Would love a hand-holding tutorial on preparing a dataset... #178

I'm doing my best to follow the wiki, but can't seem to prepare a dataset. I am trying to train/finetune some voices.

I'm running Windows 11, using Miniconda3 (so Python 3.10) and have an RTX 4080, so I used the Cuda install.

I can generate voices using models I created with DLAS, but they don't work nearly as well as they do in DLAS--not complaining, just trying to learn, and I love the singular UI.

Using the WebUI, I have placed a set of "curated" wav files in a folder in the /voices/ directory and then tried to prepare a dataset using the default settings, but only get errors in return:

Failed to transcribe: ./voices\voicetest\1.wav
*
*
*
Failed to transcribe: ./voices\voicetest\100.wav

Followed by:
Traceback (most recent call last):
File "D:\ai-voice-cloning\venv\lib\site-packages\gradio\routes.py", line 394, in run_predict
output = await app.get_blocks().process_api(
File "D:\ai-voice-cloning\venv\lib\site-packages\gradio\blocks.py", line 1075, in process_api
result = await self.call_function(
File "D:\ai-voice-cloning\venv\lib\site-packages\gradio\blocks.py", line 884, in call_function
prediction = await anyio.to_thread.run_sync(
File "D:\ai-voice-cloning\venv\lib\site-packages\anyio\to_thread.py", line 31, in run_sync
return await get_asynclib().run_sync_in_worker_thread(
File "D:\ai-voice-cloning\venv\lib\site-packages\anyio_backends_asyncio.py", line 937, in run_sync_in_worker_thread
return await future
File "D:\ai-voice-cloning\venv\lib\site-packages\anyio_backends_asyncio.py", line 867, in run
result = context.run(func, *args)
File "D:\ai-voice-cloning\src\webui.py", line 208, in prepare_dataset_proxy
message = prepare_dataset( voice, use_segments=slice_audio, text_length=validation_text_length, audio_length=validation_audio_length, progress=progress )

Thanks for any help here.

I'm doing my best to follow the wiki, but can't seem to prepare a dataset. I am trying to train/finetune some voices. I'm running Windows 11, using Miniconda3 (so Python 3.10) and have an RTX 4080, so I used the Cuda install. I can generate voices using models I created with DLAS, but they don't work nearly as well as they do in DLAS--not complaining, just trying to learn, and I love the singular UI. Using the WebUI, I have placed a set of "curated" wav files in a folder in the /voices/ directory and then tried to prepare a dataset using the default settings, but only get errors in return: Failed to transcribe: ./voices\voicetest\1.wav * * * Failed to transcribe: ./voices\voicetest\100.wav Followed by: Traceback (most recent call last): File "D:\ai-voice-cloning\venv\lib\site-packages\gradio\routes.py", line 394, in run_predict output = await app.get_blocks().process_api( File "D:\ai-voice-cloning\venv\lib\site-packages\gradio\blocks.py", line 1075, in process_api result = await self.call_function( File "D:\ai-voice-cloning\venv\lib\site-packages\gradio\blocks.py", line 884, in call_function prediction = await anyio.to_thread.run_sync( File "D:\ai-voice-cloning\venv\lib\site-packages\anyio\to_thread.py", line 31, in run_sync return await get_asynclib().run_sync_in_worker_thread( File "D:\ai-voice-cloning\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 937, in run_sync_in_worker_thread return await future File "D:\ai-voice-cloning\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 867, in run result = context.run(func, *args) File "D:\ai-voice-cloning\src\webui.py", line 208, in prepare_dataset_proxy message = prepare_dataset( voice, use_segments=slice_audio, text_length=validation_text_length, audio_length=validation_audio_length, progress=progress ) Thanks for any help here.

See #177

Dangit. I just found that one and was going to post it here... Thanks.

Labels Milestones

Would love a hand-holding tutorial on preparing a dataset... #178