Repository Issue Activity (beta)

#1315 Verify evals on Papers with Code

#1306 [Vulnerability] OS command injection in finetune_gradio.py start_training() via shell=True

#1312 中文文本中的阿拉伯数字会被转换为英文

#902 IndicF5

#1263 Short sentences glitches solved for me

#1308 fix_duration 被每个 chunk 重复使用,长文本生成 N 倍变长(倍速感来源)

#1214 My RTF is always around 0.12 with TRT-LLM docker on L20, never match the claimed benchmark result

#1300 Feature: Consider SenseVoice for training data transcription

#1305 FileNotFoundError: [WinError 2] The system cannot find the file specified

#840 32khz or 44.1khz supports

#1292 Document that F5TTS_v1_Base/model_1250000.safetensors contains only EMA-prefixed weights

#1162 Update numpy version constraint to support numpy 2.x

#955 ref text with end punctuation.

#1236 RuntimeError: HIP error: invalid device function

#1234 RuntimeError: Could not load libtorchcodec

#1293 [Vulnerability]Path Traversal Leading to Arbitrary Directory Creation and Fixed-Name File Write Outside Intended Base Directory in SWivid/F5-TTS

#1289 Workaround: UI freeze / High CPU usage when changing tabs (Gradio 6.11 bug)

#1287 Can't generate anything if text is longer than one batch

#1286 What is the best way to contribute Benchmark information?

#1283 400k steps into training, still heavy halucinating/unintelligible

#1238 mispronounces some words

#1113 知识付费，请求搭建F5-TTS-triton_trtllm

#1282 About training 1.2M updates over one week mentioned in the paper

#1277 叠字会出现吞字

#1276 📋 Documentation Enhancement Suggestion