| .. |
|
__init__.py
|
b1507687ae
Use hivemind.Optimizer in examples/albert (#426)
|
4 lat temu |
|
adaptive.py
|
ed4204009f
move PerformanceEMA to utils, TrainingAverager to optim, update utils (#405)
|
4 lat temu |
|
base.py
|
b1507687ae
Use hivemind.Optimizer in examples/albert (#426)
|
4 lat temu |
|
collaborative.py
|
b1507687ae
Use hivemind.Optimizer in examples/albert (#426)
|
4 lat temu |
|
grad_averager.py
|
25e7e1216e
Merge update
|
4 lat temu |
|
grad_scaler.py
|
6da8683975
Improve All-Reduce fault-tolerance (#423)
|
4 lat temu |
|
optimizer.py
|
795ca8c30c
stile
|
4 lat temu |
|
power_ef_averager.py
|
46d61e0a51
better naming
|
4 lat temu |
|
progress_tracker.py
|
b1507687ae
Use hivemind.Optimizer in examples/albert (#426)
|
4 lat temu |
|
simple.py
|
12099ce03c
Remove trailing dots in log messages and errors (#419)
|
4 lat temu |
|
state_averager.py
|
b1507687ae
Use hivemind.Optimizer in examples/albert (#426)
|
4 lat temu |
|
training_averager.py
|
12099ce03c
Remove trailing dots in log messages and errors (#419)
|
4 lat temu |
Artem Chumachenko
46d61e0a51
better naming