• Joined on 2023-02-23
hman360 opened issue mrq/ai-voice-cloning#430 2023-10-26 03:20:19 +00:00
American Accent training data gives British Accented results
hman360 commented on issue mrq/ai-voice-cloning#180 2023-03-27 05:55:11 +00:00
Failed to run Voicefixer

I think the voicefixer model may not have downloaded properly.

hman360 commented on issue mrq/ai-voice-cloning#173 2023-03-25 18:55:02 +00:00
Recommendations for generating latents and finetunes?

Check this out if you haven't https://git.ecker.tech/mrq/ai-voice-cloning/wiki/Training

I have; those settings don't really give me a good point of reference.

hman360 opened issue mrq/ai-voice-cloning#173 2023-03-25 08:31:28 +00:00
Recommendations for generating latents and finetunes?
hman360 commented on issue mrq/ai-voice-cloning#160 2023-03-23 05:55:06 +00:00
Can't train a single good model

I tried redoing it with commit 0231550287 from about 2 weeks ago, and the output was much better; close to the dataset voice. The training ran much faster too.

This repo itself doesn't…

hman360 commented on issue mrq/ai-voice-cloning#160 2023-03-22 08:05:27 +00:00
Can't train a single good model

Did redoing it include re-preparing the dataset using the old version? I've had terrible luck with the audio slicing in the newer versions.

Nope. I reused the exact same audio files and…

hman360 commented on issue mrq/ai-voice-cloning#160 2023-03-22 07:14:06 +00:00
Can't train a single good model

I'm having issues too. I trained a model with a single-voice dataset normalized to between 1-11 seconds, using a recent version of the repo, and got a terrible voice that was way too deep.

I…

hman360 closed issue mrq/ai-voice-cloning#143 2023-03-19 06:29:36 +00:00
Error when preparing dataset
hman360 opened issue mrq/ai-voice-cloning#143 2023-03-16 08:01:17 +00:00
Error when preparing dataset
hman360 closed issue mrq/ai-voice-cloning#140 2023-03-16 07:19:55 +00:00
Question: What is the second set of lowercase text in the training transcription for?
hman360 opened issue mrq/ai-voice-cloning#140 2023-03-15 08:13:45 +00:00
Question: What is the second set of lowercase text in the training transcription for?
hman360 commented on issue mrq/ai-voice-cloning#113 2023-03-13 19:32:22 +00:00
Generated voices from training data always garbled.... but works fine using tortoise-tts-fast ... (?)

Just to clear up my understanding: The recommendation now is to not use all-in-one files and instead make sure the audio clips post-transcription are always under 11.6s?

hman360 commented on issue mrq/ai-voice-cloning#113 2023-03-12 03:29:36 +00:00
Generated voices from training data always garbled.... but works fine using tortoise-tts-fast ... (?)

How feasible would it be to run sliced audio through Whisper a second time to see if the transcription matches the sliced audio? If the re-transcription doesn't match, you can throw out that slice.

hman360 commented on issue mrq/ai-voice-cloning#113 2023-03-11 08:35:34 +00:00
Generated voices from training data always garbled.... but works fine using tortoise-tts-fast ... (?)

The biggest problem I'm having when it comes to transcription, even when using WhisperX, is the transcribed text having a full sentence, but the audio having the first or last word or two cut off.…

hman360 opened issue mrq/ai-voice-cloning#45 2023-02-27 08:42:13 +00:00
Feature Request: Use WhisperX instead of Whisper for preparing dataset
hman360 opened issue mrq/ai-voice-cloning#44 2023-02-27 08:39:03 +00:00
Feature Request: tortoise-fast-tts