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.

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

Update logits_process.py docstrings to clarify penalty and reward cases (attempt #2) #26784

Merged
merged 2 commits into from
Oct 17, 2023
Merged
Changes from all commits
Commits
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
26 changes: 19 additions & 7 deletions src/transformers/generation/logits_process.py
Original file line number Diff line number Diff line change
Expand Up @@ -276,9 +276,14 @@ class RepetitionPenaltyLogitsProcessor(LogitsProcessor):
selected. The formula can be seen in the original [paper](https://arxiv.org/pdf/1909.05858.pdf). According to the
paper a penalty of around 1.2 yields a good balance between truthful generation and lack of repetition.

This technique can also be used to reward and thus encourage repetition in a similar manner. To penalize and reduce
repetition, use `penalty` values above 1.0, where a higher value penalizes more strongly. To reward and encourage
repetition, use `penalty` values between 0.0 and 1.0, where a lower value rewards more strongly.

Args:
repetition_penalty (`float`):
The parameter for repetition penalty. 1.0 means no penalty. See [this
penalty (`float`):
The parameter for repetition penalty. 1.0 means no penalty. Above 1.0 penalizes previously generated
tokens. Between 0.0 and 1.0 rewards previously generated tokens. See [this
paper](https://arxiv.org/pdf/1909.05858.pdf) for more details.

Examples:
Expand Down Expand Up @@ -313,7 +318,7 @@ def __init__(self, penalty: float):
def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> torch.FloatTensor:
score = torch.gather(scores, 1, input_ids)

# if score < 0 then repetition penalty has to be multiplied to reduce the previous token probability
# if score < 0 then repetition penalty has to be multiplied to reduce the token probabilities
score = torch.where(score < 0, score * self.penalty, score / self.penalty)

scores.scatter_(1, input_ids, score)
Expand All @@ -322,11 +327,18 @@ def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> to

class EncoderRepetitionPenaltyLogitsProcessor(LogitsProcessor):
r"""
[`LogitsProcessor`] enforcing an exponential penalty on tokens that are not in the original input.
[`LogitsProcessor`] that avoids hallucination by boosting the probabilities of tokens found within the original
input.

This technique can also be used to reward and thus encourage hallucination (or creativity) in a similar manner. To
penalize and reduce hallucination, use `penalty` values above 1.0, where a higher value penalizes more strongly. To
reward and encourage hallucination, use `penalty` values between 0.0 and 1.0, where a lower value rewards more
strongly.

Args:
hallucination_penalty (`float`):
The parameter for hallucination penalty. 1.0 means no penalty.
penalty (`float`):
The parameter for hallucination penalty. 1.0 means no penalty. Above 1.0 penalizes hallucination. Between
0.0 and 1.0 rewards hallucination.
encoder_input_ids (`torch.LongTensor`):
The encoder_input_ids that should be repeated within the decoder ids.
"""
Expand All @@ -342,7 +354,7 @@ def __init__(self, penalty: float, encoder_input_ids: torch.LongTensor):
def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> torch.FloatTensor:
score = torch.gather(scores, 1, self.encoder_input_ids)

# if score < 0 then repetition penalty has to be multiplied to reduce the previous token probability
# if score < 0 then hallucination penalty has to be multiplied to increase the token probabilities
score = torch.where(score < 0, score * self.penalty, score / self.penalty)

scores.scatter_(1, self.encoder_input_ids, score)
Expand Down
Loading