-
Notifications
You must be signed in to change notification settings - Fork 176
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
Split hivemind.client into hivemind.averaging and hivemind.moe #304
Conversation
Codecov Report
@@ Coverage Diff @@
## master #304 +/- ##
==========================================
+ Coverage 81.63% 82.05% +0.41%
==========================================
Files 63 65 +2
Lines 5866 5856 -10
==========================================
+ Hits 4789 4805 +16
+ Misses 1077 1051 -26
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
Please fix the docs before merging (see comment) (already fixed)
If you have no strong preferences, let's merge #298 ahead of this this
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There are no conceptual problems
Unfortunately, can't visually check if something was missed =)
# Conflicts: # benchmarks/benchmark_dht.py # benchmarks/benchmark_throughput.py # hivemind/averaging/training.py # hivemind/hivemind_cli/run_server.py
This PR refactors the hivemind.client module, since it contains two different functionalities: Mixture-of-Experts client code and decentralized averaging. To properly separate the responsibilities of these classes, I moved them to hivemind.averaging and hivemind.moe (also putting the server under the same directory).
Also, to prepare for #275, I slightly removed the number of wildcard imports and modules with class definitions in
__init__