tensorflow-wavenet
tensorflow-wavenet copied to clipboard
ibab
A TensorFlow implementation of DeepMind's WaveNet paper
Would an equivalent @karpathy's [sanity check for overfitting a small dataset](https://twitter.com/karpathy/status/1013244313327681536) for image classification for WaveNet music generation be feeding a single music file corpus and _the same_ music file...
There seems to a problem in training when using the 1080X GPU vs V100 GPU. The quality drops way below and time to training also increases with V100 when in...
In the white paper, they mention conditioning to a particular speaker as an input they condition globally, and the TTS component as an up-sampled (deconvolution) conditioned locally. For the latter,...
hi, did anyone try to use google audiosets? https://research.google.com/audioset/ If yes, can you report your experience and maybe share a howto? thanks!
Is there any parametric or concatenative TTS projects availible for comparision? Like described here: https://deepmind.com/blog/wavenet-generative-model-raw-audio/
This addresses issue https://github.com/ibab/tensorflow-wavenet/issues/364 train.py was attempting to write to the 'train' directory and throwing a `NotFoundError` here: https://github.com/ibab/tensorflow-wavenet/blob/master/train.py#L274 which could happen before the `makedirs()` line in the save function...
The WaveNet paper mentioned very briefly (without any details) the possibility of using it for audio source separation. Examples of this include: extract the vocal part out of a piece...
Hi, first, thanks for this implementation of WaveNet! I'm interested in performing feature extraction from raw audio files. this features will be used for different tasks such as voice activity...
After training the network for about 700 times, i use generate.py to generate a wav. python generate.py --wav_out_path=generated.wav --samples 16000 ./logdir/train/2019-01-03T09-10-58/model.ckpt-700 . But there is a SyntaxError in the generate.py...
Hi, I have trained wavenet on 109 files of hi quality french audio (no noise, each file is approx 5 seconds). Unfortunately after 70K steps of training, I got files...
