Skip to content

Conversation

@fcollman
Copy link
Collaborator

moved the middle_auth_client features over to the v2 graph to support public neurons for FANC. Needs testing.

@fcollman
Copy link
Collaborator Author

this is now tested and working

@akhileshh
Copy link
Contributor

Thanks! chunkMapping.proto is not required anymore though. I must have forgotten to remove the generated file.
It was more efficient to encode mapping in the components itself (see here).

@fcollman
Copy link
Collaborator Author

Gotcha. Removed the unused .proto and compiled files now.

@sdorkenw sdorkenw merged commit 7624d0d into main Nov 30, 2022
@sdorkenw sdorkenw deleted the add_public_neurons branch November 30, 2022 23:32
sdorkenw pushed a commit that referenced this pull request Jan 3, 2023
* feat: add debug functions

* fix: refactor and fix incorrect bounding l2 children due to skips, and speedups

* code cleanup

* decide if unique on number of IDs involved

* sort IDs for faster reads

* test pyfqmr

* fix: log L2 IDs being remeshed

* undo meshgen changes, handled in another pr

* ingest faster automatic enqueuing (#428)

* fix: improved auto queuing

* feat: faster tracker worker with preloaded imports

* Add public neurons (#426)

* adding public info file

* adding mesh manifest public neuron

* using config for AUTH_TOKEN

* adding more endpoints for public neurons

* fixing AUTH_TOKEN

* adding oldest timestamp

* removing node id

* removing auth required before/after request

* trying to add reset auth

* removing unused reset auth

* pinning zmesh back

* adding compiled req feature

* trying to fix tests

* compiled new protobuf files

* new chunkMapping proto files

* modifiying input reqs

* removing unused chunkMapping files

* improve mesh worker logging

* feat: add graph versioning

* feat: add debug functions

* fix: refactor and fix incorrect bounding l2 children due to skips, and speedups

* code cleanup

* decide if unique on number of IDs involved

* sort IDs for faster reads

* test pyfqmr

* fix: log L2 IDs being remeshed

* undo meshgen changes, handled in another pr

* improve mesh worker logging

* feat: add graph versioning

* ensure correct version in frontend

* fix: valid version takes precedence over whitelist

* fix: add to cache after checks pass

* fix: retain error message

* fixing whitelist logging

* fixing merge with skip connections (#432)

* fixing merge with skip connections

* fixing merge with skip connections

* fix: mistaken change

* remove hardcoded credentials from tests

Co-authored-by: Forrest Collman <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants