Open
Conversation
Member
|
@andyyPark Thanks for your entry! Ahaha, am I glad to see another JAX neural network approach! When you have them, could you add to your description a few metrics, I'm very curious to see how the CNN compares to the Dense network ;-) |
Member
Author
|
Since my original submission, I have changed my original jaxCNN to maximize the FOM_DETF score, and have implemented ResNet50 (jaxResNet.py) using jax and jax-cosmo library. Although I don't have any scores to show, below is an example of the binning generated for 5 and 6 bins using the Buzzard dataset: 5 Bins (jaxCNN)
6 Bins (jaxCNN)
5 Bins (jaxResNet)
6 Bins (jaxResNet)
I haven't tried this yet but it seems like my original CNN network (jaxCNN.py) performs better than ResNet50 with epochs ~O(100). |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This method was inspired by #9 where I exploited the auto differentiable metrics from
jax-cosmolibrary. Flax was used to implement a simple Convolutional Neural Network that assigns bins to galaxies.My network was
rizbandsI have created a jupyter notebook,
JaxCNN.ipynb, that walks through my code in thenotebooksfolder, but I still have yet to finish running the notebook on NERSC.Below is an example of the binning generated for 4 bins:
Scores.ipynbin thenotebooksfolder shows the plots of the metrics for a different number of bins.FOM_3x2
FOM_DETF_3x2
SNR_3x2