Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

rewrite codebase #80

Draft
wants to merge 30 commits into
base: dev
Choose a base branch
from
Draft

rewrite codebase #80

wants to merge 30 commits into from

Conversation

HyperCodec
Copy link
Owner

Massive changes to the codebase

@HyperCodec HyperCodec added bug Something isn't working enhancement New feature or request labels Nov 21, 2024
@HyperCodec HyperCodec self-assigned this Nov 21, 2024
@HyperCodec
Copy link
Owner Author

All the runtime things should be pretty much done, now I just need to add all the reproduction and mutation stuff.

@HyperCodec
Copy link
Owner Author

probably should get rid of all the impl AsRef<NeuronLocation> things, kind of annoying and not really that fast if it impleents Copy.

@HyperCodec
Copy link
Owner Author

turns out the bi-directional thing was not actually necessary, should remove

@HyperCodec
Copy link
Owner Author

It's possible that with yield::now, two waiting tasks in a row may spinlock and block the entire thread from executing (potentially creating the deadlock issue again). To prevent this, I might have to fork/PR rayon and add a function that actually reorders the task to the back.

@HyperCodec
Copy link
Owner Author

for whatever reason the activations registry is coming up empty.

@HyperCodec
Copy link
Owner Author

after further debugging, it seems that the registry does in fact have activations, the activations_in_scope function is what's returning empty.

@HyperCodec
Copy link
Owner Author

HyperCodec commented Nov 26, 2024

still looks like it's hanging/deadlocking

@HyperCodec
Copy link
Owner Author

thread '<unnamed>' panicked at src/neuralnet.rs:581:26:
index out of bounds: the len is 1 but the index is 1

@HyperCodec
Copy link
Owner Author

somehow it's creating input neurons that point to invalid indices in the output layer

@HyperCodec
Copy link
Owner Author

Taking a break from this for AoC and other things, should be back after Christmas break

@HyperCodec
Copy link
Owner Author

Will probably be back to working on this soon, though I might wait until after midterms due to the amount of stuff still going on.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working enhancement New feature or request
Projects
None yet
1 participant