Comments (8)
Hey, thanks for the response!
I have tried between different numbers from 1 up to 9 (9 is the default setting if I am not mistaken) for n_quantizers. (--n_quantizers argument when encoding).
However, I cannot see any impact on the compressed .dac file, its size remains the same. I would expect the file size to change depending on the compression ratio. Am I missing something?
Cheers!
from descript-audio-codec.
@ekgren thank you for raising this issue!
However, I cannot see any impact on the compressed .dac file, its size remains the same. I would expect the file size to change depending on the compression ratio. Am I missing something?
Yes, this was a bug on our end. We have release a new version that works as expected.
Please update your dac installation.
pip install descript-audio-codec==0.0.3
from descript-audio-codec.
@listener17 The default bitrate is 8kbps for the following command
python3 -m dac encode /path/to/input --output /path/to/output/codes
from descript-audio-codec.
Hello,
Thank you for providing your code!
Similar to ekgren, I was wondering how can I test different compression ratios and the impact on reconstruction quality.
My understanding is that the model needs to be retrained with a different setting for encoder and decoder rates (stride settings). Is that correct? If so, could you please provide more checkpoints for those different settings if available?
Thanks!
Mary
from descript-audio-codec.
Different compression ratios can be passed in via the n_quantizers
argument here: https://github.com/descriptinc/descript-audio-codec/blob/main/dac/model/dac.py#L268. This would be for the pretrained model. We currently don't have any other checkpoints trained, though, with different strides, but they may come if we have time and resources to train them!
from descript-audio-codec.
Different compression ratios can be passed in via the
n_quantizers
argument here: https://github.com/descriptinc/descript-audio-codec/blob/main/dac/model/dac.py#L268. This would be for the pretrained model. We currently don't have any other checkpoints trained, though, with different strides, but they may come if we have time and resources to train them!
In the script in the github readme n_quantizers doesn't seem to change the compression ratio.
from dac.utils.encode import process as encode
# Encode audio signal
encoded_out = encode(signal, 'cuda', model, n_quantizers=n_quantizers)
print(encoded_out['codes'])
The codes returne are the same regardless of the number of n_quantizers put in.
from descript-audio-codec.
@ekgren @marypilataki: if I just run the encoder as:
python3 -m dac encode /path/to/input --output /path/to/output/codes
At what bitrate is it running? Did you figure that out?
from descript-audio-codec.
Great, thanks so much for the prompt response!
from descript-audio-codec.
Related Issues (20)
- Chunked inference result depends on chunk length HOT 9
- Can't set n_quantizers for encode() HOT 2
- Fine-tuning from 44.1Khz
- 16kHz configs result in shape mismatch HOT 1
- does commit_loss and codebook_loss always be equal? HOT 2
- Different code sizes when encoding versus when compressing HOT 2
- audiotools error
- How to compress stereo sound by model.encode HOT 2
- (Paper Error?) MSD Not Used? HOT 3
- Error when set win_duration small
- Encoding new file - use of `zero_pad` HOT 2
- Inference speed
- Loading DAC files is insecure due to pickle
- Error with 16khz
- Memory leak? HOT 2
- Padding Mismatches Output Dimension in Conv1d HOT 1
- broken training: please specify versions of libraries used
- tensor shape mismatch when training on 24khz LibriTTS dataset HOT 2
- Same error in #18
- Very low bitrate models
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from descript-audio-codec.