LLMwatermark / alternative_prf_schemes.py
jianuo
first upload
c99a3a6
raw
history blame
5.97 kB
"""实现其他 PRF 函数(这些函数的不同之处仅在于如何从上下文中的令牌生成单个哈希值)。
可作为修改后的基类 WatermarkBase 挂接到现有的 WatermarkLogitsProcessor 中,请参见
extended_watermark_processor.py 中的实现。
"""
import torch
from itertools import combinations
from functools import cache
# 哈希方案的关键属性
props = {
"prf_type": str, # 基础 PRF 的字符串名称,将多个令牌 ID 映射到随机种子
"context_width": int, # 这是论文中的 h,每个 PRF 应考虑多少个先前的令牌
"self_salt": bool, # 根据鲁棒水印技术中的规则,是否使用令牌本身来生成种子,并可能拒绝其自身的列表
"hash_key": int, # 整数,大质数,用于将种子移动到上述所选 PRF 中的低熵位序列的远离位置
}
def seeding_scheme_lookup(seeding_scheme: str):
if not isinstance(seeding_scheme, str):
raise ValueError("Seeding scheme should be a string summarizing the procedure.")
if seeding_scheme == "simple_1" or seeding_scheme == "lefthash":
# 默认的简单二元哈希 # 别名为 ff-additive_prf-1-False-15485863
prf_type = "additive_prf"
context_width = 1
self_salt = False
hash_key = 15485863
elif seeding_scheme == "algorithm-3" or seeding_scheme == "selfhash":
prf_type = "anchored_minhash_prf"
context_width = 4
self_salt = True
hash_key = 15485863
elif seeding_scheme == "minhash":
prf_type = "minhash_prf"
context_width = 4
self_salt = False
hash_key = 15485863
elif seeding_scheme == "skipgram":
prf_type = "skipgram_prf"
context_width = 5
self_salt = False
hash_key = 15485863
elif seeding_scheme.startswith("ff"): # 自由形式的种子方案 API - 仅用于实验目的
# 期望形式为 ff-additive_prf-4-True-hash 或 ff-additive_prf-5-True (哈希键是可选的)
split_scheme = seeding_scheme.split("-")
prf_type = str(split_scheme[1])
context_width = int(split_scheme[2])
self_salt = split_scheme[3] == "True"
if len(split_scheme) == 5:
hash_key = int(split_scheme[4])
else:
hash_key = 15485863
else:
raise ValueError(f"Invalid seeding scheme name {seeding_scheme} given. Try 'simple_1'?")
assert prf_type in prf_lookup.keys()
return prf_type, context_width, self_salt, hash_key
def multiplicative_prf(input_ids: torch.LongTensor, salt_key: int) -> int:
return salt_key * input_ids.prod().item()
def additive_prf(input_ids: torch.LongTensor, salt_key: int) -> int:
return salt_key * input_ids.sum().item()
def minfunc_prf(input_ids: torch.LongTensor, salt_key: int) -> int:
# 对于非随机输入 id(如文本),这不是一个好主意
return salt_key * input_ids.min().item()
def simple_skip_prf(input_ids: torch.LongTensor, salt_key: int, k=2) -> int:
# k是一个跳跃的距离
return hashint(salt_key * input_ids[::k]).prod().item()
def skipgram_prf(input_ids: torch.LongTensor, salt_key: int) -> int:
# # 上下文内的最大距离跳字
return hashint(salt_key * input_ids[0]).item()
def anchored_skipgram_prf(input_ids: torch.LongTensor, salt_key: int, anchor: int = -1) -> int:
# 上下文内的最大距离跳字
return (hashint(salt_key * input_ids[0]) * hashint(salt_key * input_ids[anchor])).item()
def minhash_prf(input_ids: torch.LongTensor, salt_key: int) -> int:
return hashint(salt_key * input_ids).min().item()
def anchored_minhash_prf(input_ids: torch.LongTensor, salt_key: int, anchor: int = -1) -> int:
# 另一个关键是生成一个key
return (salt_key * hashint(input_ids) * hashint(input_ids[anchor])).min().item()
def minskipgram_prf(input_ids: torch.LongTensor, salt_key: int, k: int = 2) -> int:
# 上下文中所有跳字组合的最小值,k=2 表示所有对
skipgrams = torch.as_tensor(list(combinations(hashint(salt_key * input_ids), 2)))
return skipgrams.prod(dim=1).min().item()
def noncomm_prf(input_ids: torch.LongTensor, salt_key: int, k: int = 2) -> int:
key = torch.as_tensor(salt_key, dtype=torch.long)
for entry in input_ids:
key *= hashint(key * entry)
key %= 2**32
return key.item()
def position_prf(input_ids: torch.LongTensor, salt_key: int, k: int = 2) -> int:
return (salt_key * input_ids * torch.arange(1, len(input_ids) + 1, device=input_ids.device)).sum().item()
prf_lookup = {
"multiplicative_prf": multiplicative_prf,
"additive_prf": additive_prf,
"minfunc_prf": minfunc_prf,
"simple_skip_prf": simple_skip_prf,
"skipgram_prf": skipgram_prf,
"anchored_skipgram_prf": anchored_skipgram_prf,
"minhash_prf": minhash_prf,
"anchored_minhash_prf": anchored_minhash_prf,
"minskipgram_prf": minskipgram_prf,
"noncomm_prf": noncomm_prf,
"position_prf": position_prf,
}
# 在启动时生成全局置换表一次
rng = torch.Generator(device=torch.device("cpu"))
rng.manual_seed(2971215073)
table_size = 1_000_003
fixed_table = torch.randperm(1_000_003, device=torch.device("cpu"), generator=rng) # 这个速度很快
def hashint(integer_tensor: torch.LongTensor) -> torch.LongTensor:
return fixed_table[integer_tensor.cpu() % table_size] + 1 # 这里有一个小技巧,这个函数总是返回 CPU 的值
def _hashint_avalanche_tensor(integer_tensor: torch.LongTensor):
i = integer_tensor.to(torch.int32).clone() # or torch.int16?
i -= i << 6
i ^= i >> 17
i -= i << 9
i ^= i << 4
i -= i << 3
i ^= i << 10
i ^= i >> 15
return i.to(torch.long)
@cache
def _hashint_avalanche_int(integer: int):
i = integer % (2**32)
i -= i << 6
i ^= i >> 17
i -= i << 9
i ^= i << 4
i -= i << 3
i ^= i << 10
i ^= i >> 15
return i