Conversation
ale18V
marked this pull request as ready for review
September 23, 2026 21:41
ale18V
force-pushed
the
test/decentralized-webapp-ci
branch
from
September 24, 2026 12:38
52131a5 to
73e9130
Compare
ale18V
added this pull request to stack #1235
September 24, 2026 13:05
ale18V
force-pushed
the
test/decentralized-webapp-ci
branch
from
September 24, 2026 20:31
73e9130 to
194a7c6
Compare
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What & why
This PR implements the decentralized portion of #1196 with an end-to-end webapp test for the default MNIST task. A Chrome browser trains with two Node participants over real WebRTC connections. The test also exposed a DiscoJS race that could label outgoing weights with the next aggregation round and leave peers waiting indefinitely; this PR fixes that race.
Technical plan
A dedicated Cypress job starts a real Disco server and two Node participants, then drives the browser through collaborative training. All three participants use the repository's compact CIFAR image fixture as image input for the MNIST task, assigning the images to digit label 0. The test checks completion of all epochs and model-sharing rounds, and waits for the Node peers to close during teardown. DiscoJS captures the aggregation round once for each payload exchange so concurrent arrivals cannot change outgoing message labels.
Deviations from plan
The test uses the existing 24-image CIFAR fixture rather than MNIST samples. It validates image preprocessing, browser-to-Node WebRTC, and model sharing, rather than classification quality or multiclass MNIST accuracy. The CI run also revealed the aggregation-round race, which is fixed in this PR because it prevented reliable completion.
Todo
None.