support for voice cloning is done (+2 squashed commit)

Squashed commit:

[e730162] support for voice cloning is done

[1653c57] wip adding voice cloning

Author: LostRuins

examples/outetts/readme.md

@@ -0,0 +1,23 @@
+## KoboldCpp now supports OuteTTS Voice Cloning
+
+However, it can be slightly challenging to set it up.
+- The easiest way to get started is to use already pre-cloned voices, which you can find within the [speakers directory](speakers).
+- Simply download the .json speaker file, open KoboldCpp with an OuteTTS model and WavTokenizer loaded.
+- Paste the voice JSON into Kobold Lite UI's Settings > Media > TTS > voiceclone option.
+
+![Preview](vc1.png)
+![Preview](vc2.png)
+
+- If you don't know how to get the OuteTTS and WavTokenizer models, [please read the wiki](https://github.com/LostRuins/koboldcpp/wiki#getting-an-ai-model-file).
+- Once configured, you are good to go. Narrations will be done with that selected cloned speaker.
+
+### Cloning your own speaker
+You can also create your own cloned speaker voice.
+To do that, you will need Python 3.10+ with the `outetts` package installed. You can install it with this commands:
+```
+pip install outetts --no-deps
+pip install scipy einops pyyaml huggingface-hub encodec matplotlib transformers pytorch-lightning tensorboardX soundfile numpy jsonargparse torchcrepe librosa inflect loguru polars natsort tqdm sounddevice mecab-python3 unidic-lite uroman openai-whisper pygame
+```
+Then after that, obtain a wav file containing a sample of the voice you want to clone.
+Modify the [voice_cloning.py](voice_cloning.py) script with your input audio file, and let it generate the JSON.
+Finally, use the JSON file in the same way as the above with the precloned voices.