NeMo installation - done; performance to be evaluated

@zilles

you may like to try https://ngc.nvidia.com/catalog/containers/nvidia:jetson-voice

NeMo issues may need to be posted to there in order to be addressed https://github.com/NVIDIA/NeMo/issues/

However, what are the exact steps to reproduce the issue? do some tests fail or all of them?
" Also, you can try running inference of a small dataset like an4 (see ASR notebook tutorial)" [okuchaiev, source github]

I tried the speech to text example step by step now:

python3
Python 3.6.9 (default, Oct 8 2020, 12:12:24)
[GCC 8.4.0] on linux
Type “help”, “copyright”, “credits” or “license” for more information.

import pytorch_lightning as pl
from omegaconf import OmegaConf
from nemo.collections.asr.models import EncDecCTCModel
[NeMo W 2020-12-13 14:51:09 experimental:28] Module <class ‘nemo.collections.asr.losses.ctc.CTCLoss’> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-13 14:51:12 experimental:28] Module nemo.collections.asr.data.audio_to_text.AudioToCharDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-13 14:51:12 experimental:28] Module nemo.collections.asr.data.audio_to_text.AudioToBPEDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-13 14:51:12 experimental:28] Module nemo.collections.asr.data.audio_to_text.AudioLabelDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-13 14:51:12 experimental:28] Module nemo.collections.asr.data.audio_to_text.TarredAudioToTextDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-13 14:51:12 experimental:28] Module nemo.collections.asr.data.audio_to_text.TarredAudioToCharDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-13 14:51:12 experimental:28] Module nemo.collections.asr.data.audio_to_text.TarredAudioToBPEDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
Segmentation fault (core dumped)

I get the fault although import Nemo itself would works fine - I guess there is something wrong in the air-part of nemo?

@zilles
there is targeted in jetson NGC deployment solution that should just work on Jetsons out of the box
regarding the manual assembly NeMo at Jetson, once I have a chance I will try to test it with the latest Jetson release to see if the error will come up.
Until then I suggest to try the dockerized deployment/ posting to github
it is highly likely version issue, in my opinion
some component might need to be switched to some different version maybe
However, the voice to text example using dockerized deployment you will run with
“Streaming transcript of live microphone input or pre-recorded wav file:”

$ python3 src/test_asr.py --mic 24

Thank you for your help - but I am not familiar with docker and wanted to install locally so I can program my own training and inference code to be implemented into a n app running on the Xavier…

by looking at the versions installed pip3 list gives:
Package Version Location


absl-py 0.10.0
alabaster 0.7.12
antlr4-python3-runtime 4.8
apex 0.1
appdirs 1.4.4
apturl 0.5.2
argon2-cffi 20.1.0
asn1crypto 0.24.0
attrdict 2.0.1
attrs 20.1.0
audioread 2.1.8
Babel 2.8.0
backcall 0.2.0
beautifulsoup4 4.6.0
black 19.10b0
bleach 3.1.5
blinker 1.4
boto3 1.14.54
botocore 1.17.54
braceexpand 0.1.5
Brlapi 0.6.6
cachetools 4.1.1
certifi 2020.6.20
cffi 1.14.2
chardet 3.0.4
click 7.1.2
configparser 5.0.1
cryptography 2.1.4
cupshelpers 1.0
cycler 0.10.0
Cython 0.29.21
dataclasses 0.7
decorator 4.1.2
defer 1.0.6
defusedxml 0.6.0
Distance 0.1.3
distro-info 0.18ubuntu0.18.04.1
docker-pycreds 0.4.0
docopt 0.6.2
docutils 0.15
editdistance 0.5.3
entrypoints 0.3
feedparser 5.2.1
filelock 3.0.12
frozendict 1.2
fsspec 0.8.4
future 0.18.2
g2p-en 2.1.0
gdown 3.12.2
gitdb 4.0.5
GitPython 3.1.11
google-auth 1.21.0
google-auth-oauthlib 0.4.1
graphsurgeon 0.4.5
grpcio 1.31.0
h5py 2.10.0
html2text 2020.1.16
html5lib 0.999999999
httplib2 0.9.2
hydra-core 1.0.4
idna 2.6
imagesize 1.2.0
importlib-metadata 1.7.0
importlib-resources 3.3.0
inflect 4.1.0
iniconfig 1.0.1
ipdb 0.13.3
ipykernel 5.3.4
ipyparallel 6.3.0
ipython 7.16.1
ipython-genutils 0.2.0
ipywidgets 7.5.1
isort 4.3.21
jedi 0.17.2
Jetson.GPIO 2.0.11
Jinja2 2.11.2
jmespath 0.10.0
joblib 0.16.0
json5 0.9.5
jsonschema 3.2.0
jupyter-client 6.1.7
jupyter-core 4.6.3
jupyterlab 2.2.6
jupyterlab-server 1.2.0
kaldi-io 0.9.4
kaldi-python-io 1.1.2
keyring 10.6.0
keyrings.alt 3.0
kiwisolver 1.2.0
language-selector 0.1
latexcodec 2.0.1
launchpadlib 1.10.6
lazr.restfulclient 0.13.5
lazr.uri 1.0.3
librosa 0.7.2
linecache2 1.0.0
llvmlite 0.31.0
louis 3.5.0
lxml 4.2.1
macaroonbakery 1.1.3
Mako 1.0.7
Markdown 3.2.2
MarkupSafe 1.0
marshmallow 3.7.1
matplotlib 3.3.3
megatron-lm 1.1.5
mistune 0.8.4
more-itertools 8.5.0
msgpack 1.0.0
mypy-extensions 0.4.3
nbconvert 5.6.1
nbformat 5.0.7
nemo-toolkit 1.0.0b2 /data/NeMo-main
nltk 3.5
nose 1.3.7
notebook 6.1.3
num2words 0.5.10
numba 0.48.0
numpy 1.19.1
oauth 1.0.1
oauthlib 3.1.0
objectio 0.2.29
omegaconf 2.0.5
onboard 1.4.1
onnx 1.7.0
onnxruntime-gpu 1.4.0
oset 0.1.3
packaging 20.4
pandas 0.22.0
pandocfilters 1.4.2
parameterized 0.7.4
parso 0.7.1
pathspec 0.8.0
pathtools 0.1.2
pbr 3.1.1
pesq 0.0.1
pexpect 4.8.0
pickleshare 0.7.5
Pillow 7.2.0
pip 20.3.1
pip-api 0.0.14
pipreqs 0.4.10
pluggy 0.13.1
progressbar 2.5
prometheus-client 0.8.0
promise 2.3
prompt-toolkit 3.0.7
protobuf 3.14.0
psutil 5.7.3
ptyprocess 0.6.0
py 1.9.0
pyasn1 0.4.8
pyasn1-modules 0.2.8
pybind11 2.5.0
pybtex 0.22.2
pybtex-docutils 0.2.2
pycairo 1.16.2
pycparser 2.20
pycrypto 2.6.1
pycups 1.9.73
Pygments 2.6.1
pygobject 3.26.1
PyJWT 1.5.3
pymacaroons 0.13.0
PyNaCl 1.1.2
pyparsing 2.4.7
pypinyin 0.39.0
pyRFC3339 1.0
pyrsistent 0.16.0
PySocks 1.7.1
pystoi 0.3.3
pytest 6.0.1
pytest-runner 5.2
python-apt 1.6.5+ubuntu0.4
python-dateutil 2.6.1
python-debian 0.1.32
pytorch-lightning 1.0.8
pytz 2018.3
pyxdg 0.25
PyYAML 5.3.1
pyzmq 19.0.2
qtconsole 4.7.7
QtPy 1.9.0
rapidfuzz 0.11.1
regex 2020.7.14
requests 2.24.0
requests-oauthlib 1.3.0
requests-unixsocket 0.1.5
resampy 0.2.2
rouge-score 0.0.4
rsa 4.6
ruamel.yaml 0.16.10
ruamel.yaml.clib 0.2.0
s3transfer 0.3.3
sacremoses 0.0.43
scikit-learn 0.23.2
scipy 1.5.4
SecretStorage 2.3.1
semantic-version 2.8.5
Send2Trash 1.5.0
sentencepiece 0.1.93
sentry-sdk 0.19.4
setuptools 51.0.0
setuptools-rust 0.11.3
shortuuid 1.0.1
simplejson 3.13.2
six 1.15.0
smmap 3.0.4
snowballstemmer 2.0.0
SoundFile 0.10.3.post1
sox 1.4.0
Sphinx 3.2.1
sphinx-rtd-theme 0.5.0
sphinxcontrib-applehelp 1.0.2
sphinxcontrib-bibtex 1.0.0
sphinxcontrib-devhelp 1.0.2
sphinxcontrib-htmlhelp 1.0.3
sphinxcontrib-jsmath 1.0.1
sphinxcontrib-qthelp 1.0.3
sphinxcontrib-serializinghtml 1.1.4
ssh-import-id 5.7
subprocess32 3.5.4
system-service 0.3
systemd-python 234
tensorboard 2.3.0
tensorboard-plugin-wit 1.7.0
tensorboardX 2.1
tensorrt 7.1.3.0
terminado 0.8.3
testpath 0.4.4
testresources 2.0.0
threadpoolctl 2.1.0
tokenizers 0.9.4
toml 0.10.1
torch 1.7.0
torch-stft 0.1.4
torchtext 0.6.0
torchvision 0.8.0a0+45f960c
tornado 6.0.4
tqdm 4.48.2
traceback2 1.4.0
traitlets 4.3.3
transformers 4.0.1
typed-ast 1.4.1
typer 0.3.2
typing-extensions 3.7.4.2
ubuntu-drivers-common 0.0.0
uff 0.6.9
Unidecode 1.1.1
unittest2 1.1.0
unity-scope-calculator 0.1
unity-scope-chromiumbookmarks 0.1
unity-scope-colourlovers 0.1
unity-scope-devhelp 0.1
unity-scope-firefoxbookmarks 0.1
unity-scope-manpages 0.1
unity-scope-openclipart 0.1
unity-scope-texdoc 0.1
unity-scope-tomboy 0.1
unity-scope-virtualbox 0.1
unity-scope-yelp 0.1
unity-scope-zotero 0.1
urllib3 1.22
urwid 2.0.1
virtualenv 15.1.0
wadllib 1.3.2
wandb 0.10.12
watchdog 0.10.4
wcwidth 0.2.5
webdataset 0.1.37
webencodings 0.5
Werkzeug 1.0.1
wget 3.2
wheel 0.30.0
widgetsnbextension 3.5.1
wrapt 1.12.1
xkit 0.0.0
yarg 0.1.9
youtokentome 1.0.6
zipp 3.1.0
zope.interface 4.3.2

at least all minimum requirements for nemo and nemo asr are met…but maybe you already know some incompatibilities by looking at the list?

Hi zilles,
I got the similar error as you , i succeed install nemo on jetson nano , and “import nemo” works , but when i “import nemo.collections.asr” , it raise error "segmentation fault core dumped " and python3.6.9 crashed …
finally do you figure it out ?

@Andrey1984 could you give some advices ? many thanks !

@yipengl I would assume it is a version issue
However, it seems that the issue is being worked out in the github issues https://github.com/NVIDIA/NeMo/issues/1552
In my opinion, It is much easier to run the dockerized jetson implementation that you may like to try, as there will be no chance to run into issues with environment as it has been supplied with the context in a form of a docker container https://ngc.nvidia.com/catalog/containers/nvidia:jetson-voice
Moreover, once I have a chance to complete installation with the latest JetPack I will try to do so.
some reference https://colab.research.google.com/github/NVIDIA/NeMo/blob/main/tutorials/asr/01_ASR_with_NeMo.ipynb
https://github.com/NVIDIA/NeMo#installation
lets try with the latter url:

git clone https://github.com/NVIDIA/NeMo
cd NeMo
./reinstall.sh 

first attempt obviously fails.
I can recollect I had to install multiple packages manually including llvm also specifc versions
e.g. llvmlite
ref https://github.com/jefflgaol/Install-Packages-Jetson-ARM-Family#llvm-installation
Here we will try with these steps

wget https://github.com/llvm/llvm-project/releases/download/llvmorg-9.0.1/llvm-9.0.1.src.tar.xz
tar -xvf  llvm-9.0.1.src.tar.xz
cd llvm-9.0.1.src/
mkdir build
cd build
cmake ../ -DCMAKE_BUILD_TYPE=Release -DLLVM_TARGETS_TO_BUILD="ARM;X86;AArch64"
make -j8
sudo make install
cd bin
echo "export LLVM_CONFIG=\""`pwd`"/llvm-config\"" >> ~/.bashrc
echo "alias llvm='"`pwd`"/llvm-lit'" >> ~/.bashrc
source ~/.bashrc
sudo pip3 install llvmlite

there also seems to exist a shortcut
etc.

Hi Andrey ,
Thanks for your reply , yeah it is complex to install nemo on jetson, cos so many dependance packages need to install manually , i struggle some days succeed install it .
In my opinion mainly is version of “llvm ,llvmlite ,numba, sentencepiece and h5py” need pay more atteintion
Then i run ‘import nemo’ and ‘import nemo.core’ in python3.6.9 works well , But unfortunately ‘import nemo.collections.asr’ make python crashed with
segmentation fault core dumped error …
I raised this issue on https://github.com/NVIDIA/NeMo/issues/1552 haven’t yet find a solution.
I followed up your solution , grateful for your contribution . i will try with docker container https://ngc.nvidia.com/catalog/containers/nvidia:jetson-voice too
My purpose is similar as you to run https://colab.research.google.com/github/NVIDIA/NeMo/blob/main/tutorials/asr/01_ASR_with_NeMo.ipynb , to use nemo do more tasks [asr nlp tts ] and transfer-learning in jetson enviroment.
let us try to work it out . i will share how i install nemo on jetson after test .

@yipengl
NeMo is just a prototype of Jarvis SDK,
While the former is intended for testing /development rather than production
you should be able to use the latter

Yes , you are right good advice! if you can achieve run this https://colab.research.google.com/github/NVIDIA/NeMo/blob/main/tutorials/asr/01_ASR_with_NeMo.ipynb with nemo , please let me know too , i follow you up . very thankful!

@yipengl right now I am getting this

import nemo.collections.asr as nemo_asr
[NeMo W 2020-12-30 10:18:09 experimental:28] Module <class 'nemo.collections.asr.losses.ctc.CTCLoss'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[nltk_data] Downloading package averaged_perceptron_tagger to
[nltk_data]     /home/nvidia/nltk_data...
[nltk_data]   Unzipping taggers/averaged_perceptron_tagger.zip.
[nltk_data] Downloading package cmudict to /home/nvidia/nltk_data...
[nltk_data]   Unzipping corpora/cmudict.zip.
Traceback (most recent call last):
  File "<stdin>", line 3, in <module>
  File "/home/nvidia/NeMo/nemo/collections/asr/__init__.py", line 15, in <module>
    from nemo.collections.asr import data, losses, models, modules
  File "/home/nvidia/NeMo/nemo/collections/asr/models/__init__.py", line 16, in <module>
    from nemo.collections.asr.models.classification_models import EncDecClassificationModel
  File "/home/nvidia/NeMo/nemo/collections/asr/models/classification_models.py", line 27, in <module>
    from nemo.collections.asr.data.audio_to_text import AudioLabelDataset
  File "/home/nvidia/NeMo/nemo/collections/asr/data/audio_to_text.py", line 25, in <module>
    from nemo.collections.asr.parts.features import WaveformFeaturizer
  File "/home/nvidia/NeMo/nemo/collections/asr/parts/features.py", line 37, in <module>
    import librosa
  File "/home/nvidia/.local/lib/python3.6/site-packages/librosa/__init__.py", line 211, in <module>
    from . import core
  File "/home/nvidia/.local/lib/python3.6/site-packages/librosa/core/__init__.py", line 5, in <module>
    from .convert import *  # pylint: disable=wildcard-import
  File "/home/nvidia/.local/lib/python3.6/site-packages/librosa/core/convert.py", line 7, in <module>
    from . import notation
  File "/home/nvidia/.local/lib/python3.6/site-packages/librosa/core/notation.py", line 8, in <module>
    from ..util.exceptions import ParameterError
  File "/home/nvidia/.local/lib/python3.6/site-packages/librosa/util/__init__.py", line 83, in <module>
    from .utils import *  # pylint: disable=wildcard-import
  File "/home/nvidia/.local/lib/python3.6/site-packages/librosa/util/utils.py", line 10, in <module>
    import numba
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/__init__.py", line 196, in <module>
    import numba.typed
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/typed/__init__.py", line 3, in <module>
    from .typeddict import Dict
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/typed/typeddict.py", line 19, in <module>
    @njit
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/decorators.py", line 238, in njit
    return jit(*args, **kws)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/decorators.py", line 175, in jit
    return wrapper(pyfunc)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/decorators.py", line 191, in wrapper
    **dispatcher_args)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/dispatcher.py", line 650, in __init__
    self.targetctx = self.targetdescr.target_context
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/targets/registry.py", line 50, in target_context
    return self._toplevel_target_context
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/utils.py", line 390, in __get__
    res = instance.__dict__[self.name] = self.func(instance)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/targets/registry.py", line 34, in _toplevel_target_context
    return cpu.CPUContext(self.typing_context)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/targets/base.py", line 260, in __init__
    self.init()
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/compiler_lock.py", line 32, in _acquire_compile_lock
    return func(*args, **kwargs)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/targets/cpu.py", line 51, in init
    self._internal_codegen = codegen.JITCPUCodegen("numba.exec")
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/targets/codegen.py", line 628, in __init__
    self._init(self._llvm_module)
  File "/home/nvidia/.local/lib/python3.6/site-packages/numba/targets/codegen.py", line 637, in _init
    tm = target.create_target_machine(**tm_options)
TypeError: create_target_machine() got an unexpected keyword argument 'jitdebug'

it is also a version issue peculiar to python3.6 / [ numba, cuml, llvm]
you may also like to have a look at https://github.com/NVIDIA/tacotron2
https://developer.nvidia.com/blog/how-to-deploy-real-time-text-to-speech-applications-on-gpus-using-tensorrt/

@Andrey1984 I used llvm 0.7.1 llvmlite0.30.0 ,numba==0.40.0 then update tbb version solve my issus for building numba . but the issus from nemo.collections.asr i have no idea too right now . maybe you can raise your issues here too https://github.com/NVIDIA/NeMo/issues/1552

@yipengl
I would say it rather works with

llvmlite-0.32.1
#numba-0.49.1
>>> import nemo
>>> import nemo.collections.asr as nemo_asr
[NeMo W 2020-12-30 11:07:45 experimental:28] Module <class 'nemo.collections.asr.losses.ctc.CTCLoss'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-30 11:07:47 experimental:28] Module nemo.collections.asr.data.audio_to_text.AudioToCharDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-30 11:07:47 experimental:28] Module nemo.collections.asr.data.audio_to_text.AudioToBPEDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-30 11:07:47 experimental:28] Module nemo.collections.asr.data.audio_to_text.AudioLabelDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-30 11:07:47 experimental:28] Module nemo.collections.asr.data.audio_to_text.TarredAudioToTextDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-30 11:07:47 experimental:28] Module nemo.collections.asr.data.audio_to_text.TarredAudioToCharDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-12-30 11:07:47 experimental:28] Module nemo.collections.asr.data.audio_to_text.TarredAudioToBPEDataset is experimental, not ready for production and is not fully supported. Use at your own risk.
2020-12-30 11:07:48.293453: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.10.2
[NeMo W 2020-12-30 11:07:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text_dali.AudioToCharDALIDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
>>>

However, in my opinion, the point is that it will be more convinient to use jetson-voice container on jetson, but NeMo on Desktop with descrete GPU, as there is already dockerized container with nemo for x86_64 [NGC]
also Jasper

@Andrey1984 I try reinstall llvmlite-0.32.1 and numba-0.49.1 , i got a error as below :
Python 3.6.9 (default, Oct 8 2020, 12:12:24)
[GCC 8.4.0] on linux
Type “help”, “copyright”, “credits” or “license” for more information.

import nemo
import nemo.collections.asr as nemo_asr
Traceback (most recent call last):
File “”, line 1, in
File “/home/yipeng/.local/lib/python3.6/site-packages/nemo/collections/asr/init.py”, line 15, in
from nemo.collections.asr import data, losses, models, modules
File “/home/yipeng/.local/lib/python3.6/site-packages/nemo/collections/asr/models/init.py”, line 16, in
from nemo.collections.asr.models.classification_models import EncDecClassificationModel
File “/home/yipeng/.local/lib/python3.6/site-packages/nemo/collections/asr/models/classification_models.py”, line 24, in
from nemo.collections.asr.data.audio_to_text import AudioLabelDataset
File “/home/yipeng/.local/lib/python3.6/site-packages/nemo/collections/asr/data/audio_to_text.py”, line 25, in
from nemo.collections.asr.parts.features import WaveformFeaturizer
File “/home/yipeng/.local/lib/python3.6/site-packages/nemo/collections/asr/parts/features.py”, line 37, in
import librosa
File “/home/yipeng/.local/lib/python3.6/site-packages/librosa/init.py”, line 211, in
from . import core
File “/home/yipeng/.local/lib/python3.6/site-packages/librosa/core/init.py”, line 5, in
from .convert import * # pylint: disable=wildcard-import
File “/home/yipeng/.local/lib/python3.6/site-packages/librosa/core/convert.py”, line 7, in
from . import notation
File “/home/yipeng/.local/lib/python3.6/site-packages/librosa/core/notation.py”, line 8, in
from …util.exceptions import ParameterError
File “/home/yipeng/.local/lib/python3.6/site-packages/librosa/util/init.py”, line 83, in
from .utils import * # pylint: disable=wildcard-import
File “/home/yipeng/.local/lib/python3.6/site-packages/librosa/util/utils.py”, line 10, in
import numba
File “/usr/local/lib/python3.6/dist-packages/numba/init.py”, line 131, in
importlib.import_module(_old_mod)
File “/usr/lib/python3.6/importlib/init.py”, line 126, in import_module
return _bootstrap._gcd_import(name[level:], packageImportError: cannot import name ‘total_ordering’, level)
File “/usr/local/lib/python3.6/dist-packages/numba/typeconv/init.py”, line 1, in
from .castgraph import Conversion
File “/usr/local/lib/python3.6/dist-packages/numba/typeconv/castgraph.py”, line 6, in
from numba.utils import total_ordering
ImportError: cannot import name 'total_ordering’
you didn’t get this error with numba-0.49.1 ?
Then i found need to change version of numba-0.48.0 can solve this error , but it is return back old error segmentation fault core dumped again …
https://github.com/NVIDIA/NeMo/issues/1552

@yipengl
what are the steps to reproduce the error?
in my setup it seems working with numba 47 . However, I did not have chance to upgrade it to 49 in the last trial.
Although I am running it on Jetson AGX. I did not try on nano. But if you do it might need extra swap memory.

1 Like

@Andrey1984 finally my issue solved with numba-0.47.0 . super!thank you very much Andrey ,

Appreciate the insights from all, really helpful, trying to install Nemo on Jetson Nano, @Andrey1984, can you please tell me if there is a place wherein I can find straight forward instructions in one place for installing Nemo on Jetson Nano?

NeMo / Jarvis are primarily targeted x86_64. Howerver the former can be built from sources.
Moreover, there is also cloud-natve jetson voice container that supports only certain version of l4t but is easier in deployment
https://ngc.nvidia.com/catalog/containers/nvidia:jetson-voice
the latter will likely require to downgrade the jetpack to match the version supported by the container

Thanks @Andrey1984 , clear on the first one. Can you pass me the link for Cloud Native container? Sorry I can not find it and will save me time, TIA.