QueRE/mlp.py at main · dsam99/QueRE · GitHub

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
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
import torch
import numpy as np
from src.quere import ClosedEndedExplanationDataset, OpenEndedExplanationDataset, SquadExplanationDataset
from baselines.rep_dataset import RepDataset
from src.utils import get_mlp_results

import sys

import argparse

if __name__ == "__main__":

    # set random seed
    np.random.seed(0)
    torch.manual_seed(0)

    parser = argparse.ArgumentParser()
    parser.add_argument("--llm", type=str, default="llama-7b")
    parser.add_argument("--dataset", type=str, default="WinoGrande")
    parser.add_argument("--inv_cdf_norm", action="store_true", default=False, help="Use inverse cdf normalization")
    parser.add_argument("--random", action="store_true", default=False, help="Use random prompts")
    parser.add_argument("--gpt_exp", action="store_true", default=False, help="Use GPT explanations")
    parser.add_argument("--gpt_state", action="store_true", default=False, help="Use GPT state prompts")
    args = parser.parse_args()

    if args.dataset == "BooIQ":
        dataset = ClosedEndedExplanationDataset("BooIQ", args.llm, load_quere=True)
    elif args.dataset == "HaluEval":
        dataset = ClosedEndedExplanationDataset("HaluEval", args.llm, load_quere=True)
    elif args.dataset == "ToxicEval":
        dataset = ClosedEndedExplanationDataset("ToxicEval", args.llm, load_quere=True)
    elif args.dataset == "CommonsenseQA":
        dataset = ClosedEndedExplanationDataset("CommonsenseQA", args.llm, load_quere=False)
    elif args.dataset == "WinoGrande":
        dataset = ClosedEndedExplanationDataset("WinoGrande", args.llm, load_quere=True)
    elif args.dataset == "nq":
        dataset = OpenEndedExplanationDataset(args.llm, load_quere=True)
    elif args.dataset == "squad":
        dataset = SquadExplanationDataset(args.llm, load_quere=True)

    b = True

    rep_dataset = RepDataset(args.dataset, args.llm)
    train_rep = rep_dataset.train_rep
    test_rep = rep_dataset.test_rep

    train_data, train_labels, train_log_probs = \
        dataset.train_data, dataset.train_labels, dataset.train_log_probs

    test_data, test_labels, test_log_probs, = \
        dataset.test_data, dataset.test_labels, dataset.test_log_probs
    train_logits, train_pre_conf, train_post_conf = dataset.train_logits, dataset.train_pre_confs, dataset.train_post_confs
    test_logits, test_pre_conf, test_post_conf = dataset.test_logits, dataset.test_pre_confs, dataset.test_post_confs

    train_pre_conf = train_pre_conf.reshape(-1, 1)
    test_pre_conf = test_pre_conf.reshape(-1, 1)

    if "70b" in args.llm:
        # make all train only len 1000
        train_data = train_data[:1000]
        train_labels = train_labels[:1000]
        train_log_probs = train_log_probs[:1000]
        train_logits = train_logits[:1000]
        train_pre_conf = train_pre_conf[:1000]
        train_post_conf = train_post_conf[:1000]
        train_rep = train_rep[:1000]

    # reshape log probs
    train_log_probs = train_log_probs.reshape(len(train_data), -1)
    test_log_probs = test_log_probs.reshape(len(test_data), -1)


    #print shapes
    print(train_data.shape, train_labels.shape)
    print(train_log_probs.shape)

    results = {
        "logits_acc": [],
        "rep_acc": [],
        "logprob_acc": [],
        "preconf_acc": [],
        "postconf_acc": [],
        "exp_acc": [],
        "exp_all_acc": [],
        "logprob_f1": [],
        "rep_f1": [],
        "logits_f1": [],
        "preconf_f1": [],
        "postconf_f1": [],
        "exp_f1": [],
        "exp_all_f1": [],
        "logprob_ece": [],
        "rep_ece": [],
        "logits_ece": [],
        "preconf_ece": [],
        "postconf_ece": [],
        "exp_ece": [],
        "exp_all_ece": [],
        "logits_auroc": [],
        "rep_auroc": [],
        "logprob_auroc": [],
        "preconf_auroc": [],
        "postconf_auroc": [],
        "exp_auroc": [],
        "exp_all_auroc": [],
    }

    seeds = range(1)

    for seed in seeds:

        # set random seed
        np.random.seed(seed)
        torch.manual_seed(seed)

        # get results for logprob
        acc, f1, ece, auroc = get_mlp_results(train_log_probs, train_labels, test_log_probs, test_labels, seed=seed, balanced=b)
        results["logprob_acc"].append(acc)
        results["logprob_f1"].append(f1)
        results["logprob_ece"].append(ece)
        results["logprob_auroc"].append(auroc)

        # get results for preconf
        acc, f1, ece, auroc = get_mlp_results(train_pre_conf, train_labels, test_pre_conf, test_labels, seed=seed, balanced=b)
        results["preconf_acc"].append(acc)
        results["preconf_f1"].append(f1)
        results["preconf_ece"].append(ece)
        results["preconf_auroc"].append(auroc)

        # get results for postconf
        acc, f1, ece, auroc = get_mlp_results(train_post_conf, train_labels, test_post_conf, test_labels, seed=seed, balanced=b)
        results["postconf_acc"].append(acc)
        results["postconf_f1"].append(f1)
        results["postconf_ece"].append(ece)
        results["postconf_auroc"].append(auroc)

        # get results for logits
        acc, f1, ece, auroc = get_mlp_results(train_logits, train_labels, test_logits, test_labels, seed=seed, balanced=b)
        results["logits_acc"].append(acc)
        results["logits_f1"].append(f1)
        results["logits_ece"].append(ece)
        results["logits_auroc"].append(auroc)

        # get results for exp
        acc, f1, ece, auroc = get_mlp_results(train_data, train_labels, test_data, test_labels, seed=seed, balanced=b)
        results["exp_acc"].append(acc)
        results["exp_f1"].append(f1)
        results["exp_ece"].append(ece)
        results["exp_auroc"].append(auroc)

        # get reuslts for exp_all
        train_data_all = np.concatenate([train_data, train_log_probs, train_pre_conf, train_post_conf], axis=1)
        test_data_all = np.concatenate([test_data, test_log_probs, test_pre_conf, test_post_conf], axis=1)

        acc, f1, ece, auroc = get_mlp_results(train_data_all, train_labels, test_data_all, test_labels, seed=seed, balanced=b)
        results["exp_all_acc"].append(acc)
        results["exp_all_f1"].append(f1)
        results["exp_all_ece"].append(ece)
        results["exp_all_auroc"].append(auroc)

        # get results for rep
        acc, f1, ece, auroc = get_mlp_results(train_rep, train_labels, test_rep, test_labels, seed=seed, balanced=b)
        results["rep_acc"].append(acc)
        results["rep_f1"].append(f1)
        results["rep_ece"].append(ece)
        results["rep_auroc"].append(auroc)

    # compute means
    results = {k: np.mean(v) for k, v in results.items()}
    results = {k: round(v, 4) for k, v in results.items()}
    # for k in ["logits_f1", "rep_f1", "logprob_f1", "preconf_f1", "postconf_f1", "exp_f1", "exp_all_f1"]:
    for k in ["logits_auroc", "rep_auroc", "preconf_auroc", "postconf_auroc", "logprob_auroc", "exp_all_auroc"]:
        print(k, results[k])