-
Notifications
You must be signed in to change notification settings - Fork 0
/
game_based_dataset.py
145 lines (119 loc) · 6.58 KB
/
game_based_dataset.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
import random
import numpy as np
import torch
import json
import os
import h5py
from torch.utils.data import Dataset
class GameBased(Dataset):
"""
Custom PyTorch Dataset class for the Twitch-FIFA dataset.
Args:
tokenizer (transformers.AutoTokenizer): Tokenizer to encode chat text and other inputs.
root (str, optional): Root directory path for the dataset. Defaults to "/media/livechat/game_based_dialogue/".
dataset_json (str, optional): JSON file containing dataset information. Defaults to "train.json".
video_feature_file (str, optional): File containing video features in h5 format. Defaults to "train_video_feat.h5".
comments_padding (int, optional): Maximum length of chat comments after padding. Defaults to 50.
nb_context_comments (int, optional): Number of context comments to include in the chat context. Defaults to 30.
mode (str, optional): Mode of operation ('train', 'eval', or 'gen'). Defaults to "train".
allow_special_token (bool, optional): Whether to add special tokens during tokenization. Defaults to True.
Attributes:
root (str): Root directory path for the dataset.
video_feature_file (str): File containing video features in h5 format.
ds_json (list): Loaded JSON data from the dataset file.
video_context_id (list): List of video context IDs.
chat_context (list): List of chat contexts.
response_indexes (list): List of response indexes.
responses (list): List of response messages.
h5f (h5py.File): File object for accessing video features.
tokenizer (transformers.AutoTokenizer): Tokenizer used for encoding text sequences.
comments_padding (int): Maximum length of chat comments after padding.
nb_context_comments (int): Number of context comments to include in the chat context.
allow_special_token (bool): Whether to add special tokens during tokenization.
mode (str): Mode of operation ('train', 'eval', or 'gen').
Methods:
__len__(): Returns the number of samples in the dataset.
__getitem__(index): Returns a single sample from the dataset based on the given index.
Note:
For evaluation mode ('eval'), the dataset must contain 'label' and 'response' fields in the JSON data.
Example:
tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
livechat = GameBased(tokenizer)
print(len(livechat)) # Print the number of samples in the dataset.
item = livechat[1] # Get the second sample in the dataset.
print(item["video_features"].size()) # Print the size of video features for the second sample.
"""
def __init__(
self,
tokenizer,
root="/media/livechat/game_based_dialogue/",
dataset_json="train.json",
video_feature_file="train_video_feat.h5",
comments_padding=50,
nb_context_comments=30,
mode="train",
allow_special_token=True
):
self.root = root
self.video_feature_file = video_feature_file
with open(os.path.join(root, dataset_json), "r") as data:
self.ds_json = json.load(data)
# self.ds_json = self.ds_json[:100]
self.video_context_id = [video["video_context_id"] for video in self.ds_json]
self.chat_context = [video["chat_context"].split("\u0c09") for video in self.ds_json]
self.response_indexes = [video["label"] for video in self.ds_json]
self.responses = [video["response"] for video in self.ds_json]
self.h5f = h5py.File(os.path.join(self.root,video_feature_file),'r')
self.tokenizer = tokenizer
self.comments_padding = comments_padding
self.nb_context_comments = nb_context_comments
self.allow_special_token = allow_special_token
self.mode = mode
def __len__(self):
return len(self.video_context_id)
def __getitem__(self, index):
chat_context = self.chat_context[index]
response = self.responses[index][np.argmax(self.response_indexes[index])]
chat_context_ids = []
chat_context_am = []
for _ in range(self.nb_context_comments):
random_index = random.randint(0, len(chat_context)-1)
chat_context_input = self.tokenizer.encode_plus(chat_context[random_index], add_special_tokens=self.allow_special_token, max_length=self.comments_padding, padding='max_length', truncation=True)
chat_context_ids.append(chat_context_input["input_ids"])
chat_context_am.append(chat_context_input["attention_mask"])
response_input = self.tokenizer.encode_plus(response, add_special_tokens=self.allow_special_token, max_length=self.comments_padding, padding='max_length', truncation=True)
response_ids = response_input["input_ids"]
response_am = response_input["attention_mask"]
video_features = self.__getvideofeatures__(index)
candidates_ids, candidates_am = self.__getcandidates__(index) if self.mode=="eval" else ([0], [0])
return {
"chat_context": torch.tensor(chat_context_ids, dtype=torch.long),
"chat_context_am": torch.tensor(chat_context_am, dtype=torch.long),
"response": torch.tensor(response_ids, dtype=torch.long),
"response_am": torch.tensor(response_am, dtype=torch.long),
"video_features": torch.tensor(video_features, dtype=torch.float),
"candidates": torch.tensor(candidates_ids, dtype=torch.long),
"candidates_am": torch.tensor(candidates_am, dtype=torch.long)
}
def __getvideofeatures__(self, index):
video_features = self.h5f[self.video_context_id[index]]
return np.array(video_features)
def __getcandidates__(self, index):
assert self.mode=="eval"
candidates = self.responses[index]
candidates_ids = []
candidates_am = []
for candidate in candidates:
candidate_input = self.tokenizer.encode_plus(candidate, add_special_tokens=self.allow_special_token, max_length=self.comments_padding, padding='max_length', truncation=True)
candidates_ids.append(candidate_input["input_ids"])
candidates_am.append(candidate_input["attention_mask"])
return candidates_ids, candidates_am
if __name__=="__main__":
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
livechat = GameBased(
tokenizer
)
print(len(livechat))
item = livechat[1]
print(item["video_features"].size())