args_model_utils.py

import torch
import spacy
import en_core_web_sm
from torch import nn
import math


device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")

from transformers import AutoModel, TrainingArguments, Trainer, RobertaTokenizer, RobertaModel
from transformers import AutoTokenizer

model_checkpoint = "ehsanaghaei/SecureBERT"

tokenizer = AutoTokenizer.from_pretrained(model_checkpoint, add_prefix_space=True)
roberta_model = RobertaModel.from_pretrained(model_checkpoint).to(device)

nlp = en_core_web_sm.load()
pos_spacy_tag_list = ["ADJ","ADP","ADV","AUX","CCONJ","DET","INTJ","NOUN","NUM","PART","PRON","PROPN","PUNCT","SCONJ","SYM","VERB","SPACE","X"]
ner_spacy_tag_list = [bio + entity for entity in list(nlp.get_pipe('ner').labels) for bio in ["B-", "I-"]] + ["O"]
dep_spacy_tag_list = list(nlp.get_pipe("parser").labels)
event_nugget_tag_list = ["Databreach", "Ransom", "PatchVulnerability", "Phishing", "DiscoverVulnerability"]
arg_nugget_relative_pos_tag_list = ["before-same-sentence", "before-differ-sentence", "after-same-sentence", "after-differ-sentence"]

class CustomRobertaWithPOS(nn.Module):
    def __init__(self, num_classes):
        super(CustomRobertaWithPOS, self).__init__()
        self.num_classes = num_classes

        self.pos_embed = nn.Embedding(len(pos_spacy_tag_list), 16)
        self.ner_embed = nn.Embedding(len(ner_spacy_tag_list), 8)
        self.dep_embed = nn.Embedding(len(dep_spacy_tag_list), 8)
        self.depth_embed = nn.Embedding(17, 8)
        self.subtype_embed = nn.Embedding(len(event_nugget_tag_list), 2)
        self.dist_embed = nn.Embedding(11, 6)
        self.relative_pos_embed = nn.Embedding(len(arg_nugget_relative_pos_tag_list), 2)

        self.roberta = roberta_model
        self.dropout1 = nn.Dropout(0.2)
        self.fc1 = nn.Linear(self.roberta.config.hidden_size + 50, num_classes)

    def forward(self, input_ids, attention_mask, pos_spacy, ner_spacy, dep_spacy, depth_spacy, nearest_nugget_subtype, nearest_nugget_dist, arg_nugget_relative_pos):
        outputs = self.roberta(input_ids=input_ids, attention_mask=attention_mask)
        last_hidden_output = outputs.last_hidden_state
        
        pooler_output = outputs.pooler_output
        pooler_output_unsqz = pooler_output.unsqueeze(1)
        pooler_output_fin = pooler_output_unsqz.expand(-1, last_hidden_output.shape[1], -1)


        pos_mask = pos_spacy != -100
        pos_embed_masked = self.pos_embed(pos_spacy[pos_mask])
        pos_embed = torch.zeros((pos_spacy.shape[0], pos_spacy.shape[1], 16), dtype=torch.float).to(device)
        pos_embed[pos_mask] = pos_embed_masked

        ner_mask = ner_spacy != -100
        ner_embed_masked = self.ner_embed(ner_spacy[ner_mask])
        ner_embed = torch.zeros((ner_spacy.shape[0], ner_spacy.shape[1], 8), dtype=torch.float).to(device)
        ner_embed[ner_mask] = ner_embed_masked

        dep_mask = dep_spacy != -100
        dep_embed_masked = self.dep_embed(dep_spacy[dep_mask])
        dep_embed = torch.zeros((dep_spacy.shape[0], dep_spacy.shape[1], 8), dtype=torch.float).to(device)
        dep_embed[dep_mask] = dep_embed_masked

        depth_mask = depth_spacy != -100
        depth_embed_masked = self.depth_embed(depth_spacy[depth_mask])
        depth_embed = torch.zeros((depth_spacy.shape[0], depth_spacy.shape[1], 8), dtype=torch.float).to(device)
        depth_embed[dep_mask] = depth_embed_masked

        nearest_nugget_subtype_mask = nearest_nugget_subtype != -100
        nearest_nugget_subtype_embed_masked = self.subtype_embed(nearest_nugget_subtype[nearest_nugget_subtype_mask])
        nearest_nugget_subtype_embed = torch.zeros((nearest_nugget_subtype.shape[0], nearest_nugget_subtype.shape[1], 2), dtype=torch.float).to(device)
        nearest_nugget_subtype_embed[dep_mask] = nearest_nugget_subtype_embed_masked

        nearest_nugget_dist_mask = nearest_nugget_dist != -100
        nearest_nugget_dist_embed_masked = self.dist_embed(nearest_nugget_dist[nearest_nugget_dist_mask])
        nearest_nugget_dist_embed = torch.zeros((nearest_nugget_dist.shape[0], nearest_nugget_dist.shape[1], 6), dtype=torch.float).to(device)
        nearest_nugget_dist_embed[dep_mask] = nearest_nugget_dist_embed_masked

        arg_nugget_relative_pos_mask = arg_nugget_relative_pos != -100
        arg_nugget_relative_pos_embed_masked = self.relative_pos_embed(arg_nugget_relative_pos[arg_nugget_relative_pos_mask])
        arg_nugget_relative_pos_embed = torch.zeros((arg_nugget_relative_pos.shape[0], arg_nugget_relative_pos.shape[1], 2), dtype=torch.float).to(device)
        arg_nugget_relative_pos_embed[dep_mask] = arg_nugget_relative_pos_embed_masked

        features_concat = torch.cat((last_hidden_output, pos_embed, ner_embed, dep_embed, depth_embed, nearest_nugget_subtype_embed, nearest_nugget_dist_embed, arg_nugget_relative_pos_embed), 2).to(device)
        features_concat = self.dropout1(features_concat)

        logits = self.fc1(features_concat)

        return logits


def tokenize_and_align_labels_with_pos_ner_dep(examples, tokenizer, label_all_tokens = True):
    tokenized_inputs = tokenizer(examples["tokens"], padding='max_length', truncation=True, is_split_into_words=True)
    #tokenized_inputs.pop('input_ids')
    ner_spacy = []
    pos_spacy = []
    dep_spacy = []
    depth_spacy = []
    nearest_nugget_subtype = []
    nearest_nugget_dist = []
    arg_nugget_relative_pos = []

    for i, (pos, ner, dep, depth, subtype, dist, relative_pos) in enumerate(zip(examples["pos_spacy"], 
                                                                                examples["ner_spacy"], 
                                                                                examples["dep_spacy"], 
                                                                                examples["depth_spacy"], 
                                                                                examples["nearest_nugget_subtype"], 
                                                                                examples["nearest_nugget_dist"], 
                                                                                examples["arg_nugget_relative_pos"])):
        word_ids = tokenized_inputs.word_ids(batch_index=i)
        previous_word_idx = None
        ner_spacy_ids = []
        pos_spacy_ids = []
        dep_spacy_ids = []
        depth_spacy_ids = []
        nearest_nugget_subtype_ids = []
        nearest_nugget_dist_ids = []
        arg_nugget_relative_pos_ids = []

        for word_idx in word_ids:
            # Special tokens have a word id that is None. We set the label to -100 so they are automatically
            # ignored in the loss function.
            if word_idx is None:
                ner_spacy_ids.append(-100)
                pos_spacy_ids.append(-100)
                dep_spacy_ids.append(-100)
                depth_spacy_ids.append(-100)
                nearest_nugget_subtype_ids.append(-100)
                nearest_nugget_dist_ids.append(-100)
                arg_nugget_relative_pos_ids.append(-100)
            # We set the label for the first token of each word.
            elif word_idx != previous_word_idx:
                ner_spacy_ids.append(ner[word_idx])
                pos_spacy_ids.append(pos[word_idx])
                dep_spacy_ids.append(dep[word_idx])
                depth_spacy_ids.append(depth[word_idx])
                nearest_nugget_subtype_ids.append(subtype[word_idx])
                nearest_nugget_dist_ids.append(dist[word_idx])
                arg_nugget_relative_pos_ids.append(relative_pos[word_idx])
            # For the other tokens in a word, we set the label to either the current label or -100, depending on
            # the label_all_tokens flag.
            else:
                ner_spacy_ids.append(ner[word_idx] if label_all_tokens else -100)
                pos_spacy_ids.append(pos[word_idx] if label_all_tokens else -100)
                dep_spacy_ids.append(dep[word_idx] if label_all_tokens else -100)
                depth_spacy_ids.append(depth[word_idx] if label_all_tokens else -100)
                nearest_nugget_subtype_ids.append(subtype[word_idx] if label_all_tokens else -100)
                nearest_nugget_dist_ids.append(dist[word_idx] if label_all_tokens else -100)
                arg_nugget_relative_pos_ids.append(relative_pos[word_idx] if label_all_tokens else -100)
            previous_word_idx = word_idx

        ner_spacy.append(ner_spacy_ids)
        pos_spacy.append(pos_spacy_ids)
        dep_spacy.append(dep_spacy_ids)
        depth_spacy.append(depth_spacy_ids)
        nearest_nugget_subtype.append(nearest_nugget_subtype_ids)
        nearest_nugget_dist.append(nearest_nugget_dist_ids)
        arg_nugget_relative_pos.append(arg_nugget_relative_pos_ids)

    tokenized_inputs["pos_spacy"] = pos_spacy
    tokenized_inputs["ner_spacy"] = ner_spacy
    tokenized_inputs["dep_spacy"] = dep_spacy
    tokenized_inputs["depth_spacy"] = depth_spacy
    tokenized_inputs["nearest_nugget_subtype"] = nearest_nugget_subtype
    tokenized_inputs["nearest_nugget_dist"] = nearest_nugget_dist
    tokenized_inputs["arg_nugget_relative_pos"] = arg_nugget_relative_pos 
    return tokenized_inputs

def find_nearest_nugget_features(doc, start_idx, end_idx, event_nuggets):
            nearest_subtype = None
            nearest_dist = math.inf
            relative_pos = None

            mid_idx = (end_idx + start_idx) / 2
            for nugget in event_nuggets:
                mid_nugget_idx = (nugget["startOffset"] + nugget["endOffset"]) / 2
                dist = abs(mid_nugget_idx - mid_idx)

                if dist < nearest_dist:
                    nearest_dist = dist
                    nearest_subtype = nugget["subtype"]
                    for sent in doc.sents:
                        if between_idxs(mid_idx, sent.start_char, sent.end_char) and between_idxs(mid_nugget_idx, sent.start_char, sent.end_char):
                            if mid_idx < mid_nugget_idx:
                                relative_pos = "before-same-sentence"
                            else:
                                relative_pos = "after-same-sentence"
                            break
                        elif between_idxs(mid_nugget_idx, sent.start_char, sent.end_char) and mid_idx > mid_nugget_idx:
                            relative_pos = "after-differ-sentence"
                            break
                        elif between_idxs(mid_idx, sent.start_char, sent.end_char) and mid_idx < mid_nugget_idx:
                            relative_pos = "before-differ-sentence"
                            break
            
            nearest_dist = int(min(10, nearest_dist // 20))
            return nearest_subtype, nearest_dist, relative_pos

def find_dep_depth(token):
            depth = 0
            current_token = token
            while current_token.head != current_token:
                depth += 1
                current_token = current_token.head
            return min(depth, 16)
        
def between_idxs(idx, start_idx, end_idx):
    return idx >= start_idx and idx <= end_idx