Mirrors/Voice-Cloning-App
1
0
Fork 0
mirror of https://github.com/voice-cloning-app/Voice-Cloning-App.git synced 2026-02-24 20:20:22 +01:00
Code Issues Packages Projects Releases Wiki Activity

Fix docs links

Browse Source
This commit is contained in:
BenAAndrew
2021-10-14 19:46:06 +01:00
parent d599546259
commit 19d742b655
3 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
docs/dataset.md
View File

@@ -68,7 +68,7 @@ If your text does not accurately match your audio, the dataset cannot be built.
To build a dataset in the app you can simply upload your source audio/text and enter a name for the dataset.
If using another language you can select this here (to add a language see [Adding languages](Voice-Cloning-App/#adding-languages))
If using another language you can select this here (to add a language see [Adding languages](../#adding-languages))
![Dataset](images/dataset.PNG "Dataset")

2
docs/synthesis.md
View File

@@ -38,4 +38,4 @@ You can also modify the text at the bottom and press submit to generate more syn
![Synthesis results](images/synthesis-results.PNG "Synthesis results")
If audio could not be produced or was poor this may be due to the given text (either being too short/long or containing difficult words for your voice) or [poor voice quality](Voice-Cloning-App/training/#verifying-quality)
If audio could not be produced or was poor this may be due to the given text (either being too short/long or containing difficult words for your voice) or [poor voice quality](../training/#verifying-quality)

4
docs/training.md
View File

@@ -106,7 +106,7 @@ When you want to try the checkpoint out, simply download it and import it into t
## Verifying quality
To identify whether you have produced a good voice model you can stop training at any time and test your model in [synthesis](Voice-Cloning-App/synthesis)
To identify whether you have produced a good voice model you can stop training at any time and test your model in [synthesis](../synthesis)
However, you can also tell whether your model quality will be good by looking at a variety of other factors:
@@ -116,7 +116,7 @@ However, you can also tell whether your model quality will be good by looking at
If any of these things are not true, then your model may not be good quality. There are a number of possible reasons for this:
- Your dataset was not good enough quality (see [Verify your source](Voice-Cloning-App/dataset/#verify-your-source) for possible reasons why)
- Your dataset was not good enough quality (see [Verify your source](../dataset/#verify-your-source) for possible reasons why)
- You did not train for long enough
- Did you upload a transfer learning model?
- Did you train for 500+ epochs?