Skip to content

Ensure autograd graphs are freed between attribute calls #1347

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

Closed
wants to merge 1 commit into from
Closed

Ensure autograd graphs are freed between attribute calls #1347

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
9 changes: 5 additions & 4 deletions captum/attr/_core/llm_attr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -574,9 +574,10 @@ def attribute(
if not gen_args:
gen_args = DEFAULT_GEN_ARGS

model_inp = self._format_model_input(inp.to_model_input())
output_tokens = self.model.generate(model_inp, **gen_args)
target_tokens = output_tokens[0][model_inp.size(1) :]
with torch.no_grad():
model_inp = self._format_model_input(inp.to_model_input())
output_tokens = self.model.generate(model_inp, **gen_args)
target_tokens = output_tokens[0][model_inp.size(1) :]
else:
assert gen_args is None, "gen_args must be None when target is given"

Expand Down Expand Up @@ -605,7 +606,7 @@ def attribute(
cur_target_idx,
),
**kwargs,
)
).detach()
attr = cast(Tensor, attr)

# will have the attr for previous output tokens
Expand Down