tensorflow / tensorflow/probability
Why inverse log determinant jacobian will be broadcastrd in tfd.TransformedDistribution?
Open
@jvdillon is already working on this.
Since Sep 30, 2019.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
In previous, I Implemented Glow in this library. #545
But my training, range of loss is too large.
I try these code. https://colab.research.google.com/gist/MokkeMeguru/650a3cc41769ae8fab9be73f9826644f/matveclu.ipynb
I notice ildj is broadcasted in tfd.TransformedDistribution.log_prob.
I think there are two problems.
-
Blockwise Bijector is invalid implementation.
folding batch_size is unexpected -
tfd.TransfromedDistribution._finish_log_prob_for_one_fiber needs some assertion
I don't think ildj can be broadcast.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.