justheuristic
|
0a0e290ea3
Add per-tensor compression, make all-reduce faster and more flexible (#272)
|
4 年之前 |
justheuristic
|
053c7c7d13
Disentangle DecentralizedAverager components, add weights (#217)
|
4 年之前 |
foksly
|
edf9327e45
Support client-only participants in AllReduceProtocol (#176)
|
4 年之前 |
justheuristic
|
1d1252c30d
Load state from peers in DecentralizedAverager (#154)
|
4 年之前 |
justheuristic
|
10917b259e
Averager: update group keys after every step, infer nbits dynamically (#141)
|
4 年之前 |
justheuristic
|
8466d722da
Add Averager load balancing and public endpoints (#140)
|
4 年之前 |
justheuristic
|
e159605143
Address averaging corner cases, add benchmark_averaging.py, chunk averaged tensors, fix DHTNode get (#134)
|
4 年之前 |
justheuristic
|
eb93789ac6
Implement averaging parameters over DHT (2/3)
|
4 年之前 |
justheuristic
|
0595f4af90
Group AllReduce protocol (#119)
|
4 年之前 |