SWivid/F5-TTS

View on GitHub

Official code for "F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching"

Stars: 15.3k
Forks: 2.2k
Open beginner issues: 0
Indexed issues: 49

Dominant language: Python
License: MIT
Last GitHub push: Jul 23, 2026
Latest indexed: Sep 19, 2026
Contributing guide: No contributing guide
Code of conduct: No code of conduct
Beginner labels: help wanted
PR merge metrics: No merged PRs in 30d

Beginner friendly Everything

Open All

49 open issues indexed

请增加对拟声词的标注支持 Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 35/100

SWivid/F5-TTS#1317 · Opened Sep 10, 2026
Verify evals on Papers with Code Open

documentation

Difficulty 3/5 1-2 days Newbie friendliness 42/100

SWivid/F5-TTS#1315 · 1 reaction · Opened Aug 28, 2026
中文文本中的阿拉伯数字会被转换为英文 Open

question

Difficulty 5/5 Over a week Newbie friendliness 30/100

SWivid/F5-TTS#1312 · 1 comment · Opened Aug 11, 2026
[Vulnerability] OS command injection in finetune_gradio.py start_training() via shell=True Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 68/100

SWivid/F5-TTS#1306 · 1 comment · 1 reaction · Opened Jul 13, 2026
Feature: Consider SenseVoice for training data transcription Open

enhancement

SWivid/F5-TTS#1300 · 2 reactions · 1 assignee · Opened May 31, 2026
Document that F5TTS_v1_Base/model_1250000.safetensors contains only EMA-prefixed weights Open

Difficulty 2/5 1-3 hours Newbie friendliness 68/100

SWivid/F5-TTS#1292 · 1 comment · Opened May 9, 2026
What is the best way to contribute Benchmark information? Open

documentation question

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1286 · Opened Apr 10, 2026
400k steps into training, still heavy halucinating/unintelligible Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1283 · 1 comment · Opened Mar 28, 2026
About training 1.2M updates over one week mentioned in the paper Open

question

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1282 · 3 comments · Opened Mar 24, 2026
The order of chunked values in AdaLayerNorm_Final Open

Difficulty 2/5 1-3 hours Newbie friendliness 55/100

SWivid/F5-TTS#1272 · 2 reactions · Opened Mar 5, 2026
Short sentences glitches solved for me Open

documentation

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1263 · 1 comment · Opened Feb 5, 2026
The same gen_text sometimes outputs correctly and sometimes outputs incorrectly. Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 20/100

SWivid/F5-TTS#1253 · Opened Jan 19, 2026
I'm learing from your project. I want sone help about custom dataset, vocabulary, and retrain from zero. Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#1247 · 19 comments · Opened Jan 13, 2026
mispronounces some words Open

question

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1238 · 2 comments · Opened Dec 14, 2025
The pronunciation of a Chinese sentence Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 25/100

SWivid/F5-TTS#1237 · 10 comments · Opened Dec 13, 2025
RuntimeError: Could not load libtorchcodec Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1234 · 7 comments · Opened Dec 7, 2025
licensed training Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1228 · Opened Nov 25, 2025
FlashAttention training is unstable Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 25/100

SWivid/F5-TTS#1217 · 7 comments · Opened Nov 11, 2025
My RTF is always around 0.12 with TRT-LLM docker on L20, never match the claimed benchmark result Open

documentation enhancement help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1214 · 30 comments · 1 reaction · Opened Nov 10, 2025
Seeking Fine-tuning Guidance and Experience Sharing Open

help wanted

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1211 · 1 comment · Opened Nov 7, 2025
Prepared a REST API wrapper for F5-TTS Open

documentation

Difficulty 5/5 Over a week Newbie friendliness 10/100

SWivid/F5-TTS#1208 · 2 reactions · Opened Nov 3, 2025
finetune-gradio "test model" function has inference bug with speech & text mismatch, caused by clipped ref_audio and unchanged ref_text (sampled from dataset) Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 38/100

SWivid/F5-TTS#1207 · 8 comments · Opened Nov 3, 2025
Suggestion for generation of long text like podcast Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1205 · 1 comment · Opened Nov 2, 2025
After performing static and dynamic quantization to int8, the inference speed became slower rather than faster Open

enhancement help wanted

Difficulty 4/5 3-5 days Newbie friendliness 35/100

SWivid/F5-TTS#1203 · 1 comment · Opened Oct 31, 2025
`RuntimeError: FlashAttention only support fp16 and bf16 data type` when running training with flash_attn, log_samples: true & mixed_precision=bf16 Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 35/100

SWivid/F5-TTS#1201 · 1 comment · Opened Oct 29, 2025
Add CPU only docker support Open

enhancement

Difficulty 3/5 1-2 days Newbie friendliness 45/100

SWivid/F5-TTS#1198 · Opened Oct 16, 2025
Which approach to use for training voice with emotions? Open

question

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#1193 · Opened Oct 11, 2025
Use it as on the fly reader for audiobooks Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1186 · 2 comments · Opened Sep 24, 2025
A Todo List for TRT-LLM Recipe. Open

bug enhancement

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1182 · 4 comments · Opened Sep 19, 2025
请教一下src/f5_tts/runtime/triton_trtllm/client_grpc.py，这个加速推理支持生成音频的语速和音量的调节吗 Open

question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

SWivid/F5-TTS#1180 · Opened Sep 18, 2025
regarding speed adjustment for small length text Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1155 · 2 comments · 6 reactions · Opened Aug 7, 2025
Why Sinusoidal Positional Encoding before ConvNeXt in Text Embedding? Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1153 · 4 reactions · Opened Jul 31, 2025
AMD old cards not working Open

help wanted

Difficulty 4/5 3-5 days Newbie friendliness 20/100

SWivid/F5-TTS#1150 · 2 comments · Opened Jul 28, 2025
Has anybody tried the trt inference with onnx model exported? Open

question

Difficulty 5/5 Over a week Newbie friendliness 15/100

SWivid/F5-TTS#1142 · Opened Jul 21, 2025
Explicit pauses Open

enhancement

Difficulty 4/5 3-5 days Newbie friendliness 25/100

SWivid/F5-TTS#1069 · 7 comments · Opened Jun 2, 2025
comfyui native support please Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#1068 · 1 comment · Opened Jun 1, 2025
The beginning of the synthesized audio contains other speech Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1041 · 3 comments · Opened May 19, 2025
Request for guidance on building a native language model with natural emotional expression Open

question

Difficulty 5/5 Over a week Newbie friendliness 15/100

SWivid/F5-TTS#1035 · 1 comment · Opened May 14, 2025
I think the pause time for punctuation marks is insufficient. I want to insert a one-second or two-second pause in the text. Is there a way to achieve this? Similar to the <break> tag of azure sslm Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#1034 · 3 comments · Opened May 14, 2025
Request for Guidance on Tamil Voice Cloning – Training Data, Metadata & Multi-Speaker Support Open

question

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#989 · 5 comments · 1 reaction · Opened Apr 22, 2025
text normalizer Open

enhancement

Difficulty 4/5 3-5 days Newbie friendliness 30/100

SWivid/F5-TTS#951 · 1 comment · Opened Apr 3, 2025
Deployed on edge devices Open

documentation question

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#694 · 2 comments · 1 reaction · Opened Jan 6, 2025
Convert vocos checkpoint to bigvgan; bigvgan with vocos-style mel spectrograms Open

documentation enhancement

Difficulty 5/5 Over a week Newbie friendliness 15/100

SWivid/F5-TTS#654 · 4 comments · Opened Dec 20, 2024
LoRA support Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#621 · 3 comments · 1 reaction · Opened Dec 12, 2024
We are trying to use sentiment text to make the generated audio have different emotions Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#606 · 15 comments · Opened Dec 8, 2024
Will there be an entension model like E2 TTS X2 Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

SWivid/F5-TTS#576 · Opened Dec 3, 2024
EQ correction on the reference audio is crucial for achieving high-quality results. Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#378 · 7 comments · 5 reactions · Opened Nov 2, 2024
how to f5 tts train for other language Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

SWivid/F5-TTS#291 · 49 comments · Opened Oct 27, 2024
Supported Languages? Open

Difficulty 3/5 1-2 days Newbie friendliness 25/100

SWivid/F5-TTS#87 · 139 comments · 15 reactions · Opened Oct 15, 2024

SWivid/F5-TTS

Get new issues in your inbox