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

Recover gracefully from VRAM out of memory errors #5793

Closed
wants to merge 1 commit into from
Closed

Recover gracefully from VRAM out of memory errors #5793

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
13 changes: 12 additions & 1 deletion invokeai/backend/model_management/model_cache.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -287,6 +287,14 @@ def _move_model_to_device(self, key: str, target_device: torch.device):
if torch.device(source_device).type == torch.device(target_device).type:
return

if target_device.type == "cuda":
vram_device = (
target_device if target_device.index is not None else torch.device(str(target_device), index=0)
)
free_mem, _ = torch.cuda.mem_get_info(torch.device(vram_device))
if cache_entry.size > free_mem:
raise torch.cuda.OutOfMemoryError

start_model_to_time = time.time()
snapshot_before = self._capture_memory_snapshot()
cache_entry.model.to(target_device)
Expand Down Expand Up @@ -356,6 +364,10 @@ def __enter__(self) -> Any:
self.cache.logger.debug(f"Locking {self.key} in {self.cache.execution_device}")
self.cache._print_cuda_stats()

except torch.cuda.OutOfMemoryError:
self.cache.logger.warning("Out of GPU memory encountered.")
self.cache_entry.unlock()
raise
except Exception:
self.cache_entry.unlock()
raise
Expand Down Expand Up @@ -524,7 +536,6 @@ def _offload_unlocked_models(self, size_needed: int = 0):
break
if not cache_entry.locked and cache_entry.loaded:
self._move_model_to_device(model_key, self.storage_device)

vram_in_use = torch.cuda.memory_allocated()
self.logger.debug(f"{(vram_in_use/GIG):.2f}GB VRAM used for models; max allowed={(reserved/GIG):.2f}GB")

Expand Down
Loading