| .. |
|
client
|
0a0e290ea3
Add per-tensor compression, make all-reduce faster and more flexible (#272)
|
4 gadi atpakaļ |
|
dht
|
f0c5627139
Improve error handling, remove deprecated functionality (#261)
|
4 gadi atpakaļ |
|
hivemind_cli
|
2328ba9262
Fix device in Switch-MoE, overhaul Server architecture (#256)
|
4 gadi atpakaļ |
|
optim
|
86f3c0dd0d
Support auxiliary peers in CollaborativeOptimizer (#279)
|
4 gadi atpakaļ |
|
p2p
|
0b78a394c8
Add relay options to P2P (#268)
|
4 gadi atpakaļ |
|
proto
|
0a0e290ea3
Add per-tensor compression, make all-reduce faster and more flexible (#272)
|
4 gadi atpakaļ |
|
server
|
42b9b6cef8
Use logging in benchmarks, fix libp2p-related issues (#280)
|
4 gadi atpakaļ |
|
utils
|
0a0e290ea3
Add per-tensor compression, make all-reduce faster and more flexible (#272)
|
4 gadi atpakaļ |
|
__init__.py
|
34168aaf03
Switch CI to GitHub Actions (#285)
|
4 gadi atpakaļ |
justheuristic
0a0e290ea3
Add per-tensor compression, make all-reduce faster and more flexible (#272)