RVC-Project / Retrieval-based-Voice-Conversion-WebUI

Easily train a good VC model with voice data <= 10 mins!
MIT License
19.7k stars 3.04k forks source link

Can`t start training (It was working before) #1998

Closed Defflorstor closed 4 weeks ago

Defflorstor commented 4 weeks ago

Hello, sorry for bothering you, but like a half a year ago RVC was working perfectly, but for now when I press "One-Clock training" console showing me this:

Process Process-1: Traceback (most recent call last): File "multiprocessing\process.py", line 315, in bootstrap File "multiprocessing\process.py", line 108, in run File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\train_nsf_sim_cache_sid_load_pretrain.py", line 108, in run train_dataset = TextAudioLoaderMultiNSFsid(hps.data.trainingfiles, hps.data) File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\train\data_utils.py", line 18, in init self.audiopaths_and_text = load_filepaths_and_text(audiopaths_andtext) File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\train\utils.py", line 275, in load_filepaths_and_text filepaths_andtext = [line.strip().split(split) for line in f] File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\train\utils.py", line 275, in filepaths_andtext = [line.strip().split(split) for line in f] File "codecs.py", line 322, in decode UnicodeDecodeError: 'utf-8' codec can't decode byte 0xc8 in position 28: invalid continuation byte Traceback (most recent call last): File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\gradio\routes.py", line 321, in run_predict output = await app.blocks.processapi( File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\gradio\blocks.py", line 1006, in process_api result = await self.call_function(fnindex, inputs, iterator, request) File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\gradio\blocks.py", line 859, in call_function prediction = await anyio.to_thread.runsync( File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\anyio\to_thread.py", line 31, in run_sync return await get_asynclib().run_sync_in_workerthread( File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\anyio_backends_asyncio.py", line 937, in run_sync_in_workerthread return await future File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\anyio_backends_asyncio.py", line 867, in run result = context.run(func, args) File "C:\Users\Hyesos\Desktop\ИИ\RVC_(Training)\runtime\lib\site-packages\gradio\utils.py", line 408, in asynciteration return next(iterator) File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\infer-web.py", line 1217, in train1key faiss.writeindex( File "C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\runtime\lib\site-packages\faiss\swigfaiss.py", line 8255, in write_index return _swigfaiss.write_index(args) RuntimeError: Error in _cdecl faiss::FileIOWriter::FileIOWriter(const char *) at D:\a\faiss-wheels\faiss-wheels\faiss\faiss\impl\io.cpp:98: Error: 'f' failed: could not open C:\Users\Hyesos\Desktop\ИИ\RVC(Training)/logs/Goga/trained_IVF479_Flat_nprobe_1_Goga_v2.index for writing: No such file or directory


And also output information at launched programm:

Step 1: Processing data runtime\python.exe trainset_preprocess_pipelineprint.py C:\Users\Hyesos\Desktop\ИИ\RVC(Training)\Sources 40000 16 C:\Users\Hyesos\Desktop\ИИ\RVC_(Training)/logs/Goga False step2a:正在提取音高 runtime\python.exe extract_f0print.py C:\Users\Hyesos\Desktop\ИИ\RVC(Training)/logs/Goga 16 harvest Step 2b: Extracting features runtime\python.exe extract_featureprint.py cuda:0 1 0 0 C:\Users\Hyesos\Desktop\ИИ\RVC(Training)/logs/Goga v2 Step 3a: Model training started write filelist done runtime\python.exe train_nsf_sim_cache_sid_load_pretrain.py -e Goga -sr 40k -f0 1 -bs 18 -g 0 -te 500 -se 5 -pg pretrained_v2/f0G40k.pth -pd pretrained_v2/f0D40k.pth -l 1 -c 0 -sw 1 -v v2 Training complete. You can check the training logs in the console or the 'train.log' file under the experiment folder. (18698, 768),479 training index


And screenshot of all my settings:

image

Maybe some libraries are missing or I should delete some files (because I was trying to reinstall it several times). I really like this programm, can you help me, my Chinese brother?(

Defflorstor commented 4 weeks ago

Bro no fckng way it was russian "ИИ". So happy bro