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

Added skip_bos Argument to LLM Attribution To Enable Wider Model Support for Attributing Against a Single Token #1322

Open
wants to merge 9 commits into
base: master
Choose a base branch
from
Open

Added skip_bos Argument to LLM Attribution To Enable Wider Model Support for Attributing Against a Single Token #1322

Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
73dc20f
Update llm_attr.py
SulRash Aug 12, 2024
3b8b6aa
Update llm_attr.py
SulRash Aug 12, 2024
136b7da
Update llm_attr.py
SulRash Aug 12, 2024
fb93ce0
Merge branch 'master' into master
SulRash Aug 13, 2024
333852c
Changed error handling
SulRash Aug 13, 2024
5391085
Merge branch 'master' into master
SulRash Aug 15, 2024
f7373a3
Merge branch 'master' into master
SulRash Aug 18, 2024
382ae7a
Merge branch 'master' into master
SulRash Aug 29, 2024
90c1311
Fixed ufmt and linting check
SulRash Aug 29, 2024
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
20 changes: 15 additions & 5 deletions captum/attr/_core/llm_attr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -289,9 +289,12 @@ def _forward_func(
# 1st element is the total prob, rest are the target tokens
# add a leading dim for batch even we only support single instance for now
if self.include_per_token_attr:
target_log_probs = torch.stack(
[total_log_prob, *log_prob_list], dim=0 # type: ignore
).unsqueeze(0)
try:
target_log_probs = torch.stack(
[total_log_prob, *log_prob_list], dim=0 # type: ignore
).unsqueeze(0)
except TypeError:
raise TypeError("Try using the skip_bos argument.")
else:
target_log_probs = total_log_prob # type: ignore
target_probs = torch.exp(target_log_probs)
Expand Down Expand Up @@ -325,6 +328,10 @@ def attribute(
inp: InterpretableInput,
target: Union[str, torch.Tensor, None] = None,
num_trials: int = 1,
skip_bos: bool = True,
# pyre-fixme[24]: Generic type `dict` expects 2 type parameters, use
# `typing.Dict[<key type>, <value type>]` to avoid runtime subscripting
# errors.
gen_args: Optional[Dict[str, Any]] = None,
use_cached_outputs: bool = True,
# internal callback hook can be used for logging
Expand Down Expand Up @@ -375,8 +382,11 @@ def attribute(
assert gen_args is None, "gen_args must be None when target is given"

if type(target) is str:
# exclude sos
target_tokens = self.tokenizer.encode(target)[1:]
# exclude sos / bos
if skip_bos:
target_tokens = self.tokenizer.encode(target)[1:]
else:
target_tokens = self.tokenizer.encode(target)
target_tokens = torch.tensor(target_tokens)
elif type(target) is torch.Tensor:
target_tokens = target
Expand Down