issues
search
RobinBruegger
/
RevTorch
Framework for creating (partially) reversible neural networks with PyTorch
BSD 3-Clause "New" or "Revised" License
145
stars
18
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Allow amp training
#15
vovaf709
opened
2 years ago
0
Fixes #13
#14
vovaf709
opened
2 years ago
0
Corrupting running stats on backward
#13
vovaf709
opened
2 years ago
0
Feature Request: iRevNet
#12
etienne87
opened
3 years ago
0
The autograd might not work as expected
#11
rouniuyizu
closed
1 year ago
2
Can RevTorch work in multi-gpu?
#10
ewrfcas
opened
4 years ago
1
Feature/allow multple backward passes
#9
RobinBruegger
closed
4 years ago
2
add eagerly_discard_variables flag for ReversibleSequence to be able …
#8
lucidrains
closed
4 years ago
1
Trying to backward through the graph a second time ... Specify retain_graph=True
#7
lucidrains
closed
4 years ago
4
be able to pass additional arguments to F and G
#6
lucidrains
closed
4 years ago
8
Be able to pass additional arguments to F and G reversible blocks
#5
lucidrains
closed
4 years ago
0
update to add setting for fixing a seed before execution of f and g f…
#4
lucidrains
closed
4 years ago
4
Reproducing the same random outputs in forward and backward passes
#3
ag1988
closed
4 years ago
6
make dimension on which input is chunked and then concatted modifiable
#2
lucidrains
closed
4 years ago
2
Subtle memory leak in _ReversibleModuleFunction
#1
spezold
closed
5 years ago
4