search

jojojaeger / whisper-streamlit

this master thesis project is based on OpenAI Whisper with the goal to transcibe interviews
https://openai.com/research/whisper
Other
41 stars 16 forks source link

Speaker recognition tickbox option missing #6

Closed menelic closed 1 year ago

menelic commented 1 year ago

model selection and transcription work, but regardless which model is chosen, the speaker recognition tick box is missing:

image

jojojaeger commented 1 year ago

i moved it out of the repo - you can access it now, i will push it! but you have to install https://pypi.org/project/pyannote.audio/ for that

menelic commented 1 year ago

thanks - can you add this as a dependency so its pulled while installing? In the install instructions of pyannote I saw conda is required, so I am wondering if pyanot can be added to a pre-existing install of whisper-streamlit. I will try later, if just installing it works, but there might be issues and I might not be the only one wondering about this..... ;-)

menelic commented 1 year ago

I can see that the new whisper-streamlit-speaker folder contains all the folders and files from the main app. Do I just git pull whisper streamlit to install it or do I git pull just the whisper-streamlit-speaker folder?

in the whisper-streamlit readme you write: still experimental, available in whisper-streamlit-speaker folder

the readme in that folder does not have that note, but also does nt specify which folder the generic install instruction listed there is referring to.

Sorry for a n00b question, but please clarify in the readme so that less python savvy folk know how to install the version with diarisation.

jojojaeger commented 1 year ago

exactly you just need the whisper-streamlit-speaker folder

menelic commented 1 year ago

did that, and when I installed pyannote.audio it installed a lot of huge dependencies such as torch wheels etc. without any errors, except in the end:

ERROR: Exception: Traceback (most recent call last): File "/usr/lib/python3/dist-packages/pip/_internal/cli/base_command.py", line 165, in exc_logging_wrapper status = run_func(*args) File "/usr/lib/python3/dist-packages/pip/_internal/cli/req_command.py", line 205, in wrapper return func(self, options, args) File "/usr/lib/python3/dist-packages/pip/_internal/commands/install.py", line 389, in run to_install = resolver.get_installation_order(requirement_set) File "/usr/lib/python3/dist-packages/pip/_internal/resolution/resolvelib/resolver.py", line 188, in get_installation_order weights = get_topological_weights( File "/usr/lib/python3/dist-packages/pip/_internal/resolution/resolvelib/resolver.py", line 276, in get_topological_weights assert len(weights) == expected_node_count

it also claims ModuleNotFoundError: No module named 'pyannote.audio'

despite it having been installed. This is part of the following app exception that occurs on start (streamlit loads but display only this error message)

2023-05-31 22:06:25.511 Uncaught app exception Traceback (most recent call last): File "/home/.../.local/lib/python3.10/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 562, in _run_script exec(code, module.dict) File "/home/...//software/whisper-streamlit-speaker/Transcribe.py", line 2, in from transcriber import Transcription File "/home/.../Documents/software/whisper-streamlit-speaker/transcriber.py", line 3, in from pyannote.audio import Pipeline ModuleNotFoundError: No module named 'pyannote.audio' 2023-05-31 22:07:19.667 Uncaught app exception Traceback (most recent call last): File "/home/../.local/lib/python3.10/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 562, in _run_script exec(code, module.dict) File "/home/...//software/whisper-streamlit-speaker/Transcribe.py", line 2, in from transcriber import Transcription File "/home/...//software/whisper-streamlit-speaker/transcriber.py", line 3, in from pyannote.audio import Pipeline ModuleNotFoundError: No module named 'pyannote.audio'

a second attempt results in the same error but confirms pyannote.audio is present: Using cached pyasn1-0.5.0-py2.py3-none-any.whl (83 kB) Requirement already satisfied: oauthlib>=3.0.0 in /usr/lib/python3/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<1.1,>=0.5->tensorboard>=2.2.0->pytorch-lightning<1.7,>=1.5.4->pyannote.audio) (3.2.0)