Closed
Bug 1462107
Opened 7 years ago
Closed 7 years ago
Federated Learning server
Categories
(Data Platform and Tools :: General, enhancement, P1)
Data Platform and Tools
General
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: fhartmann, Assigned: fhartmann, Mentored)
References
Details
(Whiteboard: [DataPlatform])
Attachments
(1 file)
A web server that coordinates the Federated Learning process should be created.
It waits for weight updates from users and updates the model periodically.
The exact protocol that this is based on still needs to be decided on.
Assignee | ||
Updated•7 years ago
|
Assignee: nobody → fhartmann
Updated•7 years ago
|
Whiteboard: [DataPlatform]
Updated•7 years ago
|
Points: --- → 3
Priority: -- → P1
Assignee | ||
Comment 1•7 years ago
|
||
Arkadiusz Komarzewski and Sunah Suh ported my Python code to Scala to make it run on the server-side.
They also added additional parts for loading the data using Spark and writing it to s3.
The resulting Spark job lives in the telemetry-streaming repository [1].
[1] https://github.com/mozilla/telemetry-streaming/pull/154
Comment 2•7 years ago
|
||
:fistbump:
The server ran, did its thing, and did it right the first time! And now we have a nice codebase to work off of for future projects.
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → FIXED
You need to log in
before you can comment on or make changes to this bug.
Description
•