GPT-SoVITS
GPT-SoVITS copied to clipboard
RVC-Boss
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
**Windows11 CUDA 13.0** 进行 GPT-SoVITS-v4的SoVTIS训练提示 `D:\GPT-SoVITS-v4-20250529\runtime\python.exe" -s GPT_SoVITS/s2_train.py --config "D:\GPT-SoVITS-v4-20250529\TEMP/tmp_s2.json` ``` Traceback (most recent call last): File "D:\GPT-SoVITS-v4-20250529\GPT_SoVITS\s2_train.py", line 680, in main() File "D:\GPT-SoVITS-v4-20250529\GPT_SoVITS\s2_train.py", line 61, in main mp.spawn( File...
When I tried succesfully deployed api_v2, I found when I recieved the sound stream from the server, I got frequently blank wav files..I don't know what's wrong and does any...
我在手动清洗数据集的时候遇到了一个问题：当遇到一些文字无法标注的声音例如哈气声，呼吸声，笑声，哭声等等，是否应该保留？ 询问DeepSeek后它这么回答我 跟我觉得不行的猜想相符。如果贸然引入这类声音，应该会扰乱合成结果→不可控 然而我注意到它说对于支持的模型，用标签特殊标注这些声音是可行的→也就是说可以做到可控 虽然在绝大多数情况这类声音是用不到的，但请原谅我还是想合成这类声音
实测NVIDIA CMP 30HX和NVIDIA CMP 40HX无法生成音频。据说cmp系列显卡(矿卡)被阉割了半精度算力，我在config.py添加两处，识别cmp系显卡，不使用半精度。现在能成功生成音频了 ``` is_16_series = bool(re.search(r"16\d{2}", name)) and sm_version == 7.5 is_cmp_series = "CMP" in name.upper() # ✅ 新增 CMP 检查 if mem_gb < 4 or sm_version...
Excellent TTS! Can you tell me if you plan to support the Russian language? That would be very nice.
I using the latest 20250606v2pro [windows 7z package download] in the interface I could not find any setting to change it to dark version, so I tried to add ?__theme=dark...
fatal error displaying this function is for sm80,but was built for sm370. GPU rtx5090
I am experiencing a major performance difference when running on an NVIDIA Jetson Orin (ARM64) compared to a modern x86-64 system. The same code, with the same data input, runs...
