Copied from https://huggingface.co/marduk-ra/F5-TTS-German, added trained duration model on emilia dataset using https://github.com/eamag/f5-tts-duration

Inference with https://github.com/lucasnewman/f5-tts-mlx

python -m f5_tts_mlx.generate --model "eamag/f5-tts-mlx-german" \
--text "The quick brown fox jumped over the lazy dog." \
--ref-audio /path/to/audio.wav \
--ref-text "This is the caption for the reference audio."

Github: https://github.com/SWivid/F5-TTS
Paper: F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching

NOTE: You can set the number of nfe steps to 64 to produce better quality sound.

Downloads last month
222
Safetensors
Model size
337M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and the HF Inference API does not support f5-tts models with pipeline type text-to-speech

Dataset used to train eamag/f5-tts-mlx-german