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

[XPU] Enable empty cache on XPU device #438

Merged
merged 1 commit into from
Aug 26, 2024
Merged

[XPU] Enable empty cache on XPU device #438

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
6 changes: 3 additions & 3 deletions megatron/training.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -724,7 +724,7 @@ def train_step(forward_step_func, data_iterator,

# Empty unused memory.
if args.empty_unused_memory_level >= 1:
torch.cuda.empty_cache()
get_accelerator().empty_cache()

# Reduce gradients.
if not args.deepspeed:
Expand Down Expand Up @@ -781,7 +781,7 @@ def train_step(forward_step_func, data_iterator,

# Empty unused memory.
if args.empty_unused_memory_level >= 2:
torch.cuda.empty_cache()
get_accelerator().empty_cache()

if mpu.is_pipeline_last_stage(ignore_virtual=True):
# Average loss across microbatches.
Expand Down Expand Up @@ -1437,7 +1437,7 @@ def evaluate(forward_step_func,

# Empty unused memory
if args.empty_unused_memory_level >= 1:
torch.cuda.empty_cache()
get_accelerator().empty_cache()

if mpu.is_pipeline_last_stage(ignore_virtual=True):
# Reduce across processes.
Expand Down