FunCodec

FunCodec copied to clipboard

Feature to resume training after stopping?

1

Hi @ZhihaoDU, does this repository have a feature to resume training after stopping after a certain number of epochs or in the middle of a epoch? Thanks in advance!

pracdl314

运行run.sh在stage 4时报错

3

报错信息如下所示： run.pl: job failed, log is in /mnt/d/Work/FunCodec/egs/LibriTTS/text2speech_laura/dump/libritts/test-other/codecs//logdir/inference.1.log cat: '/mnt/d/Work/FunCodec/egs/LibriTTS/text2speech_laura/dump/libritts/test-other/codecs//logdir/output.*/codecs.txt': No such file or directory Codes are saved to /mnt/d/Work/FunCodec/egs/LibriTTS/text2speech_laura/dump/libritts/test-other/codecs//logdir/output.*/codecs.txt and collected to /mnt/d/Work/FunCodec/egs/LibriTTS/text2speech_laura/dump/libritts/test-other/codecs//codecs.txt. codec scp files are collected into...

voidpointz

Stage 3

3

/home/rian0532/anaconda3/envs/py39Ubuntu/lib/python3.9/site-packages/torch/nn/modules/conv.py:306: UserWarning: Applied workaround for CuDNN issue, install nvrtc.so (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:80.) return F.conv1d(input, weight, bias, self.stride, [DESKTOP-PQV8NDO] 2024-04-16 14:56:40,650 (codec_basic:648) INFO: Will update discriminator: forward_step=0, disc_loss=2.0000, gen_loss=0.0000 Traceback...

a897456

How to check progress？

1

 Hi @ZhihaoDU The stage3 has started training, but where can I see the progress bar? I am a 4080 card, and the training parameters have not changed, and I...

a897456

LauraTTS模型的训练花了多长时间？

5

Dinxin

zipfile.BadZipFile: File is not a zip file

1

**Issue Decription** : I followed the instructions in the `README.md` step by step, until I encountered the following problem when I executed the command in the order described under “Use...

lemonLCT

Questions about training from scratch

2

Hello, I followed the steps in run.sh to train with the LibriTTS-R dataset. Below is the training process loss. When I use the current checkpoint to synthesize speech, it is...

Levent9

[bug] encoding阶段生成的codec.txt, 无法直接读取？

1

按照提供的[encoding_decoding.sh](https://github.com/alibaba-damo-academy/FunCodec/blob/master/egs/LibriTTS/codec/encoding_decoding.sh)脚本，encoding阶段会生成codec.txt文件 这个文件的形式类似于： utts_id "空格" json.dumps(codecs) 这个形式无法被[read_text.py](https://github.com/alibaba-damo-academy/FunCodec/blob/master/funcodec/fileio/read_text.py)直接读取，需要改写“load_jsonl_trans_int”函数，如下 `def load_jsonl_trans_int(path: Union[Path, str]) -> Dict[str, np.ndarray]: d = read_2column_text(path) retval = {} for k, v in d.items(): try: value = json.loads(v) if isinstance(value,...

zyy-fc

LauraTTS: _pickle.UnpicklingError: invalid load key, 'v'.

5

#### Environment - PyTorch version: 1.12.0 - Python version: 3.8 #### Issue Description I believe I have correctly installed the required PyTorch version as per the README instructions and have...

HDUysz

Hi author, thanks for you sharing the creative project. When I read the paper and code, I found that it is no needed speaker labels when training LauraTTS. The same...

forwiat