Linux install instructions

[mael-lenoc]

Hi,
Thanks a lot for all your work.
I've been able to install noScribe and noScribeEdit on linux (ubuntu).

Here are the steps I followed : I hope it might help some people (feel free to edit/reuse as you want).
FYI, I've had to make a few very minor edits in order for the two softwares to work well together. (I might make it cleaner and submit pull requests at some points...)

Installation Instructions:

Note : The procedure below uses a python virtual environment
Make sure to have python and pip installed
Download and extract the latest NoScribe sourcecode from https://github.com/kaixxx/noScribe/releases
Download and extract the latest NoScribeEdit sourcecode from https://github.com/kaixxx/noScribeEditor
Put the noScribeEditor inside the NoScribe folder you dowloaded previously, and rename the noScribeEditor folder as noScribeEdit
Open a terminal and cd in the NoScribe directory
Enter the following command lines :

	python3 -m venv .venv #you may need to replace python3 by python, depending on the version of python installed)
	source .venv/bin/activate
	pip install -r environments/requirements_linux.txt
	pip install -r noScribeEdit/environments/requirements_macOS.txt #Yes it says mac, but there is no linux requirements file at the moment

Download all the files from https://huggingface.co/guillaumekln/faster-whisper-large-v2/tree/main to the models/faster-whisper-large-v2 to the models/faster-whisper-large-v2 within your noScribe folder

Download all the files from https://huggingface.co/guillaumekln/faster-whisper-small/tree/main to the models/faster-whisper-small to the models/faster-whisper-small within your noScribe folder

Edit the noScribe.py file as follow in order to be able to open noScribeEditor from noScribe
Edit line 566 so that it reads
program = os.path.join(app_dir, 'noScribeEdit', "noScribeEdit.py")
Edit line 578 so that it reads
Popen(['python3', program, file], **kwargs) #might need to be python instead of python3 depending on your environment
Edit line 580 so that it reads
Popen(['python3', program], **kwargs)

Launching noScribe :

Open a terminal and cd into the directory where the noScribe sources are located
Activate your virtual environment with
source .venv/bin/activate
Run
python3 noScribe.py

HiDPI issue :

If you have a HiDpi screen, noScribe window and font might be very small.
You can add the following lines to the noScribe.py file after line 70

ctk.set_widget_scaling(2)  # widget dimensions and text size / adjust the value as you see fit
ctk.set_window_scaling(2)  # window geometry dimension / adjust the value as you see fit

Sidenote : Sometimes, when you select a filename for saving, the "." is missing before the extension by default and this throws an error. Just manually add the dot.

[kaixxx]

Thank you very much for these detailed instructions! I will link to this post from the main Readme.

[mael-lenoc]

You're welcome.
And FYI, I've also encountered a "segmentation fault" error, but using the fix suggested here #75 works.

[kaixxx]

I've also encountered a "segmentation fault" error, but using the fix suggested here #75 works.

Thank you for reminding me of this problem. As you can see in the discussion about the pull request, I had some questions, but the original author never got back to me. Could you help me with that (especially the second bullet point in question 2 - checking self.log_textbox.winfo_exists() as a simple solution)?

[mael-lenoc]

Hi,
I'm not sure I fully understand how the code suggested by the original author works, but I've tried the "simple solution" that you suggested (i.e. replacing if where != 'file': by if where != 'file' and self.log_textbox.winfo_exists(): ) and this does NOT work. The segfault error still happens.
It seems to me that the following 3 lines in the original author post are what really matter
# Ensure this is done on the main thread
if not self.log_textbox.winfo_exists():
return # Exit if widget does not exist

[Lychfindel]

Hi, thank you for the instructions!
I am using Ubuntu 24.04, I followed the instructions (using conda environment instead of venv), but I'm getting some errors which, I believe, comes from some wrong version of installed packages.
In particular by following all the instructions I'm getting this error in step2:

Traceback (most recent call last):

  File "/home/ale/whisper/noScribe/diarize.py", line 24, in <module>

    from pyannote.audio import Pipeline

  File "/home/ale/miniconda3/envs/noscribe/lib/python3.9/site-packages/pyannote/audio/__init__.py", line 29, in <module>

    from .core.inference import Inference

  File "/home/ale/miniconda3/envs/noscribe/lib/python3.9/site-packages/pyannote/audio/core/inference.py", line 49, in <module>

    class Inference(BaseInference):

  File "/home/ale/miniconda3/envs/noscribe/lib/python3.9/site-packages/pyannote/audio/core/inference.py", line 533, in Inference

    missing: float = np.NaN,

  File "/home/ale/miniconda3/envs/noscribe/lib/python3.9/site-packages/numpy/__init__.py", line 397, in __getattr__

    raise AttributeError(

AttributeError: `np.NaN` was removed in the NumPy 2.0 release. Use `np.nan` instead.

Therefore I downgraded numpy to version 1.26.4, Step2 runs correctly, but then I'm getting error on Step3 (printed only on GUI):

Error in step 3 - transcription
__init__() got an unexpected keyword argument 'threshold'

and on the terminal the only printed text is:

[2024-11-28 18:10:48.101] [ctranslate2] [thread 113569] [warning] The compute type inferred from the saved model is float16, but the target device or backend do not support efficient float16 computation. The model weights have been automatically converted to use the float32 compute type instead.
INFO:faster_whisper:Processing audio with duration 01:00.000

I don't understand where exactly the threshold is given as argument, and how to fix it.

[mael-lenoc]

Hi,
Can you try to run the software on this file with the default settings, just to confirm that it's not an issue with specific settings or file related ?

If you still get the same error, here is my pip freeze, so you can see if there are any significant version differences :.

AdvancedHTMLParser==9.0.2
aiohappyeyeballs==2.4.0
aiohttp==3.10.5
aiosignal==1.3.1
alembic==1.13.2
altgraph==0.17.4
antlr4-python3-runtime==4.9.3
appdirs==1.4.4
asteroid-filterbanks==0.4.0
attrs==24.2.0
audioread==3.0.1
av==12.3.0
certifi==2024.8.30
cffi==1.17.1
charset-normalizer==3.3.2
click==8.1.7
coloredlogs==15.0.1
colorlog==6.8.2
contourpy==1.3.0
ctranslate2==4.4.0
customtkinter==5.2.2
cycler==0.12.1
darkdetect==0.8.0
decorator==5.1.1
docopt==0.6.2
einops==0.8.0
faster-whisper==1.0.3
ffpyplayer==4.5.1
filelock==3.16.0
flatbuffers==24.3.25
fonttools==4.53.1
frozenlist==1.4.1
fsspec==2024.9.0
greenlet==3.1.0
huggingface-hub==0.24.7
humanfriendly==10.0
HyperPyYAML==1.2.2
idna==3.8
Jinja2==3.1.4
joblib==1.4.2
julius==0.2.7
kiwisolver==1.4.7
lazy_loader==0.4
librosa==0.10.2.post1
lightning==2.4.0
lightning-utilities==0.11.7
llvmlite==0.43.0
Mako==1.3.5
markdown-it-py==3.0.0
MarkupSafe==2.1.5
matplotlib==3.9.2
mdurl==0.1.2
mpmath==1.3.0
msgpack==1.1.0
multidict==6.1.0
networkx==3.3
numba==0.60.0
numpy==1.26.4
nvidia-cublas-cu12==12.1.3.1
nvidia-cuda-cupti-cu12==12.1.105
nvidia-cuda-nvrtc-cu12==12.1.105
nvidia-cuda-runtime-cu12==12.1.105
nvidia-cudnn-cu12==9.1.0.70
nvidia-cufft-cu12==11.0.2.54
nvidia-curand-cu12==10.3.2.106
nvidia-cusolver-cu12==11.4.5.107
nvidia-cusparse-cu12==12.1.0.106
nvidia-nccl-cu12==2.20.5
nvidia-nvjitlink-cu12==12.6.68
nvidia-nvtx-cu12==12.1.105
omegaconf==2.3.0
onnxruntime==1.19.2
optuna==4.0.0
packaging==24.1
pandas==2.2.2
pillow==10.4.0
platformdirs==4.3.2
pooch==1.8.2
primePy==1.3
protobuf==5.28.1
pyannote.audio @ git+https://github.com/kaixxx/pyannote-audio.git@d70ce34c3130130d0353f8a231929000db685129
pyannote.core==5.0.0
pyannote.database==5.1.0
pyannote.metrics==3.2.1
pyannote.pipeline==3.0.1
pycparser==2.22
Pygments==2.18.0
pyinstaller==6.10.0
pyinstaller-hooks-contrib==2024.8
pyparsing==3.1.4
PyQt5==5.15.11
PyQt5-Qt5==5.15.15
PyQt5_sip==12.15.0
python-dateutil==2.9.0.post0
python-i18n==0.3.9
pytorch-lightning==2.4.0
pytorch-metric-learning==2.6.0
pytz==2024.2
PyYAML==6.0.2
QDarkStyle==3.2.3
QtAwesome==1.3.1
QtPy==2.4.1
QueryableList==3.1.0
requests==2.32.3
rich==13.8.1
ruamel.yaml==0.18.6
ruamel.yaml.clib==0.2.8
scikit-learn==1.5.2
scipy==1.14.1
semver==3.0.2
sentencepiece==0.2.0
setuptools==74.1.2
shellingham==1.5.4
six==1.16.0
sortedcontainers==2.4.0
soundfile==0.12.1
soxr==0.5.0.post1
speechbrain==0.5.16
SQLAlchemy==2.0.34
sympy==1.13.2
tabulate==0.9.0
tensorboardX==2.6.2.2
threadpoolctl==3.5.0
tokenizers==0.20.0
torch==2.4.1
torch-audiomentations==0.11.1
torch-pitch-shift==1.2.4
torchaudio==2.4.1
torchmetrics==1.4.2
tqdm==4.66.5
triton==3.0.0
typer==0.12.5
typing_extensions==4.12.2
tzdata==2024.1
urllib3==2.2.3
yarl==1.11.1

[gernophil]

Sorry, haven't found the time yet to check and will likely not get back to my Mac today. But I'll promise to check tomorrow. First thing I realize with your pip freeze is that you seem to run noScribe and the noScribeEditor from the same environment. There might be conflicting packages. Better make one environment for noScribe and a separate for noScribeEditor.

We use Python3.9 for noScribe and Python3.10 for noScribeEditor.

[mael-lenoc]

@gernophil, I'm not the person with the issue. In fact everything works perfectly fine on my end.
I'm the person who provided the steps to install noScribe and noScribeEditor on Linux, and so I provided my pip freeze in case that could help @Lychfindel.
(And so, in my case, I have no issue running both noScribe and noScribeEditor in the same environment).

[Lychfindel]

Hi, thanks a lot!
I confirm that the problem was the version of faster-whisper: I had version 1.1.0 and after downgrading to version 1.0.3 everything works correctly! Moreover, also noScribeEditor works fine in the same environment with python 3.9.
I suggest to modify the modify the environment file with a strict fast-whisper==1.1.0

Thanks again!

[gernophil]

Ah, Sorry @mael-lenoc. I'm on the phone and didn't see that was you and not OP. Good to know nS and nSE work from the same env. I never tried that, but maybe we can simplify it a little then.