F5-TTS

开源

"F5-TTS:以流相匹配来欺骗流利而忠实的演讲"的官方代码

15.1K stars0 点赞0 次浏览

工具介绍

"F5-TTS:以流相匹配来欺骗流利而忠实的演讲"的官方代码

F5-TTS:用流相匹配的F5-TTS:用ConvNeXt V2来传播Transformer的"活泼而忠实"的"仙女"(fairytaler),更快地被训练并推论. E2 TTS:平地-UNet Transformer,从纸上最接近复制. 斯威采样:推论-时间流程步骤采样策略,能大大提高性能. 感谢所有贡献者! News - 2025/03/12:QQ F5-TTS v1 基础模型,具有更好的训练与推论性能. - 2024/10/08: F5-TTS & E2 TTS 基模在 QHugging Face, QQ 模型范围, QQ Wisemodel上. 安装如果需要, 创建单独的环境创建有 pythonversion=3. 10 (您也可以使用 virtualenv) 的 conda 创建 -n f5-ts Python=3. 11 conda 激活 f5-tts 安装 FFmpeg 如果您还没有安装 ffmpeg 安装 PyTorch 并配对设备 < details < summaryNVIDIA GPU </summary bash # 安装 pytorch 和您的 CUDA 版本, ..

核心特点

•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.
•In order to achieve desired performance, take a moment to read detailed guidance.
•By properly searching the keywords of problem encountered, issues are very helpful.
•Basic TTS with Chunk Inference
•Multi-Style / Multi-Speaker Generation
•Voice Chat powered by Qwen2.5-3B-Instruct
•Custom inference with more language support
•"7860:7860"
•driver: nvidia