Problems when install on macOS

[paulocoutinhox]

Hi,

I have this error:

Building wheels for collected packages: torchcrepe, rvc-beta, pyworld
  Building wheel for torchcrepe (setup.py) ... done
  Created wheel for torchcrepe: filename=torchcrepe-0.0.20-py3-none-any.whl size=72326240 sha256=f1883703d9f2ae72ce72b648f4a2eff7a361b136d35976dbc9c185e186846c34
  Stored in directory: /private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-ephem-wheel-cache-9zofukgn/wheels/4a/7c/9f/170aad22b22218ca80febc6cf9f321b105adfed5c6930233fe
  Building wheel for rvc-beta (pyproject.toml) ... error
  error: subprocess-exited-with-error

  × Building wheel for rvc-beta (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [30 lines of output]
      Traceback (most recent call last):
        File "/Users/paulo/Downloads/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 353, in <module>
          main()
        File "/Users/paulo/Downloads/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 335, in main
          json_out['return_val'] = hook(**hook_input['kwargs'])
        File "/Users/paulo/Downloads/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 251, in build_wheel
          return _build_backend().build_wheel(wheel_directory, config_settings,
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/api.py", line 57, in build_wheel
          return WheelBuilder.make_in(
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/builders/wheel.py", line 87, in make_in
          wb.build(target_dir=directory)
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/builders/wheel.py", line 123, in build
          self._copy_module(zip_file)
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/builders/wheel.py", line 234, in _copy_module
          to_add = self.find_files_to_add()
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/builders/builder.py", line 200, in find_files_to_add
          if self.is_excluded(
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/builders/builder.py", line 143, in is_excluded
          if exclude_path.as_posix() in self.find_excluded_files(fmt=self.format):
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/masonry/builders/builder.py", line 111, in find_excluded_files
          vcs_ignored_files = set(vcs.get_ignored_files()) if vcs else set()
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/vcs/git.py", line 340, in get_ignored_files
          output = self.run(*args)
        File "/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-build-env-ht4n8qb8/overlay/lib/python3.10/site-packages/poetry/core/vcs/git.py", line 373, in run
          subprocess.check_output(
        File "/Users/paulo/Downloads/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/subprocess.py", line 421, in check_output
          return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
        File "/Users/paulo/Downloads/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/subprocess.py", line 526, in run
          raise CalledProcessError(retcode, process.args,
      subprocess.CalledProcessError: Command '['git', '--git-dir', '/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-install-yby4jxs1/rvc-beta_012b38e2953241258889a6c232b364c6/.git', '--work-tree', '/private/var/folders/0j/1vzxlz_d135_p0gs5jjc428c0000gn/T/pip-install-yby4jxs1/rvc-beta_012b38e2953241258889a6c232b364c6', 'ls-files', '--others', '-i', '--exclude-standard']' returned non-zero exit status 128.
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building wheel for rvc-beta
  Building wheel for pyworld (pyproject.toml) ... done
  Created wheel for pyworld: filename=pyworld-0.3.4-cp310-cp310-macosx_11_0_arm64.whl size=175721 sha256=bd570dbf3ffd57f022bfb92349e81633d42d828e319da9b2242a2cdcddbe8654
  Stored in directory: /Users/paulo/Library/Caches/pip/wheels/66/09/8a/a1d79b73d59756f66e9bfe55a199840efc7473adb76ddacdfd
Successfully built torchcrepe pyworld
Failed to build rvc-beta
ERROR: Could not build wheels for rvc-beta, which is required to install pyproject.toml-based projects

Also, how can i completely remove this?

Thanks.

[rsxdalv]

Hi, thanks for the report. It seems that git just failed in between. May I ask if git is working fine on your machine? Also to delete, one half is the one-click-installer directory, but the other part is in the cache folders. You have HuggingFace cache and torch cache. These caches are shared between projects so if you delete everything you might need to download something (though it happens automatically) for another project again.

[paulocoutinhox]

Hi,

Yes, git is working nice, i use it all days.

Yes, i want delete all these caches. Where they are?

Thanks.

[rsxdalv]

Are you able to open the settings tab in the UI? It should have the full list of model weight locations there.

[paulocoutinhox]

Hi,

It is not installing, neither opening the web ui.

[rsxdalv]

If you haven't used a single model then nothing would be downloaded. If your .cache folder has something, it's another project's files.

[jaylex32]

Can we fix this issue? I’m having the same problem running the start.sh in a macOS M1 chip in Ventura.

[rsxdalv]

Might have to give up automatic installation for RVC on Mac. Although it used to work, it sounds like now it's potentially breaking.

On Fri, Nov 3, 2023, 4:13 AM jaylex32 @.***> wrote:

Can we fix this issue? I’m having the same problem running the start.sh in a macOS M1 chip in Ventura.

— Reply to this email directly, view it on GitHub https://github.com/rsxdalv/tts-generation-webui/issues/196#issuecomment-1791815837, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTRXIZTBQW724ELH6EEH53YCRHMBAVCNFSM6AAAAAA6DAX7LWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJRHAYTKOBTG4 . You are receiving this because you commented.Message ID: @.***>

[jaylex32]

Might have to give up automatic installation for RVC on Mac. Although it used to work, it sounds like now it's potentially breaking. … On Fri, Nov 3, 2023, 4:13 AM jaylex32 @.> wrote: Can we fix this issue? I’m having the same problem running the start.sh in a macOS M1 chip in Ventura. — Reply to this email directly, view it on GitHub <#196 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABTRXIZTBQW724ELH6EEH53YCRHMBAVCNFSM6AAAAAA6DAX7LWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJRHAYTKOBTG4 . You are receiving this because you commented.Message ID: @.>

this is what I get in the terminal when the start_macos.sh complete's and it doesn't run the webui like in windows 11:Last login: Fri Nov 3 07:31:35 on ttys001 /Volumes/PROGRAMS/one-click-installers-tts-6.0/start_macos.sh ; exit; javierrosario@JayRosss-Mini ~ % /Volumes/PROGRAMS/one-click-installers-tts-6.0/start_macos.sh ; exit; Traceback (most recent call last): File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/tts-generation-webui/server.py", line 40, in from src.tortoise.generation_tab_tortoise import generation_tab_tortoise File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/tts-generation-webui/src/tortoise/generation_tab_tortoise.py", line 7, in from src.tortoise.gen_tortoise import ( File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/tts-generation-webui/src/tortoise/gen_tortoise.py", line 7, in from tortoise.api import TextToSpeech, MODELS_DIR File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/tortoise/api.py", line 10, in import torchaudio File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torchaudio/init.py", line 1, in from . import ( # noqa: F401 File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torchaudio/_extension/init.py", line 45, in _load_lib("libtorchaudio") File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torchaudio/_extension/utils.py", line 64, in _load_lib torch.ops.load_library(path) File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torch/_ops.py", line 643, in load_library ctypes.CDLL(path) File "/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/ctypes/init.py", line 374, in init self._handle = _dlopen(self._name, mode) OSError: dlopen(/Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torchaudio/lib/libtorchaudio.so, 0x0006): Symbol not found: __ZN2at4_ops6conv1d4callERKNS_6TensorES4_RKN3c108optionalIS2_EENS5_8ArrayRefIxEENSA_INS5_6SymIntEEESB_x Referenced from: /Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torchaudio/lib/libtorchaudio.so Expected in: <78DA4C67-8BC7-3995-8324-D2FD95E57BA4> /Volumes/PROGRAMS/one-click-installers-tts-6.0/installer_files/env/lib/python3.10/site-packages/torch/lib/libtorch_cpu.dylib

Done!

Saving session... ...copying shared history... ...saving history...truncating history files... ...completed.

[Process completed]

any ideas? thanks in advance

[rsxdalv]

I researched more and this seems like a tough one: https://github.com/pytorch/audio/issues/62

If this is the matching issue, it would mean that conda isn't choosing the right "platform", something that becomes very complicated to handle very quickly.

Can we fix this issue? I’m having the same problem running the start.sh in a macOS M1 chip in Ventura.

Also, just to clarify - this is a different issue you are talking about. The issue for this user was probably the XCode's crappy git that doesn't work like it should, as seen by the git command returning an error in the original log.

[rsxdalv]

@testter21

I can't guarantee it, but this might work (run within conda with cmd_macos.sh):

conda install torchaudio=2.0.0

I am seeing that the version of torchaudio no longer matches the pytorch. I will investigate further at some point, but this could fix the installation.

[testter21]

tester@testers-MBP one-click-installers-tts % chmod +x ./cmd_macos.sh
tester@testers-MBP one-click-installers-tts % ./cmd_macos.sh

The default interactive shell is now zsh. To update your account to use zsh, please run chsh -s /bin/zsh. (/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env) bash-3.2$ conda install torchaudio=2.0.0 Collecting package metadata (current_repodata.json): done Solving environment: failed with initial frozen solve. Retrying with flexible solve. Collecting package metadata (repodata.json): done Solving environment: failed with initial frozen solve. Retrying with flexible solve.

PackagesNotFoundError: The following packages are not available from current channels:

• torchaudio=2.0.0

Current channels:

• https://repo.anaconda.com/pkgs/main/osx-arm64
• https://repo.anaconda.com/pkgs/main/noarch
• https://repo.anaconda.com/pkgs/r/osx-arm64
• https://repo.anaconda.com/pkgs/r/noarch

To search for alternate channels that may provide the conda package you're looking for, navigate to

https://anaconda.org

and use the search bar at the top of the page.

(/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env) bash-3.2$

[rsxdalv]

My bad, here's the command with the relevant conda channels:

conda install -y -k torchaudio=2.0.0 -c pytorch -c nvidia/label/cuda-11.7.0 -c nvidia

[testter21]

Ok, so webui starts. Current log is this:

tester@testers-MBP ~ % cd /Users/tester/Downloads/xtts/one-click-installers-tts/ tester@testers-MBP one-click-installers-tts % chmod +x ./start_macos.sh tester@testers-MBP one-click-installers-tts % ./start_macos.sh Loading extensions: Loaded extension: callback_save_generation_ffmpeg Loaded extension: empty_extension Loaded extension: callback_save_generation_musicgen_ffmpeg Loaded 2 callback_save_generation extensions. Loaded 1 callback_save_generation_musicgen extensions. 2023-12-12 21:12:05 | INFO | fairseq.tasks.text_to_speech | Please install tensorboardX: pip install tensorboardX Failed to load musicgen demo No module named 'audiocraft' Failed to load rvc demo No module named 'rvc_beta' Failed to load demucs demo No module named 'demucs' Starting Gradio server... Gradio interface options: inline: False inbrowser: True share: False debug: False enable_queue: True max_threads: 40 auth: None auth_message: None prevent_thread_lock: False show_error: False server_name: 0.0.0.0 server_port: None show_tips: False height: 500 width: 100% favicon_path: None ssl_keyfile: None ssl_certfile: None ssl_keyfile_password: None ssl_verify: True quiet: True show_api: True file_directories: None _frontend: True Running on local URL: http://0.0.0.0:7860

Nothing is missing here in relation to multilingual voice-cloning/tts tabs?

Models download automatically, or have to be searched and downloaded manually? I'm specifically interested to have them gathered in some hassle-free path, so that I can for example backup whole project simply by compressing it and keeping copy somewhere (so that at least the project runs on the same machine in the future).

[rsxdalv]

Ok, so webui starts. Current log is this:

tester@testers-MBP ~ % cd /Users/tester/Downloads/xtts/one-click-installers-tts/ tester@testers-MBP one-click-installers-tts % chmod +x ./start_macos.sh tester@testers-MBP one-click-installers-tts % ./start_macos.sh Loading extensions: Loaded extension: callback_save_generation_ffmpeg Loaded extension: empty_extension Loaded extension: callback_save_generation_musicgen_ffmpeg Loaded 2 callback_save_generation extensions. Loaded 1 callback_save_generation_musicgen extensions. 2023-12-12 21:12:05 | INFO | fairseq.tasks.text_to_speech | Please install tensorboardX: pip install tensorboardX Failed to load musicgen demo No module named 'audiocraft' Failed to load rvc demo No module named 'rvc_beta' Failed to load demucs demo No module named 'demucs' Starting Gradio server... Gradio interface options: inline: False inbrowser: True share: False debug: False enable_queue: True max_threads: 40 auth: None auth_message: None prevent_thread_lock: False show_error: False server_name: 0.0.0.0 server_port: None show_tips: False height: 500 width: 100% favicon_path: None ssl_keyfile: None ssl_certfile: None ssl_keyfile_password: None ssl_verify: True quiet: True show_api: True file_directories: None _frontend: True Running on local URL: http://0.0.0.0:7860

Nothing is missing here in relation to multilingual voice-cloning/tts tabs?

Models download automatically, or have to be searched and downloaded manually? I'm specifically interested to have them gathered in some hassle-free path, so that I can for example backup whole project simply by compressing it and keeping copy somewhere (so that at least the project runs on the same machine in the future).

The modern models (2023+) generally download automatically (As for this project, only RVC does not download automatically). There are things that require extra installation on Macs because of limited or no official support (i.e., Facebook didn't build support for Mac in audiocraft). Bark has multiple languages, Tortoise can do many languages but it requires finding the models and setting them up, the default is English.

[testter21]

Okay, a quick question. I started webui, went to bark voice clone tab, and selected model for polish language. It looks that it's downloading, or at least the webui shows progress time and queue mark. I don't see anything changed on the terminal window however (should I? In some other apps, downloads were indicated on terminal).

And where is the download folder? I mean - is it within the tts webui folder, or somewhere outside?

[rsxdalv]

Those models are in 'data', 'models', 'hubert' Other models are in your user .cache RVC models are downloaded manually to data/models/...

[testter21]

Okay, so probably the webui is not downloading anything. The counter goes into infinity, but on the hugginggface via github https://github.com/gitmylo/bark-voice-cloning-HuBERT-quantizer?tab=readme-ov-file https://huggingface.co/Hobis/bark-voice-cloning-polish-HuBERT-quantizer/tree/main PL model is just 250MB in size if this is what is supposed to be. Also, I see no subfolder named hubert is created in /data/models/ So I'm putting it there manually and we see what happens.

[testter21]

If I put the model to destination fole and restart, nothing more happens. Queue is counting time ad infinity when selecting polish model.

In the meantime. If I just start webui and in the "Generation (Bark)" tab switch to "or Use a voice:", terminal outputs these errors. I don't know if this is normal at this point.

Running on local URL: http://0.0.0.0:7860 2023-12-13 18:46:40 | ERROR | asyncio | Task exception was never retrieved future: <Task finished name='ekdi7gy1mt_9' coro=<Queue.process_events() done, defined at /Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py:342> exception=1 validation error for PredictBody event_id Field required [type=missing, input_value={'data': ['or Use a voice...ion_hash': 'ekdi7gy1mt'}, input_type=dict] For further information visit https://errors.pydantic.dev/2.5/v/missing> Traceback (most recent call last): File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py", line 346, in process_events client_awake = await self.gather_event_data(event) File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py", line 219, in gather_event_data data, client_awake = await self.get_message(event, timeout=receive_timeout) File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py", line 448, in get_message return PredictBody(**data), True File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/pydantic/main.py", line 164, in init pydantic_self.pydantic_validator.validate_python(data, self_instance=__pydantic_self__) pydantic_core._pydantic_core.ValidationError: 1 validation error for PredictBody event_id Field required [type=missing, input_value={'data': ['or Use a voice...ion_hash': 'ekdi7gy1mt'}, input_type=dict] For further information visit https://errors.pydantic.dev/2.5/v/missing

[testter21]

From what I see on colab version, when selecting polish model on second tab, I get error (although 'queue' mark changes to 'processing' mark), plus on first tab, webui is loaded fully.

[rsxdalv]

If I put the model to destination fole and restart, nothing more happens. Queue is counting time ad infinity when selecting polish model.

In the meantime. If I just start webui and in the "Generation (Bark)" tab switch to "or Use a voice:", terminal outputs these errors. I don't know if this is normal at this point.

Running on local URL: http://0.0.0.0:7860 2023-12-13 18:46:40 | ERROR | asyncio | Task exception was never retrieved future: <Task finished name='ekdi7gy1mt_9' coro=<Queue.process_events() done, defined at /Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py:342> exception=1 validation error for PredictBody event_id Field required [type=missing, input_value={'data': ['or Use a voice...ion_hash': 'ekdi7gy1mt'}, input_type=dict] For further information visit https://errors.pydantic.dev/2.5/v/missing> Traceback (most recent call last): File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py", line 346, in process_events client_awake = await self.gather_event_data(event) File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py", line 219, in gather_event_data data, client_awake = await self.get_message(event, timeout=receive_timeout) File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/gradio/queueing.py", line 448, in get_message return PredictBody(data), True File "/Users/tester/Downloads/xtts/one-click-installers-tts/installer_files/env/lib/python3.10/site-packages/pydantic/main.py", line 164, in init pydantic_self.pydantic_validator.validate_python(data, self_instance=pydantic_self**) pydantic_core._pydantic_core.ValidationError: 1 validation error for PredictBody event_id Field required [type=missing, input_value={'data': ['or Use a voice...ion_hash': 'ekdi7gy1mt'}, input_type=dict] For further information visit https://errors.pydantic.dev/2.5/v/missing

oh no that's a bad one - that's an error in gradio. It might be a gradio bug, or something that requires a restart or a reinstall or there's a bad gradio package (gradio would work but there are two different packages)

[testter21]

It's the same after restarts.

I followed only your instructions and used scripts from your github. Nothing external added. So as of right now, this will be probably the end result when someone attempts to use this project "as is" on mac.

Any ideas on how this could be continued?

[rsxdalv]

I can research this and maybe get a workable solution. As for the packages - I know that you used the correct script and thanks to you I also have the full log for debugging. This is more of a "that shouldn't even break" situation which is why I mentioned all the possible sources of failure.

[testter21]

Okay, thanks, so I'm waiting for further info. In any case, I can help with testing.

[rsxdalv]

After a long research, I found that it could be related to this: https://github.com/oobabooga/text-generation-webui/issues/3042

They suggest two potential solutions there (which partially match what I am seeing):

• only upgrade gradio pip install 'gradio>=3.36.1'
• OR only downgrade fastapi pip install 'fastapi==0.95.2' These commands need to be run within the conda environment (i.e, by cmd_macos.sh).

Myself I have confirmed that I have the following versions:

(C:\Users\admin\Desktop\one-click-installers-tts-main\installer_files\env) C:\Users\admin\Desktop\one-click-installers-tts-main>pip show fastapi
Name: fastapi
Version: 0.100.0
Summary: FastAPI framework, high performance, easy to learn, fast to code, ready for production
Home-page:
Author:
Author-email: Sebastián Ramírez <tiangolo@gmail.com>
License:
Location: c:\users\admin\desktop\one-click-installers-tts-main\installer_files\env\lib\site-packages
Requires: pydantic, starlette, typing-extensions
Required-by: gradio

(C:\Users\admin\Desktop\one-click-installers-tts-main\installer_files\env) C:\Users\admin\Desktop\one-click-installers-tts-main>pip show gradio
Name: gradio
Version: 3.35.2
Summary: Python library for easily interacting with trained machine learning models
Home-page:
Author:
Author-email: Abubakar Abid <team@gradio.app>, Ali Abid <team@gradio.app>, Ali Abdalla <team@gradio.app>, Dawood Khan <team@gradio.app>, Ahsen Khaliq <team@gradio.app>, Pete Allen <team@gradio.app>, Ömer Faruk Özdemir <team@gradio.app>
License:
Location: c:\users\admin\desktop\one-click-installers-tts-main\installer_files\env\lib\site-packages
Requires: aiofiles, aiohttp, altair, fastapi, ffmpy, gradio-client, httpx, huggingface-hub, jinja2, markdown-it-py, markupsafe, matplotlib, mdit-py-plugins, numpy, orjson, pandas, pillow, pydantic, pydub, pygments, python-multipart, pyyaml, requests, semantic-version, uvicorn, websockets
Required-by: audiocraft, rvc-beta

If this has been the core issue, how could it happen? fastapi project might have made an update that did not match gradio's expectations.

[testter21]

Ok, I will give a try in a few days and let you know.