This Fork allow to use the German Mira TTS variant made by Sebastian Bodza https://huggingface.co/SebastianBodza/MiraToffel_miraTTS_german
With the help of Sebastian in this thread https://huggingface.co/SebastianBodza/MiraToffel_miraTTS_german/discussions/1 the logic for sentence splitting for German language required entire adjustments. Sentences are now split by punctuation marks, which are typical in German language to end sentences. Before the split was done on capital letters. With the help of GPT 5.2 the programming was adjusted and works now for German language with improve chunking.
MiraTTS is a finetune of the excellent Spark-TTS model for enhanced realism and stability performing on par with closed source models. This repository also heavily optimizes Mira with Lmdeploy and boosts quality by using FlashSR to generate high quality audio at over 100x realtime!
- Incredibly fast: Over 100x realtime by using Lmdeploy and batching.
- High quality: Generates clear and crisp 48khz audio outputs which is much higher quality then most models.
- Memory efficient: Works within 6gb vram.
- Low latency: Latency can be low as 100ms.
Simple 1 line installation:
uv pip install git+https://github.com/ysharma3501/MiraTTS.git
Running the model(bs=1):
from mira.model import MiraTTS
from IPython.display import Audio
mira_tts = MiraTTS('YatharthS/MiraTTS') ## downloads model from huggingface
file = "reference_file.wav" ## can be mp3/wav/ogg or anything that librosa supports
text = "Alright, so have you ever heard of a little thing named text to speech? Well, it allows you to convert text into speech! I know, that's super cool, isn't it?"
context_tokens = mira_tts.encode_audio(file)
audio = mira_tts.generate(text, context_tokens)
Audio(audio, rate=48000)Running the model using batching:
file = "reference_file.wav" ## can be mp3/wav/ogg or anything that librosa supports
text = ["Hey, what's up! I am feeling SO happy!", "Honestly, this is really interesting, isn't it?"]
context_tokens = [mira_tts.encode_audio(file)]
audio = mira_tts.batch_generate(text, context_tokens)
Audio(audio, rate=48000)Examples can be seen in the huggingface model
I recommend reading these 2 blogs to better easily understand LLM tts models and how I optimize them
- How they work: https://huggingface.co/blog/YatharthS/llm-tts-models
- How to optimize them: https://huggingface.co/blog/YatharthS/making-neutts-200x-realtime
- Release code and model
- Support low latency streaming
- Release native 48khz bicodec
- Support multilingual models
This fork is based on the project of https://github.com/Si-ris-B