if you use NeMo/Dockerfile
if [ "$#" -eq 1 ]
then
DECODERS_PATH=$1
fi
cd $DECODERS_PATH
# Install OpenGrm OpenFST
wget https://www.openfst.org/twiki/pub/FST/FstDownload/openfst-1.8.2.tar.gz && tar xvzf openfst-1.8.2.tar.gz && cd openfst-1.8.2 && ./configure --enable-grm && make -j4 && make -j4 install && cd ..
# Install OpenGrm Ngram
OPENFSTPREFIX=$DECODERS_PATH/openfst-1.8.2/src && wget https://www.opengrm.org/twiki/pub/GRM/NGramDownload/ngram-1.3.14.tar.gz && tar xvzf ngram-1.3.14.tar.gz && cd ngram-1.3.14 && LDFLAGS="-L${OPENFSTPREFIX}/lib" CXXFLAGS="-I${OPENFSTPREFIX}/include" ./configure --prefix ${OPENFSTPREFIX} && make -j4 && make -j4 install && cd ..
+17
View File
@@ -0,0 +1,17 @@
#!/bin/bash
# Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
pip install kaldifst kaldilm riva-asrlib-decoder
+144
View File
@@ -0,0 +1,144 @@
# Copyright (c) 2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Script to build and install decoder package.
It is used by scripts/asr_language_modeling/ngram_lm/install_beamsearch_decoders.sh to install
KenLM and OpenSeq2Seq decoder.
You can set the order of KenLM model by changing -DKENLM_MAX_ORDER=10 argument.
"""
from __future__ import absolute_import, division, print_function
import argparse
import distutils.ccompiler
import glob
import multiprocessing.pool
import os
import platform
import sys
from setuptools import Extension, distutils, setup
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
"--num_processes", default=1, type=int, help="Number of cpu processes to build package. (default: %(default)d)"
)
args = parser.parse_known_args()
# reconstruct sys.argv to pass to setup below
sys.argv = [sys.argv[0]] + args[1]
# monkey-patch for parallel compilation
# See: https://stackoverflow.com/a/13176803
def parallelCCompile(
self,
sources,
output_dir=None,
macros=None,
include_dirs=None,
debug=0,
extra_preargs=None,
extra_postargs=None,
depends=None,
):
# those lines are copied from distutils.ccompiler.CCompiler directly
macros, objects, extra_postargs, pp_opts, build = self._setup_compile(
output_dir, macros, include_dirs, sources, depends, extra_postargs
)
cc_args = self._get_cc_args(pp_opts, debug, extra_preargs)
# parallel code
def _single_compile(obj):
try:
src, ext = build[obj]
except KeyError:
return
self._compile(obj, src, ext, cc_args, extra_postargs, pp_opts)
# convert to list, imap is evaluated on-demand
thread_pool = multiprocessing.pool.ThreadPool(args[0].num_processes)
list(thread_pool.imap(_single_compile, objects))
return objects
def compile_test(header, library):
dummy_path = os.path.join(os.path.dirname(__file__), "dummy")
command = (
"bash -c \"g++ -include "
+ header
+ " -l"
+ library
+ " -x c++ - <<<'int main() {}' -o "
+ dummy_path
+ " >/dev/null 2>/dev/null && rm "
+ dummy_path
+ " 2>/dev/null\""
)
return os.system(command) == 0
# hack compile to support parallel compiling
distutils.ccompiler.CCompiler.compile = parallelCCompile
FILES = glob.glob('kenlm/util/*.cc') + glob.glob('kenlm/lm/*.cc') + glob.glob('kenlm/util/double-conversion/*.cc')
FILES += glob.glob('openfst-1.6.3/src/lib/*.cc')
FILES = [fn for fn in FILES if not (fn.endswith('main.cc') or fn.endswith('test.cc') or fn.endswith('unittest.cc'))]
LIBS = ['stdc++']
if platform.system() != 'Darwin':
LIBS.append('rt')
ARGS = ['-O3', '-DNDEBUG', '-DKENLM_MAX_ORDER=10', '-std=c++11']
if compile_test('zlib.h', 'z'):
ARGS.append('-DHAVE_ZLIB')
LIBS.append('z')
if compile_test('bzlib.h', 'bz2'):
ARGS.append('-DHAVE_BZLIB')
LIBS.append('bz2')
if compile_test('lzma.h', 'lzma'):
ARGS.append('-DHAVE_XZLIB')
LIBS.append('lzma')
os.system('swig -python -c++ ./decoders.i')
decoders_module = [
Extension(
name='_swig_decoders',
sources=FILES + glob.glob('*.cxx') + glob.glob('*.cpp'),
language='c++',
include_dirs=[
'.',
'kenlm',
'openfst-1.6.3/src/include',
'ThreadPool',
],
libraries=LIBS,
extra_compile_args=ARGS,
)
]
setup(
name='ctc_decoders',
version='1.1',
description="""CTC decoders""",
ext_modules=decoders_module,
py_modules=['ctc_decoders', 'swig_decoders'],
)