Comments (4)
Hi Jean,
Ok you are right, it is my mistake. Probably used some older version of the _likelihood() function without the scale_bound (as in the original tf_compression code, the scale lower bound does not exist in the likelihood function).
Thanks!
from compressai.
Hi, good point. Thanks for the report!
The _biases
, _factors
, _matrices
are actually updated by the auxiliary optimizer, but their gradients are computed via the "main" loss, not the auxiliary ones.
You are right to highlight it would be more correct to only update the quantiles
with the aux_optimizer
, and update all the other parameters via the main optimizer. I'm actually working on a cleaner implementation, which would also not require to override .parameters()
(which is a major issue do support multi-gpus training right now) and I'll address then.
In the meantime, the current version should still allow you to train and reproduce results. That said, please let me know if you have troubles reproducing results or training. Thanks!
from compressai.
Hi Jean,
Thank you for acknowledging this point.
Regarding being able to reproduce the training results, one interesting issue that I've encountered and had to solve, concerns MeanScaleHyperprior and JointAutoregressiveHierarchicalPriors models. In both of them, the estimated bpp on the latent entropy model (i.e. the gaussian conditional) started very high when training (around 20-30 bpp) and hasn't gone down to the the proper values (e.g. 1-2bpp). So although D loss term gone down, R remained high resulting in a poor model.
After investigation I found out that the root cause of this, is the fact that some "scale_hat" values (as predicted by the hyper-synthesis subnetwork) were negative (e.g -0.3) at the beginning of training, as there is no ReLU at the end of h_s (or context_prediction in case of JAHP). Those negative scales disrupted the entropy model in a way that the optimizer couldn't get out of.
My solution is to lower bound the predicted "scale_hat" to zero using LowerBound class. After this, training behaves normally and bpp values are as expected.
It's probably not the most elegant or correct solution, but it worked for me. One other option is to add a final ReLU to h_s (or context_prediction) but apply it only on the chunk that belongs to scale.
(Note that the ScaleHyperprior model doesn't suffer from this issue, since it has a ReLU at the end of h_s.)
If you managed to train successfully without encountering this issue at all, I'd be curious to understand what makes the difference. Maybe some other training hyper-params, since we are probably using different training scripts...
Anyway, it is a pleasure to work with your great framework. Thanks again.
Danny
from compressai.
Hi Danny, mmh this is weird.
So I've seen some papers and implementations where authors used ReLU or an exponential function to prevent the scales values going below zero. I've experimented a bit with torch.exp
for the scales but then I had convergence issues.
Anyway, regarding the GaussianConditional
class, there's already a scale_bound
parameter that uses the LowerBound
op to clip the scales to a minimum values (https://github.com/InterDigitalInc/CompressAI/blob/master/compressai/entropy_models/entropy_models.py#L496 and https://github.com/InterDigitalInc/CompressAI/blob/master/compressai/entropy_models/entropy_models.py#L567).
How do you initialize the GaussianConditional
class? Can you share your lambda and hyper-parameters values?
from compressai.
Related Issues (20)
- When will the compressai 1.2.5 be submitted to PyPI?
- Adding new codec into CompressAI HOT 4
- Lambda values missing for neural codecs HOT 1
- The reconstructed image using codec.py is inconsistent with the original image HOT 2
- How can I finetune the large bitrate model (λ=0.05) to obtain the low bitrate model (i.e. λ=0.013) HOT 2
- training SSF2020 model with quality 8 HOT 1
- The Variable-Rate model training script HOT 1
- from compressai.ans import BufferedRansEncoder, RansDecoder HOT 1
- 'utf-8' UnicodeDecodeError: codec can't decode byte oxd6 in position 1: invalid continuation byte When export to ONNX HOT 1
- quality level argument missing in train.py? HOT 1
- Issue Regarding the Use of GaussianMixtureConditional HOT 6
- seems something worng with "python -m compressai.utils.eval_model --help " HOT 2
- Issue with Using CheckerboardMaskedConv2d from layer.py HOT 1
- sensetime and google HOT 8
- When I wanted to migrate that program to sequence signal data, I found that @register_model("mbt2018") The compress and decompress methods of class JointAutoregressiveHierarchicalPriors are very complex HOT 1
- On the Compatibility of GAN HOT 2
- Request ONNX export support for Cheng2020 model HOT 1
- error when trying to use entropy bottleneck of compressai in my code HOT 1
- Module Not Found Error: 'compressai' HOT 1
- How to convert compress and decompress into two ONNXs? Normally, compress and decompress are placed on two devices
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 compressai.