watermark_processor.py

# coding=utf-8
# Copyright 2023 Authors of "A Watermark for Large Language Models"
# available at https://arxiv.org/abs/2301.10226
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import annotations
import collections
from math import sqrt

import scipy.stats

import numpy as np
import torch
from torch import Tensor
from tokenizers import Tokenizer
from transformers import LogitsProcessor

from nltk.util import ngrams

from normalizers import normalization_strategy_lookup
import torch
class WatermarkBase:
    def __init__(
        self,
        vocab: list[int] = None,
        gamma: float = 0.5,
        delta: float = 2.0,
        seeding_scheme: str = "simple_1",  # mostly unused/always default
        hash_key: int = 15485863,  # just a large prime number to create a rng seed with sufficient bit width
        select_green_tokens: bool = True,
    ):

        # watermarking parameters
        # self.tokenizer = tokenizer
        # self.cluster_k = cluster_k
        self.vocab = vocab
        self.vocab_size = len(vocab)
        self.gamma = gamma
        self.delta = delta
        self.seeding_scheme = seeding_scheme
        self.rng = None
        self.hash_key = hash_key
        self.select_green_tokens = select_green_tokens

    def _seed_rng(self, input_ids: torch.LongTensor, seeding_scheme: str = None) -> None:
        # can optionally override the seeding scheme,
        # but uses the instance attr by default
        if seeding_scheme is None:
            seeding_scheme = self.seeding_scheme

        if seeding_scheme == "simple_1":
            assert input_ids.shape[-1] >= 1, f"seeding_scheme={seeding_scheme} requires at least a 1 token prefix sequence to seed rng"
            prev_token = input_ids[-1].item()
            self.rng.manual_seed(self.hash_key * prev_token)
        else:
            raise NotImplementedError(f"Unexpected seeding_scheme: {seeding_scheme}")
        return

    def _get_greenlist_ids(self, input_ids: torch.LongTensor) -> list[int]:
        # seed the rng using the previous tokens/prefix
        # according to the seeding_scheme
        self._seed_rng(input_ids)

        greenlist_size = int(self.vocab_size * self.gamma)
        vocab_permutation = torch.randperm(self.vocab_size, device=input_ids.device, generator=self.rng)
        if self.select_green_tokens: # directly
            greenlist_ids = vocab_permutation[:greenlist_size] # new
        else: # select green via red
            greenlist_ids = vocab_permutation[(self.vocab_size - greenlist_size) :]  # legacy behavior
        return greenlist_ids

    def _cluster_get_greenlist_ids(self, input_ids, similar_matrix, final_choices):
        
        self._seed_rng(input_ids)
        
        total_greenlist_ids = []
        input_cluster_tensor = similar_matrix
        final_cluster_tensor = final_choices

        greenlist_ids = input_cluster_tensor
        total_greenlist_ids.append(greenlist_ids.flatten())

        final_greenlist_size = int(self.similar_matrix.shape[0] * self.gamma - greenlist_ids.unique().shape[0])
        final_vocab = torch.randperm(len(final_cluster_tensor), device=self.encoder_input_ids.device, generator=self.rng)
        if self.select_green_tokens:
            greenlist_ids = final_cluster_tensor[final_vocab[:final_greenlist_size]]
        else:
            greenlist_ids = final_cluster_tensor[final_vocab[(len(final_cluster_tensor) - final_greenlist_size) :]]
        total_greenlist_ids.append(greenlist_ids)

        final_greenlist_ids = torch.cat(total_greenlist_ids, dim=0)
        
        return final_greenlist_ids

class WatermarkLogitsProcessor(WatermarkBase, LogitsProcessor):

    def __init__(self, *args, **kwargs):
        self.encoder_input_ids = None
        self.similar_matrix = None
        self.distance_matrix = None
        self.cluster_k = None
        self.tokenizer = None
        self.cluster_k = None
        super().__init__(*args, **kwargs)

    def _calc_greenlist_mask(self, scores: torch.FloatTensor, greenlist_token_ids) -> torch.BoolTensor:
        # TODO lets see if we can lose this loop
        green_tokens_mask = torch.zeros_like(scores)
        for b_idx in range(len(greenlist_token_ids)):
            green_tokens_mask[b_idx][greenlist_token_ids[b_idx]] = 1
        final_mask = green_tokens_mask.bool()
        return final_mask

    def _bias_greenlist_logits(self, scores: torch.Tensor, greenlist_mask: torch.Tensor, greenlist_bias: float) -> torch.Tensor:
        scores[greenlist_mask] = scores[greenlist_mask] + greenlist_bias
        return scores

    def init_setting(self, encoder_input_ids, generation_config):
        self.num_beams = generation_config.num_beams
        self.encoder_input_ids = encoder_input_ids

        # for cluster.
        self.batch_similar_matrix = []
        self.batch_final_indices = []

        if self.similar_matrix is not None:
            assert self.encoder_input_ids != None
            
            total_indices = np.arange(self.similar_matrix.shape[0])

            for item in self.encoder_input_ids:
                all_input_tokens = []
                unique_item = torch.unique(item)
                for id in unique_item:
                    all_input_tokens.append(id.item())
                
                cur_choice_indices = self.similar_matrix[all_input_tokens][:, :self.cluster_k]
                self.batch_similar_matrix.append(torch.from_numpy(cur_choice_indices).to(self.encoder_input_ids.device))

                final_choice_indices = np.setdiff1d(total_indices, np.unique(cur_choice_indices), assume_unique=True)
                self.batch_final_indices.append(torch.from_numpy(final_choice_indices).to(self.encoder_input_ids.device))

    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> torch.FloatTensor:

        # this is lazy to allow us to colocate on the watermarked model's device
        if self.rng is None:
            self.rng = torch.Generator(device=input_ids.device)

        # NOTE, it would be nice to get rid of this batch loop, but currently,
        # the seed and partition operations are not tensor/vectorized, thus
        # each sequence in the batch needs to be treated separately.
        batched_greenlist_ids = [None for _ in range(input_ids.shape[0])]

        for b_idx in range(input_ids.shape[0]):
            # org
            if self.encoder_input_ids is not None:
                if self.similar_matrix is not None:
                    # use cluster to enhance.
                    beam_b_idx = b_idx // self.num_beams
                    greenlist_ids = self._cluster_get_greenlist_ids(input_ids[b_idx], self.batch_similar_matrix[beam_b_idx], self.batch_final_indices[beam_b_idx])
                    batched_greenlist_ids[b_idx] = greenlist_ids
                else:
                    # only use input to enhance.
                    greenlist_ids = self._get_greenlist_ids(input_ids[b_idx])
                    beam_b_idx = b_idx // self.num_beams
                    batched_greenlist_ids[b_idx] = torch.cat((greenlist_ids, self.encoder_input_ids[beam_b_idx]), dim=0)
            else:
                greenlist_ids = self._get_greenlist_ids(input_ids[b_idx])
                batched_greenlist_ids[b_idx] = greenlist_ids

        green_tokens_mask = self._calc_greenlist_mask(scores=scores, greenlist_token_ids=batched_greenlist_ids)

        scores = self._bias_greenlist_logits(scores=scores, greenlist_mask=green_tokens_mask, greenlist_bias=self.delta)
        return scores


class WatermarkDetector(WatermarkBase):
    def __init__(
        self,
        *args,
        device: torch.device = None,
        tokenizer: Tokenizer = None,
        z_threshold: float = 4.0,
        normalizers: list[str] = ["unicode"],  # or also: ["unicode", "homoglyphs", "truecase"]
        ignore_repeated_bigrams: bool = False,
        **kwargs,
    ):
        super().__init__(*args, **kwargs)
        # also configure the metrics returned/preprocessing options
        assert device, "Must pass device"
        assert tokenizer, "Need an instance of the generating tokenizer to perform detection"

        self.tokenizer = tokenizer
        self.device = device
        self.z_threshold = z_threshold
        self.rng = torch.Generator(device=self.device)

        self.similar_matrix = None
        self.cluster_k = None

        if self.seeding_scheme == "simple_1":
            self.min_prefix_len = 1
        else:
            raise NotImplementedError(f"Unexpected seeding_scheme: {self.seeding_scheme}")

        self.normalizers = []
        for normalization_strategy in normalizers:
            self.normalizers.append(normalization_strategy_lookup(normalization_strategy))
        
        self.ignore_repeated_bigrams = ignore_repeated_bigrams
        if self.ignore_repeated_bigrams: 
            assert self.seeding_scheme == "simple_1", "No repeated bigram credit variant assumes the single token seeding scheme."

    def init_setting(self, encoder_input_ids):
        # for cluster.

        assert self.similar_matrix is not None
        assert encoder_input_ids is not None
        
        total_indices = np.arange(self.similar_matrix.shape[0])
            
        self.cur_choice_indices = self.similar_matrix[encoder_input_ids][:, :self.cluster_k]
        self.final_choice_indices = np.setdiff1d(total_indices, np.unique(self.cur_choice_indices), assume_unique=True)
        self.cur_choice_indices = torch.from_numpy(self.cur_choice_indices).to(self.device)
        self.final_choice_indices = torch.from_numpy(self.final_choice_indices).to(self.device)

    def _compute_z_score(self, observed_count, T):
        # count refers to number of green tokens, T is total number of tokens
        expected_count = self.gamma
        numer = observed_count - expected_count * T
        denom = sqrt(T * expected_count * (1 - expected_count))
        z = numer / denom
        return z

    def _compute_p_value(self, z):
        p_value = scipy.stats.norm.sf(z)
        return p_value

    def _score_sequence(
        self,
        input_ids: Tensor,
        return_num_tokens_scored: bool = True,
        return_num_green_tokens: bool = True,
        return_green_fraction: bool = True,
        return_green_token_mask: bool = True,
        return_z_score: bool = True,
        return_p_value: bool = True,
    ):
        if self.ignore_repeated_bigrams:
            # Method that only counts a green/red hit once per unique bigram.
            # New num total tokens scored (T) becomes the number unique bigrams.
            # We iterate over all unqiue token bigrams in the input, computing the greenlist
            # induced by the first token in each, and then checking whether the second
            # token falls in that greenlist.
            assert return_green_token_mask == False, "Can't return the green/red mask when ignoring repeats."
            bigram_table = {}
            token_bigram_generator = ngrams(input_ids.cpu().tolist(), 2)
            freq = collections.Counter(token_bigram_generator)
            num_tokens_scored = len(freq.keys())
            for idx, bigram in enumerate(freq.keys()):
                prefix = torch.tensor([bigram[0]], device=self.device) # expects a 1-d prefix tensor on the randperm device
                greenlist_ids = self._get_greenlist_ids(prefix)
                bigram_table[bigram] = True if bigram[1] in greenlist_ids else False
            green_token_count = sum(bigram_table.values())
        else:
            num_tokens_scored = len(input_ids) - self.min_prefix_len
            if num_tokens_scored < 1:
                raise ValueError((f"Must have at least {1} token to score after "
                                f"the first min_prefix_len={self.min_prefix_len} tokens required by the seeding scheme."))
            # Standard method.
            # Since we generally need at least 1 token (for the simplest scheme)
            # we start the iteration over the token sequence with a minimum 
            # num tokens as the first prefix for the seeding scheme,
            # and at each step, compute the greenlist induced by the
            # current prefix and check if the current token falls in the greenlist.
            green_token_count, green_token_mask = 0, []
            for idx in range(self.min_prefix_len, len(input_ids)):
                curr_token = input_ids[idx]
                if self.similar_matrix is not None:
                    greenlist_ids = self._cluster_get_greenlist_ids(input_ids[:idx], self.cur_choice_indices, self.final_choice_indices)
                else:
                    greenlist_ids = self._get_greenlist_ids(input_ids[:idx])
                if curr_token in greenlist_ids:
                    green_token_count += 1
                    green_token_mask.append(True)
                else:
                    green_token_mask.append(False)

        score_dict = dict()
        if return_num_tokens_scored:
            score_dict.update(dict(num_tokens_scored=num_tokens_scored))
        if return_num_green_tokens:
            score_dict.update(dict(num_green_tokens=green_token_count))
        if return_green_fraction:
            score_dict.update(dict(green_fraction=(green_token_count / num_tokens_scored)))
        if return_z_score:
            score_dict.update(dict(z_score=self._compute_z_score(green_token_count, num_tokens_scored)))
        if return_p_value:
            z_score = score_dict.get("z_score")
            if z_score is None:
                z_score = self._compute_z_score(green_token_count, num_tokens_scored)
            score_dict.update(dict(p_value=self._compute_p_value(z_score)))
        if return_green_token_mask:
            score_dict.update(dict(green_token_mask=green_token_mask))

        return score_dict

    def detect(
        self,
        text: str = None,
        tokenized_text: list[int] = None,
        input_tokenize_text: str = None,
        return_prediction: bool = True,
        return_scores: bool = True,
        z_threshold: float = None,
        **kwargs,
    ) -> dict:

        assert (text is not None) ^ (tokenized_text is not None), "Must pass either the raw or tokenized string"
        if return_prediction:
            kwargs["return_p_value"] = True  # to return the "confidence":=1-p of positive detections

        # run optional normalizers on text
        for normalizer in self.normalizers:
            text = normalizer(text)
        if len(self.normalizers) > 0: 
            print(f"Text after normalization:\n\n{text}\n")

        if tokenized_text is None:
            assert self.tokenizer is not None, (
                "Watermark detection on raw string ",
                "requires an instance of the tokenizer ",
                "that was used at generation time.",
            )
            tokenized_text = self.tokenizer(text, return_tensors="pt", add_special_tokens=False)["input_ids"][0].to(self.device)
            if tokenized_text[0] == self.tokenizer.bos_token_id:
                tokenized_text = tokenized_text[1:]
        else:
            # try to remove the bos_tok at beginning if it's there
            if (self.tokenizer is not None) and (tokenized_text[0] == self.tokenizer.bos_token_id):
                tokenized_text = tokenized_text[1:]

        # call score method
        output_dict = {}
        if self.similar_matrix is not None:
            encoder_input_ids = set()
            for token in input_tokenize_text:
                encoder_input_ids.add(token)
            self.init_setting(list(encoder_input_ids))
            self.encoder_input_ids = torch.tensor(list(encoder_input_ids), dtype=torch.int32, device=self.device)
        score_dict = self._score_sequence(tokenized_text, **kwargs)
        if return_scores:
            output_dict.update(score_dict)
        # if passed return_prediction then perform the hypothesis test and return the outcome
        if return_prediction:
            z_threshold = z_threshold if z_threshold else self.z_threshold
            assert z_threshold is not None, "Need a threshold in order to decide outcome of detection test"
            output_dict["prediction"] = score_dict["z_score"] > z_threshold
            if output_dict["prediction"]:
                output_dict["confidence"] = 1 - score_dict["p_value"]

        return output_dict