Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

setting PGU device #1128

Closed
wants to merge 16 commits into from
Closed

setting PGU device #1128

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
16 commits
Select commit Hold shift + click to select a range
5c554a1
SA: for #958: set torch cuda device when finding root
shubhamagarwal92 Mar 8, 2020
2f17b2f
SA: for #958: removing root gpu hack in trainer/evaluation_loop
shubhamagarwal92 Mar 8, 2020
6d89505
SA: setting torch cuda device
shubhamagarwal92 Mar 8, 2020
54e9a5e
comment line too long
shubhamagarwal92 Mar 8, 2020
83c291d
check if root gpu exists or available
shubhamagarwal92 Mar 9, 2020
8a66308
SA: for #958: set torch cuda device when finding root
shubhamagarwal92 Mar 8, 2020
ae30b27
SA: for #958: removing root gpu hack in trainer/evaluation_loop
shubhamagarwal92 Mar 8, 2020
67072ad
SA: setting torch cuda device
shubhamagarwal92 Mar 8, 2020
c684cc9
comment line too long
shubhamagarwal92 Mar 8, 2020
3b42a40
check if root gpu exists or available
shubhamagarwal92 Mar 9, 2020
30f0bf3
Merge branch 'master' of https://github.com/shubhamagarwal92/pytorch-…
shubhamagarwal92 Mar 12, 2020
9473303
Related to #609. Filter params for tensorboard logging.
shubhamagarwal92 Mar 12, 2020
10a5b05
comment line too long
shubhamagarwal92 Mar 12, 2020
dd8a7c2
Apply suggestions from code review
Borda Mar 18, 2020
2acd85d
revert to master
Borda Mar 25, 2020
08bb574
revert to origin
Borda Mar 25, 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
4 changes: 4 additions & 0 deletions pytorch_lightning/trainer/distrib_parts.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -640,4 +640,8 @@ def determine_root_gpu_device(gpus):
# set root gpu
root_gpu = gpus[0]

# set cuda device to root gpu
root_device = (torch.device("cuda", root_gpu) if root_gpu >= 0 else torch.device("cpu"))
torch.cuda.set_device(root_device)

return root_gpu
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To me it looks like this should not go here, because the method is called "determine_root_gpu_device", and the added code also checks if the device is cpu. This code should probably go to the place where we call determine_root_gpu_device.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Where do you suggest to place this?

11 changes: 10 additions & 1 deletion pytorch_lightning/trainer/evaluation_loop.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -421,9 +421,18 @@ def evaluation_forward(self, model, batch, batch_idx, dataloader_idx, test_mode:
# single GPU data transfer
if self.single_gpu:
# for single GPU put inputs on gpu manually
root_gpu = 0

if isinstance(self.data_parallel_device_ids, list):
root_gpu = self.data_parallel_device_ids[0]

# set cuda device to root gpu
root_device = (torch.device("cuda", root_gpu)
if root_gpu >= 0 else torch.device("cpu"))
torch.cuda.set_device(root_device)
else:
raise RuntimeError(
'Expected `data_parallel_device_ids` as a list, cannot determine root gpu.'
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Above we have if self.single_gpu, so why should device ids be a list?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

There was already an if clause:if isinstance(self.data_parallel_device_ids, list) , I raised the runtime error as a safenet because root_gpu = 0 was used earlier by default.

)
batch = self.transfer_batch_to_gpu(batch, root_gpu)
args[0] = batch

Expand Down