issues
search
lucidrains
/
audiolm-pytorch
Implementation of AudioLM, a SOTA Language Modeling Approach to Audio Generation out of Google Research, in Pytorch
MIT License
2.36k
stars
255
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
EOFError: Ran out of input when trying to train CoarseTransformer
#103
adamfils
closed
1 year ago
1
Soundstream failes to load and resume training
#102
adamfils
closed
1 year ago
1
typical data length and train steps of the transformers?
#101
cyanbx
closed
1 year ago
4
Modify default SoundStream strides to 480 to match default 24kHz sampling rate
#100
smcio
closed
1 year ago
3
Should Attention class compute separate keys & values across heads?
#99
LWprogramming
closed
1 year ago
4
How can I Increase Length of SoundStream Output?
#98
adamfils2
closed
1 year ago
9
Soundstream saving function is broken again...
#97
ckwdani
closed
1 year ago
5
The code cannot use the “load” function to load ckpt normally
#96
MingHui-Fang
closed
1 year ago
1
Use correct default feedforward dropout
#95
LWprogramming
closed
1 year ago
3
remove unused pad_id
#94
LWprogramming
closed
1 year ago
1
SoundStreamTrainer tweaks
#93
ilya16
closed
1 year ago
1
Audio downsampling for MultiScaleDiscriminators
#92
ilya16
closed
1 year ago
3
When training soundstream, how to continue training on a model pt file that has been trained to a certain number of steps
#91
So-Fann
closed
1 year ago
4
How to use multi-GPU training? I can't use CUDA_VISIBLE_DEVICES to implement multi-GPU
#90
lzl1456
closed
1 year ago
4
soundstream.load() no longer works in 0.12.3
#89
djqualia
closed
1 year ago
7
Error training soundstream on 0.12.1
#88
djqualia
closed
1 year ago
3
Can you please implement 'SingSong' by Google?
#87
adamfils2
closed
1 year ago
4
Adapting AudioLM to support SingSong style accompaniment generation
#86
smcio
opened
1 year ago
10
Remove print statement from debugging
#85
djqualia
closed
1 year ago
1
support SPEAR-TTS
#84
Rongjiehuang
closed
11 months ago
8
ComplexConv2d in ComplexSTFTDiscriminator gives RuntimeError
#83
lg-lg
closed
1 year ago
8
Consider default of normalized=True for all STFT and Mel transforms
#82
turian
closed
1 year ago
7
l2norm should compute the norm, not normalize
#81
zhvng
closed
1 year ago
3
typical range of `num_train_steps`?
#80
naotokui
opened
1 year ago
32
Update README.md
#79
eltociear
closed
1 year ago
2
hubert instead of w2v-bert ?
#78
jackieassa
opened
1 year ago
3
Training VALL-E
#77
pashanitw
opened
1 year ago
2
multi-scale spectral reconstruction loss
#76
AndreyBocharnikov
closed
1 year ago
6
small bug fixes
#75
zhvng
closed
1 year ago
1
[feature request] add trained model with saved weights
#74
amirgamil
opened
1 year ago
5
batch unique consecutive in CoarseTransformer
#73
zhvng
closed
1 year ago
1
I can't make it work
#72
olainid
closed
1 year ago
2
Fix training output sample hz in non-default cases
#71
djqualia
closed
1 year ago
1
fix generate
#70
zhvng
closed
1 year ago
0
bugfix - forgetful mask was not used
#69
zhvng
closed
1 year ago
1
Convert stereo/multi-channel audio to mono
#68
djqualia
closed
1 year ago
2
Prefix context in CoarseTransformer and FineTransformer
#67
zhvng
opened
1 year ago
13
Wave discriminator?
#66
turian
closed
1 year ago
2
less is more
#65
lucidrains
closed
1 year ago
2
use_complex_stft_discriminator = False
#64
inspirit
closed
1 year ago
4
Add demo Jupyter notebook to run end-to-end
#63
LWprogramming
closed
1 year ago
1
Support our open source music pretrained Transformer
#62
a43992899
opened
1 year ago
6
Soundstream loss doesn't decrease after 1167 steps - version 0.7.1
#61
yigityu
closed
1 year ago
51
Consider adding a loss balancer?
#60
turian
opened
1 year ago
5
Special reason for this single kernel to be 3?
#59
inspirit
closed
1 year ago
4
Activation units position
#58
inspirit
closed
1 year ago
6
Soundstream does not get better in training after update
#57
ckwdani
closed
1 year ago
10
Global num_workers?
#56
turian
closed
1 year ago
3
RuntimeError: stack expects each tensor to be equal size, but got [5440] at entry 0 and [5120] at entry 2
#55
turian
closed
1 year ago
5
How do I know SoundStream is properly trained?
#54
adamfils2
closed
1 year ago
8
Previous
Next