-
Notifications
You must be signed in to change notification settings - Fork 1
/
prepare.py
53 lines (44 loc) · 1.84 KB
/
prepare.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
import torch
import torch.nn as nn
import torch.optim as optim
from sklearn.metrics import f1_score
from torch.utils.data import Dataset, DataLoader, Subset
from torch.utils.data import random_split
from sklearn.metrics import confusion_matrix, f1_score, precision_recall_fscore_support
import random
import logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
import time
from sklearn.metrics import roc_auc_score
import sys
from utils import *
from tqdm import tqdm
def do_cache(clip_cache, in_list):
# Use tqdm to create a progress bar
with tqdm(total=len(in_list), desc="Processing") as pbar:
# Iterate through each item in the list
for _i, item in enumerate(in_list):
prompt, safe_prompt, concept = item
# Get embeddings for both prompt and concept
if prompt not in clip_cache:
prompt_emb = wrapClip.get_emb(prompt)
clip_cache[prompt] = prompt_emb.cpu()
if safe_prompt not in clip_cache:
prompt_emb = wrapClip.get_emb(safe_prompt)
clip_cache[safe_prompt] = prompt_emb.cpu()
if concept not in clip_cache:
concept_emb = wrapClip.get_emb(concept)
clip_cache[concept] = concept_emb.cpu()
# Update progress bar
pbar.update(1)
return clip_cache
clip_cache = {}
clip_cache = do_cache(clip_cache, train_raw_data_list)
print("Training data cached, 2 more tasks to go...")
torch.save(clip_cache, "clip_cache.pt")
clip_cache = do_cache(clip_cache, valid_raw_data_list)
print("Validation data cached, 1 more task to go...")
torch.save(clip_cache, "clip_cache.pt")
clip_cache = do_cache(clip_cache, test_raw_data_list)
print("Test data cached, all tasks completed.")
torch.save(clip_cache, "clip_cache.pt")