+12
日增长
15,597
stars
2,281
forks
130
watchers
253
prs
898
issues
22
releases
README
F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching -- F5-TTS: Diffusion Transformer with ConvNeXt V2, faster trained and inference. E2 TTS: Flat-UNet Transformer, closest reproduction from paper. Sway Sampling: Inference-time flow step sampling strategy, greatly improves performance Thanks to all the contributors ! News - 2025/03/12: 🔥 F5-TTS v1 base model with better training and inference performance. Few demo. - 2024/10/08: F5-TTS & E2 TTS base models on 🤗 Hugging Face, 🤖 Model Scope, 🟣 Wisemodel. Installation Create a separate environment if needed bash Create a conda env with python_version=3.10 (you could also use virtualenv) conda create -n f5-tts python=3.11 conda activate f5-tts Install FFmpeg if you haven't yet conda install ffmpeg Install PyTorch with matched device NVIDIA GPU bash Install pytorch with your CUDA version, e.g. pip install torch==2.8.0+cu128 torchaudio==2.8.0+cu128 --extra-index-url https://download.pytorch.org/whl/cu128 And also possible previous versions, e.g. pip install torch==2.4.0+cu124 torchaudio==2.4.0+cu124 --extra-index-url https://download.pytorch.org/whl/cu124 etc. AMD GPU bash Install pytorch with your ROCm version (Linux only), e.g. pip install torch==2.9.1+rocm7.2 torchaudio==2.9.1+rocm7.2 --extra-index-url https://download.pytorch.org/whl/rocm7.2 For older GPUs (RDNA1/2/3 only): pip install torch==2.5.1+rocm6.2 torchaudio==2.5.1+rocm6.2 --extra-index-url https://download.pytorch.org/whl/rocm6.2 Note: RDNA 3.5 and RDNA 4 GPUs (Radeon 8050S/8060S, RX 9060/9070 series) require ROCm 7.x — these architectures (gfx1151/gfx1201) are not included in ROCm 6.x (6.2 compatibility matrix vs 7.2 compatibility matrix). Using ROCm 6.x on these GPUs causes HIP error: invalid device function (1236). Intel GPU bash Install pytorch with your XPU version, e.g. Intel® Deep Learning Essentials or Intel® oneAPI Base Toolkit must be installed pip install torch torchaudio --index-url https://download.pytorch.org/whl/test/xpu Intel GPU support is also available through IPEX (Intel® Extension for PyTorch) IPEX does not require the Intel® Deep Learning Essentials or Intel® oneAPI Base Toolkit See: https://pytorch-extension.intel.com/installation?request=platform Apple Silicon bash Install the stable pytorch, e.g. pip install torch torchaudio Then you can choose one from below: 1. As a pip package (if just for inference) bash pip install f5-tts 2. Local editable (if also do training, finetuning) bash git clone https://github.com/SWivid/F5-TTS.git cd F5-TTS git submodule update --init --recursive (optional, if use bigvgan as vocoder) pip install -e . Docker usage also available bash Build from Dockerfile docker build -t f5tts:v1 . Run from GitHub Container Registry docker container run --rm -it --gpus=all --mount 'type=volume,source=f5-tts,target=/root/.cache/huggingface/hub/'
Top issues
#87Supported Languages?139 comments · 2024-10-15#291how to f5 tts train for other language49 comments · 2024-10-27#1214My RTF is always around 0.12 with TRT-LLM docker on L20, never match the claimed benchmark result30 comments · 2025-11-10#1247I'm learing from your project. I want sone help about custom dataset, vocabulary, and retrain from zero.19 comments · 2026-01-13#606We are trying to use sentiment text to make the generated audio have different emotions15 comments · 2024-12-08#1237The pronunciation of a Chinese sentence10 comments · 2025-12-13#1207finetune-gradio "test model" function has inference bug with speech & text mismatch, caused by clipped ref_audio and unchanged ref_text (sampled from dataset)8 comments · 2025-11-03#1234RuntimeError: Could not load libtorchcodec7 comments · 2025-12-07
Discussions
#1344Adding emotion/style control when fine-tuning for a new language (Uzbek, custom Latin vocab)0 comments · 2026-11-12#1330Permission to use bundled English reference voice in a non-commercial podcast1 comments · 2026-10-18#1324Invitation to try RVCBench: voice-cloning utility, robustness and automated evaluation0 comments · 2026-10-10#1303Using iPhone voice messages thoughts, ideas, recommendations.1 comments · 2026-07-07#1295Commercial Use1 comments · 2026-05-17#1291Menu or Models Button2 comments · 2026-04-30#1278A lightweight, 150-million-parameter Arabic and English TTS model based on the F5-TTS architecture0 comments · 2026-03-15#1275Text-Normalizer deactivate? Other Languages1 comments · 2026-03-12
Latest release
1.1.22 1.1.22 2026-07-23What's Changed fix: allocate fix_duration across chunks to avoid N× duration blow-up by @Wing900 in https://github.com/SWivid/F5-TTS/pull/1309 New Contributors @Wing900 made their first contribution in https://github.com/SWivid/F5-TTS/pull/1309 Full Changelog: https://github.com/SWivid/F5-TTS/compare/1.1.21...1.1.22