PaddleSpeech
PaddleSpeech copied to clipboard
PaddlePaddle
Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation a...
### PR types New features ### PR changes APIs ### Describe sox io and kaldi feat
### Discussed in https://github.com/PaddlePaddle/PaddleSpeech/discussions/1326 Originally posted by **yt605155624** January 12, 2022 1. 写完代码之后可以用我们的 pre-commit 检查一下代码格式，注意只改自己修改的代码的格式即可，其他的代码有可能也被改了格式，不要 add 就好 ``` pip install pre-commit pre-commit run --file 你修改的代码 ``` 2. 提交 commit 中增加必要信息跳过不必要的...
## General Question 使用Paddle2ONNX对声纹模型转存，报错失败 [Paddle2ONNX] Start to parsing Paddle model... [Paddle2ONNX] Failed to parse paddlepaddle model from read content. [Paddle2ONNX] Failed to load program of PaddlePaddle model. [Paddle2ONNX] Paddle model...
### PR types One of Bug fixes ### PR changes ### Describe change the parameters and code logic , so as to make the benchmark can finish training in 5...
Hello，我在使用conformer_wenetspeech模型测试时发生Skip loading for encoder.embed.out.0.weight.的报错， 猜测是某一层的权重维度与模型不匹配，研究了很长时间都没找到错误原因，请帮忙指点下啊，感谢。 报错和日志如下， 报错: ``` 2022-09-09 16:17:06.267 | INFO | paddlespeech.s2t.exps.u2.model:setup_model:263 - Setup model! D:\Program\miniconda3\envs\paddle\lib\site-packages\paddle\fluid\dygraph\layers.py:1492: UserWarning: Skip loading for encoder.embed.out.0.weight. encoder.embed.out.0.weight receives a shape [9728, 512], but...
使用 AIshell3 训练合成器，效果如下：  与第三方TTS效果对比：  自己训练的语音质量不高，有沙沙的声音、字之间的清晰度较差。 小白请教下各位大佬，想要提高语音质量，我该如何优化？ 注：复制链接 wget 下，并将jpg 后缀需要改为 wav 谢谢！！！
### PR types New features ### PR changes Models ### Describe wav2vec2 pipeline
装好环境之后，通过aishell/asr1里面的脚本run.sh进行训练, 在创建模型参数的时候提示config里面没有num_conv_layers这个选项，见图1 **图1 报错信息** 然后看了一下对应的run.sh里的config，见图2 **图2 run.sh里面的配置参数** 发现默认提供的conformer.ymal里面没有对应的num_conv_layer的参数，请问是我的配置文件选择出了问题吗？
### PR types ### PR changes ### Describe 使用CSMSC数据集训练了transformer声源模型，waveflow作为声码器进行语音合成
## General Question I try to train a s2t model by paddlespeech on our owns machine with 7 gpus, but always report error about having no aviliable GPU memory, I...
Metadata
Owner
Metadata
Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation a...