Skip to content
This repository has been archived by the owner on Sep 18, 2024. It is now read-only.

Fix Error in examples/nas/spos/supernet.py #2961

Merged
merged 2 commits into from
Oct 19, 2020
Merged

Fix Error in examples/nas/spos/supernet.py #2961

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f110540
Merge pull request #2959 from microsoft/v1.9
liuzhe-lz Oct 15, 2020
32c9206
Fix Error in SPOS Example supernet.py
HeekangPark Oct 15, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion examples/nas/spos/supernet.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,14 +45,15 @@
torch.backends.cudnn.deterministic = True

model = ShuffleNetV2OneShot()
flops_func = model.get_candidate_flops
if args.load_checkpoint:
if not args.spos_preprocessing:
logger.warning("You might want to use SPOS preprocessing if you are loading their checkpoints.")
model.load_state_dict(load_and_parse_state_dict())
model.cuda()
if torch.cuda.device_count() > 1: # exclude last gpu, saving for data preprocessing on gpu
model = nn.DataParallel(model, device_ids=list(range(0, torch.cuda.device_count() - 1)))
mutator = SPOSSupernetTrainingMutator(model, flops_func=model.module.get_candidate_flops,
mutator = SPOSSupernetTrainingMutator(model, flops_func=flops_func,
flops_lb=290E6, flops_ub=360E6)
criterion = CrossEntropyLabelSmooth(1000, args.label_smoothing)
optimizer = torch.optim.SGD(model.parameters(), lr=args.learning_rate,
Expand Down