-
Notifications
You must be signed in to change notification settings - Fork 0
/
Captum_BERT.py
157 lines (115 loc) · 8.22 KB
/
Captum_BERT.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
import os
import sys
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import torch
import torch.nn as nn
from transformers import BertTokenizer, BertForQuestionAnswering, BertConfig, PreTrainedTokenizerBase
from captum.attr import visualization as vis
from captum.attr import IntegratedGradients, LayerConductance, LayerIntegratedGradients, InterpretableEmbeddingBase
from captum.attr import configure_interpretable_embedding_layer, remove_interpretable_embedding_layer
# from transformers import AutoTokenizer, AutoModelForMaskedLM
# tokenizer = torch.hub.load('huggingface/pytorch-transformers', 'tokenizer', 'bert-base-uncased')
# model = torch.hub.load('huggingface/pytorch-transformers', 'model', 'bert-base-uncased')
# print(torch.__version__) # hx_pc -> 1.6.0 + cu101
# tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
#
# model = AutoModelForMaskedLM.from_pretrained("bert-base-uncased")
device: torch.device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
model_path: str = 'saved_models/'
# load model
model: nn.Module = BertForQuestionAnswering.from_pretrained(model_path)
model.to(device)
model.eval()
model.zero_grad()
# load tokenizer
tokenizer: PreTrainedTokenizerBase = BertTokenizer.from_pretrained(model_path)
def predict(inputs: list, token_type_ids: list=None, position_ids: list=None, attention_mask: any=None) -> nn.Module:
return model(inputs, token_type_ids=token_type_ids, position_ids=position_ids, attention_mask=attention_mask)
def squad_pos_forward_func(inputs:list, token_type_ids:list=None, position_ids:list=None, attention_mask: any=None,\
position:int=0) -> torch.Tensor:
pred: torch.Tensor = predict(inputs,
token_type_ids=token_type_ids,
position_ids=position_ids,
attention_mask=attention_mask)
pred = pred[position]
return pred.max(1).values
# Optional[int]
ref_token_id = tokenizer.pad_token_id
# Optional[int]
sep_token_id = tokenizer.sep_token_id
# Optional[int]
cls_token_id = tokenizer.cls_token_id
interpretable_embedding: InterpretableEmbeddingBase = configure_interpretable_embedding_layer(model, \
'bert.embeddings')
interpretable_embedding1: InterpretableEmbeddingBase = configure_interpretable_embedding_layer(model, \
'bert.embeddings.word_embeddings')
interpretable_embedding2: InterpretableEmbeddingBase = configure_interpretable_embedding_layer(model, \
'bert.embeddings.token_type_embeddings')
interpretable_embedding3: InterpretableEmbeddingBase = configure_interpretable_embedding_layer(model, \
'bert.embeddings.position_embeddings')
def construct_input_ref_pair(question: str, text: str, ref_token_id: int | str, sep_token_id: int | str, \
cls_token_id: int | str) \
-> (torch.Tensor, torch.Tensor, int):
question_ids: list = tokenizer.encode(question, add_special_tokens=False)
text_ids: list = tokenizer.encode(text, add_special_tokens=False)
input_ids: list = [cls_token_id] + question_ids + [sep_token_id] + text_ids + [sep_token_id]
ref_input_ids: list = [cls_token_id] + [ref_token_id] + len(question_ids) + [sep_token_id] + \
[ref_token_id] * len(text_ids) + [sep_token_id]
return torch.tensor([input_ids], device=device), torch.tensor([ref_input_ids], device=device), len(question_ids)
def construct_input_ref_token_type_pair(input_ids: torch.Tensor, sep_ind:int = 0) -> (torch.Tensor, torch.Tensor):
seq_len: int = input_ids.size(1)
token_type_ids: torch.Tensor = torch.tensor([[0 if i <= sep_ind else 1 for i in range(seq_len)]], device=device)
ref_token_type_ids: torch.Tensor = torch.zeros_like(token_type_ids, device=device) # * -1
return token_type_ids, ref_token_type_ids
def construct_input_ref_pos_id_pair(input_ids: torch.Tensor) -> (torch.Tensor, torch.Tensor):
seq_length: int = input_ids.size(1)
position_ids: torch.Tensor = torch.arange(seq_length, dtype=torch.long, device=device)
ref_position_ids: torch.Tensor = torch.zeros(seq_length, dtype=torch.long, device=device)
position_ids = position_ids.unsqueeze(0).expand_as(input_ids)
ref_position_ids = ref_position_ids.unsqueeze(0).expand_as(input_ids)
return position_ids, ref_position_ids
def construct_attention_mask(input_ids: torch.Tensor) -> torch.Tensor:
return torch.ones_like(input_ids)
def construct_bert_sub_embedding(input_ids: any, ref_input_ids: any, \
torken_type_ids: any, ref_token_type_ids: any, \
position_ids: any, ref_position_ids: any) \
-> ((torch.Tensor, torch.Tensor),(torch.Tensor, torch.Tensor),(torch.Tensor, torch.Tensor)):
input_embeddings: torch.Tensor = interpretable_embedding1.indices_to_embeddings(input_ids)
ref_input_embeddings: torch.Tensor = interpretable_embedding1.indices_to_embeddings(ref_input_ids)
input_embeddings_token_type: torch.Tensor = interpretable_embedding2.indices_to_embeddings(torken_type_ids)
ref_input_embeddings_token_type: torch.Tensor = interpretable_embedding2.indices_to_embeddings(ref_token_type_ids)
input_embeddings_position_ids: torch.Tensor = interpretable_embedding3.indices_to_embeddings(position_ids)
ref_input_embeddings_position_ids: torch.Tensor = interpretable_embedding3.indices_to_embeddings(ref_position_ids)
return (input_embeddings, ref_input_embeddings), (input_embeddings_token_type, ref_input_embeddings_token_type), \
(input_embeddings_position_ids, ref_input_embeddings_position_ids)
def construct_whole_bert_embeddings(input_ids: any, ref_input_ids: any, \
token_type_ids: any=None, ref_token_type_ids: any=None, \
position_ids: any=None, ref_position_ids:any=None)\
-> (torch.Tensor, torch.Tensor):
input_embeddings: torch.Tensor = interpretable_embedding.indices_to_embeddings(input_ids, \
token_type_ids=token_type_ids, \
position_ids=position_ids)
ref_input_embeddings: torch.Tensor = interpretable_embedding.indices_to_embeddings(ref_input_ids, \
token_type_ids=token_type_ids,\
position_ids=position_ids)
return input_embeddings, ref_input_embeddings
question, text = "What is important to us?", "It is important to us to include, empower and support humans of all kinds."
input_ids, ref_input_ids, sep_id = construct_input_ref_pair(question, text, ref_token_id, sep_token_id, cls_token_id)
token_type_ids, ref_token_type_ids = construct_input_ref_token_type_pair(input_ids, sep_id)
position_ids, ref_position_ids = construct_input_ref_pos_id_pair(input_ids)
attention_mask = construct_attention_mask(input_ids)
indices = input_ids[0].detach().tolist()
all_tokens = tokenizer.convert_ids_to_tokens(indices)
ground_truth = 'to include, empower and support humans of all kinds'
ground_truth_tokens = tokenizer.encode(ground_truth, add_special_tokens=False)
ground_truth_end_ind = indices.index(ground_truth_tokens[-1])
ground_truth_start_ind = ground_truth_end_ind - len(ground_truth_tokens) + 1
start_scores, end_scores = predict(input_ids, \
token_type_ids=token_type_ids, \
position_ids=position_ids, \
attention_mask=attention_mask)
print('Question: ', question)
print('Predicted Answer: ', ' '.join(all_tokens[torch.argmax(start_scores) : torch.argmax(end_scores)+1]))